The Royal Observatory Greenwich

…where east meets west

Home
Brief History
The Greenwich Meridian
Greenwich (1675–1958)
Herstmonceux (1948–1990)
Cambridge (1990–1998)
Outstations (1822–1971)…
- – Chingford (1822–1924)
- – Deal (1864–1927)
- – Abinger (1923–1957)
- – Bristol & Bradford on Avon (1939–1948)
- – Bath (1939–1949)
- – Hartland (1955–1967)
- – Cape of Good Hope (1959–1971)
Administration…
- – Funding
- – Governance
- – Inventories
- – Pay
- – Regulations
- – Royal Warrants
Contemporary Accounts
People
Publications
Science
Technology
Telescopes
Chronometers
Clocks & Time
Board of Longitude
Libraries & Archives
Visit
Search

Extracts from the Reports of the Astronomer Royal relating to the Time Service (1912–1964)

The following extracts are taken from the Annual Reports of the Astronomer Royal to the Board of Visitors. Other extracts are also available for Chronometers and the Buildings and Grounds (1836–1964).

The reports were presented at the annual visitation, which normally took place in the first week of June. The first was produced in 1836. The report for the following year established a format that remained much the same until 1963, (the year before the Board was abolished).

The reports from 1837 onwards were grouped into sections dealing with different topics. The bulk of material comes from the section dealing specifically with the time service. Also included are extracts below from other sections of the Reports that are relevant. Within these, there may be inadvertent omissions.

Whilst every effort has been made to check for transcription for errors, the text is not guaranteed error free. Most of the reports as published can be accessed on-line via this link. During the Second World War, a confidential addendum was tabled to each report containing information (relating mainly to chronometers, the time-service, the Nautical Almanac Office and magnetic and meteorological data) that might have been useful to the enemy. They were never published. Copies for the years 1940–42 are held at the National Archives in ADM190/19, ADM190/20 and ADM190/21 respectively.

The following items of possible relevance have been omitted

Transit A. This was on loan to the Cape Observatory until about 1950 when it was presumably returned. It is not mentioned as still being on loan in the 1951 and later Reports.
Transit E: the 1926 inventory states that Transit E was sold on 18 February 1929 for £100 to the Egytian Government with a 1902 stand (described as Stand 4)(RGO39/5/44)

1912

The Report here presented refers to the year from 1911 May 11 to 1912 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The strong room to the west of the Lower Computing Room was altered between July and November to make it serve as a clock room. A second wall was built separated from the outside walls to the West and South by an air space, and double doors were provided. The necessary extensions and alterations of the electric cables were carried out by the Engineering Department of the General Post Office during the latter part of October and early part of November.

The principal movable instruments are thus distributed: –

On loan-

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Dent No. 2010, at Devonport ; Arnold 2, at the University Observatory, Oxford.

II. Astronomical Observations: –

Clocks and Chronographs. – The Sidereal Standard Clock [Dent 1906] was dismounted from the Magnet House Basement on 1911 September 8, and taken away by Messrs. E. Dent & Co. for complete overhauling, preparatory to its erection in the New Clock Room. It was returned on October 20, and was brought into use again as the Standard Clock on December 4. During the period September 8-December 4 the clock Dent 2 was used as Standard.

The Mean Solar Clock, Dent 2012, was dismounted from the Ball Lobby on November 13, and cleaned by E. Dent & Co. at the Royal Observatory. It was re-erected in the New Clock Room and brought into use again on November 14

VIII. Time Signals: –

The Greenwich time-ball was not raised on two days (November 5 and December 10) owing to the violence of the wind, and could not be raised on one day (January 18) owing to accumulation of ice on the mast.

On November 13 no signal left the Royal Observatory at 13h owing to failure in the Mean Solar Clock springs. The Time Ball failed to drop, but was released by a false signal at three and a half minutes after 13h. It was raised again and dropped correctly by signal at 14h.

The general performance and accuracy of the drop of the time-ball at the Admiralty Signal Stations, and or the going of the Westminster Clock, is shown by the following table of the signals received at Greenwich: –

It should be borne in mind in comparing these results that the Deal time-ball is dropped directly from Greenwich at 13h; that the clocks at Portsmouth, Portland, and Devonport are corrected daily by the Greenwich lOh time-signal; and that the Westminster clock is not corrected daily.

Apparent Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Deal.	Westminster.
	Days.	Days.	Days.	Days.	Days.
Not greater than 0.2	230	209	250	–	58
Between 0.2 and 0.5	48	24	25	–	26
Between 0.5 and 1.0	3	18	9	–	72
Greater than 1.0	1	2	1	–	104*
Average error of signal for the year	±0.15s	±0.17s	±0.13s	–	±0.97s
Failures of return signal on Sunday	35	43	25	4	–
Total failures	64	87	66	14	6
Interuptions by telegraph signals on Sunday	6	6	5	3	–
Total interuptions	16	15	15	9	–

*Never exceeding 3.0s.

The above figures do not include the period January 5 to 8 (Portsmouth) and September 17 to 30 (Portland), when owing to special causes large errors occurred, the inclusion of which would give a wrong idea of the general accuracy.

As I was informed by the General Post Office that it would be more convenient for them if the return from Deal was sent at 10 seconds past one o'clock instead of at 5 seconds past, the time of this return signal has been changed. To effect this a wheelwork mechanism was designed which is set in motion by the actual drop of the ball, and which after a fixed interval transmits the-signal to Greenwich. It has acted very satisfactorily.

While these alterations were being made it was discovered that certain bearings and pinions of the raising apparatus, which were put up in 1853, showed considerable wear. These have been renewed and the machinery made quite effective, by the Works Department.

Lieutenant F. C. Hanning-Lee attended at the Observatory 'on 'the 9th, 10th, and 11th of May for a short course of practice with the transit instrument, before proceeding to Malta to take charge of the Chart and Chronometer Depôt there.

X. General Remarks: –

Special attention may be drawn to the successful conclusion of the long continued observations on the influence of magnetic fields on chronometers and watches. The Time Department of the Observatory is peculiarly well suited to carry out such research.

1913

The Report here presented refers to the year from 1912 May 11 to 1913 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford. (Dent No. 2010 has been transferred to Devonport on permanent. charge.)

VIII. Time Signals: –

An apparatus was set up on July 5 for the reception of the wireless time-signals from the Eiffel Tower and Norddeich. The signals have been constantly observed by Mr. Lewis and Mr. Bowyer since this date, the morning signals being observed each day (except Sundays). The night signals from the Eiffel Tower have been observed on 128 and the rhythmic signals on 82 occasions: The night signals from Norddeich have been observed on 124 occasions. The morning signals from the Eiffel Tower were observed by both Mr. Lewis and Mr. Bowyer on 167 days: there is a mean difference L - WB = -.066s in their times of observation with an accidental discordance of ± .06s between the observers.

Similarly in the receipt of the Norddeich Signals the two observers showed a mean personal difference L - WB = - . 043s and a similar accidental discordance. Thus the ordinary signals are observed by either observer with a mean error of less than ± 0.05s. The rhythmic signals are apparently received with an error of less than ± 0.01s, and the mean discordances between these and the ordinary signals are less than ± 0.05s.

As regards the actual difference between the time sent out by the Eiffel Tower and that of the Greenwich 10h and 1h signals, from 184 observations Mr. Lewis makes the Eiffel Tower Signal 0·256s late on Greenwich, and Mr. Bowyer from 234 observations makes it 0·313s late. It is supposed that the difference is mainly due to the difference of personal errors of the Standard Observers at Paris and Greenwich. The mean discordance after allowing for this constant difference is ± ·11s.

The following table shows the actual differences found during May, the mean of the two observers being taken after allowance for their constant difference: –

May 1	+0.35s	May 9	+0.46s	May 17	+0.20s	May 25	-
2	.42	10	.39	18	.28	26	+0.31s
3	.35	11	-	19	.26	27	.33
4	-	12	-	20	.28	28	.32
5	.46	13	.35	21	.28	29	.21
6	.43	14	.32	22	.29	30	.37
7	.30	15	.18	23	.28	31	.40
8	.34	16	.13	24	.29

The Norddeich Signals are according to 160 observations or Mr. Lewis 0.297s late on the Greenwich Time Signals, and 0.340s late according to 229 observations of Mr. Bowyer. Allowing for this. there is an accidental discordance of ± 0·23s.

On January 19 the Greenwich Time Ball was dropped 7.7s early, and the Time Signal was also this amount in error.

The general performance and. accuracy of the drop of the time-ball at the Admiralty Signal Stations, and of the going of the Westminster Clock, is shown by the following table of the signals received at Greenwich: –

Apparent Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Deal.	Westminster.
	Days.	Days.	Days.	Days.	Days.
Not greater than 0.2	239	220	261	–	53
Between 0.2 and 0.5	31	10	8	–	48
Between 0.5 and 1.0	4	1	4	–	83
Greater than 1.0	0	0	1	–	91*
Average error of signal for the year	±0.14s	±0.07	±0.07	–	±0.88
Failures of return signal on Sunday	43	49	23	17	–
Total failures	76	118	64	64	8
Interuptions by telegraph signals on Sunday	6	7	3	5	–
Total interuptions	14	15	25	11	–

* Exceeded 3.0 on one occasion.

It should be borne in mind in comparing these results that the Deal time-ball is dropped directly from Greenwich at I3h; that the clocks at Portsmouth, Portland, and Devonport are corrected daily by the Greenwich 10h time-signal; and that the Westminster clock is not corrected daily.

In consequence, of the fire at the G.P.O, the total failures are in excess of the average. In the case of Deal, the time was supplied direct from the Observatory by Telephone.

A few failures can also be ascribed to change of electrical apparatus in the G,P.O. and at Deal.

The electrical apparatus used for dropping the Time Ball at Deal has been completely overhauled. The wiring has been, renewed, the old instruments in use since 1853 replaced, and the arrangements in general brought more into line with modern methods.

The Westminster clock stopped at about 10h 40m a.m. on December 6 and 6h 52m a.m. on December 23. These stoppages were due to temporary adjustment of the clock.

X. General Remarks: –

The daily comparison of the Eiffel Tower signal affords a useful regular check on the time as determined at Greenwich. At the request of the Director of the Paris Observatory, this comparison has, since the installation of the receiving apparatus, been forwarded daily to him as a check on the rate of the clock at Paris. These comparisons are specially serviceable in cloudy weather. In October the Astronomer Royal attended the "Conference internationale de l'heure" as one of the British delegates, where the further development of the wireless times-service was discussed. The distribution of time in this way is of great value to navigators and is likely to be of importance in the determination of longitude on land.

1914

The Report here presented refers to the year from 1913 May 11 to 1914 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford.

VIII. Clocks and Time Service: –

The Clock room near the base of the pier of the Astrographic telescope has proved very satisfactory. It is in a sheltered part of the Observatory, and is insulated by double walls, so that there is little variation of temperature beyond a steady annual change from about 50° to 60° Fahrenheit. The Sidereal Standard and Mean Time Clocks have performed very satisfactorily since their removal to this room, owing to freedom from damp and the stability of the suspension of the pendulums. The Mean Solar Clocks, Shepherd and Graham, have been recently removed to this room.

The renewal of the Time-Ball and the general overhaul and reconstruction of its mechanism has been sanctioned. A new teak mast is being constructed for the Observatory by the Works Department of the Admiralty. Messrs. E. Dent & Co. are supplying a new aluminium ball, and a new winch for winding, and are repairing and adapting parts of the existing machinery. The winch will be removed from its present position near the Octagon Room to the Ball Lobby. (It may be noted that the Time Ball was first erected in 1833. An electric current from the clock was first used to drop it at 1 o'clock in 1852. In 1855 the Ball was blown down into the courtyard. Some repairs were made in 1895 when the chain broke during winding, and again, in August 1913, some temporary repairs were made to the ball.)

The signals from the Eiffel tower and Norddeich have been regularly received and compared with the Greenwich Time. The results to 1914 May 10, using the impersonal micrometer of the altazimuth as the standard for the personal equations of the Greenwich Observers, are as follows: –

Observer	Eiffel Tower				Norddeich
	No. of Obs.	Signal late onG.M.T.	Personal Equation	Mean Discordance	No. of Obs.	Signal late onG.M.T.	Personal Equation	Mean Discordance
		s	s	s		s	s	s
L (Morn,)	412	0.002	-	±.110	336	.016	-	±.225
WB (Morn.)	539	.044	-	±.114	526	.060	-	±.209
L-WB	375	-	.057	±.060	300	-	.051	±.075
Rhymic signals	175	.041		±.065	-	-	-	-
WB (Night)	256	.050		±.122	242	.062	-	±.214

A “field set" of Mr. S. G. Brown's relays and syphon recorder were obtained in March, and very satisfactory records have been obtained on 25 occasions.

There have been no large errors in the time signal sent to the G.P.O. during the year. The Greenwich Time-Ball was not dropped on March 16 and April 6 owing to the violence or the wind. On January 11 and January 31 it was not dropped owing to defects in the mechanism.

The accuracy of the time-balls at the Admiralty Signal Stations and of the Westminster Clock are shown by the following table of the errors of the return-signals received at Greenwich.

Error of Return Signal. s	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2	271	267	286	47
Between 0.2 and 0.5	14	9	10	48
Between 0.5 and 1.0	3	1	3	73
Greater than 1.0	0	1	0	87

The Westminster Clock was only on one occasion found to be more than 3.0s in error.

The return-signals are seldom received from Portsmouth and Portland on Sundays, and there are in the year a number of days on which some of the return-signals fail or are interrupted by telegraph signals. This year the Devonport clock was under repair during April.

The time-ball at Deal is dropped directly from Greenwich; the clocks at Portsmouth, Portland, and Devonport are corrected daily by the Greenwich 10h signal; the Westminster Clock is not corrected daily.

1915

The Report here presented refers to the year from 1914 May 11 to 1915 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The lamp lobby and a small room adjoining it, previously used as a store room, have been converted into a room for the transit-circle chronograph which has been transferred from a room on the North side of the court-yard. The external painting of the domes was carried out during February and March.

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford.

VIII. Clocks and Time Service: –

The Clock Grabam 2 has been fitted with an electro-magnetic corrector and contact springs to give a signal every minute. The Mean Solar Clock Shepherd has been dismounted from the Ball Lobby and sent to be overhauled. Dent 2014 and Dent 2017 have been cleaned.

The renewal of the Time-Ball and the general overhaul and reconstruction of its mechanism, which was sanctioned last year, has not yet been completed. The new teak mast has been constructed for the Observatory by the Works Department of the Admiralty. Messrs. E. Dent & Co. have supplied a new aluminium ball, and a new winch for winding, and are adapting parts of the, existing machinery.

The daily time signals from the Eiffel tower have been regularly observed by Mr. Lewis and Mr. Bowyer. The mean of the times as observed by Mr. Lewis is + 0·026s late on Greenwich Time from 209 observations, and by Mr. Bowyer +0.043s late from 280 observations. The difference of a quarter of a second between the Eiffel Tower signal and the time as determined by the transit circle which existed two years ago appears to have been the personal equation of the Standard Observer as compared with the new impersonal micrometer.

There have been no large errors in the time signal sent to the G.P.O. during the year. The Greenwich Time-Ball was not dropped on December 4 owing to, the violence or the wind.

The accuracy of the time-balls at the Admiralty Signal Stations and of the Westminster. Clock is shown by the following table of the errors of the return-signals received at Greenwich

Error of Return Signal. s	Portsmouth.	Portland.	Devonport.	Westminster.

Not greater than 0.2	271	251	293	21
Between 0.2 and 0.5	13	12	14	28
Between 0.5 and 1.0	1	0	4	65
Greater than 1.0	0	0	0	74

The Westminster Clock was on two occasions only found to be more than 3.0s in error. The clock was under repair from March 17 to May 5.

The clocks at Portsmouth, Portland, and Devonport are corrected daily by means of the Greenwich 10h signal; the Westminster Clock is not corrected daily.

The Time-Ball at Deal which is released by the 13h signal from Greenwich failed to drop accurately on four occasions. On three occasions the Ball was not raised. On 39 occasions the return-signal failed to reach Greenwich or was interrupted.

1916

The Report here presented refers to the year from 1915 May 11 to 1916 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford.

II. Astronomical Observations: –

Transit Circle. – The travelling wire micrometer has very largely freed the clock errors from personal equation.

The 28-inch Refractor. – The suspension springs of the pendulum of the water clock broke on January 20. New springs were applied and the clock was overhauled

VIII. Clocks and Time Service: –

The renewal of the Time-Ball mentioned m last year's report is again necessarily deferred.

The daily time signals from the Eiffel tower have been observed regularly by Mr. Lewis and Mr. Bowyer. The mean of the times as observed by Mr. Lewis is + 0·011s late on Greenwich Time from 221 observations, and by Mr. Bowyer + 0·003 from 291 observations. The accordance between the telegraphic signals from Greenwich and the wireless signals from the Eiffel Tower has been quite satisfactory.

On four occasions, December 10, December 17, January 1 and February 16, the Greenwich Time-Ball was not raised on account of the violence of the wind. On November 26, owing to failure in the action of the clips, the ball did not drop at 13h. It was dropped correctly at l4h.

The Time-ball at Deal, which is released by the I3h signal from Greenwich, failed to drop accurately on six occasions, and could not be raised on three days owing to wind.

The performance of the time-balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return-signals received at Greenwich: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	280	235	266	40
Between 0.2s and 0.5s	14	12	16	53
Between 0.5s and 1.0s	3	1	1	66
Greater than 1.0s	0	0	0	91

On the 250 days of observation the error of the Westminster Clock never exceeded 3.0s. The clocks at Portsmouth, Portland and Devonport are corrected daily by the 10h signal; the Westminster Clock is not corrected daily.

1917

The Report here presented refers to the year from 1916 May 11 to 1917 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford.

VIII. Clocks and Time Service: –

The renewal of the Time-Ball mentioned in last year's report is again necessarily deferred.

The daily time signals from the Eiffel tower have been observed regularly by Mr. Lewis and Mr. Bowyer. The mean of the times as observed by Mr. Lewis is 0.021s early on Greenwich Time from 199 observations, and by Mr. Bowyer 0.026s from 302 observations. The accordance between the telegraphic signals from Greenwich and the wireless signals from the Eiffel Tower has been quite satisfactory.

On two occasions, November 5 and March 7, the Greenwich Time-Ball was not raised on account of the violence of the wind. On November 12, owing to failure in the action of the clips, the ball did not drop at 13h. It was dropped correctly at 14h.

The Time-ball at Deal, which is released by the 13h signal from Greenwich, failed to drop accurately on four occasions, and could not be raised on two days owing to wind.

The performance of the time-balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return-signals received at Greenwich: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	261	250	297	64
Between 0.2s and 0.5s	18	20	10	73
Between 0.5s and 1.0s	1	4	3	111
Greater than 1.0s	0	0	0	33

On the 281 days of observation the error of the Westminster Clock once exceeded 3.0s. The clocks at Portsmouth, Portland and Devonport are corrected daily by the 10h signal; the Westminster Clock is not corrected daily.

1918

The Report here presented refers to the year from 1917 May 11 to 1918 May 10, and exhibits the state of the Observatory on the last named day.

I. Buildings and Grounds, Movable Property, and Library: –

Several small repairs have been made to …, the Astrographic dome, and that of the longitude pavilion.

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford.

VII. Time Service: –

Daily time signals have been sent to the General Post Office without failures. The signals from the Eiffel Tower have been received regularly, and a comparison with the Greenwich Clock sent daily to Paris.

The Sidereal Standard Clock was overhauled between July 23 and November 6, the Clock Dent II [2] being used as standard for this period.

On 7 occasions – August 28, September 29, October 7, October 25, February 9, February 28 and March I – the Greenwich Time-Ball was not raised on account of the violence of the wind. The renewal of the Time-Ball previously referred to is again necessarily deferred.

The Time-ball at Deal, which is released by the 13h signal from Greenwich, failed to drop accurately on eight occasions, and could not be raised on one day owing to wind.

The performance of the Time-balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return-signals received at Greenwich: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	286	256	299	53
Between 0.2s and 0.5s	8	31	11	68
Between 0.5s and 1.0s	4	5	2	109
Greater than 1.0s	1	0	0	49

On the 279 days of observation the error of. the Westminster Clock never exceeded 3.0s. The clocks at Portsmouth, Portland and Devonport are corrected daily by the 10h signal; the Westminster Clock is not corrected daily.

VIII. Personal Establishment: –

Mr. Lewis retired on 1917 September 30. Since 1882 he had had charge of the Time and Chronometer Department, and carried out his responsible duties with great efficiency. He also had charge of the 28-inch equatorial since its erection, and made important contributions to the study of double stars.

Mr. Bowyer was appointed Acting Assistant in charge of the Time and Chronometer Department on Mr. Lewis’ retirement.

1919

The Report here presented refers to the year from 1918 May 11 to 1919 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The renewal of the Time-Ball, referred to in the Report for 1914, but postponed on account of the war, has now been taken in hand, together with structural repairs needed to the Ball Turret. A new ball of aluminium, a new winch and part of the raising machinery, supplied by Messrs. E. Dent & Co., and a new teak must, constructed by the Works Department of the Admiralty, have been at the Observatory since 1914. The installation of the new apparatus and repairs of the turret are now in progress, the old Time-Ball having been put out of use on April 14 and dismounted.

The principal movable instruments are thus distributed: –

On loan –

Clocks. – Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford; Graham 2, for use in connection with the Time Ball at Rosyth.

VII.-Time Service: –

Daily time signals have been sent to the General Post Office without failures.

The signals from the Eiffel Tower have been received regularly, and a comparison. with the Greenwich Clock sent daily to Paris.

On 3 occasions – May 23, June 22, July 18 – the Greenwich Time-Ball was not raised on account of the violence or the wind. On June 16, owing to failure to set the clips in position, the ball would not remain at the top of the mast, and was in consequence lowered slowly by hand.

On February 18, although the clips were released by the 13-hour signal, the ball failed to drop. It was lowered slowly by hand about 13h 9m 30s.

Owing to the installation of a new Time-Ball and apparatus, the use of the time-Ball has been discontinued since April 14.

The Time-Ball at Deal, which is released by the 13h signal from Greenwich, failed to drop accurately on seven occasions.

The performance of the Time-Balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return-signals received at Greenwich: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	271	259	290	50
Between 0.2s and 0.5s	12	24	10	46
Between 0.5s and 1.0s	2	4	0	55
Greater than 1.0s	0	1	0	79

On the 232 days of observation the error of the Westminster Clock at mid-day exceeded 3.0s on one occasion. The clocks at Portsmouth, Portland and Devonport are corrected daily by the 10h signal; the Westminster Clock is not corrected daily.

IX.-General Remarks: –

A Conference, appointed by the Admiralty under the Chairmanship of the Hydrographer of the Navy, was held on 1917 June 21 to consider the advisability of adopting a system of Time Zones for time-keeping at sea. The Astronomer Royal attended this Conference as one of the representatives of the Admiralty. Acting upon a recommendation of the Conference, the Lords of the Admiralty have decided to establish in H.M. Navy a uniform method of time-keeping by zones, generally similar to that used on land. This decision was announced by Admiralty Order No. 1,291, dated 1919 April 9.

The Conference also put forward a recommendation that a day commencing at 0 hours midnight should be substituted in all nautical publications for the astronomical day commencing at 0 hours noon. The Admiralty [after consultation with the Royal Astronomical Society] has decided to adopt this recommendation. The necessary alterations will be made in the Nautical Almanac commencing in the year 1925, and in the Admiralty Tide Tables for 1920.

Much time has been spent during the last few months on the sidereal standard clock. The room in which it is mounted is dry, and a constant temperature is satisfactorily maintained. The barometric adjustment has been overhauled and appears to be working satisfactorily. But the going of the clock is not up to the standard of the best modern clocks, and it will be necessary to obtain a new one. Another instrumental change, which should be made as soon as possible, is a weight-driven and electrically-controlled clock for the 28-inch refractor.

1920

The Report here presented refers to the year from 1919 May 11 to 1920 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The external painting of the Astrographic, Sheepshanks, Altazimuth, and Thompson domes, the Transit Pavilion, the New Building, and the covered way of the Front Court, was commenced on April 10 and is nearly completed.

A number of structural repairs have been made in the Octagon Room building. The ball turret has been strengthened and partly rebuilt. New struts have been fixed to support the mast of the time-ball. The woodwork at the top of the staircase leading into the Octagon Room was found to have rotted. The lintel has been replaced by one of concrete, and the side walls have been replastered.

With a view to extending the range of reception of the Wireless Time Signals a new aerial was made at the Observatory during August. The aerial is of the two-wire type and consists of 7 strand No. 19 gauge phosphor-bronze wire with separators three feet long. Its erection was supervised by the Civil Engineer of the Royal Naval College. The length is 160 yards from the western corner of the New Observatory to the Octagon Room and a lead-in of 40 yards to the wireless room under the Sheepshanks dome.

The new aluminium time-ball was fitted in position on the turret on June 16, and the first trials of raising and dropping the ball were made on August 15. The winch for raising the ball has been removed from the small recess near the Octagon Room and is placed in the small room on the ground floor near the trigger apparatus. Part of the dead weight of the ball has been taken off the locking surfaces of the clips by introducing a counterpoise of 1 cwt., arranged in the ball turret. The new ball was brought into use on August 23, and the installation completed by Messrs. E. Dent & Co. on September 12.

The principal movable instruments are thus distributed: –

On loan –

On loan to the Royal Observatory –

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.
One seven-valve Amplifier and two Condensers from the Wireless Station at the Admiralty.

VII. Time Service: –

After sanction had been given by the Admiralty for the purchase of a new Standard Clock of British manufacture, an order was placed with Mr. E. T. Cottingham. The design of the clock includes an air-tight case, an invar pendulum, and a self-winding movement.

Daily time signals have been sent to the General Post Office without failure. The new time-ball, as previously stated, was brought into use on August 23. It has been in daily use since that date and has worked satisfactorily.

On 2 occasions – January 8 and January 11 – the time-ball was not raised on account of the violence of the wind.

The time-ball at Deal, which is released by the 13h signal at Greenwich, failed to drop accurately on 7 occasions, and on 3 days was not raised on account of the violence of the wind.

The performance of the time-balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	265	223	280	69
Between 0.2s and 0.5s	10	8	16	50
Between 0.5s and 1.0s	0	2	1	65
Greater than 1.0s	0	0	1	55

Of the 260 days of observation, the error of the Westminster Clock at mid-day exceeded 3.0s on one occasion only. The clocks at Portsmouth, Portland, and Devonport are corrected daily by the 10h Signal; the Westminster Clock is not corrected daily.

Wireless time signals have been received regularly from the Eiffel Tower throughout the year.

Since December, signals from Nauen and Annapolis have been received regularly, and those from Lyons and Darien are observed whenever possible. Many unsuccessful attempts have been made to detect the time signals sent from Honolulu and San Diego. An account of the reception of these signals was communicated to the Royal Astronomical Society, and will be published in the Monthly Notices. It is proposed in future to publish regularly in the Greenwich Volume the times of reception of all wireless time signals.

IX.-General Remarks: –

Mr. Sargent, of the Durham University Observatory, was at Greenwich from June 23 to July 17, studying time determination and the methods employed In deriving latitude variation from the observations with the Zenith telescope.

1921

The Report here presented refers to the year from 1920 May 11 to 1921 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The external painting of the Astrographic, Sheepshanks, Altazimuth, and Thompson domes, the Transit Pavilion, the New Observatory, and the covered way of the Front Court, was completed on May 25.

The internal wiring of the Observatory has been taken over by the Admiralty from the General Post Office as from April 1.

The principal movable instruments are thus distributed: –

On loan –

On loan to the Royal Observatory –

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.
One seven-valve Amplifier and two Condensers from the Wireless Station at the Admiralty.

The Observatory is indebted to Lieut.-Commander R. T. Gould, R.N., of the Hydrographic Department, for the restoration of Harrison's first Time machine. In connection with work on the history of the chronometer, permission was given him to examine records and horological instruments of historical interest at the Observatory. He kindly volunteered to take to pieces, clean and reassemble the instrument in a condition to prevent future deterioration. This historic instrument is now placed in the Octagon Room.

VII. Time Service: –

The new Standard Clock, mentioned in the last report, was delivered at the Observatory by Mr. Cottingham on February 26 and is now under test. A clock made by Riefler, of Munich, purchased from the University of Manchester, was set up on July 30.

Daily time signals have been sent to the General Post Office without failure.

On one occasion – December 4 – the time-ball was not raised on account of the violence of the wind. On six occasions – May 13, 15, June 19, 20, September 12 and December 12 – the ball was not dropped correctly at 13h owing to various causes. From September 13 – September 16 the time-ball was out of use.

The 13h signal from Greenwich failed to release the Deal time-ball on one occasion. On four days –June 14, November 14, January 19 and May 10 – the Greenwich signal failed to reach Deal. On these dates the ball was released by hand.

The performance of the time-balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	277	236	276	88	54
Between 0.2s and 0.5s	3	9	12	8	62
Between 0.5s and 1.0s	1	2	2	5	82
Greater than 1.0s	0	0	0	1	73

Of the 274 days of observation, the error of the Westminster Clock at midday never exceeded 3.0s. The clocks at Portsmouth, Portland, Devonport and Rosyth are corrected daily by the 10h Signal; the Westminster Clock is not corrected daily.

The connection between Greenwich and Rosyth for return signal was brought into operation during the latter part of October.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, and Annapolis throughout the year, and observations have been made of signals from Lyons and Darien as often as possible.

Special series of rhythmic signals from Lyons, for longitude determination, were observed during the periods June 21–July 4, November 16–23, December 5–20 and April 20–May 10. During the last period special signals from Annapolis and Bordeaux were also observed.

A very low frequency amplifier, designed for operating a syphon recorder for registering time signals, was brought into use during November. The apparatus is suitable for the reception of wireless signals if transmitted by spark, or continuous wave from high frequency alternator. The Paris and Nauen signals are usually registered and the last three special series of rhythmic signals sent from Lyons for longitude determination in Australia were recorded.

VIII. Personal Establishment: –

Mr. Bowyer was promoted Assistant on June 11 with effect from January 10.

IX. General Remarks: –

During the year much time has been occupied with the repair of instruments and the installation of new ones. … . Two new clocks, one by Mr. Cottingham and the Riefler clock bought from the Manchester University, have been set up in the Clock Room.

1922

The Report here presented refers to the year from 1921 May 11 to 1922 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

In the Report for 1921 it was mentioned that the maintenance of batteries and internal wiring of the Observatory for clock, observation, and time-signal circuits had been transferred from the Post Office to Admiralty charge. During the year the electrical engineer, Chatham Dockyard, has carried out the renewal of the wiring and substituted a central accumulator battery in place of separate dry cell batteries for each circuit.

Certain portions of the wiring of the Observatory had been in use nearly forty years, from the time. the G.P.O. took over maintenance, and the last renewal of any part of the wiring was carried out in 1905. At that time it was found necessary to renew the wires between the S.E. Dome and the Altazimuth. The cables then brought into use were carried in casing on the top of the closed wooden fence dividing the footpath connecting the old and new buildings from the Observatory grounds. The cables now brought into use have been placed in an underground duct through this section, and advantage has been taken of this to remove the fence, so leaving the whole Observatory grounds more open.

The wiring in the Clock Room has been left as installed in 19l1, the connections to the new wiring being made on a small through junction board near the door of the room. With the exception of the short length of wire in the Clock Room, the conductors throughout the Observatory are homogeneous and are carried in lead-cased cables, the insulation being composed of an enamel coating and a wrapping of waxed cotton.

The general scheme of the circuits has not been altered to any considerable extent, but the substitution of higher resistance relays in two of the primary clock circuits should be advantageous.

A small room on the ground floor, under the photoheliograph dark room, has been utilised for the battery room instead of the basement under the Ball Lobby.

Two sets of eighteen accumulators have been installed, arranged so that one set can be substituted for the other as occasion requires. The total voltage of 36 is used hourly for the Post Office signal and twice a day for the signal from the Westminster clock. For the Observatory Circuits lower voltages are employed and these are obtained by tappings taken off the battery at suitable positions. The grouping of the cells has been arranged to secure as far as possible an equal output from each cell. A compound change-over switch is provided for changing one set of accumulators for the other, the contacts being arranged that no break in circuits occurs during the operation. A control board for charging the accumulators in position has been provided, the necessary direct current being supplied by the motor generator already in use at the Observatory.

The various battery leads are connected to blocks mounted on a test frame in the Lower Computing Room near the Time Desk. From this point the supply of current is obtained for the various circuits operating. In addition to carrying strip connections to all circuits, sixteen relays are also mounted on the test frame. Normally twelve of these are in use on the Standard Sidereal Clock, hourly signal, minute signal and acknowledgment signal circuits.

The principal movable instruments are thus distributed: –

On loan –

On loan to the Royal Observatory –

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.
One seven-valve Amplifier and two Condensers, from the Wireless Station at the Admiralty.
One three-valve Amplifier from Mr. Witt Burnham.
A battery of 72 Leelanche cells, from the Director of Scientific Research of the Admiralty.

In the previous report it was recorded that Lieut.-Commander R. T. Gould, R.N. had restored Harrison's first Time machine. The Observatory again records its indebtedness to this officer for cleaning Harrison's original chronometer. The chronometer, along with the copy made by Larcum Kendal, is now preserved at the Observatory in a special case.

VII. Time Service: –

The new Standard Clock, mentioned in the last report, which was delivered at the Observatory by Mr. Cottingham on 1921 February 26, is still under test. During the year the clock Dent, 2,016 [2016], has been fitted with contact springs and correcting apparatus for use as a duplicate controlling clock for mean time circuits.

Daily time signals have been sent to the General Post Office without failure, exoept on one occasion, at 10 a.m.

On two occasions, Feb. 4 and April 3, the time-ball was raised, but could not be operated properly owing to the accumulation of snow at the top of the mast. On Feb. 5 the ball was not raised for a similar reason. On April 2, owing to improper working of the raising gear, the ball was not dropped correctly at 13h.

On two days, March 8 and 22, the Deal Time Ball was not raised on account of the violence of the wind. On six days, May 2, 10, Oct. 23, Dec. 6, Feb. 19, April 14, the Greenwich signal failed to reach Deal. On these dates the ball was released by hand, except on Oct. 23.

The performance of the time balls at the Admiralty Signal Stations and of the Westminster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	282	258	294	206	61
0.3s to 0.5s	1	8	8	14	49
0.6s to 1.0s	1	3	3	1	62
Greater than 1.0s	0	0	0	0	65

On the 237 days of observation, the error of the Westminster Clock at midday never exceeded 3.0s. The clocks at Portsmouth, Portland, Devonport and Rosyth are corrected daily by the l0h signal; the Westminster Clock is not corrected daily.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen and Annapolis throughout the year, and observations have been made of signals from Lyons, Moscow and Bordeaux as often as possible.

A special series of rhythmic signals from Lyons, for longitude determination, was observed during the period June 20 – July 12.

Signals from Eiffel Tower, Nauen, Bordeaux, Lyons and Moscow are usually automatically recorded on a syphon recorder operated by means of an oscillating valve relay system.

A summary of all Wireless Time Signals observed from 1921 July to 1921 December was communicated to the Royal Astronomical Society in 1922 January. In future it is intended to publish the results of the time signals in the annual volume.

IX. General Remarks: –

The re-wiring of the Observatory referred to in the first section of the report involved an extensive reorganisation of the existing arrangements. The Electrical Engineer, Chatham Dockyard, gave careful consideration to the requirements of the Observatory, and with the co-operation of his department a satisfactory modern installation has been introduced.

1923

The Report here presented refers to the year from 1922 May 11 to 1923 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The external painting of the domes of the 28··inch, Astrographic and Sheepshanks equatorials" and of the Transit Pavilion was completed on 1922 October 31. The inside of the 28-inch dome was painted in November.

The principal movable instruments are thus distributed: –

On loan –

On loan to the Royal Observatory –

VII.-Time Service: –

The new Standard Clock, mentioned in the last report, which was delivered at the Observatory by Mr. Cottingham on 1921 February 26, was brought into use on October 24 and has performed quite satisfactorily.

The driving mechanism of the Transit Circle Chronograph was thoroughly overhauled and repaired during January and February,

Daily time signals have been sent to the General Post Office without failure.

The time-ball has been dropped correctly at 13h, except on May 28, when it failed owing to improper working of the raising gear, and on June 14 owing to the failure of the signal current to release the trigger apparatus.

On six days, May 5, October 19, 29, February 7, 18, April 8, the Deal Time Ball was not raised on account of the violence of the wind. On six days, May 9, September 19, 20, December 10, February 4, March 29, the Greenwich signal failed to reach Deal. On these dates the ball was released by hand, except on September 19 and 20. From May 1–3 the ball was out of action on account of repairs, and on two days, May 20 and December 6, the ball was released by hand on account of telegraphic interruptions on the line.

The performance of the time balls at the Admiralty Signal stations and of the Westminster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	204	255	254	41	48
0.3s to 0.5s	8	4	4	11	58
0.6s to 1.0s	5	2	1	3	97
Greater than 1.0s	0	0	0	0	90

During the period 1922 August to 1923 March only a few signals were received from Rosyth owing to difficulties in transmission and alteration in route of the telegraphic line.

Signals were received from the Westminster Clock on 293 days and the error exceeded3.0s on three occasions. The clocks at Portsmouth, Portland, Devonport and Rosyth are corrected daily by the 10h signal; the Westminster Clock is not corrected daily.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis throughout the year, and from Moscow during the winter months.

Signals from the Eiffel Tower, Nauen, Bordeaux, .and Moscow are automatically recorded on a syphon recorder operated by means of an oscillating valve relay system, and since the beginning of August records of the Annapolis signals have been obtained on fifty per cent. of the possible number of days.

The concluded monthly means of comparisons of the various wireless time, signals, using automatic registration at Greenwich, for the year 1922 are as follows: –

Month.	Paris.	Nauen	Annapolis.	Bordeaux.
	10h Rhytmic.	(Noon).	17hr.	20h Rhythmic.
	s.	s.	s.	s.
January	-0.01	-0.05	-	+0.06
February	+0.10	-0.02	-	+0.11
March	+0.09	0.00	-	+0.12
April	+0.05	-0.04	-	+0.09
May	+0.11	+0.01	-	+0.12
June	+0.12	-0.02	-	+0.20
July	+0.11	-0.01	-	+0.19
August	+0.14	+0.08	-+0.05	+0.19
September	+0.13	+0.04	+0.07	+0.18
October	+0.12	+0.01	+0.01	+0.15
November	+0.16	+0.04	+0.02	+0.17
December	+0.11	0.00	0.00	+0.12
_____	_____	_____	_____	_____
Average number of comparisions per month	21	21	16	17

The sign + means late on Greenwich.

1924

The Report here presented refers to the year from 1923 May 11 to 1924 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Clocks.-Dent No. 1916 and Dent No. 2013, at the Cape Observatory; Dent No. 2011, at the Kew Observatory; Arnold 2, at the University Observatory, Oxford; Graham 2, for use in connection with the Time Ball at Rosyth.

On loan to the Royal Observatory-

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.

The Observatory is indebted to Mr. Witt Burnham, who has kindly presented the three-valve Amplifier which had been on loan from him.

Reference has been made in previous reports to the generosity of Lieut-Commander R. T. Gould, R.N., in overhauling Harrison's first time machine and the chronometer which obtained the prize offered by the Board of Longitude. Commander Gould now has in hand the two machines made by Harrison subsequent to his first attempt, and expects to complete them in a few months.

VII. Time Service: –

Two new sets of accumulators in banks of eighteen each were installed in 1923 October for use in the Time Service.

The gearing of the drums of the recording chronograph has been renewed. The Standard Sidereal Clock Dent 1906 was cleaned during October 1923. The clock Dent 2 used as the controlled clock with the Cottingham Standard Clock is now under repair.

The Mean Time Clock Dent 2016 has been fitted with suitable control switches for the transmission of time signals to the British Broadcasting Company.

Signals can be sent to the Company's studio at 2, Savoy Hill, every half-hour, and are broadcast, usually, once during the afternoon and once during the evening. The broadcasting of time from the Royal Observatory commenced on February 5.

A precision clock of Riefler pattern, in air-tight case, and a clock with Graham Escapement, also in an air-tight case, were under test from 1923 February to 1924 March. These clocks, required for the Time Service at Mauritius, were tested at the request of the Crown Agents for the Colonies.

Both clocks have given a satisfactory performance, and after inspection by Mr. WaIter, the Director of the Royal Alfred Observatory, during his visit to England, were despatched for shipment to Mauritius.

Daily time signals have been sent to the General Post Office without failure. The Greenwich Time Ball has been dropped correctly at 13h with six exceptions. On August 16 and September 15 the current failed to release the trigger apparatus; on September 16 and September 18 the ball was not raised owing to defects in the machinery; on January 9 it could not be raised owing to severe frost; and on April 7 an error was made in working the raising gear.

The Deal Time Ball was not raised owing to the violence of the wind on six days, November 10, November 15, February 13, February 14, April 2, April 3. On ten days the current from Greenwich failed to reach Deal. On these days the ball was released by hand except on two occasions. On two days, November 19 and April 24, the ball was released by hand on account of telegraphic interruptions on the line.

The performance of the time balls at the Admiralty Signal stations and of the Westminster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	253	239	273	164	77
0.3s to 0.5s	11	15	5	15	73
0.6s to 1.0s	1	4	1	4	70
Greater than 1.0s	0	0	0	0	72

The Westminster Clock was under repair during the periods May 18 – May 30 and June 23–28. The error never exceeded 3.0 on any of the 292 days on which signals were received. The clocks. at Portsmouth, Portland, Devonport and Rosyth are corrected daily by the 10h signal from Greenwich.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis, and from Moscow during the winter months. They are registered automatically on a syphon recorder operated by an oscillating valve relay system. The signals from Annapolis were recorded on 60 per cent. of the possible number of days during 1923.

The monthly means of the differences from Greenwich shown by the various wireless time-signals, using the final clock errors, and in addition correcting the Annapolis signals for lag on the land line from Washington to Annapolis, 0.02s for time of travel, and 0.05s for difference in the adopted right ascensions of the clock stars, are as follows: –

Month	Rhytmic	Time Signals.	Ordinary	Time	Signals.
	Paris	Bordeaux	Paris	Nauen	Annapolis
	10h.	20h.	10h 45m	Noon.	17hr.
	s	s	s	s	s
January	+.012	+.014	+.07	+.07	+.06
February	+.04	+.05	-.03	+.02	.00
March	+.02	+.02	-.09	-.02	.00
April	+.02	+.03	-.04	-.01	+.01
May	+.05	+.07	-.02	-.02	+.06
June	+.02	+.06	-.04	-.09	.00
July	+.08	+.12	+.02	.00	+.09
August	+.12	+.12	+.02	.00	+.09
September	+.13	+.13	+.04	+.06	+.08
October	+.13	+.15	+.08	+.02	+.11
November	+.17	+.18	+.11	+.06	+.03 (3)
December	+.14	+.15	+.09	+.08	+.04 (6)
_____	_____	_____	_____	_____	_____
Mean	+.090	+.103	+.021	+.014	+.046
Number of comparisions	289	319	298	265	242

The sign + means late on Greenwich.

1925

The Report here presented refers to the year from 1924 May 11 to 1925 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

A cellar underneath the Octagon Room has been rendered suitable for a standard clock referred to in a later section. This room was chosen on account of its very thick walls and freedom from temperature changes. The room was rather damp, the moisture apparently coming through the thick outside wall. It was thought that this was due to wet percolating between the tiles on the north terrace. These were taken up, and a layer of concrete 6 inches thick placed underneath and the tiles relaid. The work was completed by February 13, but the room is not yet sufficiently dry.

The principal movable instruments are thus distributed: –

On loan –

Harrison's time machine No. 2. to the British Empire Exhibition, Wembley.

Larcum Kendall's Chronometer (copy of the one for which Harrison was awarded the Board of Longitude's prize) also to the Exhibition.

On loan to the Royal Observatory-

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.

Harrison's second time machine has been completely taken to pieces, every part cleaned and repaired, and the whole of this very delicate instrument put together again by Lieut.-Commander Gould, R.N. The instrument has been' going well for weeks, and is now on exhibition at Wembley. The Observatory is greatly indebted to Commander Gould's skill and enthusiasm for the restoration of this beautiful piece of mechanism.

VII.- Time Service: –

In the course of the year a new standard clock was purchased. The installation consists of a master pendulum controlling a slave clock, the system being designed by Mr. W. H. Shortt and manufactured by the Synchronome Company. The master pendulum is mounted in a chamber two storeys below the Octagon Room and the slave clock in the usual clock room.

The system has been in use since January 1 and the results are very satisfactory. The Standard Clock, Cottingham, has been repaired recently.

The Mean Time Clock Dent 2012 has been fitted with suitable control switches for the transmission of time signals to the British Broadcasting Company. Thus both Mean Time Clocks are now available for this service. Signals can be sent to the Company's studio at 2, Savoy Hill, every half-hour, and are broadcast, usually twice during the afternoon and once during the evening.

A precision clock of Riefler pattern, in air-tight case, was under test from 1924 June to 1924 September. This clock, required for the Time Service at Hong Kong, was tested at the request of the Crown Agents for the Colonies. The clock gave a satisfactory performance, and has been despatched for shipment to Hong-Kong.

A clock by J. Shelton, lent by the Hydrographer, has been fitted with suitable contact springs and controlling apparatus and has been installed for use as standard clock in the time scaling of records at the Magnetic Station at Abinger.

Daily time signals have been sent to the General Post Office without failure.

The Greenwich Time Ball has been dropped correctly at 13h without exception. On three occasions, November 27, December 27, and January 2 the Time Ball could not be raised on account of high wind.

The Deal Time Ball was not raised owing to the violence of the wind on eight days, May 24, October 30, December 27, January 2, January 4, January 29, March 3, April 16. On six days the current from Greenwich failed to reach Deal. On these days the ball was released by hand. On one day, January 3, the ball was released by hand on account of telegraphic interruptions on the line, and on two days, November 18 and December 6, the Greenwich current failed to release the ball.

The performance of the time balls at the Admiralty Signal stations and of the Westminster Clock is shown by the following table of the errors of the return signals: -

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	225	259	272	126	82
0.3s to 0.5s	15	7	7	7	85
0.6s to 1.0s	9	1	1	0	86
Greater than 1.0s	2	0	0	0	49

The Westminster Clock was under repair during the period, January 20 – January 25. The error exceeded 3.0s on two days of the 304 days on which signals were received. The clocks at Portsmouth, Portland0, Devonport and Rosyth are corrected daily by the 10h signal from Greenwich.

The acknowledgment signal from Rosyth has not been received since the end of December.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis, They are registered automatically on a syphon recorder operated by an oscillating valve relay system. The signals from Annapolis were recorded on 50 per cent. of the possible number of days during 1924.

The monthly means of the differences from Greenwich shown by the various wireless time-signals, using the final clock errors, and in addition correcting the Annapolis signals for lag on the land line from Washington to Annapolis, 0. 02s for time of travel, and 0.05s for difference in the adopted right ascensions of the clock stars, are as follows: –

Month	Rhytmic	Time	Signals.	Ordinary	Time	Signals.
	Paris	Nauen	Bordeaux	Paris	Nauen	Annapolis
	10h.	12h.	20h.	10h 45m	Noon.	17h.
	s		s	s	s	s
January	+0.06	-	+0.06	+0.05	+0.05	+.0.05
February	+.06	-	+.07	+.03	.00	+.03
March	+.08	-	+.09	+.03	+.03	+.03
April	+.09	-	+.10	+.07	+.04	+.06
May	+.15	-	+.14	+.12	+.08	+.13
June	+.13	-	+.13	+.11	+.07	+.12
July	+.01	+.05	+.12	+.09	+.05	+.14
August	+.08	+.06	+.09	+.08	+.04	+.09
September	+.08	-.02	+.07	+.07	-.03	+.03(3)
October	.00	00	+.01	-.01	-.01	+.05
November	+.03	-.02	+.04	+.03	-.01	+.07
December	+.05	-.01	+0.6	+.07	.00	+.05
_____	_____	_____	_____	_____	_____	_____
Mean	+.076	+.003	+.081	+.061	+.024	+.067
Number of comparisions	286	119	278	297	278	194

The sign + means late on Greenwich. The reference point for the rhythmic SIgnals is the commencement of the signals and not the middle as in previous reports.

1926

The Report here presented refers to the year from 1925 May 11 to 1926 May 10, and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The walls and floor of the cellar under the Octagon room, in which the Standard Clock is mounted, have been rendered with a waterproof cement. Ventilation by a four-inch shaft extending to the roof of the building has also been arranged. Nevertheless, the cellar remains damp. As the clock is in a sealed case, moisture in the atmosphere, though inconvenient, has no effect on the clock. The room has been found very suitable for a standard clock both on account of the substantial foundation on which the clock is mounted, and also of its freedom from rapid changes of temperature.

The masts supporting the wireless aerial have been renewed.: The aerial has been raised about 6 feet and a single wire has been substituted for a double one.

The principal movable instruments are thus distributed: –

On loan –

Harrison's Time Machine No. 2, Whewell's Anemometer, Airy's Dip Circle, and the large Bar Magnets in use from 11841 to 1915, at the Science Museum.

Clocks.-Dent No. 1916 and Dent 2013 at the Cape Observatory; Arnold 2 at the University Observatory, Oxford.

Clock Dent 2011 has been permanently transferred to the Air Ministry in exchange for magnetic apparatus of equivalent value. Clock Graham 2 has been returned by the Admiralty in consequence of the closing of the Chart and Chronometer Depot at Rosyth Dockyard.

On loan to the Royal Observatory-

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.

To meet the requirements. of the Exchequer and Audit Department a new inventory of instruments and movable property has been made.

VII. Time Service: –

An additional standard clock similar in all essentials to the clock purchased during 1924-25 has been ordered. The master pendulum will be mounted in the cellar under the Octagon Room and the slave clock in the clock room.

The daily time signals to the General Post Office and the British Broadcasting Company have been transmitted regularly throughout the year.

The Time Balls at Greenwich and Deal have been operated regularly.

The performance of the time balls at the Admiralty Signal stations and of the Westniinster Clock is shown by the following table of the errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Rosyth.	Westminster.
Not greater than 0.2s	272	267	242	44	95
0.3s to 0.5s	0	5	4	4	98
0.6s to 1.0s	0	0	2	0	81
Greater than 1.0s	0	0	5	0	29

The error of the Westminster Clock never exceeded 2.0s on any of the 306 days on which signals were received. The clocks at Portsmouth, Portland, Devonport and Rosyth are corrected daily by the 10h signal from Greenwich.

The acknowledgment signal from Rosyth was in operation only between October and January and has now been discontinued.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. They are registered automatically on a syphon recorder operated by an oscillating valve relay system. The signals from Annapolis were recorded on 49 per cent. of the possible number of days during 1925.

Month	Rhytmic	Time	Signals.	Ordinary	Time	Signals.
	Paris	Nauen	Bordeaux	Paris	Nauen	Annapolis
	10h.	12h.	20h.	10h 45m	Noon.	17h.
	s	s	s	s	s	s
January	+0.17	+0.03	+0.15	+0.16	+0.01	+0.06
February	+.08	+.02	+.06	+.06	.00	+.02
March	+.05	+.05	+.04	+.04	+.03	+.05
April	+.15	+.10	+.13	+.13	+.09	+.10
May	+.14	+.07	+.14	+.13	+.10	+.10
June	+013	+.07	+.09	+.12	+.08	+.07
July	+.14	+.04	+.10	+.13	+.04	+.08
August	+.14	+.06	+.11	+.14	+.06	+.10
September	+.17	+.11	+.05	+.21	+.10	+.12
October	+.10	+.04	+.08	+.03	+.02	-
November	+.05	+.01	+.03	+.05	-.01	+.04
December	+.06	-.00	+.04	+.04	.00	.00
_____	_____	_____	_____	_____	_____	_____
Mean	+.114	+.050	+.092	+.101	+.041	+.064
Number of comparisions	277	272	310	286	270	179

The sign + means late on Greenwich. The reference point for the rhythmic signals is the commencement of the signals.

IX. General Remarks: –

Observations for a redetermination of the longitude Greenwich-Pulkovo were made by MM. Beljaieff and Dnieprovsky during the months September to November. Their . instruments, Repsold reversible transits with axis view and self-registering micrometers, were mounted in the transit pavilion. A large number of full nights of observation were secured, but the result of the determination has not yet been published.

The second Free Pendulum Clock mentioned earlier in the Report was delivered on May 20, and is now being adjusted. The Free Pendulum Clock Shortt No. 3 brought into use on January 1, 1925 has been very satisfactory. The daily losing rate has increased slightly during this period. From September 1 to the present time this increase has been very uniform and amounts to 0.011s per day per month. With the exception of a period during February and March when work was in progress in the cellar in connection with the erection of the second clock, the daily rate calculated from 5-day intervals has not varied more than 0.01s from the mean value. Owing to the reliability of this Standard Sidereal Clock and to the frequent comparisons made with it a very accurate daily time service has been maintained.

1927

The Report here presented refers to the year from 1926 May 11 to 1927 May 10,. and exhibits the state of the Observatory on the last-named day.

I. Buildings and Grounds, Movable Property, and Library: –

The domes of the Photoheliograph, Transit Pavilion, Altazimuth, Sheepshanks and 28-inch Equatorials were painted during the summer of 1926.

The principal movable instruments are thus distributed: –

On loan –

Harrison's Time Machine No. 2, Whewell's Anemometer, Airy's Dip Circle, and the large Bar Magnets in use from 11841 to 1915, at the Science Museum.

Clocks.-Dent No. 1916 and Dent 2013 at the Cape Observatory; Arnold 2 at the University Observatory, Oxford.

On loan to the Royal Observatory-

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.

VII. Time Service:–

The Clock Shortt No. 3 has been used as Standard Sidereal Clock throughout the year with the exception of the period June 11 to July 1. It has been regularly compared with Shortt No. 11 which was obtained last year and brought into use in July. Both clocks have gone with very great regularity; the air-leakage into the clock cases has been negligible. The rate of No. 3 in July was O·12s gaining per day and it has gradually decreased to zero. Similarly No. 11 has decreased from 0.16s gaining in July to zero. During the last 50 days when the rates have been nearly coincident the clocks have never differed by more than 0.05s.

The daily time signals to the General Post Office and the British Broadcasting Company have been transmitted regularly throughout the year.

The Time Ball at Greenwich has been operated regularly and also that at Deal until February 25 when the establishment was closed down.

The performance of the time balls at Admiralty Signal Stations is shown by the following table of errors of the return signals: –

Error of Return Signal.	Portsmouth.	Portland.	Devonport.
Not greater than 0.2s	242	230	234
0.3s to 0.5s	1	7	4
0.6s to 1.0s	2	0	0
Greater than 1.0s	1	0	0

The Westminster Clock was purposely kept approximately 2s fast as the chimes were used as a warning for the Greenwich Broadcast Time Signal. On 2 days it was less than 1s fast, on 164 days between 1s and 2s fast, on 122 days between 2s and 3s fast, and exceeded 3s on 9 days. Now that the Chimes are no longer used as a warning for the Greenwich signal it is no longer necessary to keep the clock fast.

Longitude Observations. – At the Meeting of the International Astronomical Union at Cambridge in 1925 it was decided to carry out an extensive scheme of longitude determinations in October and November 1926. About 50 observatories throughout the world took part. At Greenwich special series of observations for time were made with the small reversible transit B which is fitted with a self-registering micrometer. Observations with this instrument were secured on 34 nights. On account of the great regularity of the clock a smooth clock error was deduced from a combination of the whole series of time determinations, giving a reliable clock error for each night whether observations of stars were obtained or not.

Eleven series of time signals from Eiffel Tower, Bordeaux, Nauen and Annapolis, were registered on most days during October and November. When rhythmic signals were received, the Greenwich Sidereal times of the first and last signals were deduced, and when ordinary time signals were received the Greenwich Sidereal time of the reception of the standard signal was determined. The results have been printed and distributed to the observatories which took part in the operations. When the data from all the participating observatories are ready, valuable longitude determinations should result.

Preliminary comparisons from figures communicated from Paris and Washington point to corrections of less than 0.02s to the adopted longitudes of these Observatories. The following figures for the apparent longitude of Greenwich-Washington are of interest as showing the difference of time of travel from Annapolis and Bordeaux to Greenwich and Washington.

	h	m	s	s
October	5	8	15.737 (Bordeaux)	15.784 (Annapolis)
November			15.720 "	15.772 "

The differences for the two months are ·047s and .052s respectively and are closely explained by the known velocity of propagation of wireless waves, and are approximately twice the travel time across the Atlantic.

The results of the observations made in the autumn of 1925 by MM. Beljajeff and Dnieprovsky for the determination of the longitude Poulkovo-Greenwich has been communicated by the Director, Poulkovo Observatory. The final difference of longitude is 2h 1m 18.572s.

IX. General Remarks: –

The observations for longitude during the months of October and November were carried out very successfully, thanks to the continuous astronomical observations of Dr. Jackson and the regular recording of a large number of wireless time signals daily by Mr. Bowyer.

The Admiralty, with the co-operation of the Board of Trade, has arranged with the Post Office for the distribution of Time Signals from the Wireless Station at Rugby. The signals will be sent at 10h and 18h. They will be of rhythmic type, the separate signals being at intervals of 60/61 seconds with prolonged signals marking the exact minutes of Greenwich Mean Solar Time. The apparatus will include a master pendulum of the Shortt type rated to Mean Time, the slave clock usual for this system, and in addition, a pendulum adjusted to transmit 61 signals in one minute. The latter pendulum will be controlled by the master by an action similar to that employed to keep the slave in phase. This apparatus has been ordered from the Synchronome Co., who have promised delivery by the end of July. The installation and adjustment of the apparatus at Greenwich will probably take at least two months before the service can be inaugurated.

1928

The Report here presented refers to the year from 1927 May 11 to 1928 May 10. and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The room under the Shuckburgh Dome has been converted into a satisfactory chamber for the apparatus for the transmission of time signals to Rugby. Most of the structural alterations had been made at the date of the last report but the wiring and painting were completed later.

The principal movable instruments are thus distributed : –

On loan –

Harrison's Time Machine No. 2, Whewell's Anemometer, Airy's Dip Circle, and the large Bar Magnets in use from 11841 to 1915, at the Science Museum.

Clocks. Dent 1916 and Dent 2013 at the Cape Observatory; Arnold 2 at the University Observatory, Oxford.

On loan to the Royal Observatory-

One seven-valve Amplifier and one C.W. Syntoniser, from the Royal Air Force.

VII. -Time Service:-

The longitude observations made in October and November 1926, indicated a systematic difference of .09s between the transit circle and the small reversible "transit instrument B." This difference is fairly steadily maintained but as a reversible instrument is more suited for time-determinations, the practice has been started of obtaining clock errors at intervals of about a week with the small reversible instrument and basing the time-service upon these observations. The number of nights in the year under report on which observations were made is 44.

The Clock Shortt No. 3 has been used as Standard Sidereal Clock throughout the year and Shortt No. 11has been compared regularly with it. Both clocks have maintained very steady rates; the air leakage into the clock' cases has been negligible.

A discussion of the performance of the two clocks has been published in Monthly Notices of the Royal Astronomical Society, Vol. 88, No. 5. It has been shown that the progressive change found in the clock rates can be accounted for by an increase in the length of the pendulum rod which is made of invar. The progressive increase of 0. 011s per: day per month in the rate corresponds to a lengthening of the pendulum rod at the rate of one micron (0·001 mm.) in 118 days. The clocks also show a temperature inequality of 0. 003s per day for 1° F. change, or approximately 1 second per year for one degree (Fahr.) change of temperature.

The daily time signals to the General Post Office and the British Broadcasting Corporation have been transmitted regularly. The service of Rhythmic Time Signals transmitted through the Post Office Wireless station at Rugby, mentioned in last year's Report, was commenced on 1927 December 19, signals being sent out daily at 10h and 18h. The signal apparatus is controlled by a Free Pendulum clock, Shortt No. 16, which also controls the mean time clocks used for the Post Office and Broadcast signals.

The corrections to the transmitted times of the rhythmic 'signals for adjustment to the final time determinations are published and distributed periodically by the Hydrographer of the Navy in "Notices to Mariners."

The Time Ball has been dropped accurately throughout the year, with two exceptions, viz.: May 29, when a false drop was made before the signal, and on February 11 when the Ball was not raised on account of high wind.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	254	207	267	103
0.3s to 0.5s	4	12	5	88
0.6s to 1.0s	0	0	0	63
Greater than 1.0s	0	0	0	22

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals have been received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. They are registered automatically on a syphon recorder operated by a relay, controlled by an amplifying system. A four valve receiving set by the Marconi Co., Ltd., type R.P. 11, designed for continuous wave working from 15,000 to 22,000 metres, was purchased in 1927, July. The set is arranged to operate from a small frame aerial with a phasing unit connected to an open aerial in conjunction. By this means a high degree of selectivity is obtained.

The monthly means of the differences from Greenwich shewn by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument "B" are given in the following table. The Annapolis signals have been corrected for lag of land line to Washington and 0.02s for time of travel.

1927.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	+.044	-.023	+.022	+.042
February	+.051	-.031	+.009	+.028
March	+.017	-.096	-.005	+.012
April	+.048	-.051	+.021	+.021
May	+.102	+.031	+.086	+.087
June	+.014	-.114	-.010	+.022
July	+.081	-.055	+.068	+.085
August	+.009	-.037	+.054	+.014
September	+.052	-.039	+.008	+.056
October	+.068	-.074	+.008	+.073
November	*	-.058	-.024	+.034
December	*	-.064	-.071	+.023
_____	_____	_____	_____	_____
Mean	-	-.050	+.010	+.042
Number of comparisions	226	278	264	262

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals.

* Final corrections not yet available.

IX.-General Remarks :-

The distribution of Greenwich Time from the Wireless Station at Rugby should be of service to Astronomers, Surveyors and Navigators in all parts of the world. It may be of interest to note that the possible use of "galvanism" for the distribution of time from Greenwich was referred to by Airy in the Report to the Board of Visitors for 1849, and in 1852 he states that he "has entered into communication with the authorities of the South Eastern Railway (whose line of galvanic communication will shortly pass within nine furlongs of the Observatory)." On August 28, 1852, a time-ball was dropped in the Strand by signals transmitted through an electric clock at Greenwich made by Shepherd. Shortly afterwards a clock at London Bridge Railway Station was made to distribute hourly time signals to the Electric Telegraph Company's system and to different branches of the South Eastern Railway. When the telegraphs were taken over by the General Post Office in 1870, a complete system for telegraphic distribution of time-signals from Greenwich was gradually developed. The transmission of wireless time-signals from Greenwich by the British Broadcasting Company was commenced in January, 1924, but has only a limited range. The present service brings Great Britain into line with France, Germany and the United States. The times of transmission have 'been chosen after consideration of the hours at which signals are sent from other stations, and after consultation with the Hydrographer of the Navy and the Secretary of the Royal Geographical Society. The installation of these signals, as of the " Shortt " clocks, owes much to the enthusiasm, care and patience of Mr. Bowyer.

1929

The Report here presented refers to the year from 1928 May 11 to 1929 May 10, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The principal movable instruments are thus distributed: –

On loan –

Harrison's Time Machine No. 2, Whewell's Anemometer, Airy's Dip Circle, and the large Bar Magnets in use from 11841 to 1915, at the Science Museum.

Clocks.-Dent 1916 and Dent 2013 at the Cape Observatory; Arnold 2 at the University Observatory, Oxford.

VII.-Time Service: –

The Time determinations have been made throughout the year by means of the reversible transit instrument "B" fitted with impersonal micrometer. When possible, observations are made at intervals of not more than 5 days. In the year, 1928 May 11 – 1929 May 10, observations have been made on 71 nights.

The Clock Shortt No. 3 has been used as Standard Sidereal Clock throughout the year and Shortt No. 11 has been compared regularly with it. Both clocks have maintained very steady rates; the air leakage into the clock cases has been negligible. The temperature in the clock cellar is now maintained at 62.8° F.

An extension of the discussion of the performance of the two clocks has been published in Monthly Notices, Vol. 89, No. 3. The progressive increase of losing rate of the clocks shows no diminution. It seems possible that this effect may be due in part to the pendulum bobs, which are made of type-metal, and it is proposed to substitute a bob of invar on one of the clocks.

The service of Rhythmic Time Signals at 10h and I18h transmitted through the Post Office Wireless station at Rugby, and also the daily time signals to the General Post Office and the British Broadcasting Corporation, have been transmitted regularly. The Free Pendulum Clock, Shortt No. 16, controls the mean time clocks used for the Post Office and Broadcast signals, in addition to the Rhythmic signal transmitter.

The corrections to the transmitted times of the rhythmic signals for adjustment to the final time determinations are published and distributed periodically by the Hydrographer of the Navy in "Notices to Mariners."

The Time Ball failed to drop accurately on one occasion, November 11, when the ball was not released at 13h. On November 25.the Ball was not raised on account of high wind.

From July 9 – July 11 and November 12–14 the Time Ball was not raised on account of repairs to the lifting gear being in progress. From August 24 – September 21 the Ball was out of action during progress of repairs to the wind vane and crosstrees at the top of the Mast.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	211	178	281	109
0.3s to 0.5s	7	5	2	86
0.6s to 1.0s	0	0	0	87
Greater than 1.0s	0	0	0	8

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published in “Notices to Mariners,” together with the corrections to the Rugby signals.

The monthly means of the differences from Greenwich shewn by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument "B," are given in the following table. The Annapolis signals have been corrected for lag of land line to Washington and 0·02s for time of travel.

1928.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	.000	-.059	-.072	-.020
February	+.038	-.064	-.037	+.032
March	+.064	-.033	+.011	+.060
April	+.036	-.013	-.045	+.035
May	+.088	+.018	+.011	+.092
June	+.090	+.035	+.005	+.086
July	+.011	-.050	-.016	+.019
August	+.030	+.030	-.022	+.040
September	-.011	-.054	-.039	-.001
October	+.013	-.060	-.030	+.019
November	*	-.069	-.038	*
December	*	-.037	-.028	*
_____	_____	_____	_____	_____
Mean	(+.036)	-.030	-.021	(+.036)
Number of comparisions	232	292	339	280

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals

* Final corrections not yet available.

1930

The Report here presented refers to the year from 1929 May II to 1930 May 10, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Science Museum – Harrison’s Time Machines No. 2 and a number of other instruments of historical interest.

VII. Time Service: –

The Time determinations have been made throughout the year by means of the reversible transit instrument "B" fitted with impersonal micrometer. When possible, observations are made at intervals of not more than 5 days. In the year 1929 May 11 – 1930 May 10, observations have been made on 72 nights.

The Clock Shortt No. 3 was used as Standard Sidereal Clock until July 31. No. 11 was used for the period 1929 August 1 to 1930 March 31, and No. 3 replaced as Standard on April 1. Shortt No. 3 was dismounted in October, overhauled and some refinements made to the mechanism. An invar bob was also fitted in place of the original type-metal bob. This alteration was made for the purpose of testing whether the increasing losing rate found in the clocks could be attributed to " settling" of the pendulum bob which contains a large percentage of lead. With the invar bob, which is considerably heavier than the type-metal bob, it is found that the clock still increases its losing rate, but by an amount which is half the increase shown in a corresponding time, when the type-metal bob was in use. The progressive increase of losing rate of the Clock No. 11 shows no diminution.

The service of Rhythmic Time Signals at 10h and 18h transmitted through the Post Office Wireless station at Rugby, and also the daily time signals to the General Post Office and the British Broadcasting Corporation, have been transmitted regularly. The Free Pendulum Clock, Shortt No. 16, controls the mean time clocks used for the Post Office and Broadcast signals, in addition to the Rhythmic signal transmitter.

The Time Ball was not raised on account of high wind on five occasions, viz.: May 6, December 7, 9, 25 and 29.

The performance of the time balls at the Admiralty Signa] stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	176	157	180	83
0.3s to 0.5s	6	2	9	80
0.6s to 1.0s	2	2	4	104
Greater than 1.0s	0	0	10	21

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published monthly for distribution in "Notices to Mariners," together with the corrections to the Rugby signals.

The monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument "B," are given in the following table. The Annapolis signals have been corrected for lag of land line to Washington and 0.02s for time of travel.

1929.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	...	-.061	-.053	-.010
February	...	-.002	-.050	-.025
March	...	+.013	-.016	+.028
April	+.045	+.042	-.020	+.021
May	+.070	-.028	+.046	+.075
June	+.082	+.038	+.049	+.086
July	+.059	+.035	+.055	+.071
August	+.020	+.031	+.034	+.033
September	+.054	+.025	+.045	+.057
October	+.054	+.047	+.022	+.059
November	-.020	-.038	-.008	-.007
December	+.031	-.038	+.016	+.038
_____	_____	_____	_____	_____
Mean	+.044	+.006	+.007	+.035
Number of comparisions	208	289	301	341

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places.

1931

The Report here presented refers to the year from 1930 May 11 to 1931 May 10, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The external painting of the Astrographic, Sheepshanks, Altazimuth, Photoheliograph, and Transit Pavilion domes was completed on July 10, 1930.

The large wireless aerial, stretching from the Octagon Room to the South Building, was taken down in October as it was found no longer necessary.

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Science Museum – Harrison’s Time Machines No. 2 and a number of other instruments of historical interest.

Reference has been made in previous reports to the restoration of Harrison's Time Machines, Nos. 1, 2 and 4. The more complicated machine No. 3, offered peculiar difficulties which have been surmounted by the care and skill of Commander Gould. The Observatory is greatly indebted to him for the trouble and time he has generously given to the repair and preservation of these historic instruments.

VII.-Time Service:-

The Time determinations have been made throughout the year by means of the reversible transit instrument" B" fitted with impersonal micrometer. When possible, observations are made at intervals of not more than 5 days. In the year 1930 May 11-1931 May 10, observations have been made on 63 nights.

The Clock Shortt No. 3 was used as Standard Sidereal Clock throughout the year. No. 11 has been kept going as reserve standard, and compared regularly with No. 3. The progressive increase of losing rate in clock No. 11 shows no diminution.

The Time Ball was not raised on account of high wind on two occasions, viz.: September 20 and October 8.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	239	217	210	101
0.3s to 0.5s	1	0	9	86
0.6s to 1.0s	2	0	9	96
Greater than 1.0s	0	0	4	23

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published monthly for distribution in "Notices to Mariners," together with the corrections to the Rugby signals.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument "B " at Greenwich, and the corrections for final time determinations published by Paris, Hamburg and Washington Observatories.

The Annapolis signals have been corrected for lag of land line to Washington and 0·02s for time of travel.

1930.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	+.004	+.107	-.021	+.008
February	+.046	+.106	+.021	+.054
March	-.021	+.045	+.004	+.001
April	-.004	+.040	+.023	+.011
May	+.005	+.015	+.064	+.023
June	+.006	+.017	+.043	+.014
July	+.013	+.013	+.074	+.034
August	+.027	+.024	+.070	+.051
September	+.027	+.030	+.046	+.041
October	+.042	+.079	+.071	+.052
November	+.007	+.029	+.046	+.026
December	+.069	+.022	+.026	+.087
_____	_____	_____	_____	_____
Mean	+.016	+.044	+.039	+.033
Number of comparisions	303	322	327	332

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places.

IX.-General Remarks :-

Lieut.-Commanders F.C. Woodhouse and A. Jones, R.N., made a series of observations in September to examine the performance of a new 45° prismatic astrolabe constructed for the Hydrographic Department. Using the standard sidereal clock their observations gave the known latitude and longitude of the station with great accuracy.

1932

The Report here presented refers to the year from 1931 May 11 to 1932 May 10, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Science Museum – Harrison’s Time Machines No. 2 and a number of other instruments of historical interest.

A suggestion was made by Commander Gould to complete his work on Harrison’s Time Machines by restoring the first machine. He had already cleaned it and prevented further deterioration but had not restored it to going order. The proposal is to replace some missing parts which will be carefully marked as not original. Sanction was given by the Admiralty on June 22 and the instrument placed in Commander Gould’s hands on October 2. The work is difficult and complicated but it is anticipated that the restoration will be completed by the end of the year.

VIII.-Time Service: –

The Time determinations have been made throughout the year by means of the reversible transit instrument ”B“ fitted with impersonal micrometer. When possible, observations are made at intervals of not more than 5 days. In the year 1931 May 11 – 1932 May 10, observations have been made on 63 nights.

The Clock Shortt No. 3 was used as Standard Sidereal Clock until February 18 when No. 11 was substituted. Comparisons between the two clocks have been made regularly throughout the year.

A new pendulum of invar with a 14-lb. invar bob and a steel suspension spring has been fitted to Shortt No. 3.

A short series of warning signals before the Rugby Rhythmic signals was introduced in June last. The apparatus used is automatically controlled and transmits in morse during the minute preceding the Rugby series the code signal “G.B.R. G.B.R. Time” repeated four times. In the construction of the automatic morse signal apparatus we are indebted to the Telegraph Section of the Engineer-in-Chief’s Department of the General Post Office.

The Time Ball was not raised on account of snow on the mast on two occasions, viz., February 10 and 11. On October 7 the Ball did not drop correctly at 13h owing to faulty manipulation of the lifting gear, and on December 29 the Ball did not drop at 13h owing to accumulation of snow holding it at the top of the mast

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	213	173	214	121
0.3s to 0.5s	3	10	9	98
0.6s to 1.0s	1	3	7	111
Greater than 1.0s	0	0	10	9

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published monthly for distribution in “Notices to Mariners,” together with the corrections to the Rugby signals.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument “B“ at Greenwich, and the corrections for final time determinations published by Paris, Hamburg and Washington Observatories.

The Annapolis signals have been corrected for lag of land line to Washington and by 0.02s for time of travel.

1931.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	+.018	-.003	+.011	+.030
February	+.002	+.039	+.050	+.015
March	+.006	+.055	+.047	+.024
April	+.004	+.036	+.038	+.013
May	+.047	+.053	+.050	+.056
June	+.052	+.049	+.047	+.063
July	+.050	+.033	+.063	+.062
August	+.007	+.029	+.012	+.012
September	-.028	+.037	+.006	-.022
October	-.007	+.030	+.020	-.003
November	-.033	+.012	-.020	-.035
December	-.028	-.001	-.021	-.025
_____	_____	_____	_____	_____
Mean	-.007	+.031	+.026	+.015
Number of comparisions	229	296	346	338

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places.

X. General Remarks: –

As this is the last Report I shall have the honour of presenting to the Board of Visitors, I venture to review the work of the Observatory since my appointment in 1910.

A new type of clock, the free pendulum designed by Mr. W.H. Shortt and constructed by the Synchronome Company, was introduced to the Observatory in 1924 as Standard Sidereal Clock. The results were so satisfactory that a second clock was installed in 1926, and a similar clock for mean solar time in 1927. Broadcasting of time from the Observatory by wireless was commenced in 1924, by the British Broadcasting Corporation. Transmission of Rhythmic series of time signals from the Post Office Station at Rugby was begun in December, 1927.

1933

The Report here presented refers to the year from 1932 May 11 to 1933 May 10, and exhibits the state of the Observatory on the last-named day.

Sir Frank Dyson retired from the office of Astronomer Royal on February 28.

II. Buildings and Grounds, Movable Property, and Library: –

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Science Museum – Harrison's Time Machines Nos. 2 and 3, and a number of other instruments of historical interest.

The restoration of Harrison's Time Machine No. 1, referred to in last year's Report, has been completed. The machine was returned in going order on February 3 together with a complete set of drawings, prepared by Commander Gould, on which all repaired parts, replacements and modified winding apparatus are distinctly marked for future guidance. In connection with the restoration of the mechanism Messrs. T. and F. Mercer, of St. Albans, very generously, at appreciable cost to themselves, rendered valuable assistance in making and supplying the special set of balance springs required and in carrying out Commander Gould's design for winding the machine.

VIII. Time Service: –

The time determinations were made by means of the reversible transit instrument B, fitted with impersonal micrometer, until April 5. A similar instrument, transit D, was substituted and used subsequent to April 7. The pivots of transit B have been re-ground. When possible, observations are made at intervals of not more 'than 5 days. In the year 1932 May 11-1933 May 10, observations have been made on 81 nights.

The Clock Shortt No. 11 was used as standard sidereal clock until June 1 when No. 3 was substituted and used until November 22 when it was replaced by No. 11.

The service of rhythmic time signals at 10h and 18h transmitted through the Post Office wireless station at Rugby, and also the daily time signals to the General Post Office and the British Broadcasting Corporation, have been transmitted regularly. The Free Pendulum Clock, Shortt No. 16, controls the mean time clocks used for the Post Office and Broadcast signals, in addition to the rhythmic signal transmitter.

At the request of the Director of the Bureau International de l'Heure a signal of ten seconds duration is now transmitted immediately following the Rugby series at 10h and 18h. The purpose of these signals is the determination of time-lags in receiving and recording sets.

The Time Ball did not drop at 13h on September 27 owing to failure of the clock signal contact.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	205	199	217	114
0.3s to 0.5s	1	4	4	83
0.6s to 1.0s	0	1	3	97
Greater than 1.0s	0	1	9	9

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Wireless time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published monthly for distribution in “Notices to Mariners,” together with the corrections to the Rugby signals.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from

observations made with the small reversible transit instrument B at Greenwich, and the corrections for final time determinations published by Paris, Hamburg and Washington Observatories.

The Annapolis signals have been corrected for lag of land line to Washington and by 0.02s for time of travel.

1932.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	-.034	-.016	-.006	-.036
February	-.009	-.020	-.006	-.007
March	-.012	+.023	+.017	-.011
April	+.022	+.024	+.025	+.026
May	+.014	+.003	+.021	+.021
June	+.010	+.013	+.030	+.012
July	+.031	+.054	+.047	+.035
August	+.057	+.050	+.085	+.062
September	+.023	+.041	+.060	+.024
October	+.021	+.016	+.059	+.025
November	-.001	+.017	+.065	+.004
December	-.025	+.033	+.069	-.013
_____	_____	_____	_____	_____
Mean	+.008	.020	+.039	+.012
Number of comparisions	303	301	343	351

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places.

IX. Personal Establishment: –

Dr. Jackson resigned on March 1st the post of Chief Assistant, which he had held since 1914 October 8, in consequence of his appointment as His Majesty's Astronomer at the Cape. Dr. Jackson had associated himself particularly with the meridian and double-star observations and had made a detailed study of the performance of the free-pendulum clocks. He was in charge of three expeditions from the Observatory, to observe a total eclipse of the sun. The vacancy caused by Dr. Jackson's resignation has not yet been filled.

X. General Remarks: –

With the completion of the work on Harrison's Time Machine No. 1, referred to previously, Lieutenant-Commander Gould has completed his work on Harrison's four timekeepers. For the machines Nos. 1 and 3, Commander Gould made fullsized drawings of all parts. These drawings, together with his notes and workbooks, he has presented to the Observatory. If at any subsequent time it should become. necessary to give further attention to the machines, the work will thereby be much facilitated. The Observatory is deeply indebted to Commander Gould for the great amount of time which he has voluntarily expended in the intricate task of restoring these historical time-pieces and for the care with which the restoration has been carried out, with the result that the machines can now be seen in going order and practically as they were left by Harrison. As an outcome of the work on these machines, it has been possible for Commander Gould to place on record many of Harrison's principles and designs, of which little was hitherto known.

1934

The Report here presented refers to the year from 1933 May 11 to 1934 May 10, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The cellar below Flamsteed House, adjacent to the existing clock cellar, has been prepared as a second clock cellar. The walls and ceiling have been rendered with cement, a new concrete floor and steps have been provided and a thick lagged door, to reduce leakage of heat, has been made.

Internal re-decoration of the Transit Pavilion, Time Department Office, and Photoheliograph dome, has been carried out.

The dome of the Transit Pavilion had become very difficult to open owing to the rusting of the iron work in places where it was impossible to keep it painted; this caused the angle iron to twist and resulted in considerable friction on the running rails. The two halves of the dome were therefore jacked up, the bad portions of ironwork were cut away and new angle iron pieces were bolted on.

Owing to a crack appearing in the stone on which the small reversible transit, used for time observations, is mounted, a new stone has been substituted. At the same time, the top portion of the brick work of the pier was rebuilt.

A new electric main cable has been laid between the basement housing the transformer and the courtyard. This cable is supplementary to the old cable and its installation was necessitated by the increase in the amount of electric current used for heating and power in the older portion of the Observatory. Electric heating has been installed in the Transit Pavilion and will be used in damp weather for keeping the small reversible transit and its accessories dry.

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Science Museum – Harrison's Time Machines Nos. 2 and 3, and a number of other instruments of historical interest.

VI.-Time Service:-

Determinations of time have been made by a rota of three observers with the reversible transit instrument D fitted with an impersonal micrometer. Observations were secured on 142 nights, including 45 nights during the period covered by the special longitude observations described below. The transits are now recorded on a syphon recording tape chronograph with a scale of 2·5 cm. per sec. The chronograph can be started or stopped by the observer at the instrument by means of a switch, so that the chronograph is not in operation during the interval between the observation of two stars. The mechanical relay previously in use for converting the breaks in the micrometer circuit into makes has been replaced by a thermionic valve circuit, to eliminate the time-lag of the mechanical relay.

A special programme of observations was carried out during the period 1933 September 15 to December 15, in connection with the International Longitude Programme planned by the International Astronomical Union and the International Union for Geodesy and Geophysics. During this period, observations for time were secured on 45 nights, stars in the list prepared by the Joint Longitude Commission of the two Unions being observed. The results were corrected for the relative, personal equations of the three observers, but no determination has been made of the absolute personal equation of the mean observer. The probable error of the time determination on a single night as judged both by internal evidence and by intercomparison of different nights was 0.010s. The results of the time determinations in connection with this programme and the observed times of reception of various time-signals have been communicated to the Bureau International de l'Heure, Paris, where the results obtained by the various co-operating observatories will be co-ordinated and discussed.

The free pendulum clock Shortt No. 3 was overhauled by the Synchronome Company in February and coils of higher insulation were fitted in all its circuits. The upper face of the clock casing being far from plane and never having been ground to the glass bell-jar, difficulty had always been experienced in securing an air-tight joint after the clock case had been opened. The opportunity provided by the overhaul of the clock was used to have the bearing face properly surfaced and the bell-jar ground to the face. This was done by the workshop staff and an air-tight joint is now easily obtained.

Creed relays were installed in November for operating all standard sidereal clock circuits. These relays have a very small time-lag and they have all been adjusted to give as nearly as possible equal time-lag,

The service of rhythmic time signals at 10h and 18h transmitted through the Post Office wireless station at Rugby, and also the daily time signals to the General Post Office and the British Broadcasting Corporation, have been transmitted regularly. The free pendulum clock, Shortt No. 16, controls the mean time clocks used for the Post Office and Broadcast signals, in addition to the rhythmic signal transmitter.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Portland.	Devonport.	Westminster.
Not greater than 0.2s	223	51	232	105
0.3s to 0.5s	0	0	2	86
0.6s to 1.0s	0	1	1	74
Greater than 1.0s	2	0	0	27

The return signal from Portland was discontinued after August 31 when the Time Ball was dismantled.

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Radio time signals are received regularly from the Eiffel Tower, Nauen, Bordeaux and, when possible, from Annapolis. These are registered automatically, and the corrections found for the Nauen, Bordeaux and Annapolis signals are published monthly for distribution in “Notices to Mariners “, together with the corrections to the Rugby signals.

A cathode ray oscillograph apparatus embodying mains unit for operation of tube, square wave oscillator for time scale, and recording camera of cinema type has been obtained and is now in process of assembling for the purpose of eliminating time-lags in the recording of radio time signals.

Preliminary work in this connection carried out in November with a Duddell oscillograph lent by the Professor of Physics of the Royal Naval College, Greenwich, showed that with the present syphon recorder a time-lag of 0.01s existed in recording a radio time signal relative to the recording of the standard clock beat.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument D at Greenwich, and the corrections for final time determinations published by Paris, Hamburg and Washington Observatories.

The Annapolis signals have been corrected for lag of land line to Washington and by 0.02s for time of travel. A correction of 0.01s has been applied to all signals for lag in reception. Relative personal equations have been applied to the time observations from September 15.

1933.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	-.033	+.005	+.034	-.017
February	-.020	+.001	+.029	+.004
March	-.046	-.005	-.005	-.005
April	-.027	-.011	-.007	-.012
May	-.030	-.015	-.018	-.014
June	-.041	+.012	-.024	-.032
July	+.012	+.011	-.070	+.017
August	-.028	+.003	-.010	-.016
September	-.044	-.001	-.018	-.036
October	-.031	.000	-.043	-.020
November	-.000	+.017	-.031	+.006
December	-.041	+.018	-.012	-.031
_____	_____	_____	_____	_____
Mean	-.024	.000	-.007	-.013
Number of comparisions	302	292	231	355

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places.

X.-General Remarks:-

Special attention has been given during the year to the improvement of the time determinations. Relays of uniform pattern and of extremely small time-lag have replaced a miscellaneous series of relays whose time-lags were not negligible and were probably variable. The mechanical relay used for converting the break circuit from the transit micrometer into a make circuit for chronographic registration has been replaced by a valve relay. A syphon recording tape chronograph has been substituted for the barrel chronograph, the moving parts of which possessed considerable inertia. Preliminary determinations of the time-lag in the reception of the radio time signals have been made. Observations for time determination have been made more frequently and shared by three observers with a view to detecting the irregular variations of small amplitude in the errors of the standard clocks.

A cathode ray oscillograph and a moving film camera will in future be the standard method of recording the radio time signals and comparing with the clock signals. Valuable assistance has been received from the Superintendent of the Radio Department of the National Physical Laboratory in the choice of the best method of eliminating the lag in reception of the radio time signals; the time marker for impressing suitable time marks at intervals of 0.01s on the moving film was designed and constructed in the Radio Department. It is intended to construct in the workshop a personal equation machine for the determination at frequent intervals of the personal equation of the mean observer.

1935

The Report here presented refers to the period from 1934 May 11 to 1935 April 30, and exhibits the state of the Observatory on the last-named day.

II. Buildings and Grounds, Movable Property, and Library: –

The wiring of the new clock room, housing the two mean-time free pendulums, has been completed, and the necessary wires for connecting the two free pendulums with their respective slave clocks in the Rugby Signal Room have been run. An additional length of cable was laid across the Courtyard for carrying the current for heating the clock cellar.

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford – Clock Arnold No. 2.

To the Hydrographic Department of the Admiralty – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

VI. Time Service.

A generous offer to present to the Observatory a free pendulum installation has been made by Mr. H. R. Fry, F.R.A.S., and has been accepted by the Admiralty. The two sidereal free-pendulum clocks which the Observatory at present possesses are required for routine purposes. One is used as the standard clock; the second is kept available for adoption as standard whenever it may be required. The new clock will not be required for routine work, but will be available for experimental purposes, though it will also serve to improve the accuracy of the time distributed from the Observatory. The need for a third sidereal clock of this nature has often been felt. The new installation will include an improved slave clock, and a special time indicating and distributing clock, in an air-tight case, which will be synchronised by the free pendulum within narrower limits than in the ordinary installations; time will also be taken directly off the free pendulum by a photoelectric method. The specification for the new clock has been prepared.

Determinations of time have been made by a rota of observers with the reversible transit D fitted with an impersonal micrometer. On April 11 Transit B was used, whilst the pivots of Transit D were re-ground.

Observations were secured on 106 nights, the transits being registered on a syphon recording tape chronograph operated by a thermionic valve relay. The determinations of time have been corrected for the relative personal equations of the three observers, but no determination has been made of the absolute personal equation of the mean observer. A personal-equation machine, designed for this purpose, is under construction in the workshop.

The free pendulum clock Shortt No. 11, used as sidereal standard throughout 1934, was dismounted for overhaul early in January after a continuous run of 8½ years. High insulation coils were fitted in all circuits, and during the period of overhaul the opportunity was taken to grind the surfaces of the case, the glass bell jar and bottom plate. The clock was re-assembled in February, and is now being rated as reserve standard. The free pendulum clock Shortt No. 3 has been used as sidereal standard since January 1.

A second mean time free pendulum clock was ordered from the Synchronome Company and was delivered in November. This clock (Shortt No. 49) was mounted in the cellar under Flamsteed House adjacent to that containing the sidereal standard clocks. After adjustment it was brought into use on February 14. It controls the mean time clocks used for Post Office and broadcast signals, and also the rhythmic signal transmitter.

The clock Shortt No. 16, which until February 14 had controlled the mean time services, was then dismounted for overhaul. The free pendulum was remounted in the same cellar as No. 49, the slave clock and other accessories being left in the Clock Room under the Shuckburgh Dome [the Rugby Room]. Clock No. 16 was fitted with high insulation coils similar to the sidereal clocks, and the free pendulum case surfaces were ground flat while the clock was under repair. The clock was re-assembled and re-started early in April, and is now under adjustment.

The Time Ball was released prematurely on December 29. It was raised again and dropped correctly at 14h.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Devonport.	Westminster.
Not greater than 0.2s	225	261	80
0.3s to 0.5s	1	2	68
0.6s to 1.0s	1	0	83
Greater than 1.0s	0	0	19

The Westminster Clock was under repair from May 1 till July 3.

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Radio time signals are received regularly from the Eiffel Tower, Nauen and Bordeaux. The long wave signals from Annapolis were received, when possible, until the discontinuance of these signals at the end of June. The signals are registered automatically, and the corrections found for the Nauen and Bordeaux signals, together with the corrections to the Rugby signals, are published monthly in the Admiralty Notices to Mariners.

A cathode ray oscillograph apparatus embodying mains unit for operation of tube, square wave oscillator for time scale, and recording camera of cinema type; was used during November for determining time-lags in the recording of radio time signals. Modification in the recording circuit and various causes have interrupted the use of this instrument, but regular comparisons between the syphon recorder and the oscillograph are now in progress.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock errors derived from observations made with the small reversible transit instrument D at Greenwich, and the corrections for final time determinations published by Paris, Hamburg and Washington Observatories.

1934.	Paris 9½ h Rhythmic.	Nauen 12 h Rhythmic.	Annapolis 17h.	Bordeaux 20 h Rhythmic.
	s.	s.	s.	s.
January	-.026	.000	+.016	-.004
February	-.046		+.006	-.025
March	-.046		-.021	-.029
April	-.012	-.017	-.014	+.002
May	-.022	+.007	-.012	.000
June	-.023	+.024	-.010	-.001
July	+.008	+.033		+.037
August	-.001	+.017		+.025
September	-.021			+.014
October	-.017			+.009
November	-.015	+.008		+.007
December	+.005	+.012		+.016
_____	_____	_____	_____	_____
Mean	-.018	-.013	-.009	+.004
Number of comparisions	301	178	56	356

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places. The Annapolis signals have been corrected for lag of land line to Washington and by 0.02s for time of travel. A correction of 0.005s has been applied to all signals for lag in reception. Relative personal equations have been applied to the time observations.

1936

The Report here presented refers to the period from 1935 May 1 to 1936 April 30 and exhibits the state of the Observatory on the last-named day.

I. New Instruments: –

Free Pendulum Clock. – The construction of this clock, which is being presented to the Observatory by Mr. H. R. Fry, F.R.A.S., is approaching completion. The slave clock is of more refined construction than the standard pattern and is fully jewelled throughout. Experience has shown that the performance of one free pendulum may be consistently better than that of another; thus, the sidereal free pendulum No. 3 performs consistently better than No. 11 and the mean time free pendulum No. 16 performs consistently better than No. 49. Mr. Fry's own free pendulum, No. 40, having given a high standard of performance for 4 years, Mr. Fry has generously allowed it to be used for the new clock. The bob has been repinned to change from mean seconds to sidereal seconds and a steel suspension spring, similar to that in use in Shortt No. 3, has been substituted for the elinvar spring.

Time will be taken directly off the free pendulum by a photo-electric method. The pendulum carries, below the bob, a plate in which there is a narrow rectangular aperture, 1· 5 mm. in width. An image of the filament of a low-voltage lamp is thrown on to a narrow slit by means of a condensing lens. An image of this slit is formed in the plane of the plate on the pendulum by a specially constructed lens, achromatised for λ 7000, the light passing through an optically plane parallel window in the case. A second window is fitted diametrically opposite and the photo-electric cell is mounted behind this window. The current from the photo-electric cell is amplified by a thermionic relay and the output from this relay can be used for the distribution of time. The time of rise to maximum of each signal is one milli-second, this time being determined by the width of the slit; the duration of the signal, about 20 milliseconds, is determined by the width of the aperture in the pendulum plate.

As, with this design, the time taken directly from the free pendulum can be distributed throughout the observatory, the proposed special time indicating and distributing clock is no longer required. This portion of the original plan has therefore been dropped.

A specially designed case is being made for the free pendulum by Messrs. Metropolitan Vickers Ltd. All glass to metal joints will be finished to optical limits and the case will be submitted to the tests customary in high-vacuum technique. For convenience in mounting the optical system, the pendulum will swing at right angles to the wall. The case will be supported by two trunnions, which will be carried by two triangular self-aligning brackets, fixed to the wall. A third bracket, fixed to the wall, defines the position of the lower end of the case. This method of mounting the case was designed by Mr. Shortt.

II. Buildings and Grounds, Movable Property, and Library: –

The following instruments have been lent by the Royal Observatory :-

To the Royal Observatory, Cape of Good Hope- Clocks, Dent 1916 and Dent 2013.

To the University Observatory, Oxford Clock Arnold No. 2.

To the Hydrographic Department of the Admiralty- Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall, (For exhibition at the Science Museum.)

VI.-Time Service.

Determinations of time have been made by a rota of observers with the reversible transit D until July 24 when transit B was remounted.

Observations were secured on 153 nights, the transits being registered on a syphon recording tape chronograph operated by a thermionic valve relay. The determinations of time have been corrected for the relative personal equations of the three observers, but no determination has been made of the absolute personal equation of the mean observer. A personal-equation machine, designed for this purpose, has been constructed in the workshop and will shortly be brought into use. This machine will be housed in a small hut on the roof of the Octagon Room, adjacent to the Ball Turret. The transit instrument will be mounted, for the purpose of observations with the personal-equation machine, in the Tank Room of the Main Building, where a brick pier has been erected to carry the instrument. The observations with the transit instrument and the concurrent signals from the personal-equation machine will be recorded, together with signals from the standard clock, on a two-pen syphon chronograph, placed in the Time Department. Arrangements are being made for this chronograph to be started and stopped by distant control.

The free pendulum clock Shortt No. 3 has been used as sidereal standard throughout the year; and Shortt No. 11 has been compared regularly with it daily.

The mean time free pendulum clock Shortt No. 49 was used to control the mean time clocks used for Post Office and broadcast signals and also the rhythmic signal transmitter until October 29, when Shortt No. 16 was substituted.

The sidereal clock circuits have hitherto employed relays biased for single current working. It has been found possible to employ unbiased relays by the use of transformers. At each stage the current is fed into the primary of a transformer, and the secondary feeds the unbiased relay. Only one side of the relay output is utilised. The resulting avoidance of relay bias removes a troublesome cause of time-lag. The modified system was brought into operation on April 15.

The Time Ball failed to drop at 13h British Summer Time on June 23. It was dropped correctly at 14h B.S.T.

The performance of the time balls at the Admiralty Signal stations and also the Westminster Clock is shown by the following table of the errors of the return signals.

Error of Return Signal.	Portsmouth.	Devonport.	Westminster.
Not greater than 0.2s	218	234	120
0.3s to 0.5s	2	2	110
0.6s to 1.0s	1	1	70
Greater than 1.0s	0	0	15

The Westminster Clock was found to be 20s in error on January 17 owing to an accumulation of snow on the hands.

The clocks at the Admiralty stations are corrected daily by the 10h signal, but the Westminster Clock is not so corrected.

Radio time signals are received regularly from the Eifel Tower and Bordeaux. The Nauen long wave signals have been received since March 5 only. During the remainder of the year they were transmitted on a wave length outside the range of the Marconi Receiver. No long wave signals were sent out from Annapolis, the station being under reconstruction. The signals are registered automatically, and the corrections found for the Nauen and Bordeaux signals, together with the corrections to the Rugby signals, are published monthly in the Admiralty Notices to Mariners.

The cathode ray oscillograph apparatus embodying mains unit for operation of tube, square wave oscillator for time scale, and recording camera of cinema type, has been used for occasional determination of the time lags in recording radio time signals.

The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock corrections derived from observations made with the small reversible transit instruments at Greenwich, and the corrections for final time determinations published by the Paris and Hamburg Observatories.

1935.	Paris 9½ h Rhythmic.	Nauen 12 h* Rhythmic.	Bordeaux 20 h Rhythmic.
	s.	s.	s.
January	+.003	+.018	+.006
February	-.023	-.017	-.019
March	-.009	-.019	-.009
April	+.002	-.001	+.004
May	-.045	-.032	-.045
June	-.047	-.013	-.048
July	+.009	+.025	+.007
August	+.008	+.005	+.008
September	+.001	-.005	+.003
October	-.001	-.021	...
November	-.004	-.033	-.003
December	-.001	-.025	-.004
_____	_____	_____	_____
Mean	-.009	-.010	-.009
Number of comparisions	309	302	302

* The comparisons with Nauen from March 13 were derived via Paris signals.

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The Nauen times have been referred to the Eichelberger system of star places. The reception time of the signals has been corrected for time lag in the recording apparatus. Relative personal equations have been applied to the time observations.

1937

The Report here presented refers to the period from 1936 May 1 to 1937 April 30 and exhibits the state of the Observatory on the last-named day.

1. New Instruments:-

Free Pendulum Clock. – The free pendulum sidereal clock, Shortt No. 40, presented to the Observatory by Mr. H. R.. Fry, F.R.A.S., was delivered during the year. The slave clock is of a higher grade of construction than the standard slave clock; it is fully jewelled throughout and the consequent reduction in friction permits of a much lighter half-minute impulse. This makes it possible to run the slave-clock within much closer limits than the normal. The slave-clock is adjusted to have a natural losing rate of 1.5 seconds a day; in order that synchronization may occur at alternate half-minutes, the acceleration at synchronization is required to be about one millisecond, as compared with 4 milliseconds in the standard form of slave. To ensure a consistent acceleration on synchronization of one millisecond, the synchronizing spring of the standard form of slave has been replaced by a gravity arm, carrying a small sliding weight. By adjusting the position of this weight, the acceleration on synchronization can be varied, permitting an exact adjustment to the desired value.

It is expected that the improvements in the design of the slave-clock will be reflected in a higher standard of performance of the free pendulum itself.

The slave clock was delivered in May and was carefully rated as an independent clock from May 19 to July 24. The special case for the free pendulum, supplied by Messrs. Metropolitan Vickers, Ltd., was meanwhile mounted and tested. The rating of the free pendulum was commenced on August 12. After some adjustments for rate of the free pendulum, the case was finally sealed in October.

The clock has shown a large secular change of rate, causing a change in clock error in 100 days, after allowing for the mean rate, of nearly two seconds. When allowance was made for the secular change of rate and for the mean rate, there was a range in the residual errors, over a three months’ run, of about one-tenth of a second. The subsequent development of irregularities in both rate and arc was traced to variable friction of a blank ratchet wheel, introduced to prevent the gathering wheel of the half-minute circuit being pushed backwards on the return swing of the pendulum. The blank wheel has recently been replaced by a toothed wheel.

The photoelectric apparatus, for the purpose of taking time directly from the free pendulum, described in last year’s report, has been constructed in the workshop and mounted in position. It is expected that the final adjustments, for centring and focus, will be completed in the near future.

II. Buildings and Grounds, Movable Property, and Library: –

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

IV. – Time Service :-

Determinations of time have been made throughout the year by a rota of observers with the reversible transit instrument B. Observations were secured on 106 nights, the transits being registered on a syphon recording tape chronograph operated by a thermionic valve relay.

The personal equation machine, referred to in last year’s report, was-completed during the early summer, and a series of observations were made. Experience showed that certain modifications to the machine were required, and these have been made. For these observations, it was necessary to dismount the transit instrument from the transit pavilion, to remount it in the tank room on the roof of the Main Building and to refocus it on the artificial star of the personal equation machine. Whenever observations with the personal equation machine were in progress, time observations with the small transit instrument were therefore impossible. To avoid this inconvenience, an impersonal micrometer was accordingly constructed in the workshop and fitted to the similar transit instrument D. The transit instrument D is now mounted in the tank room on the roof of the Main Building, so that observations with the personal equation machine can be made when desired. Such observations will normally be made only on cloudy nights, when time determinations are not possible, and from time to time the instruments B and D will be interchanged. It is expected that a new series of measurements will be commenced in the near future.

The free pendulum clock Shortt No. 3 has been used as sidereal standard throughout the year and regular daily comparisons have been made with the reserve standard, Shortt No. 11. Steel suspension springs were fitted during the year to the Shortt free pendulum clocks Nos. 11 and 49. All the free pendulum clocks at the Observatory are now provided with steel suspension springs, in place of elinvar springs, whose behaviour was suspected to be capricious. The mean time free pendulum Shortt No. 16 was used throughout the year to control the mean time clocks, which provide the Post Office and broadcast signals, and also the rhythmic signal transmitter. This transmitter has been overhauled and has been fitted with a new synchronizing spring. It was found that the synchronizing spring was too strong, so that the acceleration of the transmitter due to the action of the synchronizer was 0s.010. Synchronization normally occurs once during the emission of the rhythmic signals and, when it occurred, an undesirably large discontinuity was produced in the signals. The new spring produces an acceleration of 0s.004.

The Hardy sidereal clock, which was formerly the sidereal standard clock and is now employed in the Airy Transit Circle observing room, has been cleaned and overhauled. The clock Dent 2 is being renovated and fitted with its original mercurial mean time pendulum. When finished, it will be placed in the Octagon Room. This clock was converted from mean time to sidereal time and for many years was used as a reserve sidereal standard and was later used as a subsidiary clock, controlled by the Cottingham-Riefler sidereal clock.

The service of rhythmic radio time signals at 10h and 18h. transmitted through the Post Office radio station at Rugby and also the hourly time signals to the General Post Office and the six-dot signals to the British Broadcasting Corporation, sent every quarter-hour, have been transmitted regularly. The hourly signals to the Post Office have been utilised for the automatic correction of the Post Office Talking Clock.

The radio time signals from Paris (Eiffel Tower), Bordeaux and Nauen have been regularly received, in addition to the Rugby signals sent out from Greenwich, and the corrections determined. The corrections obtained for the time signals from Nauen and Bordeaux have been published monthly in the Notices to Mariners together with the corrections to the times of the Rugby signals. The cathode ray oscillograph apparatus is employed for the determination of recorder lag. At least one of the time-signals sent out by the British Broadcasting Corporation is recorded daily, as a control on the time of emission. The following table gives the monthly means of the differences from Greenwich shown by the various time signals, using the final clock corrections derived from observations made with the small reversible transit instrument at Greenwich and the corrections for final time determinations published by the Paris and Hamburg Observatories:-

1936	Paris 9½ hr. Rhythmic. (s.)	Nauen 12 hr. Rhythmic.(s.)	Bordeaux 20 hr. Rhythmic.(s.)
January	-0.021	–	-0.019
February	-0.042	–	-0.039
March	-0.012	-0.015	-0.007
April	-0.006	-0.002	-0.009
May	-0.007	+0.014	-0.008
June	+0.013	+0.032	+0.007
July	+0.043	+0.036	+0.040
August	+0.027	+0.013	+0.024
September	+0.030	+0.016	+0.029
October	+0.023	-0.013	+0.016
November	+0.024	0.000	+0.022
December	+0.009	-0.017	+0.006
_____	_____	_____	_____
Mean	+0.007	+0.007	+0.005
Number of comparisions	299	243	344

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The times of the Nauen signals have been referred to the Eichelberger system of star places. The reception time of the signals has been corrected for time lag in the recording apparatus. Relative personal equations have been applied to the time observations.

A new radio time signal receiver has been constructed during the year by Messrs. Haynes Radio to the design of the Radio Department of the National Physical Laboratory. In general principles it resembles the receiver used at the National Physical Laboratory and described in the Report of the National Physical Laboratory for 1935, special attention having been given in the design to the constancy of the time lag in reception on all wavelengths. The receiver is designed for the long waves only, having ranges of 1,200 to 3,000 metres, 3,000 to 7,800 metres and 12,000 to 20,000 metres, but the addition of a short wave converter is now under consideration. In particular, it is desired to receive the American short wave signals, as no direct comparisons with Washington .have been possible since the long wave signals formerly sent out from Annapolis were discontinued.

Parallel tests are at present being made with the new receiver and the old Marconi receiver, in order that any discontinuity on change over may be avoided. The principle of recording the Greenwich sidereal standard clock signals simultaneously with the incoming radio signals is being used, as was done with the Marconi set. The clock contacts control a radio frequency oscillator, which acts as a low power, transmitter. The receiver picks up both the clock signals and the radio time signals and passes them through the set to the recorder.

The performance of the Westminster Clock during the year is shown by the following table giving the errors of the clock signals :-

	Error of Signal.	Number of Days.

	Not greater than 0s.2	118
	Not greater than 0s.2 to 0s.5	105
	Not greater than 0s.5 to 1s	49
	Greater than 1s	5

The clock is not corrected by any signals from the Observatory.

Return signals from the clocks at the Depots at Portsmouth and Devonport have been discontinued.

The Time Ball was raised each day and automatically released at 13h Civil Time, except on December 14, when it was not raised owing to a high wind, and on March 7, due to an accumulation of snow on the mast. It was also dropped in accordance with custom at 11h on November 11.

X. – Personal Establishment:-

Mr. W. Bowyer, Assistant, retired on October 31 after more than 44 years’ service at the Observatory, 40 of which were in an established capacity. During the Great War Mr. Bowyer dealt with an enormous amount of chronometer work in a highly successful manner. Since 1917 he had been in charge of the Time Department, the activities of which expanded considerably during his regime, of particular importance being the development of radio time signal work, and the introduction of free pendulum clocks. He has been succeeded by Mr. H. M. Smith.

XI. – General Remarks:-

At the invitation of the Postmaster-General, the Astronomer Royal attended the ceremony for the inauguration of the Post Office Talking Clock, at the Holborn Telephone Exchange on July 24, and made the first “Time” call. The Talking Clock, which is controlled by hourly signals from the Observatory, makes accurate time widely and continuously available and provides a service that is much used and appreciated.
…

Approval has been given for the supply of a quartz crystal oscillator clock and arrangements have been made for the necessary equipment to be provided, and assembly to be undertaken, at the National Physical Laboratory, where the work will be under the supervision of Dr. E. H. Rayner. The quartz vibrator will have a frequency of vibration of 100 kilocycles per sidereal second. Demultiplier circuits from the controlled oscillator will provide a frequency of 500 cycles per sidereal second, which will be amplified to drive a phonic motor, designed to give impulses at one-hundredth of a second, one-tenth of a second and one second intervals. A temperature variation of less than one-hundredth of a degree Centigrade will be ensured by means of a thermostat, and the quartz vibrator will be adjusted to have a temperature coefficient not greater than 5 parts in 100 million.

It is expected that the quartz oscillator clock will enable the Observatory to distribute time with an accuracy sufficient to meet the present-day demands for high precision in frequency control. The distribution of time necessarily depends upon predicted clock-errors and, in spells of cloudy weather, the clock errors may have to be carried forward, without observational check, for two weeks or even longer. The most accurate pendulum clocks yet constructed (the Shortt free-pendulum clocks) are liable to small erratic changes of rate, of the order of 0·01 or 0·02 seconds per day. Even with three standards for inter-comparison, an error of 0.05 seconds is easily possible at the end of a long cloudy spell. Corrections to the times of signals are determined and published monthly when the final clock error curve has been constructed. Though for some purposes these corrections enable the highest accuracy to be obtained, there are other purposes for which it is necessary that the errors of the time signals should be kept as small as possible. Quartz oscillator systems have proved stable, over limited periods of time, to the order of 0·001 seconds per day, though it is doubtful whether they are superior over long periods of time to free-pendulum clocks. It is probable that this type of clock will ultimately be considered as an essential part of the equipment of any observatory that is responsible for the determination and distribution of time.

1938

The Report here presented refers to the period from 1937 May 1 to 1938 April 30 and exhibits the state of the Observatory on the last-named day.

II. – Buildings and Grounds, Movable Property and Library:-

The old battery basement has been converted into a thermostatically-controlled room for the new crystal clock. A large capacity battery has been obtained from surplus stock at the National Physical Laboratory and work is in progress on a new battery room for it. The battery will be able to drive a small 220 volt A.C. motor generator in case of mains failure, so as to keep the crystal clock in operation and also hold the temperature of its cabinet constant for a limited period.

The adaptation of the old battery basement necessitated the transference of part of the apparatus for lifting the time ball to the ball lobby; at the same time, a new braking system was arranged, as well as an indicator to show the position of the ball, The coils of the electro-magnets in the release apparatus have been rewired.

The old watch room has been converted into a computing room for the use of the time department. The new watch room is now complete; cupboards for storing watch boxes have been fitted, and extra heating and lighting have been provided.

The charging panel for small batteries has been removed from-the wireless room and installed in a small cupboard near the time battery room. A new 7-pair cable has been laid from the lower computing room to the wireless room.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall. Harrison No.4 and the Larcum Kendall copy are being exhibited at the Empire Exhibition Glasgow.

IV. – Time Service:-

The pivots of the Small Transit (B) used for time determinations had become worn by the striding level. As they were made of gunmetal, it was decided to trim them down and fit stainless steel sleeves on to them. This was done by Messrs. Cooke, Troughton and Simms, Transit C taking the place of Transit B meanwhile. Subsequently Transit C was similarly fitted. In both cases the new pivots are somewhat conical and a little elliptical, as shown by tests with the mikrotast gauge; it is planned to lap them into shape shortly. The present performance is, however, at least as good as that of the old pivots.

Determinations of time have been made throughout the year by a rota of observers with the Transit Instrument B, except during the period September 9 to December 5 when steel sleeves were being fitted to the pivots; during this interval Transit C was substituted.

Observations were secured on 109 nights, the transits being registered on a syphon recording tape chronograph operated by a thermionic valve relay. Except during a repair period two pens were used for clock and transit respectively.

In previous years it has been the practice to employ as sidereal standard simply the free pendulum clock against which the transits were being directly recorded. During periods when observations were not obtainable the only method of detecting a change in rate was by comparisons with the reserve standards; such tests are often inconclusive. All pendulum clocks appear to be liable to sudden small changes of rate which, during cloudy weather, may give rise to considerable discrepancies between the true clock error and that predicted. With the object of minimising such effects, a change in procedure was adopted in 1937.

The mean of all available free pendulum clocks is now computed and fitted to an ephemeris. Astronomical observations give the error of one clock; by adding the difference between this clock and the mean clock the error of the mean clock is obtained; these observed errors are plotted against the ephemeris, and a smooth curve, drawn as nearly as possible through the points representing the observations, can then be extrapolated to give a forecast of the errors of the mean clock. The effect of a change of rate of anyone clock will thus be reduced in a ratio corresponding to the number of clocks incorporated in the mean. In so far as the smooth curve provisionally adopted for the mean clock does represent its real error, it is also thus possible to determine the real variations of anyone of the constituents.

Thus the error of anyone clock can be predicted with greater certainty, so that the time of transmission of signals is more nearly correct; and the final corrections, published after all the observations have been obtained, are also better determined.

For some years, by courtesy of the Director of the National Physical Laboratory, the daily errors of the free pendulum clock at Teddington, relative to the l0h Rugby rhythmic signal, have been forwarded to Greenwich. Since the error of the signal is determined at Greenwich, a comparison between the National Physical Laboratory clock and the Greenwich clocks is thus possible. From 1937 June 1 this clock has been incorporated in the mean clock, because it is of advantage to incorporate a clock that is not affected by influences local to Greenwich, which might possibly perturb all the Observatory clocks at the same time.

The sidereal clocks Shortt Nos. 3, 11 and 40 and the mean time clocks Shortt Nos. 16 and 49, together with the National Physical Laboratory Shortt clock, or those of them that are in use, are combined to give the mean clock.

The chronograph and sidereal circuits of the Observatory have been operated throughout the year by clock Shortt No. 3, and Shortt No. 16 has been used as standard mean time clock. For the purposes of incorporation in the mean clock, the rate of No. 16, which is corrected daily in order to be used as mean time standard, has been deduced from the amounts of correction required each day to keep it to time; on this basis, it has proved to be as accurate as any of the other clocks in use. The photoelectric cell apparatus on Shortt No. 40 has been in operation throughout the year and regular comparisons have been obtained.

During the summer it was noticed that the pallet wheels of both No. 3 and No. 11 showed signs of serious wear on the pivots. No. 11 was opened, and a new wheel, presented by the Synchronome Company, was mounted. It comes to rest exceptionally smoothly, after impulsing, but the performance of the clock was unreliable after the change; it proved difficult to seal the case up satisfactorily, and, during the manipulation involved, the rate several times changed suddenly by such large amounts that the case had to be opened again for re-rating. Finally a brass rod was passed down inside the case, until it rested on the top of the bob, and it was then struck smartly down on to the bob several times. This was done to ensure that the bob, which happens to be a very tight fit on the stem, really was properly bedded down on the retaining pin. Since then, the case has several times been tapped with a light mallet without producing any changes of rate greater than the normal and spontaneous ones. The pressure is still rising very slowly, but the clock will be taken into use again shortly.

As in previous years, the mean time standard clock Shortt No. 16 has controlled the clocks which provide the Post Office and the British Broadcasting Corporation time signals and the Rugby rhythmic time signals. All these clocks are magnetically corrected four times a day, as a result of inter-comparisons between No. 16 and the other free pendulums.

The Rugby time signal transmitter has undergone further modification to improve the regularity of the dots of the rhythmic series. A relay has been fitted to cut out the synchronizing coil, by which No. 16 controls it, during transmission of the series, so as to prevent a discontinuity during the emission of the signal due to the synchronizing action. Formerly, the signal was passed on to London by means of a biassed relay sending a single current signal. This relay has been replaced by a more modern type, and the circuit arrangements re-designed to allow this to be used unbiassed, and driven by the induced transients in the secondary of a transformer whose primary is in the clock circuit. The relay is used with a double current output, which means that the Post Office relay in London can also be used unbiassed.

The renovation of the clock Dent 2 has been completed and it has been installed in the Octagon Room. The Graham regulator has also been cleaned and repaired arid placed in the visitors' room. The clocks Earnshaw, Arnold 2 and Riefler 50 have been sold.

The service of rhythmic radio time signals at l0h and 18h transmitted through the Post Office Radio Station at Rugby and also the hourly time signals to the General Post Office and the six dots signals to the British Broadcasting Corporation, sent every quarter hour, have been transmitted regularly. The hourly signals to the Post Office have been utilised for the automatic correction of the Talking Clock, which has proved a remarkable success.

The radio time signals from Paris (FLE) , Bordeaux (FYL) and Nauen (DFY) have been regularly received in addition to the Rugby signals sent out from Greenwich,

The corrections to the transmitted times of the rhythmic signals for adjustment to the final time determinations are published and distributed monthly by the Hydrographer of the Navy in Notices to Mariners, together with the corrections obtained for the time signals from Nauen and Bordeaux.

The cathode-ray oscillograph has been employed for the determination of recorder lag. The six dots time signals radiated by the British Broadcasting Corporation have also been recorded in order to check the time of emission from Greenwich.

The following table gives the monthly means of the differences from Greenwich shown by the principal foreign time signals, and is computed by differencing the corrections to the signals as determined at Greenwich and at the observatory of origin.

1937	Paris 9½ hr. Rhythmic. (s.)	Nauen 12 hr. Rhythmic.(s.)	Bordeaux 20 hr. Rhythmic.(s.)
January	+0.014	-0.014	+0.006
February	-0.026	-0.032	-0.026
March	+0.012	+0.004	+0.008
April	+0.013	-0.015	+0.008
May	+0.050	+0.026	+0.046
June	+0.041	+0.010	+0.039
July	+0.036	+0.018	+0.031
August	+0.003	-0.024	0.000
September	+0.033	+0.002	+0.031
October	-0.001	-0.024	-0.006
November	+0.005	-0.020	+0.002
December	-0.021	0.000	-0.021
_____	_____	_____	_____
Mean	+0.014	-0.006	+0.009
Number of comparisions	313	296	344

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The times of the Nauen signals have been referred to the Eichelberger system of star places. The reception time of the signals has been corrected for time lag in the recording apparatus at Greenwich, but it is not known what corrections are applied at the foreign stations.

The radio time signal receiver constructed by Messrs. Haynes Radio to the design of the Radio Department at the National Physical Laboratory is not yet in regular use, as a number of small modifications and adjustments which have been found to be necessary have been held up by pressure of other work. The Marconi set, which has been in use for many years, has therefore been employed during the past year, but arrangements for a change over are approaching completion.

Considerable inconvenience has been occasioned by unbiassed relays acquiring a bias in use. It has been found necessary to keep a constant check on all relays employed on circuits of major importance. Investigations are being made with relays employing compliant tongues and stops with a view to their adoption on the time circuits, and it is also planned to reorganize the circuits so that there shall be no point at which a lag in any relay could produce an error in signal transmission that cancelled out on signal reception; at present it is possible in principle for undetectable errors of this nature to occur, but the actual lags are all small.

A short-wave wireless receiver has been ordered, but has not yet been delivered. It is hoped to receive the American short-wave time signals and thus to continue direct comparisons between Greenwich and Washington which have been in abeyance since June 1934, when Annapolis ceased to transmit on the long wave-length.

The performance of the Westminster clock during the year is shown by the following table giving the errors of the clock signals:-

	Error of Signal.	Number of Days.

	Not greater than 0s.2	131
	Not greater than 0s.2 to 0s.5	92
	Not greater than 0s.5 to 1s	52
	Greater than 1s	18

The clock is not corrected by any signals from the Observatory.

During the year trouble was experienced with the time ball apparatus and it was accordingly dismantled and overhauled and a number of parts were replaced.

The time ball was out of action on July 14, from August 17–26 and from September 28–October 8.

The ball failed to drop on September 5 due to an earth on the electrical release circuit. It was dropped in accordance with custom at 11h on November 11th.

XI. – General Remarks:-

The Astronomer Royal visited the Bureau de I'Heure at the Paris Observatory in December for the purpose of studying some of the methods and apparatus used in the determination and transmission of time and in the reception of time signals.

Considerable progress has been made at the National Physical Laboratory with the construction of the quartz crystal oscillator clock, referred to in the last report. Preparations, in anticipation of its completion, have been made at the Observatory and it is hoped that the clock will be installed during the present year.

1939

The Report here presented refers to the period from 1938 May 1 to 1939 April 30 and exhibits the state of the Observatory on the last-named day.

I. – Instruments and Apparatus:-

Quartz Crystal Clock.-The crystal clock referred to in several previous Reports has finally been delivered, after a short period of testing at the National Physical Laboratory. It has been installed in the cellar prepared for it, and is in process of being wired up.

II. – Buildings and Grounds, Movable Property and Library:-

In Flamsteed House the larder has been converted for the accommodation of batteries for the regular time service and also of the high-capacity cells required to provide current for the crystal clock in the event of an interruption of the mains supply. … ... New wires have been laid in conduits to the Rugby room, new battery room, and small transit pavilion.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

IV. – Time Service.

Determinations of time have been made throughout the year with a rota of observers with the Transit Instrument B. Observations were secured on 114 nights altogether, but 12 of these, made by observers who had recently been added to the rota, were not used in determining final values of clock correction. Further determinations of the absolute personal equation for the various regular observers are to be desired, but have not been possible on account of pressure of other work. Relative personality corrections, determined on the basis of the observational results, are applied. The Cooke level normally used with the instrument was broken on June 25, and was replaced by a Kern level which proved appreciably less satisfactory. A new Cooke level was substituted on October 17. The comparison of results obtained with these different striding levels indicated that, in the past at least, errors of level may well have been the most important single factor in the scatter of the observed clock errors. The present level is more stable and more rapid than any previous one, and the scatter is appreciably smaller than it has been before.

The principle of employing a mean clock, as described in last year’s report, has been further extended and has proved of great assistance. Comparisons of individual clocks with the mean show that in addition to sudden changes of rate they also have a tendency to wander from a steady rate and then to return; thus a linear interpolation between two clock errors, determined say a week apart, may give results several hundredths of a second out. This effect is greatly reduced by using the mean of a number of clocks, but it is hoped that the quartz crystal clock, recently delivered at Greenwich, will enable the effects of the fluctuations and discontinuities in rates of the pendulum clocks to be completely eliminated.

As in previous years, the daily errors of the free pendulum clock Shortt No. 13 at Teddington, relative to the 10h. Rugby Rhythmic Signal have been forwarded to Greenwich by courtesy of the Director of the National Physical Laboratory. In addition, the Astronomer Royal for Scotland arranged to communicate the errors of the standard clock, Shortt No. 4, of the Royal Observatory, Edinburgh. Thus there are now two clocks, which are free from any disturbance local to Greenwich, incorporated in the mean clock.

The following table shows the composition of the mean clock at different dates throughout the year:-

		Shortt Nos.
	May 1 to Aug. 31	3, 11, 13, 16, 40, 49.
	Sept. 1 to Dec. 28	3, 4, 11, 13, 16, 40, 49.
	Dec. 28 to Dec.30	3, 4, 11, 13, 16, 40.
	Dec. 30 to Jan. 31	3, 11, 16, 40.
	Jan. 31 to Mar. 27	3, 4, 11, 13, 16, 40, 49.
	Mar. 27 to Apr. 30	3, 4, 11, 13, 16, 40.

The free pendulum remontoire of clock Shortt No. 49 failed on two occasions, December 28 and March 27. In view of the erratic behaviour of this clock, the Synchronome Company has been asked to carry out a complete overhaul of the master and slave.

Shortt Nos. 4 and 13 were dropped from the mean during January on account of electrical faults which rendered their rates uncertain. All the clocks at Greenwich were affected by earth tremors at mid-day on June 11. The sidereal clocks in the inner cellar were all affected on April 2 when owing to a failure of the thermostat relay the temperature rose to over 100º F. The normal controlling point was not reached until the following morning. An improved type of thermostat relay was fitted on April 3.

The chronograph and sidereal circuits of the Observatory have been operated throughout the year by clock Shortt No. 3, and Shortt No. 16 has been used as standard mean time clock. The photoelectric cell apparatus on Shortt No. 40 has been out of action pending the provision of improved relay equipment.

The service of rhythmic radio signals at 10h. and 18h. transmitted through the Post Office Radio Station at Rugby and also the hourly time signals to the General Post Office and the six dot signal to the British Broadcasting Corporation sent every quarter hour, have been transmitted regularly. The hourly signals to the Post Office are utilised for the automatic correction of the Talking Clock, a service which has been further extended during the year. As in previous years, Shortt No. 16 has been used to control the transmitting clocks for all these signals; they are all corrected magnetically, as necessary, throughout the day as a result of inter-comparisons between No. 16 and the other free pendulums.

The radio time signals from Paris (FLE) , Bordeaux (FYL) and Nauen (DFY) have been regularly received, in addition to the Rugby signals sent out from Greenwich.

A short-wave receiver has been installed in order to continue direct comparisons between Greenwich and Washington which have been in abeyance since June, 1934, when Annapolis ceased to transmit on the long wavelength. The Arlington (NAA) signal on 9.425 megacycles has been recorded regularly since the beginning of January, the signals being taken at the times of best reception, which occur when there is darkness over the Atlantic. During January and February the 9h. signal was obtained with comparative ease; in early March it was found necessary to receive the 8h. signal which, as conditions deteriorated, was replaced by the 7h. signal. Satisfactory records of this signal were no longer possible in April, and the 23h. and Oh. signals are taken at present. This work has involved appreciable dislocation of schedules, and self-sacrifice on the part of the observers concerned; it has only been made possible by the great keenness of the junior members of the Time Department. To provide for its continuance on a more official basis is hardly possible with the present staff.

The corrections to the transmitted times of the rhythmic signals for adjustment to the final time determinations are published and distributed monthly by the Hydrographer of the Navy in Notices to Mariners, together with the corrections obtained for the time signals from Pans, Nauen, Bordeaux and Arlington.

The cathode-ray oscillograph has been employed for the determination of recorder lag. The six dots time signals radiated by the British Broadcasting Company have also been recorded in order to check the time of emission from Greenwich.

1938	Paris 9½ hr. Rhythmic. (s.)	Nauen 12 hr. Rhythmic.(s.)	Bordeaux 20 hr. Rhythmic.(s.)
January	-0.022	+0.008	-0.022
February	-0.019	-0.001	-0.017
March	+0.012	+0.003	-0.005
April	+0.006	+0.012	+0.006
May	+.027	+0.020	+0.026
June	+.014	+0.004	+0.011
July	+.006	+0.007	-0.007
August	+.002	+0.007	+0.002
September	+.001	-0.003	+0.003
October	0.000	-0.025	+0.007
November	-0.004	-0.035	-0.002
December	+0.003	-0.011	+0.002
_____	_____	_____	_____
Mean	+0.001	-0.001	+0.001
Number of comparisions	341	301	354

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The times of the Nauen signals have been referred to the Eichelberger system of star places. The reception time of the signals has been corrected for time lag in recording apparatus at Greenwich, but it is not known what corrections are applied at the foreign stations.

Further investigation of the magnitude of lags which may occur in any system employing mechanical relays has suggested the desirability of using thermionic relays exclusively on all time circuits. A thermionic relay may be designed to have a lag which for all practical purposes is negligible. The design of such relay units has been discussed with representatives of the research staff of the General Electric Company and sample units have also been submitted by Messrs. Marconi-Ekco. The co-operation of these- companies is appreciated. These developments, together with the further investigation and improvement of accuracy in the apparatus for the reception of signals, have been seriously delayed by the present conditions in the Time Department where the work has been increasing for several years; during the past year exceptional additional demands have been made. The increase of load which has resulted has been partly relieved by temporary transfer of staff from other departments, but an appreciable increase in staff appears unavoidable.

It has been decided to install at the Abinger Magnetic Observatory the necessary apparatus for the maintenance of a subsidiary time service which could take over, whenever required, the more essential duties of time signal distribution normally carried out from Greenwich. The transit instrument D, which had been used in connection with the personal equation machine, has been mounted there, and pending the provision of additional equipment, the Dent clock formerly used as the sidereal standard clock at Greenwich and the mean time Dent regulator 2016, previously kept as reserve time signal transmitter, have been erected in the old battery room, which was cleared and partitioned to enable a reasonable temperature control to be maintained. Temporary chronograph equipment has also been installed.

The performance of the Westminster clock during the year is shown by the following table giving the errors of the clock signals :-

	Error of Signal.	Number of Days.

	Not greater than 0s.2	93
	Not greater than 0s.2 to 0s.5	95
	Not greater than 0s.5 to 1s	86
	Greater than 1s	16

The clock is not corrected by any signals from the Observatory.

The Time Ball was raised and dropped correctly throughout the year except on October 4, November 20, on account of high winds; December 23, January 25, snow on mast; November l4, January 22, technical failures. It was dropped in accordance with custom at 11h. on November 11.

1940

I.-Instruments and Apparatus:-

The clock errors derived from observations with this instrument [the Cooke Reversible Transit Circle]are now being used for the determination of time, in addition, of course to the usual observations by the Small Transit. There is a small irregular but apparently systematic difference between the two instruments as measured over a period of 20 months. More data will be required before it can be decided if this is a regular seasonal variation. The observations of clock error by the Reversible Transit Circle show slightly less scatter than those made on the Small Transit, and agree better with the foreign radio time signal. This is a very satisfactory result, in view of the fact that large instrument are not usually considered as accurate as small reversible one.

II.-Buildings and Grounds, Movable Property and Library:-

A concrete housing for the new time desk and terminal frame was erected on the site of the cold greenhouse; a new greenhouse has been erected adjoining this room.

A concrete cable trench for new time circuits has been laid through the Time Department, the Airy Transit Circle house and the passage.

At Abinger, a wooden housing for the small transit instrument has been erected. Special care was taken in the design of this building to secure adequate ventilation and to avoid stratification of the air. A portion of the old battery room has been divided off by insulated plaster board for use as a clock room and thermostatically controlled electric heating has been installed. The former engine room has been divided off by glazed partitions and a block flooring has been laid.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

IV. Time Service

Determinations of time have been made throughout the year by a rota of observers with the small transit instrument B. Observations were secured on 121 nights. The application of relative personality corrections determined on the basis of the observational results has been continued. The new Cooke level referred to in the previous report has been in use throughout the year. A number of observations secured with the reversible transit instrument have also been employed as an additional check. In accordance with the recommendations of the International Astronomical Union, star places based on the FK3 catalogue have been employed for the reduction of time observations since January 1. The universal adoption for time determinations of the same fundamental system of star places will make the times determined at different observatories more directly comparable than hitherto.

The principle of employing a mean clock has been continued, and by the courtesy of the Astronomer Royal for Scotland, and the Director of the National Physical Laboratory, it has been made possible to continue the use of clocks Shortt Nos.4 and 13 an the mean. The following table shows the composition of the mean clock at different dates throughout the year.-

		Shortt Nos.
	May 1 to Jul. 31	3, 4, 11, 13, 16, 40.
	Aug. 1 to Jan. 7	4, 11, 13, 16, 40, 49.
	Jan. 8 to Jan. 31	3, 4, 11, 13, 16, 40.
	Feb.1 to Apr. 30	3, 4, 11, 13, 16, 40.

Shortt 49 after having been overhauled and cleaned, was put into the mean on August 1. Shortt 3 was fitted with a new impulse wheel and bracket on August 3 and the weight tray was turned down to compensate for the additional weight of the new bracket. Shortt 49 was found stopped On January 7, and was cleaned and restarted. Shortt 3 was opened on February 1 for cleaning the contacts, and was again opened on April 24 for resealing.

The main battery employed for all the time circuits has given rise to some concern for the past few years, and pending the installation of a new alkaline battery, a number of cells were replated and put back in circuit on November 29. Some units of the alkaline battery were hurriedly put into use for a short time, but the discrepancy in the voltage per cell combined with the inadvisability of placing alkaline and acid batteries in close proximity, rendered it necessary for the Electrical Engineering Department, Chatham Dockyard, to supply a new reserve battery. The new alkaline battery still awaits installation, and none of the auxiliary equipment is yet to hand.

The chronograph and sidereal time circuits of the Observatory were operated by Shortt 3 until July 31, since which date they have been controlled by Shortt 11. Shortt 16 has been used throughout the year to control the mean time circuits. The quartz crystal controlled oscillator clock is running satisfactorily, but with rather a large rate. It is not proposed to stop it for readjustment. The phonic motor has proved to be the most troublesome part of the equipment and it is apparent that the clock is not giving as good a performance as it should. At the completion of the present run it may be necessary to have a fresh crystal cut, and advantage will be taken of the opportunity to employ a fundamental frequency of 100 kilocycles per mean solar second, instead of per mean sidereal second as at present.

The service of rhythmic time signals at 10h and 18h, radiated by the Post Office long wave GBR .transmitter on a frequency of 16 kilocycles per second, had been maintained except for the period January 31, 18 hours, to February 19, 10 hours, when the transmitter was not available. From February 15 the signals were transmitted simultaneously from the stations GBV (78 kc/s), GIA (19·640 mc/s) and GIM (12·975 mc/s) at 10 hours and from GBV, GIM, and GIR (10·650 mc/s) at 18 hours. The signals from GBV were discontinued from the end of March, and the short wave signals from Apr11. Advantage has been taken of the opportunity thus afforded to request reports on the reception of the short wave signals as this is the first occasion on which Greenwich time signals have been transmitted on the short wave band. The B.B.C. six dots time signals and the hourly signals to the Post Office for controlling the Talking Clock have been transmitted regularly. As in previous years, the signals have been sent out from Dent 2012 controlled by the mean time standard. The reserve signal transmitting clock, Dent 2016, is on loan to the emergency time station pending the installation of Dent 2009, which is undergoing modification. When Dent 2009 is completed, Dent 2016 will resume its duties as the reserve at Greenwich.

The rhythmic time signals from Paris (FLE), Bordeaux (FYLl and Nauen (DFY) have been recorded regularly. From June 1 the Ameican long wave transmitter NSS (17.6 kc/s) recommenced radiating the Washington time signals. The resumption of these signals which had been discontinued June 1934, was made possible through the co-operation of the U.S. Navy Department, in response to a request from the Committee on Cosmic Terrestrial Relationships of the American Geophysical Union, in order that regular intercomparison of times signals between the Royal Observatory Greenwich, and the. U.S. Naval Observatory, Washington, should be made on low frequency waves which are less subject to ionospheric disturbances than waves of high frequency.

The American transmissions were recorded from the first day of their inception, and apart from occasional bad records, regular comparisons have been made.

The 0h or 23h signals from Washington (NAA) were recorded by means of the R.M.E. short wave receiver until May 25, when the receiver was transferred to the emergency station. Since the new Skyrider SX 17 short wave receiver became available on February 19, the American 14h or 15h signals have been recorded.

Until February, the FIE time signals were recorded on the equipment previously employed. The Marconi r-receiver was used for GBR throughout the year, and for FYL, DFY and NSS until February. The special long wave receiver designed at the National Physical Laboratory and constructed by Messrs. Haynes Radio has been used since February for the signals from GBV, FYL, DFY, NSS, and FLE, for GBR since the cessation of transmissions from GBV. Two Hallicrafter Super Skyrider Special SX 17 receivers have been on loan from Messrs. Webb’s Radio since February, and a new National HRO is on order and should arrive shortly. The wireless recording apparatus was removed from the wireless room in September to a better protected position, and is now assembled in a very temporary manner. Despite the make-shift aerial system, the performance of the Skyrider sets has been good, and little difficulty is experienced in obtaining excellent records from stations as far apart as Moscow, Rio de Janeiro and Arlington. It is planned to install the receiving and recording apparatus as soon as possible in a room below the old wireless room, and to erect efficient aerial systems designed to minimise interference from the clock circuits of the Time Department. The cathode-ray oscillograph apparatus for the determination of recorder lag has not been in use since September, but it is proposed to install it in an improved form when the removal of the wireless apparatus is made. The need for additional chronographs has become acute; a modified design is being drawn up in consultation with the Foreman of the Observatory and the chronographs will be constructed in the workshop.

The corrections to the transmitted times of the rhythmic signals for adjustment to the final time determinations, together with the correct ions obtained for the foreign time signals recorded at Greenwich were published and distributed monthly by the Hydrographer of the Navy in Notices to Mariners, until March, 1939. Subsequently they have been circulated direct from Greenwich.

The following table gives the monthly means of the differences from Greenwich shown by the principal foreign time signals, and. is computed by differencing the corrections to the signals as determined at Greenwich and at the observatory of origin.

1939	Paris 9½ hr. Rhythmic. (s.)	Nauen 12 hr. Rhythmic.(s.)	Bordeaux 20 hr. Rhythmic.(s.)
January	-0.008	+0.001	-0.006
February	-0.004	+0.018	-0.003
March	+0.010	+0.028	+0.009
April	-0.005	+0.008	-0.007
May	-0.007	-0.033	-0.008
June	+0.016		+0.011
July	+0.035		+0.030
August	+0.028		+0.029
September	+0.017		+0.012
October	+0.063		+0.068
November	+0.093		+0.091
December	+0.033		+0.031
_____	_____	_____	_____
Mean	+0.022	+0.003	+0.022
Number of comparisions	324	128	340

The sign + means late on Greenwich. The reference point for all comparisons is the commencement of the signals. The times of the Nauen signals have been referred to the Eichelberger system of star places. The reception time of the signals has been corrected for time lag in recording apparatus at Greenwich, but it is not known what corrections are applied at foreign stations.

Further investigation of thermionic relays has led to the design of suitable units which have been tested experimentally. The complete installation of these units has been held up by pressure of other work, and by the delay in the carrying out of the rewiring scheme. A new room has been constructed for the central control desk, but only a part of the equipment is so far to hand. In order to cause minimum interruption of the Time Service, it is proposed to wire up all the new system to the new control desk and new main battery before removing any of the existing wiring. It may, however, be necessary to run for a short time from the reserve station.

Despite some additions to staff coupled with continued temporary loans from other departments, the staffing question is still not satisfactory. The splitting up of the Department caused by the establishment of a reserve time station and the transference of the chronometer and watch work, has not made for economy of operation.

The setting-up of a duplicate time service is now almost completed. The equipment includes a small transit instrument in a newly-erected observing hut, two new free pendulum clocks – Shortt 66 (Sidereal) and Shortt 67 (Mean Time), a rhythmic signal transmitter and a time signal transmitting clock on loan from Greenwich. Mr. H. R. Fry has kindly loaned his free pendulum clock, Shortt 61, a chronograph, a wireless receiver, and a calculating machine. A short wave wireless receiver is on loan from Greenwich. Thermionic relays are being employed, and special lines have been installed to connect the emergency station with the Post Office, GBR transmitter, and the B.B.C. This station will shortly be capable of taking over the Time Service duties from Greenwich. The clocks Shortt 61, 66, and 67 will be incorporated in the Greenwich mean clock.

The performance of the Westminster Clock during the year is shown by the following table giving the error of the clock signals:-

	Error of Signal.	Number of Days.

	Not greater than 0s.2	46
	Not greater than 0s.2 to 0s.5	58
	Not greater than 0s.5 to 1s	58
	Greater than 1s	44

The clock is not corrected by any signals from the Observatory.

The Time Ball was raised and dropped correctly throughout the year except on October 24, when it was raised half way and lowered by hand; and on January 17, owing to snow on the mast, and on March 25, when it was raised and dropped at 14h. It was dropped in accordance with custom at 11h. on November 11.

X. Personal Establishment [p.28]:-

The further improvement in accuracy of the time service provided by the Observatory has been under consideration. For purposes where high precision is needed, corrections to the time signals are published at approximately monthly intervals, after the errors of the mean clock have been satisfactorily determined from a large number of observations. The great developments in recent years in precision frequency standards and in their use have created a need, which is extending rapidly, for time signals of such precision that an interval of 24 hours can be measured to an accuracy of one thousandth of a second, without the application of corrections published after a considerable interval. This precision is not attainable with pendulum clocks, which are less accurate than the precision frequency standards themselves. The installation of a battery of quartz crystal oscillator clocks of the highest precision is required, in order that the frequency stability of modern frequency standards may be satisfactorily controlled. Conferences on this subject have been held with the Radio Section of the General Post Office and the Research Department of the British Broadcasting Corporation. The help and collaboration of these bodies and also of the National Physical Laboratory are greatly appreciated and should go far towards enabling a satisfactory scheme to be drawn up.

Confidential addendum to Report:-

In the early days of the war, an ad hoc time-service was set up at the Magnetic Observatory, Abinger. This was intended to supply a time-service of a somewhat lower order of accuracy than that provided from Greenwich, with hourly signals to the Post Office for the control of the Talking Clock and for other purposes and with quarter-hour signals to the B.B.C. The generous loan by Mr. H.R. Fry of his Shortt mean-time free-pendulum clock (which has been converted to a sidereal standard), a long wave radio-receiver and a tape chronograph made it possible to have a skeleton time-service available within a few weeks of the outbreak of war.

A complete time-service has now been installed at Abinger which will be able to take over all the functions of the time-service from Greenwich, including the sending out of the Rugby rhythmic time-signals, with about the same accuracy. A permanent housing for the small-transit instrument has been erected. It is planned to operate the times-service from Abinger whilst the change-over of all the time circuits to the new terminal board is being made, in order to try out the service.

The Post Office lines from both Greenwich and Abinger pass through the Control Room at the G.P.O. London. If this room were put out of action, time-signals could not be distributed from either Greenwich or Abinger. Arrangements are in train for sending signals via Birmingham; meanwhile, arrangements have been made with the Astronomer Royal for Scotland by which time-signals (with the exception of the Rugby rhythmic signals) can be sent from the Royal Observatory, Edinburgh, whenever this should prove necessary.

1941

The Report here presented refers to the period from 1940 May 1 to 1941 April 30 and exhibits the state of the Observatory on the last-named day.

I. Buildings, Grounds, Movable Property and Library:-

Two of three two-pen chronographs have been constructed in the Workshops. These chronographs are fitted with synchronous induction motors, for conveying the tape at a uniform speed. The third chronograph is now under construction, together with a portable double spring gramophone motor as an alternative for conveying the tape on either machine.

A new relay control panel for the clock circuits and chronograph has been completed for use at one of the emergency time stations.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

III. Time Service:-

The conditions prevailing during the past year have necessitated a number of changes in the arrangements for the Time Service. The setting-up of a duplicate time-service, referred to in last year’s report, was completed. The determination of time and the distribution of time signals from Greenwich was continued until September. The continuous all-night air raids then made night observations impossible and the service was accordingly transferred to the emergency station A. For some weeks time observations were made at this station whilst time signals continued to be sent out from Greenwich. The transmissions from Greenwich were finally terminated towards the end of November and all signals.have since been sent out from the emergency stations. The small reversible transit B remained in use at Greenwich until observations ceased; from October 19, it replaced instrument D at the emergency station.

As equipment was not available at the emergency station for the determination of relative personal equations of observers, and as it was necessary to introduce certain new observers at this station, observers were employed in pairs, both making observations, until sufficient material for the determination of relative personal equations was available. A provisional longitude of the emergency station, based on observations made during the period of overlap, when observations were being made both at this station and at Greenwich, has been applied to subsequent observations to reduce them to the meridian of Greenwich.

A second emergency station B .has been equipped to ensure maintenance of the time service in the event of damage to station A or to Post Office lines used in the distribution of signals. The Shortt mean time clock No. 16 and sidereal clock No.ll, together with Rugby diminished seconds transmitter, have been transferred from Greenwich and mounted at this station. Time observations are made with a reversible transit instrument and radio receivers have been transferred from Greenwich for the reception of time signals. The station B has been available for the emission of signals since the end of December and has taken over the-Rugby rhythmic time signal service during a number of periods when, for one cause or another, it was not possible to transmit them from station A. The two stations are now being run in close collaboration and arrangements are being made for time signals to be radiated from that station at which conditions are most stable at the time. Every precaution has been taken to minimise the discontinuities when the change from one station to the other is made, and a continuous check on the signals is maintained at both stations.

Time determinations were made with the small transit instrument at Greenwich on 63 nights from 1940 May 1 to September 4. In addition 29 determinations with the reversible transit circle were included in the derivation of the final clock curves. After September 4 time determinations were secured at the emergency station A on 74 nights and from December 11 at station B on 26 nights. The observations at the two stations are combined for the derivation of the final clock curves.

The errors of the Shortt clocks at the Royal Observatory, Edinburgh and at the National Physical Laboratory are still determined by means of the Rugby time signals, and the results are communicated daily. A mean clock, employing all the available data, is used in establishing final corrections to the times of reception of radio time signals.

The National HRO short wave receiver, referred to in the last report, has been received and has been in regular use. Its performance has proved to be highly satisfactory. A Skyrider SX17 receiver was also purchased and installed as a reserve and a further similar receiver is being purchased. These short wave receivers have made it possible to secure more regular reception of American time signals. The records of the Nauen long wave time signals have also been supplemented by regular reception of the corresponding short wave transmissions from DFC and DGZ. The long wave American time signals from NSS and the long wave signals from Bordeaux are also received. The time signals are recorded at both emergency stations.

It is not known whether the Bureau de l’Heure continues to function and corrections to continental time signals have not been received. It is therefore not possible to give the monthly means of the differences from Greenwich shown by foreign time signals.

Two new chronographs have been completed in the workshop, one of which is in use at station A and the other at station B. A third chronograph is approaching completion. Since September it has not been possible to arrange for the release of the Greenwich time-ball, nor to maintain the check on the Westminster clock. The well known Shepherd clock near the main gate of the Observatory, which for many years has been controlled by the mean time system, has sustained damage. Both the Time-ball and the Shepherd clock have been popular features among members of the public visiting Greenwich Park, but otherwise they have ceased to be of importance.

General research and development work has naturally been indefinitely postponed. The quartz clock together with its standby electrical equipment have been stored until more favourable conditions permit various modifications to be made. The proposals for the installation of a number of quartz clock units and ancillary equipment have of necessity been held over. As soon as the situation permits, however, it will be necessary to embark on a bold policy in order to make up for the present period of stagnation, since the standard of the service at present possible falls short of the high precision required for the adequate control of precision frequency standards.

Confidential addendum to Report:-

With the co-operation of the Astronomer Royal for Scotland, a complete time-service is now provided by the Royal Observatory, Edinburgh (Station B of the report). Mr. Symms, Junior Assistant, is stationed at Edinburgh, to supervise this service and to assist the staff at the Observatory. The determinations of time are made with the reversible transit circle and the time is kept by the standard clocks of the Observatory, well-mounted in temperature controlled clock-cellars, which have been supplemented by one Shortt sidereal and one Shortt mean-time clock from Greenwich. As no spare preamble apparatus for giving the GBR Rugby call-sign before the Rugby time-signals, was available for sending to Edinburgh, a preamble apparatus has been constructed by the Post Office and has been installed at Edinburgh. The co-operation of the Astronomer Royal for Scotland in helping to ensure the uninterrupted maintenance of the time service if the country is greatly appreciated.

1942

I. Buildings, Grounds, Movable Property and Library:-

A building has been erected at Time Station A to provide suitable and adequate housing for three Shortt free-pendulum clocks and for the quartz clock, with a minimum expense and labour. It consists of two separate clock chambers, inside a main brick and concrete room, approximately 25 square feet in area. The outer room is heated by hot-water radiators, the clock chambers being electrically heated, with thermostatic control. The slave clocks and ancillary equipment will be installed in the outer room. Progress on the building, which has now been completed, was slow; the installation of the necessary wiring, preparatory to the erection of the standard clocks, is in hand.

The third two-pen chronograph, mentioned in last year's report, has been completed in the workshop. A vibration machine, for testing the rates of aircraft watches under vibrations similar to those to which they are subjected in aircraft, has been constructed in the workshop.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

A Shortt free-pendulum clock, No.61, a radio receiving set, a chronograph and a Brunsviga calculating machine, from Mr. H. Fry.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall.

III. Time Service:-

The two time-service stations, referred to in the previous report as Stations A and B, have been maintained in full operation throughout the year. Each station maintains a complete time service, involving transit observations, clock comparisons and transmission of time signals along the Post Office lines to Rugby Radio and to the British Broadcasting Corporation. The time signals from Rugby are normally radiated from the station at which conditions happen to be most stable; instructions are given to the Engineer-in-Charge, Rugby Radio, which signals are to receive preference. If, because of any technical trouble developing at a station or on the line between it and Rugby, signals from only one station are received at Rugby, they are transmitted, whether that station was previously sending out the signals or not. If the clocks are running equally well at the two stations, it is intended that the two stations shall alternate at approximately monthly intervals in sending out the signals.

During the period covered by this report, the Rugby time service has been operated continuously from station B, where the clocks are better mounted than at Station A, except on a few occasions when technical trouble at the station or line faults made it necessary for the signals from station A to be transmitted.

The “six-dots” time service to the British Broadcasting Corporation and the hourly signal transmissions to the Post Office, which also control the Talking Clock, have been normally operated from Station A. The service from station B has continued in full operation, although these signals have only been required occasionally.

Transit observations for the determination of clock error were made at station A on 116 nights and at station B on 137 nights during the year. The observations at the two stations have been combined for the determination of clock error; so that a total of 253 observations during the year is available. The relative personalities of the observers at station A have been determined from a special series of observations, in which observers were employed in pairs. The two observers observed the same transits, symmetrically on either side of the meridian, before and after reversal of the instrument. Errors in star places and errors of instrumental origin were thus eliminated. The personalities so determined were in good agreement with those obtained from the comparison of the residuals of the observations by different observers from the smoothed clock curves, Weighted means of the results of the two determinations have been adopted and applied from January 1.

Approximate personalities of the observers at station B have been determined from the residuals from the smoothed clock curves and have been adopted provisionally, pending the accumulation of further data.

The usual regular check on the rhythmic time signals transmitted from Rugby at 10h and 18h has been maintained at both stations. The comparisons of the clocks obtained by means of the reception times of the 10h signals have served as the link between the two stations.

Both stations have recorded, whenever· possible, the American and German time-signals on short waves as well as on long waves. Station A has also maintained a check on the French signals at 8h. The published final corrections have been based on the reception times at the two stations, when both are available.

At each time station, the Rugby signals are recorded against three standard clocks and the signal transmitter. Two sets of ten signals, preferably near the beginning and end of the transmission, are measured for each clock, the chronograph pen units being interchanged in the middle of each set. This method reduces the effects of short period erratics in the clocks and provides a strong link between the two stations. The same method has been extended to the receptions of foreign time signals.

A regular check is kept on the B.B.C. six-pips time signals; also on the Westminster Clock, by recording the radio broadcasts of 'Big Ben'.

The errors of the clock Shortt 4 at the Royal Observatory, Edinburgh, and of the clocks Shortt 13, Quartz 2 and Quartz 6 at the National Physical Laboratory are determined by means of the Rugby time signals and the results are communicated daily. This co-operation by the Astronomer Royal for Scotland and the Director of the National Physical Laboratory is of great assistance and is much appreciated.

The published corrections to the times of the Rugby time-signals and to the foreign time signals are based on some or all of nine clocks, – seven Shortt clocks and two Quartz clocks. The continuous operation of two stations and the large number of clocks concerned has greatly increased the amount of computation involved in the preparation of these corrections, which is undertaken at Station A. A simplified and improved method of dealing with the data has therefore been introduced. The transit observations from both stations are now referred to each clock in turn. The observed errors are plotted against an ephemeris for each clock. The clocks that have the most uniform rates are selected to form the mean clock. Smooth curves are drawn through the observed points as they appear on each chart. The several slightly differing systems of time defined by these curves are finally combined into a mean system, from which the error of any clock at any instant can be deduced. The process of determining the final corrections to the times of the signals has been made largely differential and graphical methods have appreciably reduced the amount of computation. It is also possible to remove one or more clocks from the mean or to introduce additional clocks without difficulty.

Some changes in the routine of signal transmission and reception at station B have resulted in a distinct advance in the standard of accuracy of the Rugby rhythmic time-signals. The changes of signal rate intentionally introduced have been restricted to a maximum figure of 3 milliseconds per day per day and have normally not exceeded 2 milliseconds per day per day. By working on a “mean clock” it has been found possible to impose this restriction and at the same time to keep the signals sufficiently close to the times indicated by the transit observations. The value of the uncorrected signals as a standard of frequency has thus been increased. Attention has been paid to the question of variation of lag of long-wave wireless reception and, by taking appropriate precautions, it has been found possible to reduce the probably error of such variations to less than 0.5 milliseconds. Provision has been made by the Post Office for a return signal by wire along the Post Office Lines between the station and Rugby. The reception of these signals has given an extremely satisfactory means of checking the time of transmission of the signals along the lines. It is now possible to make an accurate allowance for line-lag in setting the transmitter clock before the time signals are sent out. Provision of a similar return signal to station A has not been possible.

At both stations, the action of the "hit and miss" sy[n]chroniser of the transmitting clock is suspended during the signal period and an appropriate current through the magnetic corrector coils compensates for the normal losing rate of the transmitting clock, when removed from the control of the free pendulum.

The measurement of the various time signals against the available clocks and the intercomparisons of clocks at both stations are made with the aid of thermionic panel units, developed and constructed at station A. These units eliminate the variations of lag that are associated with mechanical relays.

It has been mentioned that the clocks at station B have performed consistently better during the year than the clocks at station A. This latter station was brought into use at short notice when the time service had to be transferred from Greenwich at the time of the heavy raids on London in the autumn of 1940. The accommodation for the clocks is very restricted and there is serious mutual interference between them, giving rise to fluctuations in rate. The new building, referred to at the beginning of this report, will eliminate these troubles and enable a time service of satisfactory accuracy to be provided by this station.

At station B, an investigation has been made of the various types of clock erratics, which has given some results of interest. Amongst these, it has shown clearly the superiority of quartz clocks over the best free-pendulum clocks in the matter of small erratic changes of rate.

Confidential addendum to Report:-

The two stations that provide the time service are the Magnetic Observatory, Abinger (Station A) and the Royal Observatory, Edinburgh (Station B.) The provision of a complete time service from Edinburgh has made heavy demands upon the Astronomer Royal for Scotland and the staff of the Royal Observatory Edinburgh, who have been assisted by Mr. Symms, Junior Assistant, Higher Grade (Acting), on loan from the Royal Observatory, Greenwich. During a period when the clocks at Abinger have performed indifferently, this assistance has been of the greatest value in maintaining a series of rhythmic time signals from Rugby of the high accuracy that is now demanded. Professor Greaves has introduced various refinements in technique, which have made it possible to ensure that the interval between corresponding signals on consecutive days does not normally differ from 24 hours by an amount exceeding 0.003 seconds. The co-operation of the Astronomer Royal for Scotland is greatly appreciated.

The continuous operation of two separate time service stations has been advantageous in increasing considerably the number of determinations of time during the year, as not infrequently determinations can be made at one station during spells of cloudy weather at the other. The total of 253 time determinations is greater than in any previous year; the combined determinations at the two stations are used for checking the clocks at each station. When the Board of Visitors approved in principle the removal of the Royal Observatory and of the Magnetic Observatory to new sites, a recommendation was made that an auxiliary time station should be provided at the new site for the Magnetic Observatory. The experience that has been obtained in the operation of two separate time service stations has shown that the duplication is advantageous not merely for security purposes but also for increasing the number of time determinations, and that it is of material assistance in maintaining a time service of the required accuracy.

1943

The Report here presented refers to the period from 1942 May 1 to 1943 April 30 and exhibits the state of the Observatory on the last named day.

I. Buildings, Grounds, Movable Property and Library.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

A Shortt free-pendulum clock, No.61, a radio receiving set, a chronograph and a Brunsviga calculating machine, from Mr. H. Fry.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

Commander Gould conducted, in November, a thorough examination of their condition, with the permission and cooperation of the authorities with whom they are stored. He reports that No.1 is slightly affected by rust in certain parts and will require overhauling after the war, although no replacements will be necessary. Nos, 2 and 4 and the Kendall copy were in excellent condition.

III. Time Service.

Throughout the year, the practice has been continued of operating the time service from two stations, each of which forms a self contained unit fully equipped to maintain the service in the event of serious dislocation at the other. Further experience with this dual system, which was introduced as a war-time measure, has clearly demonstrated the desirability of providing for two stations as a permanent measure. The potentialities of the arrangement have been greatly increased by the provision, during the past few months, of a direct link between the two stations, a facility which has been made possible by the ready co-operation of the Post Office Telegraph Branch. The lines which enable each time service to transmit hourly signals to the Post Office, are automatically switched over to give intercommunication except for a period of three minutes at each hour, and apparatus has been installed to permit clock signals to be sent each way. Teleprinters have also been connected for the exchange of information. The direct link has already become invaluable and has led to the accumulation of useful data, which promises to be of considerable assistance in elucidating some of the uncertainties of comparisons by means of a radio link. In normal conditions it is also possible for both stations to contribute data from their best clocks to a common pool on which the actual transmission from each station can be based.

During the period covered by this report the time signals radiated from Rugby have normally been controlled by station B. Apart from the few occasions when, owing to technical faults, transmissions have not been possible from this station, it has also been the practice during the last three months to transmit one 18.00 hour signal per week from station A in order to ensure that the system and lines are in order. As mentioned last year, it has been possible to exercise a regular control of the transmission from station B by the utilisation of a return signal to detect changes in the lag of the land line. The provision of a similar loop circuit to station A presented great practical difficulties, but owing once again to the co-operation which has been established with both the Radio and Telegraph branches of the Post Office a return line was connected in March. Although not on the same basis as the return line to station B, it nevertheless provides information by means of which changes in the lag of the sending line may be detected and compensated. The arrangement by which the actual transmission was made from that station at which the standard clocks were more reliable has given way to the principle of giving preference for the 10.00 hour transmission to the station with the more stable transmitting clock and the better conditions on the sending line to the radio transmitters, as comparisons with the best standard clocks may be made by means of the inter-station line from either end. It is necessary for some of the 18.00 hour signals to be transmitted from the station acting as reserve in order to maintain a check on the sending line lag in case of failure of the signal from the station which has been given preference.

As a result of experience gained in 1940 when the 10.00 and 18.00 hour signals were transmitted by a group of short wave stations and primarily in order to meet the needs of survey workers abroad, short wave signals were introduced from July 1942, and have been radiated simultaneously with the long wave emissions from GBR. The particular stations and frequencies employed at any time are dependent on conditions and requirements, but in most cases both time service stations are within the skip distance and reception for monitoring purposes is unreliable. Final corrections to these signals have therefore been given to the nearest hundredth of a second.

The “six dots” time service to the BBC has normally been operated from station A, and the service of hourly signals to the Post Office, which also controls the talking clock, has been taken from both stations.

169 Transit observations were made at Station A on 142 nights and 157 at station B on 156 nights during the year.

As before, all observations have been combined for the determination of clock errors. The approximate personalities, determined as described in the last report, have been applied throughout. During the year it was found that the adopted scale value of the striding level used at station A was in error, the effect having been to increase the scatter of the observations. A careful redetermination of the scale value was made in December, and observations were adjusted from July 1.

At station A a new observing hut has been erected to accommodate the Bamberg transit instrument loaned by the Astronomer Royal for Scotland. It is proposed to equip this instrument with an electrically-driven micrometer but pressure of urgent work has delayed the completion of this scheme.

The usual regular check on the rhythmic time signals transmitted from Rugby at 10.00 and 18.00 hours has been maintained at both stations. Before the installation of the land line connection, the comparisons of the clocks obtained by means of the reception times of the 10.00 hour signals served as the link between the two stations. Both stations have recorded, whenever possible, the American and German time signals on both long and short waves, and station A has also maintained a check on the French signals at 08.00 hours. The published final corrections have been based on the reception times at the two stations, when both are available. A modification in signal transmission procedure has been introduced to increase the value of the signals for the purposes of frequency checking. Every endeavour is now made to keep the rate of the signals as near zero as data permit, i.e. to give successive signals at intervals of 24 hours with the greatest possible accuracy. In the event of subsequent data indicating that the signals exceed a certain specified error in absolute time, the return will not be made gradually, but in a jump, the signals then continuing with a zero rate as nearly as possible.

Identical long wave receivers, of standard Admiralty pattern, have been installed at both stations, and modified for the reception of time signals. A continuous dash, introduced between the morse preamble and the commencement of the rhythmic signals, has been instituted to facilitate setting all receivers to a standard output in order to minimise variations in receiver lag. At first this was hand operated, but it has now been incorporated in new equipment installed for the automatic transmission of the morse preamble. For the reception of short wave signals, station A employs a National HRO receiver or a Hallicrafter SX28 with an RME 69 as reserve. Station B has a similar HRO with a Hallicrafter SXl7 as reserve. To facilitate the reception of short wave signals at station A, the Post Office Radio Department has erected two rhombic aerials and four dipole cage aerials with the necessary open wire lines and concentric cables leading to the Time Department buildings.

The free pendulum clocks Shortt 40 and Shortt 49 were erected in the new building at station A in November and Shortt 66 was transferred from the old clock room in December. The quartz controlled frequency clock Q3, previously installed at Greenwich, was overhauled at the National Physical Laboratory and re-erected in the new building in January.

The errors of the clock Shortt 4 at the Royal Observatory Edinburgh, and the clocks Shortt 13, Quartz 2 and Quartz 6 at the National Physical Laboratory, are determined by means of the 10.00 hour Rugby time signals and the results are communicated daily. This continued co-operation by the Astronomer Royal for Scotland and the Director of the National Physical Laboratory is of great assistance and is much appreciated. Commencing in August, similar comparisons with the quartz clocks designed and installed by the Post Office Radio Department have also been made available by the courtesy of the Engineer-in-Chief, Post Office Radio Branch, whose great personal interest in the work of the Time Department has resulted in a degree of co-operation which has played no small part in the considerable advances in accuracy which have been attained during the past year. Data obtained from direct intercomparisons between the Post Office clocks have been supplemented by information on relative rates.

A fundamental change in the operation of the Time Service, which has affected not only the preparation of final corrections but also prediction methods and reception technique, has been brought about by the fuller use of clocks of the quartz-controlled frequency type. Comparisons between clocks of the free pendulum type, of which there are ten available, and quartz clocks, of which there are at least six available, show the undoubted superiority of the latter. Pendulum clocks appear to be subject to a more or less continuous accidental change of rate, and the integration of these changes produces “wandering”. Quartz controlled clocks on the other hand are subject to changes of rate which are definitely recognisable and may in extreme cases be large, but the epochs, and to a large extent the amounts of these, may be detected by intercomparisons between the quartz clocks themselves. In addition they may be subject to a continuous accidental daily change of rate, which would by integration produce wandering; such erratics, if they exist, are certainly much smaller than in the case of pendulum clocks.

In the transmission of time signals it is necessary, both for prediction and for the preparation of final corrections, to determine the errors of a clock by drawing a smooth curve through points on a chart representing its errors as indicated by transit observations. Any attempt to deduce the daily errors of a pendulum clock in this manner is bound to lead to a smoothing of the real wanderings of the clock owing to the scatter in the transit observations due to inherent and inevitable errors of observation. The complete disappearance of clocks of the free pendulum type from the group used in the determination of final corrections has been gradual, Shortts 4, 13 and 16 having each for periods of some months put up a very fine performance, but it at last became apparent that the inclusion of even the best of these was of no value in smoothing out such irregularities as may have been present in the quartz clocks themselves, or may have been introduced by the methods of intercomparison, since the pendulum clock might itself have introduced greater errors.

At present the quartz controlled clocks are mostly situated at institutions other than the two time service stations but, by using the comparisons based on the 10.00 hour Rugby signals, transit observations of the errors of the observatory clocks may be transferred so as to give determinations of quartz clock errors. In view of the undoubted superior performance of these clocks, a fairly drastic smoothing process can then be applied to the individual determinations of clock error, and reliable values for quartz clock errors thus obtained in which the accidental errors of transit observations have been to a large extent eliminated by meaning. Since the Observatory clocks can be compared daily with the quartz primary standards, again through the medium of the Rugby signals, their daily errors can be deduced. The procedure therefore has been to base the time service fundamentally on quartz clocks available at other institutions and to use the Observatory clocks as intermediary substandards for extrapolating the time as given by the quartz standards over 24 hours. For this extrapolation process, free pendulum clocks are singularly ill-suited. It is thus a matter of pressing urgency to replace the existing free pendulum clocks by quartz clocks at the time service stations, so that quartz primary standards can be used to control the service direct. As regards the use of the service for the provision of a fundamental standard of frequency, the present mean error in a 24 hour interval as defined by two successive signals at 10.00 hours U.T. is about ± 4 milliseconds. The use of quartz clocks directly controlling the service should materially reduce this figure. As regards the use of the service for navigational and surveying purposes the increasing utilisation of quartz oscillators will produce a precision which will be a great advance on the past, because these clocks are capable of efficiently maintaining the accuracy of the service during prolonged periods of cloudy weather to an extent that was impossible when reliance had to be placed on free pendulum clocks alone. It must be borne in mind, however, that the inferiority of existing pendulum clocks rests in their tendency to wander and this tendency may be partially due to variable lag in the mechanical devices used to take the time off the free pendulum. The present standard of accuracy calls for the elimination of existing pendulum clocks from the operation of the time service at the earliest possible moment, but only experiment can show whether the best free pendulums would be worthy of retention provided a photo electric method of taking the time directly off the free pendulum were employed.

Further improvements should come from the replacement of the present standard type of transit instrument by one of improved design, possibly of the photographic zenith instrument type. There is evidence that the standard type of transit is subject to certain systematic and quasi-systematic errors and its accidental errors certainly are much greater than those of a photographic instrument. Some preliminary investigations and enquiries have been made.

The results from both stations have been correlated at station A and final corrections to the times of the Rugby time signals, the short wave signals, and to the foreign time signals, have been published as usual. The amount of computing work involved in this and cognate matters has shown a steady increase throughout the year, and the more elaborate methods of extrapolation and prediction which have been introduced now provide full-time occupation for one man at each station; as the stations must be capable of independent operation, much duplication is involved. The direct line link between the stations has increased the quantity of data to be dealt with, and it has been impossible to take full advantage of the facilities offered by the line because of the heavy pressure of urgent routine work. Preparation of the MSS. for “Greenwich Observations” from 1937 onwards is still in arrears.

General development work at station A has made satisfactory progress, especially during the latter half of the period under review. Unfortunately this progress has only been possible by cutting routine maintenance work below the level of safety; dislocations, impossible under normal conditions, have sometimes occurred, necessitating transfer of all services to the other station for a short period. A number of clocks have been cleaned and installed in the new building and the majority of the wiring has been completed. The photo-electric cell equipment for Shortt 40 is being redesigned and a new case for Shortt 3 has been obtained so that this may be similarly equipped. A jack panel to give greater flexibility of operation has been designed, constructed and installed in the new building. New valve relay units employing mains pentodes in an improved circuit have been tested and found free from some of the disadvantages associated with the previous equipment which employed battery triodes. It has also been found possible to design valve relays of a similar type to replace the majority of the few remaining mechanical relays where a double current output is required. In all these units great attention has been paid to the question of lag and variations of lag. A rapid and accurate method has been devised for the determination of the Lag on the line between the two stations and the results obtained have shown an unexpected standard of consistency. The increase in the number of electronic units has led to the design of a standard form of mains-operated unit for stabilised high tension supplies. Six units have been completed and further units are under construction.

A valve unit to lengthen the seconds signals from Q3 has been designed, and a number of other urgent investigations are in hand, but still in an early stage. Considerable attention is also being given to the problem of lag in radio reception. The increasing tendency for the practical work of the department to become electronic rather than electro-mechanical has rendered acute the problem of lack of test equipment. A general purposes beat frequency oscillator and an additional universal meter have been obtained, but the delay in the delivery of other instruments and the difficulties of obtaining components, of which there is a negligible stock in hand, have been serious handicaps. Progress has also been hampered by the cramped surroundings, and experimental work has been carried out in the very limited accommodation available for housing and operating equipment on routine time service, with consequent constant danger to the latter; a small hut for experimental work will shortly be available. The absence of office accommodation is a further serious disadvantage and this matter is receiving attention. The equipment for providing an emergency A. C. supply in the event of mains failure, originally ordered for Greenwich, has been installed at station A but is not yet complete. An A.C. supply is available at station H, secondary to the main supply.

The urgent need for precision quartz clocks should be alleviated to a limited extent by the delivery towards the latter part of 1943 of three clocks of the Post Office group 4 type. These clocks are now under construction by the Post Office Radio Department and it is proposed to equip two of them, one for each time service station, with phonic motors carrying gearing to enable them to be used as rhythmic signal transmitters. The present transmitting clocks, which are of the slave clock type, suffer from many defects such as irregularity of contacts and wandering of phase and their replacement is a matter of urgency. All three oscillators will be of the highest possible stability and will therefore be equally valuable as standard clocks. The assistance of the Post Office in placing their experience and facilities at the disposal of the Observatory in this matter is greatly appreciated. It is apparent however that the provision of two quartz clocks at one station and one at the other is only a step towards the inevitable replacement of the pendulum clocks by groups of quartz standards. A number of groups of oscillators, three to a group, will be required, but only for certain groups will divider and amplifier stages with phonic motors be necessary. For other primary standards, continuous records of relative rates at the fundamental frequencies will suffice, since the integration of frequency comparisons is more accurate than chronographic comparisons of seconds impulses.

One of the weakest links at the moment, apart from the clocks, is in the matter of chronographic comparison. This urgent problem has received considerable attention and arrangements have been made with the Ministry of Supply for the early installation of two decimal counter chronometers which are entirely valve operated, and capable of an accuracy which should meet the needs of the time service for many years to come. It is not anticipated, however, that these devices will be applicable to radio comparisons, and for this purpose the possibility of obtaining two chronographs of high precision has been under consideration. Work is also in progress on the design of cathode ray equipment with photographic recording. It is apparent that in this matter an increase in accuracy will entail a big increase in work, but that is considered to be inevitable.

1944

The Report here presented refers to the period from 1943 May 1 to 1944 April 30 and exhibits the state of the Observatory on the last named day.

I. Buildings, Grounds, Movable Property and Library.

A new workshop was erected at Station A in October to cope with the increased demand for constructional work and apparatus needed by the Time Department.

Enlargement of the accommodation needed for the Time Department at Station A has continued throughout the year. One building equipped for use as a laboratory and for temporary office accommodation has been completed and. put into use, and the construction of a second building to contain clock cellars, for the new quartz clock equipment, and permanent office accommodation, has been approved. It is now in course of erection and completion within the next few months is anticipated.

A bench grinder has been added to the equipment at Station A and a portable high speed drill with flexible drive has been constructed in the workshop.

One of the two-pen chronographs, constructed in the workshop, has been converted to give double speed, making the scale 100 mm. per sec.

Various mains units and chassis with component parts for rack mounting have been constructed in the workshop as equipment for the Time Department.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

A Shortt free-pendulum clock, No.61, a radio receiving set, a chronograph and a Brunsviga calculating machine, from Mr. H. Fry.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

II. Astronomical Observations

Photographic Zenith Tube. The construction of a photographic zenith tube has been sanctioned by the Admiralty and careful consideration has been given to the design of the instrument. The design is a considerably modified version of that of the photographic zenith tube of the U.S. Naval Observatory, Washington. This instrument, designed in the first place for the determination of the variation of latitude, has been adapted at Washington for the determination of time. The essential principle of the instrument is the same as that of the reflex zenith tube, designed by Airy and used at Greenwich for many years; the converging light from a star is reflected at a mercury surface, whose position is such that the light is brought to a focus at the second Gaussian point of the objective. The instrument is then insensitive to errors of level. By placing the flint component of the objective uppermost and by suitable separation of the two components, the second Gaussian point is brought below the objective, where a photographic plate can be placed. The objective and plate can be rotated through 180° and, from observations made symmetrically before and after meridian transit, measurements in the two co-ordinates enable both time and latitude variation to be determined. The primary purpose of the new instrument for the Royal Observatory is to enable time to be determined with a much higher precision than is possible with a transit instrument. The Superintendent of the U.S. Naval Observatory has kindly communicated full details of the adaptations made at Washington to enable the photographic zenith tube to be utilised for time determinations. The Greenwich instrument will have a clear aperture of l0-inches and a focal length of 11 ft. 3 ins. giving a scale of one minute of arc to one millimetre. The general principles of the design have been decided, but various points of detail have to be settled. The telescope will be constructed by Sir Howard Grubb, Parsons and Co.

III. Time Service.

Throughout the year the time service has again been maintained from two stations, designated A and B. Each of the two stations is a self-contained unit, but the work of the two stations is closely co-ordinated. Signals from the clocks at either station can be recorded at the other station by means of the inter-station line, which forms a direct link between the two stations. For this purpose it was necessary to devise a rapid and accurate method of determining the lag of the inter-station line. This line consists of one channel each way in an 18-channel voice frequency carrier system, normally used for teleprinter traffic. At each station a key is fitted, by means of which the send and receive lines may be connected: an incoming signal is then returned to its station of origin. By direct looping of the line at station B and by passing the signal at station A through a relay, so that a signal of normal strength is sent to station B, a signal introduced into the loop will continue to circulate, provided the adjustments on the Post Office carrier system are accurate. By monitoring the signals at both stations over 100 complete loops, the average time for one loop is obtained with high precision. It is then assumed that half the loop lag is the lag involved in sending the signal either way: this may not be strictly accurate, but the main purpose is to control variations in lag. The standard of line adjustment required to enable the signal to continue to circulate is higher than for normal traffic, and the Telegraph Branch of the Post Office has been very helpful in keeping the channel well adjusted.

During the year, three quartz controlled frequency standards have been installed. These are of the Post Office group IV type, in which a GT-cut crystal is maintained in oscillation at 100 kilocycles per second by a bridge drive circuit, which reduces to a minimum the effects of variations in the supply voltages. The crystals are mounted in thermostatically controlled ovens, and the temperature range permitted by the modified Turner Circuit employed ensures that variations in frequency from this cause may be neglected. Regenerative dividers are used to provide an output at 1000 cycles per second to operate phonic motors designed and constructed by Muirhead and Co., Ltd. One standard installed at each station is equipped with motors having a 60:61 gearing, to provide rhythmic contacts for the purposes of the vernier time signal transmissions. These motors are also fitted with phasing controls, by which the rate of the controlling frequency standard may be compensated and the contacts set early to allow for the travel time of the signals over the land lines to the radio transmitting stations. The phasing controls have open scales graduated in tenths of a millisecond. The third standard, installed at station A, has a plain phonic motor, provided with seconds contacts.

The 10h Rugby signals are normally controlled from the station at which the transmitting. clock and the line to the radio station were the more stable; the 18h signals are controlled from the station that acts as reserve for the 10h signals. The first phonic motor transmitter to be received was installed at station B, and was brought into use almost immediately for the control of the rhythmic time signals. The 10h signals have been operated from phonic motors since 26th February 1944 and the 18h signals from 10th March. The introduction of phonic motor transmitting clocks has provided accurately spaced signals. The irregularities in the spacing of the signal dots from the pendulum transmitters previously used was such that it was necessary to measure identical sets of signals in the five minute series at the two stations in order to separate the discrepancies arising from irregular spacing of the signals from the erratics of radio reception and measurement. This expedient permitted more accurate comparisons between the two stations, but did not assist other users of the signals. A more serious source of error in the pendulum transmitting clocks was their tendency to wander by an unpredictable amount after they had been set and checked, thus putting the signals in error by an amount that might be as large as five milliseconds; the high short-period stability of the quartz controlled phonic motors has reduced such errors to negligible amounts.

In order to provide a regular check on line-lag, the station that did not control the actual time signal from the Rugby station sends out a short series of dots immediately after the conclusion of the time-signal. These dots are checked back against the transmitting motor and thus give a measure of the lag of the line.

During the period from 1st April to 14th November, the build-up time of the transmitted signal was longer than normal and both field strength and build-up time were subject to greater variations, with consequent increase in reception errors. From 3rd to 6th July and from 24th September to 14th November, the signals were transmitted on a frequency of 15.46 kc/s., instead of the customary 16.00 kc/s. The rhythmic time-signals at 10h and 18h have been regularly transmitted on a group of short-wave transmitters, in addition to the long-wave transmissions.

The six-dots time-signals radiated by the British Broadcasting Corporation have been normally operated from station A, with station B acting as reserve. In addition to many transmissions on short waves, the number of transmissions on the medium wave programmes has been substantially increased in the current schedules. These signals are widely used and serve many useful purposes at the present time. The six-dots signals are also radiated hourly, with the exception of 10h and 18h, by three Admiralty stations on medium waves, to which are coupled stations on short waves, for the use of coastal shipping. The service of hourly signals to the Post Office, which provide also the control of the Talking Clock, has been maintained from both stations.

The number of transit observations during the year, which have been used for the determination of the finally adopted time, was l84 at station A and 116 at station B. The observations have been combined for the determination of the clock errors at both stations. With better clocks available the scatter of the transit observations has become more apparent and transits have been combined in groups of five for convenience.

The regular check on the 10h and 18h Rugby signals has been maintained at both stations. The short-wave signals have been recorded mainly at station A, where better aerial facilities are available and where reception is more reliable. Both stations have recorded, when possible, the short-wave American signals from NSS, which are controlled by the U.S. Naval Observatory; the German time-signals radiated at 12h from DFY on the long wave-band and from DFC on the short wave-band have also been recorded. Station A has maintained a check on the French signals at 8h. Some receptions have also been obtained of the Russian time-signals from Sverdlovsk and Tashkent, at 17h from RPT and at 16h from RSN. The American broadcasts from WWV, which are in the form of a standard frequency transmission on frequencies of 2.5. 5, 10 and 15 megacycles per second, modulated at 4,000 and 440 cycles per second, with seconds interval signals of five milliseconds duration, have been recorded at station A at intervals throughout the day whenever reception conditions permitted. These signals, controlled by the Bureau of Standards, Washington, are radiated all day. A remarkable degree of agreement has been noted in comparisons of both frequency and time; as a time Signal, this form of transmission appears to be capable of a very high standard of precision and to be well adapted for accurate comparisons.

The determination of final corrections to the received times of British and foreign time-signals has been continued at station A, but the publication of the results has been delayed in order that full advantage might be taken of the improved running of the clocks, and the consequent smoothing of the effects of the scatter of the astronomical observations which is thus made possible. The preparation of these corrections is now complete to the end of 1943.

The time service is still being based fundamentally on the comparisons between the transit observations and the Post Office group IV standards, because the three quartz controlled frequency standards installed at the two time stations cannot be used as primary standards until the ageing of the crystals is complete. Intercomparisons between the Post Office clocks and the Observatory clocks are still possible only by means of the radio-link provided by the Rugby signals at l0h. In order to eliminate the errors of radio reception, a direct line link has been installed between station A and the Radio Branch Laboratories. Unexpected difficulties have been encountered in using this line for clock-signals, but it is hoped that these will be shortly overcome. The figures communicated daily by the Radio Branch have been supplemented by information on relative rates, derived from automatic beat counters in operation between each pair of standards. These relative rates give, by integration, accurate intercomparisons of clock errors. Similar beat counters, of Post Office design and construction, will be incorporated in the Observatory installation. The Post Office standards are mains operated and complete stoppages of all the standards have been frequent. For long period prediction work, uninterrupted runs of many months are essential. The Radio Branch have now transferred their clocks to floating battery operation and it is hoped that stoppages involving all the standards will rarely occur. In order to provide the Post Office Radio Branch with the latest possible information on the performance of their primary standards of frequency, estimates of the rate of the standard IVA are communicated weekly and are based on the current charts of the performance of the various standards and on the most up to date transit observations.

The errors of the clock Shortt 4 at the Royal Observatory, Edinburgh, and of the clocks Shortt 13, Quartz 2, and Quartz 6 – until the latter stopped in July – at the National Physical Laboratory, have been determined by means of the Rugby signals at 10h and the results communicated daily. These results have been of the greatest value during the period that the service has been based on a mean of Shortt clocks. The results for Shortt 4 are now communicated monthly. The results for the clock Quartz 2 at the National Physical Laboratory are still received daily; this clock, though more liable to slight variations of rate than the group IV standards, has been very helpful in the determination of discontinuities, when the Post Office standards have stopped. Its value for this purpose has been enhanced by the considerable improvement effected in the reception and recording arrangements at the Laboratory. It has also been of much value in the determination of final corrections.

Some of the free pendulum clocks have been stopped; others, that have been left running, are not regularly compared. As time signals are required to be transmitted at definite mean times, the fundamental time standards for time service purposes are rated to mean time. Clocks rated to sidereal time are required only for comparison with transit observations, where the use of meantime clocks would be inconvenient. A number of free pendulum clocks and also the quartz clock Q3 at station A are used for this purpose but, by means of comparisons with the mean time standards taken immediately before and after each time observation, the observed clock errors are transferred to the mean time standards. Phonic motors now under construction are being provided with gearing which will operate contacts every sidereal second and every mean time second, when the motor is driven by a mean time standard. An exact gearing is not required; an approximate ratio merely introduces a small uniform rate into the sidereal clock thus made available for transit observing.

Two decimal counter chronographs, of the type referred to in last year’s report, have been received and installed at station A. These counters give an immediate indication, to an accuracy of ten microseconds, of the time interval between two electrical impulses. Subsidiary trigger circuits were necessary before the counters would operate successfully on signals from Synchronome clocks and another type of drive circuit has been made for operating the counters on received wireless time signals. Both counters have been modified to enable them to operate from a 100 kc/s output from one of the quartz standards, instead of from their internal source. They have also been adapted to sum up a series of ten intervals and thus to give a mean measure in units of a microsecond. They have recently been applied to the measurement of received wireless time signals. One unexpected result has been noted; signals, which have given very poor records on the standard form of tape chronograph, have operated the counter satisfactorily.

A new building is in course of construction at station A to house the quartz standards already delivered (which are temporarily housed in an existing building) and others under construction. These standards will be operated in groups, each of three primary oscillators. Each group will be provided with a separate dual-float battery system as a precaution against stoppages affecting more than one group at a time. To guard against long stoppages of the mains supply, a petrol-electric set is available at station A and a diesel-electric set at station B. The emergency A.C. supply system at station A, referred to in the previous report, is still incomplete, but it is expected that it will shortly be finished. It will then serve as a stand-by plant for the operation of the oven circuits, for which alternating current is essential.

It is of interest to review the improvement in the Time Service which has already been effected by the introduction of quartz controlled clocks at the Observatory stations, and to attempt to forecast the prospects. of further improvement. The performance of the service of signals radiated by the Post Office 16 kilocycle transmitter at Rugby can conveniently be separated into long term performance and superimposed fluctuations of short period. With the errors at present inherent in transit observations good long term performance requires that the service should be based on a battery of clocks which are so good that over periods of the order of three months all changes of rate in the clocks can be detected and evaluated by clock inter-comparisons alone, and that transit observations should be employed only for determining errors and rates, but not changes of rate. It is also necessary that the frequency drifts of the crystals should remain sensibly constant over much longer periods. If these requirements are satisfied, then by averaging signal receptions over periods of a fortnight or more any good substandard of frequency could be calibrated with an accuracy of one part in 10⁸ But if this accuracy is to be achieved it is essential that precautions should be taken against stoppages of the crystals operating the primary clocks, as such stoppages are apt to produce changes both of rate and of frequency drift. Up to now all the primary clocks have been situated at other institutions, most of them have been mains driven, and stoppages from mains failures have been frequent, with the result that an accuracy of one part in 10⁸ (corresponding to 0.9 milliseconds per day in mean signal rate) has not always been realised. In recent months the error of an interval of time as determined by averaging signal receptions over a period of a fortnight or more has occasionally been as great as three parts in 10⁸ or even more than this. But battery flotation has now been provided on the clocks used at present as primary standards, and quite apart from the introduction of quartz controlled clocks at the Observatory stations improvement in the long term performance of·the signal service is to be expected.

The recent installation of quartz controlled clocks at the Observatory stations has not yet in itself materially improved long term performance. The crystals driving these clocks have not yet “aged”, and consequently the new clocks cannot yet be treated as primary standards. During the “ageing” stage, which may extend over a period of the order of six months, the frequency of a crystal changes progressively in an unpredictable manner, and so no reliable extrapolation of the clock errors as determined from transit observations can be made. But when “ageing” is complete the crystal should settle to a steady frequency drift, that is, to a condition in which the rate of the clock changes slowly and uniformly; it then becomes possible to extrapolate the clock error in a satisfactory manner, and so to employ the clock as a primary standard. The present indication is that the ageing stage is now approaching completion and with a prospective increase in the number of primary clocks there ought to be a consequential improvement in the long period performance of the Time Signal service during the coming months.

But as regards short term performance the improvement in the service has been immediate. Before the introduction of the new quartz controlled clocks the time signals were actually operated by secondary clocks of the free pendulum type, and on the occasion of each signal the errors of these clocks had to be extrapolated from the results of comparisons between them and the primary clocks. These comparisons had to be obtained via the GBR signals themselves, that is, from measures involving a radio link, and they were affected by errors of radio reception and measurement (which however in the case of the Observatory stations and of the National Physical Laboratory have new been reduced to quantities whose average values irrespective of

sign are of the order of ±0.5 milliseconds). A more serious circumstance was that considerable uncertainty was introduced into the extrapolation of the errors of the secondary clocks by the erratics to which the best pendulum clocks are subject. The result was that if the time signals were used to determine individual 24 hour intervals, errors whose average values irrespective of sign were of the order of ±5 parts in l08 were superimposed on the long term errors of rate of the signals. But the new clocks have materially improved this position. They cannot yet, as has been indicated above, be employed as primaries, and their errors still have to be extrapolated from comparisons with the primary clocks, but the period of extrapolation has been extended from 24 hours to five days, and the recent average error irrespective of sign in a 24 hour interval as defined by successive 10h GBR signals has been reduced to about one millisecond, that is, to ±12 parts in 10⁹. This average error (which is additional to any error of rate in the long term run of the signals) now arises mainly from the combination of variations in transmission time along the Post Office lines, and of variations in lag introduced at the radio transmitting station. As long as these variations remain at their present level it does not seem probable that there will be any further substantial improvement in short term performance, and there appears to be little prospect of the average error of individual 24 hour intervals defined by successive 10h GBR signals being reduced materially below a value of ± one part in 10⁸. This average value means that extreme errors of two or three parts in 10⁸ must be expected occasionally in addition to the error of long term performance.

Returning to long term performance it is intended to install additional quartz controlled clocks at the Observatory stations; it is expected that, when all the crystals have “aged”, the equipment, used in conjunction with existing transit instruments, will ensure that the general run of the signals will specify intervals of time with a maximum error of about one part in 10⁸. But it is probable that improvement beyond this figure will not be possible until the present types of transit instrument are replaced by a photographic zenith tube. The provision of the battery of clocks now contemplated, together with improved ancillary apparatus for the measurement of clock comparisons, will probably bring the Time Service to a point at which no further material improvement can be effected by developments on the purely horological side alone. Up to now the provision of better clocks and of more refined methods of radio reception and recording has improved the service, even although it is ultimately based on transit observations made with instruments of existing standard type. The provision of more accurate horological equipment has in fact made possible a more drastic long term smoothing of the transit observations, which appear to be subject to rather large accidental and quasi-systematic errors. But there is necessarily a practical limit to the improvement that can be effected by purely horological development, and although this limit has not yet been reached, it is being approached. The provision of an accurate standard of frequency is only one aspect of the Time Service which must also aim at the supply of correspondingly accurate absolute time. Both aspects of the work demand. improved star places and substantial development in methods of observing transits of stars. Much thought has been given to the design of the photographic zenith tube, which is expected materially to improve the accuracy of the time determinations. Meanwhile consideration has been given to possible improvements in the accuracy of the time determinations with the small transit instrument. The chief source of error seems to be the determination of the level of the instrument. Laboratory tests have been carried out by an optical autocollimation method, which would enable the error of level to be determined for each star observation, with the telescope in the position in which the observation is made. In the laboratory a precision of about 0”.1 in level error determination is easily attainable; no tests have yet been made in the instrument hut, where large variations of temperature make the conditions less favourable.

1945

The Report here presented refers to the period from 1944 May 1 to 1945 April 30 and exhibits the state of the Observatory on the last named day.

I. Buildings, Grounds, Movable Property and Library.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

A Shortt free-pendulum clock, No.61, a radio receiving set and a chronograph from Mr. H. Fry.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No. 4 by Larcum Kendall, nominally on loan to the National Maritime Museum, were removed to and stored in a place of safety on the outbreak of war, with the exception of No.3, which Lieut.-Comdr. Rupert T. Gould R.N. (Ret.) has dismantled and is engaged, as opportunity permits, in the process of cleaning and re-assembling. Recent examination having revealed traces of rust on teh movement of No.4 and the Larcum Kendall, they have both been sent to the Chronometer Depot, Bradford-on-Avon, for overhauling and cleaning in the watch repair shop.

II. Astronomical Observations

Photographic Zenith Tube. It was reported last year that details of design of a photographic zenith tube, to be constructed by Sir Howard Grubb, Parsons and Co., were under consideration. The instrument will be used for the double purpose of the determination of time and of the variation of latitude. During~ the year the design of the essential parts of the instrument has been under consideration; the main details have been settled and construction is now proceeding. It was felt that the design of the Washington instrument could be considerably improved and many radical changes have been made. The aim has been to ensure accurate geometrical location of the essential parts and to simplify the design as far as possible. Kinematical principles have been followed throughout.

Some of the changes may be mentioned. (1) The conical bearings in which the rotary portion was located has been replaced by a ball-race, in which the functioning surfaces on the rotary and its support are horizontal plane annuli. The friction and consequent strains are thus greatly reduced and, because of the large number of balls employed, weight relief becomes unnecessary. (2) The exact horizontal location of the rotary is irrelevant in principle, but in the Washington instrument it is required for the subsidiary purpose of enabling the orientation to be measured by means of two diametrically opposite microscopes. In the new instrument the orientation will be determined by autocollimation from a plane parallel vertical plate of glass attached to the rotary. It thus becomes possible to avoid constraining the rotary in the horizontal plane during rotation and to allow a slight play in the motion, though this is kept small enough to prevent undesirable bumping. (3) This horizontal play makes it possible to constrain the rotary to adopt a specific orientation by means of two diametrically opposite stops. (4) In the Washington instrument the rotation is performed mechanically by a special arrangement of gears. This arrangement is so designed that the angular velocity is initially small, builds up to a maximum, and dies away as the rotation approaches the required 180°. One undesirable feature is that the force is applied at only one point of the rotary. In the present design, the rotary is actuated by a pure torque applied by means of wires. The use of wires has the further advantage of simplicity over the complex gear system. The principle of a gradual build-up and dying away of the angular velocity is retained and is achieved by making use of the principle of dead-centre. (5) The plate-carriage has been radically re-designed and the conventional form of guiding ways has been discarded. The plate holder is supported from an annulus of greater diameter than the object glass. Symmetrical images, free from troublesome diffraction effects, will thereby be obtained. The annulus is carried on three balls, running between hardened flats and is located horizontally by two further balls at the extreme ends of its diameter. It is thus possible to ensure the desired accuracy of motion with relative freedom from friction and with great simplicity of manufacture. (6) The carriage is to be traversed, not directly by a screw, but by an intermediary wedge or sine-bar. By varying the angle of the sine-bar, the speed of travel can be varied to suit stars of different zenith distances. The functioning surfaces of this linkage can be made with extreme accuracy and by the use of this method it is hoped to reduce the effects of any screw-errors by a large factor. (7) The method previously employed to set the height of the mercury surface to a specific value corresponding to the best focus was to use a pointer on the end of a rod and adjust the mercury surface in height till it touched the pointer. One disadvantage of this system is that the rod has to be swung out of position when using the instrument. It is intended to replace this method by a purely optical method, which has been tried and shown to give more than adequate precision and requires no parts to be moved out of the way. Provision is made on the setting scale for varying the height to correspond with varying temperature so that the scale value of the instrument can be kept very nearly constant. (8) The shutter is made to cover the entire object glass instead of the photographic plate only. (9) Consideration has been given to the question of improving the method of emission of electrical signals from the moving carriage. Hitherto mechanically operated electrical contacts have been used. It is hoped that, by a suitable photoelectric method, the signals can be sent directly from the carriage without any intermediate mechanical linkage and the attendant uncertainties of the nature of back lash. Experiments on these lines are at present in progress in the Observatory.

III. Time Service.

Throughout the year the time service has again been maintained from the two stations at the Abinger Magnetic Observatory and at the Royal Observatory, Edinburgh, which were designated A and B respectively in previous reports. Each station has been available for sending out the Rugby signals, the B.B.C. six-pips time signals, and the hourly signals to the Post Office, for the control of the speaking clocks and other services.

The Rugby time signal at 10h G. M. T. , which is used for the checking of precision frequency standards and for which the highest possible accuracy is therefore required, has normally been sent from Edinburgh, as the line conditions from Edinburgh to the Rugby Radio transmitting station are more stable than those from Abinger to Rugby. The Rugby time signal at 18h G.M.T. was also normally sent from Edinburgh until the 11th February; since that date it has been sent from Abinger. The B.B.C. six-pips time signal has normally been sent from Abinger.

When the time service station was installed at Edinburgh, it was decided to transfer to Edinburgh certain general work of the Time Department. It was found most convenient to transfer computing work, such as predictions and the investigation and examination of results. This division of work was facilitated by the installation of a direct teleprinter line between the two stations. The time service station at Abinger has remained responsible for the main bulk of the Time Department work, including development work, the routine reception of radio time signals and the general computing and preparation of final corrections to the times of emission of signals.

Stellar observations for the determination of clock errors have been obtained at both stations during the year. At Abinger, observations were made with the small reversible transit instrument on 110 nights and at Edinburgh they were made with the reversible transit circle on 125. Observations were made at Abinger on a number of additional nights for the training of new observers, but have not been used because the personal equations of the observers have not been sufficiently well determined. The personal equations of the regular observers have been re-determined; as these are found to be in close agreement with previous determination, no changes in the adopted values have been made.

Progress in the installation of the new quartz clocks has been slow. As mentioned last year, contracts have been placed with the Post Office Radio Branch for the supply and installation of a total of eighteen primary standards, together with certain ancillary equipment. The two clocks at Abinger and the one at Edinburgh, already in use last year, are still the only clocks available. Some progress has been made at Abinger with the installation of further equipment, but no part of the new installation is sufficiently complete to enable any useful results to be obtained. Two groups of three standards each are in temporary operation on the mains, but the divider and beat counter equipment is incomplete, so that no useful data are available for time service operational work. Another group of three oscillators, but without crystals, has been received. Other work in connection with the quartz clock installation at Abinger has been pushed ahead in the hope that the Post Office equipment would have been delivered. The battery installation at Abinger has been completed by the Electrical Engineering Department, H.M. Dockyard, Chatham, and the equipment originally prepared for Edinburgh, and now to be installed at Greenwich, is ready for erection and wiring up. Because of the delay in the supply of instrument racks from the Post Office, the Admiral Superintendent, Chatham was asked to assist, and he has agreed to have twelve racks constructed to Post Office specification in the workshops at the Dockyard. These racks are required for mounting general time service equipment at Abinger and Greenwich.

The three Observatory standards, A1at Edinburgh and B1 and B2 at Abinger, are still mains operated, since it is not practicable to transfer them to the new battery equipment. The crystals from Bl and B2 will later be transferred to one of the new groups now being installed, but this transfer cannot be made until some of the other new standards have been completed and brought into use. The best performance has been given by A1, which stopped, however, in July, September, December and January: it has shown a certain non-uniformity of frequency drift, due probably in part to these stoppages. Mains interruptions have been frequent at Abinger, especially during the period of flying-bomb and rocket attacks. B1 has stopped on 44 occasions and B2, which is apparently more susceptible to momentary flickering of the mains supply voltage, on 60 occasions. On the scanty evidence available, BI appeared to give the better performance, but early in January the frequency drift of this standard became rapidly worse, and the crystal and holder were returned to the Post Office Radio Branch Laboratories for examination. It subsequently transpired that the crystal holder had developed a leak. The trouble was rectified, and the crystal is now undergoing a period of fairly rapid ageing. When time signals are recorded against it, they are transferred on to B2 by means of a comparison taken at the time of reception of the signal.

Throughout the past year, therefore, the three Observatory standards could not be used as primary standards and it has again been necessary to base the time service on quartz standards in establishments outside the observatory. The three primary standards at the Post Office Radio Branch Laboratories at Dollis Hill have continued to give a good performance, especially IVC. The assistance of the Engineer-in-Chief, Radio Branch, in making the results of these standards available to the Time Department, is much appreciated. The daily comparisons at the National Physical Laboratory between the Rugby l0h signals and the quartz clock Q2 have been communicated by the courtesy of the Director of the National Physical Laboratory. The value of these comparisons has been much increased by the considerable improvement which has been effected at the Laboratory in the methods of reception and measurement of the signals. The N.P.L. standard, Q2, appears to be subject to slight changes of rate, which occur at intervals of from a few days to several weeks. For this reason it is less suitable for use as a long-term standard than the Post Office clocks, though it has been of great service in bridging discontinuities caused by the stoppage of other standards.

For the transmission of the l0h rhythmic signals from Rugby. preference has been given to the signals emanating from Edinburgh, because the conditions for transmission of the signals by land-line between Edinburgh and Rugby have been more stable than those between Abinger and Ruby. The Edinburgh signals are sent over a comparatively short line to the Edinburgh Repeater Station, then over a normal traffic V.F. system to Birmingham, and thence to Rugby. From Birmingham a return signal was sent to the Observatory over the return channel of the same V.F. system, and a regular check on the lag of this return signal enabled changes in the transmission time over the send circuit to be detected and allowed for. More recently, a spare V.F. channel became available between Edinburgh and Birmingham; this enabled a channel to be allocated exclusively for time signal purposes instead of switching~ a normal traffic channel. A high standard of maintenance could be ensured and the results have shown the value of this. The signals from Abinger are sent over a long land-line from an electronic relay to the terminal station of the V.F. channels in London. They are then sent to Rugby over a unidirectional carrier circuit. A return signal is received from Rugby over a separate circuit; when a change occurs in the total time of travel of the outgoing and return signals, it is not known how much of the change occurs on the send line. Because of the loss in accuracy occasioned by an unpredictable variation in the time of transit over the land-lines, a number of possible schemes for ensuring greater accuracy have been considered during the past two years. The method eventually adopted was to send out, prior to the actual time-signals, a long tuning note followed by a short series of dots: these signals are radiated by the Rugby transmitter, recorded at the transmitting station and the transmitting clock then adjusted to allow for the measured lag. The special signals from Edinburgh are sent out at 09.30h and radiated from Rugby when (as is usually the case) the transmitter is free. The corresponding signals from Abinger are sent out at 09.45h i.e. some ten minutes before the normal l0h signal. The use of the decimal counter chronometer enables the signals to be measured and the transmitting clock to be adjusted for the derived lag in this interval. The results from Abinger are not altogether satisfactory, as it is found that the transmission over the land-lines from Abinger introduces irregularities in the spacing of the individual signals; when the radiated signals (originating from Abinger or Edinburgh) are compared with the signals direct from the contacts of the Abinger transmitting clock, it is found that the erratics are less when the signals are sent from the phonic motor of Al at Edinburgh than when they are sent from the phonic motor of B2 at Abinger. The irregularities are therefore produced on the land-line between Abinger and Rugby. This matter is being investigated with the co-operation of the Post Office Telegraph Branch.

The present procedure, in which dependence has to be placed on the Post Office clocks, is to use the transit observations to determine the general run of these clocks; the accuracy of the reception of the time signals at Dollis Hill is insufficient to enable day-to-day figures to be used. The prediction line is then transferred to A1, which has been adopted as a secondary standard because it has proved to be the most reliable of the Observatory clocks. A1 is used to control the 10h signals. The predictions on Al are transferred to B2, so that B2 can be set to time to provide an alternative signal from Abinger, in the event of signals from Edinburgh failing for any reason to reach the Rugby station: B2 also controls the 18h signals.

The 10h and 18h GBR time signals from Rugby have been regularly recorded throughout the year at both Abinger and Edinburgh, with the exception that the recording of the 18h signal at Edinburgh was discontinued from 14th February. The short-wave signals, radiated simultaneously with the long-wave GBR signals, have been monitored as previously, mainly at Abinger. Since Abinger is within the skip-distance of most of these transmissions, it was expected that their reception would be uncertain. The final corrections to the times of emission are therefore given only to the nearest hundredth of a second. The uncertainty of the reception has proved, however, to be much smaller than had been expected.

Both stations have maintained a check on the German signals from stations DFY and DFC, except for the period June 22 to October 17, when these signals were not recorded at Abinger. In March, French time signals were resumed, from station TMA2 at 8hand from XWX at 20h: these signals have been received when possible at Abinger, but reception is not good. A special effort has been made to obtain good comparisons with the American short-wave time signals. Records have been obtained at Abinger of the NSS time signals, controlled by the U.S. Naval Observatory on 713 occasions and of the WWV signals from the National Bureau of Standards on 1,200 occasions. The regularity of the signal dots of the WWV transmissions has permitted the use of the selection technique used for the British short-wave signals. The decimal counter chronometer is started by the radio signals and stopped by the seconds contacts of Bl. The consistent readings obtained with the clearest signals are taken as correct. Larger but not consistent readings are assumed to be caused by an atmospheric operating the counter before the actual signal; smaller readings may be obtained with poor signals. The method has proved to give a high standard of accuracy. The NSS signals are not observed in this way, because it is not known whether they are controlled from a phonic motor. The variations between individual readings may be due to irregular spacing of the signals: in that case the best result is given by the mean of a number of individual readings, obviously discordant readings being rejected.

From the beginning of 1944 the U.S. Naval Observatory has published corrections to the times of reception of the British short-wave signals GIH at 18h. The adopted time at Washington can therefore be compared with the adopted time at Greenwich through the intermediary either of GIH signals or of the NSS signals, when the two signals are observed at both stations. 35 days were selected on which both the 16h NSS signal and the 18h GIH signals had been recorded at Washington and at Abinger and for which the corrections to the times of the two signals, as determined at both stations, were available. Applying a travel time of 20 milliseconds on the assumption that the signals travel with the speed of light along a great circle route, the amounts by which the Washington finally adopted time appeared slow on the Greenwich finally adopted time were derived, using the GIH and NSS signals separately. The mean difference between the comparisons by the two separate signals was practically zero (- 0.4 ± 0. 5 milliseconds). The probable error of a single difference (involving errors of reception of two signals at each station) was ± 3.1 milliseconds; if it is assumed that the errors of reception are equal at the two stations, the probable error of a single reception is ± 1.6 milliseconds. Since Abinger is within the skip distance of GIR, this result is encouraging. Furthermore, the small mean difference between the comparisons indicates that the travel lines of the two signals are equal, that the signals can not depart materially from the great circle route, and that the speeds of travel can not differ appreciably from the speed of light. When figures covering a substantially longer period are available, an accurate determination of the speed of travel will be possible.

There is further evidence that the short-wave signals can be received with a higher degree of accuracy than had been expected. The reception at Abinger of WWV time signals under the most varied conditions, by different observers, and by both day and night, show a remarkable degree of consistency in the comparisons between the WWV standards and the Observatory standards. It is not possible to make a strict comparison of Greenwich and Washington time by means of these signals, since Washington only checks the transmissions at 03.00 and 15.00 hours, but it may be noted that the comparisons between the Abinger clocks and the uncorrected WWV time signals shows a higher degree of consistency than those between the Abinger clocks and the NSS signals to which the final Washington corrections have been applied. The WWV type of time signal would appear to be very well suited for precision comparisons, especially where a counter chronometer is available.

Following a special request from Russia, efforts have been made to receive and measure the Russian time signals radiated from RSN and RPT. Reception conditions are not very favourable, but since May a total of 104 signals have been recorded at Abinger.

The service of six-pips signals has been extended during the year. In addition to the usual transmissions by the B.B.C., these signals have been sent over special lines to the Admiralty and to Fighter Command. They were also used for the synchronization of artillery barrages on the Western Front. Because of the various operational uses, it was important to ensure that no erroneous signal would go out. A continuous 24-hour watch each day was therefore maintained for a considerable portion of the year.

An alarm system has now been designed and constructed to guard against any failure of the six-dots transmitting clock. The operation of the alarm performs the dual function of ringing a warning bell, and cutting the signal lines. The alarm is set off if the transmitting clock gets out of control from the master clock, or if it is allowed to run down. It also operates if the magnetic corrector is left on inadvertently, or if the alarm unit itself fails, either owing to a mains interruption, or to a failure of the high tension supplies. The valves of the alarm unit are duplicated, and the correct operation of the alarm unit may be easily tested. A member of the staff sleeps at the station and is available to correct the trouble in the event of the alarm bell ringing during the night.

Various countries sent out time signals from transmitters operating on the long wave band, with low carrier frequencies. These signals have the advantage of reliable world wide reception, with relative freedom from fading and from varying path. The design of a suitable receiver for long-wave reception of time signals has therefore been one of the subjects of experimental investigation at the Observatory. It is essential to ensure that the lag introduced by the receiver shall be as constant as possible, and that the form of the build-up curve of the signal dots shall be kept as steep as possible. A fairly high degree of selectivity is essential, as the transmitting stations are very closely spaced in the frequency band; but high selectivity is associated with large delay in the receiver, and very careful design is called for. Variation in delay can be caused by variations in signal strength, so that it is essential to adopt some means of ensuring that the signal peak is adjusted to the same level for each signal. The actual field strength in the neighbourhood of the receiver cannot be relied upon to remain constant even over the five minutes of a signal period, so that continuous monitoring of the signal peaks is necessary. Modifications to various commercial or standard Admiralty receivers have been made but the results have not been entirely satisfactory. After considerable experimental work, a receiver comprising three sections has been designed and constructed. These are (1) a two-stage amplifier with negative feedback independent of frequency and positive feedback dependent on frequency. The positive feedback increases the selectivity, the negative feedback increases the stability of the circuit. The negative feedback always exceeds the positive, even at resonance, so that the stability is better than that of an amplifier without any feedback. (2) Two stage amplifier with negative feedback with gain control in the feedback circuit. This makes it possible to vary the gain at a point of relatively low impedance where wire-wound and fixed stabilised carbon resistors can be used instead of carbon track variable resistors, which had been found to cause considerable variations in amplification. (3) Cathode follower buffer stage.

Side-band splash was troublesome when the station DFY, whose frequency is close to that of GBR, was operating. It was therefore decided to use a directional aerial. An old frame aerial, erected on the roof, was tried with marked improvement. A balanced and screened frame aerial, which can be rotated from within the building, has been designed and construction is proceeding.

The receiver unit feeds two ancillary units. One is a monitoring unit, consisting of an audio monitor in which a source of 1,000 c.p.s. tone is applied to a built-in loudspeaker whenever a received signal is applied, and a peak voltmeter, with a rapid build up and slow decay time-constant of 30 seconds. This gives readings on a large scale meter corresponding to the output voltage from the receiver; because of the large delay time-constant, the indicating meter gives a sensibly constant reading, although the applied signal consists of dots of only one tenth of a second duration separated by intervals of nearly nine tenths of a second. The meter will, however, follow any changes of peak signal level due to field strength variation, during the five minutes signal period, and the gain may be adjusted to keep the output level constant. The second ancillary unit is a pulse producing unit, which selects a particular point on the build-up curve of the signal dots, and produces a very sharp pulse whenever this point is reached. This operative point may be selected by means of a switch, to be at 10%, 20%, 30%, 40% or 50% of the peak amplitude of the signal dots. The sharp impulses can then be fed to a decimal counter chronometer. For the production of these sharp impulses at pre-determined levels, trigger circuits, consisting of amplifiers with two stable states, were first tried, but the stability of the triggering point was not considered to be sufficiently stable over long periods of time. A circuit employing straight amplification and chopping with a biased diode is now being used, and has been found to be preferable. The complete receiver is being tested on G.B.R. signals. Reception of other signals will be by switched pre-tuned circuits, but although provision has been made for these, the results of the preliminary trials are awaited.

Provision has also been made for the decimal counter comparisons to be supplemented by film records of cathode ray tube signals. The received signal is applied to a rack-mounted double beam oscilloscope, one beam being used to record the time signals, and the pulses from the transmitting clock, and the other beam for timing pulses. A unit to produce the timing pulses has been constructed in the laboratory, and is based on a multivibrator circuit developed at the Radio Branch Laboratories of the Post Office. The output fed to the oscillograph consists of pulses every thousandth of a second, the fifth and tenth pulses being distinguished by larger amplitudes. The film records should be of great value in determining the performance of the receiver and in investigating the changes in the build-up of the signal dots throughout the time signal series. They can also be used during the preliminary test period for comparing the received signal with the time of the impulses passed on the decimal counters at various percentage settings of the pulse producing unit. A further unit is under consideration, which will produce a square-wave modulated test signal to be fed into the receiver in order to check up the build-up characteristics of the receiver. When tests are completed, it is proposed to build further receivers of the same pattern in order that duplicate receivers may be available at the two time service stations.

An improved type of electronic send relay has been designed and five units have been constructed. They give a faster build-up of the line signal and limit the amplitude so as to conform to normal Post Office practice. The temporary trigger units which were made for the decimal counter chronometers have been rebuilt in an improved form. Routine tests have been devised for the counters and for their ancillary equipment in order to ensure constancy in operation. Both counters have required a fair amount of skilled attention and some spare scale-of-ten units have been ordered to facilitate quick repair in case of failure.

The illumination of the field of the small transit instrument has been improved. In the previous method, a diffusing disk was mounted in front of the centre of the object glass at an angle of 45° to its axis, and illuminated by a low-power lamp, mounted outside the telescope tube. The disk, about one inch in diameter, reduced the effective aperture and white fringes, at the edges of the micrometer wires, made bisection of a star image difficult. A small disk, 2mm in diameter, is now placed on the optic axis in a plane which is conjugate to the focal plane with respect to the surface of the object glass acting as a concave mirror. The disk is illuminated from outside the tube and a very uniform illumination over the whole field is obtained. A blue filter has been incorporated in the illuminating system, giving a field illumination approximating to daylight. The intensity of the light is controlled by an iris stop, which enables the intensity to be varied without the change of colour that results when a lamp runs at a reduced voltage.

Two years ago, it was reported that photoelectric equipment was being fitted to a Shortt master clock. This work was shelved for a time for more urgent work, but some progress has been made. After a careful series of experiments to determine the optimum slit widths for the primary and secondary slits, the tolerance of focus, a means of focusing with suitable accuracy, the requisite degree of stability of the light source, and other details, an experimental optical system was constructed and has been tried out. One of the most serious difficulties has been the insufficiently rigid mounting of the clock, and it is hoped to try the system out with a clock mounted on a more stable support. It also became quite apparent that further refinement in the method of taking the time from the pendulum was of little value unless some improved method of impulsing the pendulum could be developed. Preliminary experiments have been carried out on a method of electrostatic impulsing, but little can be said except that it is possible to maintain the arc by this means. Further work on this subject is required, but for the present it is not of vital importance and is only receiving attention when more important work permits. The electronic unit for the impulsing has been completed, and also a pulse unit for operating from the photoelectric cell timing device.

X. General.

At the beginning of the war, steps were taken to install a skeleton time service at the Abinger Magnetic Station which would provide a service adequate for normal navigational and domestic needs in the event of serious dislocation at Greenwich. A hut was built for a small transit instrument. Two additional free pendulum clocks, Shortt 66 and 67, the latter being equipped for sending rhythmic time signals, were obtained and installed, together with the clock Shortt 61, which was loaned by Mr. H. R. Fry. In the autumn of 1940, the continuous air raids on London made night observations at Greenwich impossible. The entire time service was accordingly transferred to Abinger. As all the time lines from Abinger passed through London and Abinger itself was in a part of the country suffering from air attack, a second time service station became essential. With the co-operation of the Astronomer Royal for Scotland this was set up at the Royal Observatory, Edinburgh, where the transit circle was available for time determinations. Two free pendulum clocks, together with other equipment, were transferred from Greenwich to Edinburgh, and staff from Greenwich was sent to assist in the installation. This was just completed in time to take over control of all signals when the service from Abinger was temporarily put out of action as the result of the destruction of the Central Telegraph Office and the complete severance of all the time signal lines from Abinger. In the intervening period the staff of the Edinburgh Observatory has been engaged almost exclusively on time service work, assisted by one member of the Greenwich staff. The assistance and co-operation of the Astronomer Royal for Scotland and of his staff have proved of the greatest value and have materially helped in enabling a considerable improvement in the accuracy of the time service to be achieved under difficult wartime conditions. With the end of the war in Europe, the time service equipment at Edinburgh will be brought back to Greenwich, as soon as some necessary structural alterations required for its accommodation can be completed.

Abinger will remain for the present the main time service station of the Royal Observatory. There is no space on the restricted site at Greenwich for the time service installation at Abinger. When a new site has been obtained for the Royal Observatory, the portion of the time service at Abinger will be transferred to the new site. Observations both at the new site and at Greenwich will then require to be continued for a sufficient period to enable the longitude of the new site to be determined with accuracy.

1946

The Report here presented refers to the period from 1945 May 1 to 1946 April 30 and exhibits the state of the Observatory on the last named day.

1. Buildings, Grounds, Moveable Property and Library.

The Shepherd 24-hour clock and the public barometer, near the Main Entrance, were considerably damaged when the. Entrance was destroyed. As both of these were always of great interest to the public, repairs have been undertaken. The public barometer has been reinstated; the repairs to the 24-hour clock have been completed but the clock is awaiting a new dial.

A Nissen hut to serve as a temporary workshop for the Time Department, a small brick building to house a Diesel generating set and a temporary garage for Service vehicles have been erected at the Abinger Magnetic Station.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum: – Harrison's Time Machines Nos. 1, 2 and 3.

Harrison's Time Machines No. 4 and the copy of No. 4 by Larcum Kendall are at present udergoing repair and cleaning by the Chronometer Section at Bradford-on-Avon.

II. Astronomical Observations

Photographic Zenith Tube. Construction of this instrument by Sir Howard Grubb, Parsons and Co. is proceeding. In the report for last year, the principal features of the design were enumerated. Certain modifications to the details of that design have since been introduced. (a) The use of balls to constrain the carriage in a horizontal plane has been discarded; a special spring-strip constraint has been designed by means of which it is hoped to obtain greater perfection in rectilinearity of the motion. The system is compensated with respect to uniform temperature variation; to ensure sufficient uniformity of temperature during the recordings made on any one star essential. components are to be immersed in oil. (b) A differential-roller mechanism has been designed for driving the carriage. If tests on this satisfy the requirements, it will replace both the screw and the intermediary wedge. It appeared doubtful whether these would give the required precision of movement. Moreover the use of the wedge presented difficulties in attaining the extended range of motion that is requisite in connection with the adjustment of the instrument in azimuth. A further advantage of the new system is the considerably greater simplicity of construction. The variation of carriage-speed required to cover the range of declination employed will be made electrically instead of mechanically, by varying the frequency of the input to the phonic motor. Methods for adjusting the frequency have been investigated. A slave-model incorporating the essential features of the new systems of constraint and drive has been constructed and will shortly be tested. (c) Work on the photoelectric method previously proposed for the emission of electrical signals from the moving carriage has been brought to a successful conclusion and an account of it will shortly be published. But though the method has ample sensitivity and is satisfactory in operation in a fixed location in the laboratory, its application to the P.Z.T. involving, as it does, the mounting of the lamp, the optical system, the two photocells and the preamplifier, on the rotary, is not only somewhat cumbersome but introduces requirements of rigidity in the mounting of the photocells and other components, which it would require considerable precautions to satisfy. Moreover some doubt was felt about the effect of atmospheric heating from the lamp and valves on account of their proximity to the aperture of the instrument. The necessity of attempting to overcome any of these difficulties has now been removed by the introduction of a new photographic method of timing the carriage. This method, which was originally devised in connection with the Horizontal Transit Instrument, has the further considerable advantage of eliminating the use of a precise chronograph such as would be required in conjunction with the photoelectric (or mechanical contacts) method, since the recording system itself acts as a chronograph., Preliminary experimental tests indicate that a precision in timing of the carriage of better than one millisecond can be attained without difficulty. It is hoped shortly to publish a more detailed description of the method.

In addition to these modifications two further points may be mentioned. (a} An optical method of levelling the photographic plate in relation to the rotary has recently been proposed. (b) The arrangements of the phonic-motor-drive have been modified, with resulting improvement of the details of the parts to be mounted on the rotary, and considerable simplification of the gearing.

The construction of the special travelling microscope for measuring the photographic records has now been completed by the Cambridge Instrument Company, and the making of certain final adjustments is in progress.

III.Time Service.

The time service station at the Royal Observatory, Edinburgh, was maintained in operation, on a gradually reducing scale, until the end of January, when the equipment was dismantled and returned to Greenwich and Abinger. The Shortt clocks 3, 11, and 16, which were in use at Edinburgh, have been cleaned and overhauled, and will be re-mounted at Greenwich.

A secondary time service station is being installed at Greenwich. This station will provide a reserve time service, while the main station remains at Abinger. During the period of transfer of the time service equipment from Abinger to the new site for the Royal Observatory, it will take over the full duties of time service operation. On the completion of this transfer, time determinations will be continued at Greenwich for the period required for an accurate determination of the longitude of the new site [Herstmonceux].

The Shortt clocks 66 and 67, previously mounted at Abinger, have been thoroughly overhauled and have been remounted. Shortt 61 has been returned to Mr. H.R. Fry, who had generously loaned it to the Royal Observatory for the duration of the war. One of the Dent Regulators from Abinger has been fitted with new contacts, and installed at Greenwich, under the control of Shortt 67, for the emission of the 6-dots time signals for the British Broadcasting Corporation, and of the hourly signals to the Post Office for the control of the Speaking Clock and of other services. These signals are available from Greenwich as a reserve in the event of any failure of the signals from Abinger. The covering of the Altazimuth Dome; which was destroyed when the building was hit by a bomb, has been renewed, in preparation for installing a small transit instrument for time determinations.

Considerable rearrangement of the equipment at Abinger has been made, coincident with the installation of new equipment. The need to ensure uninterrupted operation of the time service has necessitated the rearrangement being made in several stages. All time service equipment has been removed from the two rooms previously used by the Magnetic Department as an engine room and battery room, which were adapted at the beginning of the war to provide a skeleton time service, in the event of interruption of the service from Greenwich. These rooms are being converted for use as dark rooms and for housing the continuous film developer and other equipment, needed for the increasing use of photographic and film technique in. the time service. The Dent regulator used for the B.B.C. and Post Office signals has been removed to the main clock building, and placed under the control of Shortt 49. The various clock circuits, which previously terminated in the engine room, have been transferred. All operational work is at present being confined, as far as possible, to the main clock building, while the newer equipment is being installed in parallel in the new building. When the new installation has been completed, the operational equipment will be removed to the new building to make room for various experimental units and quartz clocks undergoing tests.

The new installation has been planned as a series of rackmounted units. Care has been taken in the design to make operation, adjustment, and servicing as convenient as possible. The radio reception and comparison equipment will be mounted on eight standard 6ft. 6in. racks, and the send relays and ancillary equipment on another four. Five racks are now in use for the quartz clock divider stages and beat counters. Progress in installation work has been slow, because of shortage of staff and of the constant demands for maintenance and servicing of the operational equipment.

The installation of twelve quartz clocks at Abinger is approaching completion. The C and D groups, each comprising three oscillators, have been running for some time on the battery installation. The E group is of a slightly different type, employing crystals with soldered connections; this group has no provision for rate adjustment, which may eliminate one possible source of variation. The original mains-operated B1 and B2 crystals have been disconnected and a new battery operated B group has been formed, incorporating a new crystal as B1, the old B2 crystal, and the A1 crystal (formerly at Edinburgh) as B3. None of the standards has been in operation for a sufficiently long period to permit a proper determination of the ageing characteristics. The performance of C3, D3, El and E2, which have had the best runs, is determined in terms of the Post Office standards at Dollis Hill, and they are thus available for short-term prediction.

The performance of the Dollis Hill standards continues to be good, but IV C, which had run steadily with a uniform frequency drift for some years, suddenly changed its rate of drift after a stoppage in the autumn. The change was partially marked by a number of changes in other standards, and it is strongly suspected that at about this time there was also a change in the rate of rotation of the earth. The scatter of time determinations with small transit instruments is such that changes of the rate of frequency drift can only be determined over comparatively long periods, and then providing only that the change is from one steady value to another value, which is adhered to for many months. The much higher precision in time determination, which will be attained when the photographic zenith tube has been brought into operation, will make possible a considerably closer control over the performance of the standards.

The ancillary equipment for the quartz clock installation at Abinger is not fully completed. Five divider chains are in use, and the full complement of beat counters is connected. Various modifications to and readjustments of the different units have been made from time to time by the engineers of the Post Office Radio Branch Laboratories. The multivibrators, which provide an output of ten pulses per second from the 1,000 cycles per second output, remain to be aligned and adjusted.

Five phonic motors, and one phonic clock, are at present in operation. The phonic clock, which has contacts every second and every tenth of a second and is fitted with a dial and hands was originally operated by the 500 cycles per sidereal second output from the first quartz clock installed at the Observatory, which incorporated a Dye-Essen ring crystal. It has now been converted to operate on a 1,000 cycles per mean time second supply, which is the frequency of the output used for driving all the phonic motors at the Observatory. Two of the phonic motors are of a simple type with seconds contacts only. Two others are provided with a gearing of 365/366 to give contacts at every mean time second and at every sidereal second: the gearing ratio is not sufficiently exact to give correct sidereal seconds, but the inexactness in the ratio merely involves a small additional constant term in the rate of the sidereal time derived from the motor. One further motor of this type has been completed, but is at present being used by the makers in an investigation on lubrication. The fifth phonic motor is provided with a 60/61 gearing for the emission of rhythmic signals; a second motor of this type, formerly in use at Edinburgh, has been sent to the makers for minor alterations. The rhythmic signal phonic motors are phaseable, so that, by appropriate adjustment of phase, the rhythmic time-signals can be sent out exactly at the desired time. At present the D3 crystal output drives both the phonic clock and one of the phaseable phonic motors; using decimal counter chronometer technique, it is convenient to measure a rhythmic signal against the diminished seconds contact of the rhythmic phonic motor, and signals of American type (which consist of dots spaced at intervals of a mean-time second) against seconds contacts, allowance being made for the correction applied to the phaseable motor.

Although all rhythmic signals from the Royal Observatory are now controlled from phonic motors, the B.B.C. six-dots signals, the Post Office hourly signals, and various switching functions are still dependent on contacts on Dent regulator or Synchronome slave clocks. Some alterations are being made in the method of sending the six-dots signals. As originally designed, the Observatory clock merely closed a circuit at each of the six seconds of the signal, the battery and relay both being installed at Broadcasting House. Thus the monitoring meter at the Observatory showed no current flowing, except for the duration of the actual signal dots. It was not possible to detect any line faults until a signal had been missed. This circuit is now being brought into line with the other time-signal circuits operated from the Observatory, and doublecurrent electronic send-relays are being installed at the two time service stations. These relays send a current of a few milliamperes along the line continuously, the current direction being reversed for the signal dots. A constant check on the line conditions is possible. The circuits from Abinger and Greenwich will both be connected through to Broadcasting House and the B.B.C. engineers will be able to tell at a glance whether they are in correct working order.

A new phonic motor transmitting unit has been designed in co-operation with Messrs. Muirhead and Co. Ltd., which will control all signals and switching. Four of these units have been ordered, two for each station. These units will provide a safeguard against the transmission of signals at an incorrect time, which is not impossible with the present Dent regulator control. The phonic motor will either remain in synchronism with the frequency of its drive, in which case signals will be sent out at the correct times, or it will drop out of synchronism, in which case it will stop and outgoing signals will be automatically suppressed. The two motors at each station will each feed double-current, electronic send-relays, whose output will be taken by a pair of wires to the B.B.C. control room, thus providing a further safeguard against the possibility of incorrect time-signals being broadcast.

Astronomical observations for the determination of clock error were made on 78 nights at the Edinburgh station before it was closed down, and on 108 nights at Abinger. The observations at Abinger were formerly recorded against the sidereal clock Shortt 40, which was compared before and after each night’s observation with one of the mean time quartz standards. The introduction of phonic motors with gearing to give sidereal time has simplified the procedure, as no comparisons are now required. Because of the scatter of the transit observations, tape chronographs are still employed for recording transit observations, although they have been entirely superseded by the decimal counter chronometers for all other comparison purposes.

The conditions for transmission of signals by land line between Edinburgh and Rugby were consistently more stable than those between Abinger and Rugby. With the impending closure of the Edinburgh station, efforts were made, in co-operation with the Post Office Engineering Department, to locate and correct the various sources of variation in the lag of the line from Abinger to Rugby. A number of modifications were carried out, various items of Post Office equipment were disconnected from the line, and the service was rerouted on a different circuit to London. Since the introduction of the pre-signal test at 09.45h, as mentioned in last year’s report, it had proved possible to diminish the day-to-day erratics in the 10.00h GBR transmissions due to changes in the line lag, and, by arrangement with the Post Office, a similar facility was made available at 17.45h to permit readjustment before the 18.00h signal. The effects of short-period changes of lag remain. A number of special tests were radiated from the Rugby transmitter with the object of locating the cause of the variations, and for a short while the signals were sent from London to Rugby over a physical line, instead of over the usual voice frequency carrier circuit. The results were promising, and further tests are planned. It is hoped that the full co-operation of the Post Office Engineer-in-Chief’s Department will clear up these problems.

Intercomparison between the two time service stations is possible either by line link or via the Rugby signals. A direct radio link would provide a more reliable connexion. Some experiments were made, by the co-operation of the Director of Radio Equipment, Admiralty, with the aid of the Admiralty Signals Establishment, using Navy mobile radio equipment. Two-way communication was established on a frequency in the 85–95 Mcs. band, and tests were made to ascertain if communication could be established on a near-optical path at a frequency of the order 300 Mcs. Test transmissions were also made with a standard service-type transmitter, working on a frequency of approximately 2 Mcs. Useful information was obtained.

As a result of these tests application was made to the Wireless Telegraphy Board for permission to operate a test transmitter from Abinger. A frequency of 2 Mes. was allocated, and regular transmissions have been made since October for half an hour per day following the 10.00h time signals. Two multiplier stages, with factors of 4 and 5 respectively, were designed and constructed in the electronics laboratory, and are controlled from one of the 100 kcs. crystal oscillators; a power output stage feeds the temporary “Adastra” aerial used for the intercommunication tests. No attempt has been made so far to provide any means of adjusting the frequency of the transmissions to the nominal value of 2 Mcs., but the actual frequency can be estimated with a normal accuracy of one or two parts in one hundred million. For half the transmission period, a 1,000 cycles per second modulation, of the same accuracy and stability, is applied, and this is interrupted at seconds intervals, and is also used to give the morse call-sign, G.M.T. The transmissions are received regularly at Dollis Hill and the radiated frequency is compared with the Post Office Radio Branch Laboratory frequency standards. A very high standard of accuracy of intercomparison has been attained. The results are communicated daily to the Observatory, and have proved to be of the greatest value in providing an alternative link between the two sets of standards. In addition, the transmissions have been received regularly by a number of other establishments, including the B.B.C. receiving station at Tatsfield, and the Aeronautical Inspection Directorate Test House at Harefield, both of whom are regularly informed of the departure of the radiated frequency from the nominal value. Although on a small scale, and of insufficient power to reach a wide range of establishments, these tests have thus developed into a service which is very greatly appreciated, and which has proved to be of material assistance. Further tests are contemplated, and these transmissions may do much to meet the need for such a service until a more comprehensive series of transmissions can be planned and inaugurated. The Director of Radio Equipment, Admiralty, has co-operated with the Observatory in this and many other radio matters, and it has now been approved by the Board of Admiralty that D.R.E. and other technical departments of the Admiralty should co-operate to the full with the Royal Observatory in the carrying out of its work. This formal sanction is greatly appreciated, and will ensure that the Observatory can call upon the departments concerned for expert advice and assistance, an arrangement which should materially assist in development work being accelerated, and the maximum efficiency being assured.

Apart from these special transmissions, the only means hitherto available for checking the frequency of frequency standards has been by the reception of the 10h Rugby 16 kcs. signals, special care being taken to transmit the signals so that the interval between corresponding signals on consecutive days is 24 hours to a high order of accuracy. These signals, with their slow build-up time and low frequency, are not well adapted for this purpose. Frequency transmission of the type sent out from the WWV station of the Bureau of Standards, on frequencies of 2.5, 5, 10 and 15 Mcs., with superimposed time signals in the form of short pulses at a repetition rate of one second, provide a control over both frequency and time. Since, in order to meet the needs for increased accuracy in the time service, the Observatory had found it necessary to plan for a comprehensive installation of quartz controlled frequency standards, much of the equipment, required for the control of a service of standard frequency transmissions, is available. The observational and prediction work required for the control of the time signals will serve for the control of frequency transmissions also. The question of the inauguration of standard frequency transmissions from Great Britain has been under discussion interdepartmentally and it has been agreed that, if such transmissions are undertaken, the responsibility for the control both of the frequency and of the superimposed time signals shall rest with the Royal Observatory.

The petrol driven generator set installed to provide an alternative supply of alternating current in the event of a failure of the supply mains had long become inadequate to carry the full essential load, and it had also proved unsatisfactory both in starting and running. A new diesel generator set, of larger capacity, and fitted with self-starting equipment, has now been installed, but is not yet connected.

Much work has been done during the past year in the electronics laboratory, though progress has been limited by the lack of adequate staff. In addition to the work on the 2 Mcs. transmitter, referred to above, many units have been designed and constructed for incorporation in the rack mounted installation now in process of assembly in the new clock building. The long wave receiver, described in the previous report, has been in regular use. Two further receivers have been constructed, and these have embodied slight modifications which have appeared to be desirable as a result of experience with the first model. The ancillary equipment for photographic recording of cathode ray oscilloscope traces of the received signal is now being entirely redesigned, since difficulties had been experienced in operating the recording camera at an adequate speed for the original method proposed. The system now under construction in an experimental form is designed to make use of a time-base circuit applied to the tube, thus overcoming the need for high film speeds. The screened frame aerial, together with its control mechanism, constructed in the Observatory Workshop, has been mounted and brought into use.

X. General.

Mr. Smith, Assistant in Charge of the Time Department, visited Germany in October, with a representative from the Post Office Radio Branch Laboratories, to investigate the quartz clocks developed by the Physikalische Technische Reichanstalt and by the firm of Rohde and Schwartz at Munich. In order that a thorough investigation could be made of its performance, one of the clocks from the P.T.R. was brought to this country. Because of the present shortage of staff and of accommodation at Abinger, arrangements have been made for the crystal and drive circuits to be tested at the Post Office Radio Laboratories; after these tests have been completed, the clock will be returned to the Observatory for investigation of its long-term performance and for comparison with the Observatory standards. Mr. Smith also visited a number of German astronomical institutions to obtain reports of their wartime activities.

The administrative and other staff of the Royal Observatory, who were moved from Greenwich to Abinger Hammer in the autumn of 1940, returned to Greenwich in July.

The premises at Abinger Hammer, which had been occupied by this staff, have been retained for use as an Admiralty Civilian Hostel, because of the serious difficulty of finding living accommodation in the Abinger area for staff employed in the Time Department

The time service station at the Royal Observatory, Edinburgh, was closed at the end of January after having been in operation for a period of five years. The co-operation and assistance of the Astronomer Royal for Scotland and of the staff of the Royal Observatory, Edinburgh, greatly helped in meeting the requirements for a considerable improvement in the accuracy of the time service under difficult wartime conditions, and ensured the maintenance of the service during the period when south-east England was subject to heavy attacks from the air. The provision of an exclusive line between the Royal Observatory, Edinburgh, and the Post Office Radio Station at Rugby, made possible a very high standard of smoothness in the radiated GBR time signals controlled from Edinburgh. In addition, the transit observations obtained at Edinburgh were used in conjunction-with those obtained at Abinger for the determination of clock errors; these additional time determinations were of great value, because the scatter of the astronomical observations made with the conventional type of transit instrument is the most serious limitation on the accuracy now obtainable in the time service.

1947

The Report here presented refers to the period from 1946 May 1 to 1947 April 30 and exhibits the state of the Observatory on the last named day.

I. Buildings, Grounds, Moveable Property and Library.

A new dial for the Shepherd 24-hour clock near the main entrance has been supplied by Messrs. James Cooke and Son, of Stechford, Birmingham. An exact copy has been made of the original dial, which was damaged when the entrance was destroyed during the war; the makers of the dial have generously presented it to the Royal Observatory without any charge.

At the Abinger Magnetic Station the entrance roadway has been resurfaced with concrete, the gateway has been widened, and the drive has been resurfaced with tarmac. The entrance gates and adjacent fencing, which were damaged by the fall, during a heavy gale, of two large beech trees near the entrance, have been replaced.

The rooms at the Magnetic Station, originally used to house batteries and engines and subsequently for time service equipment, have been converted for use as dark rooms. The fittings, including wet and dry benches, were designed by the Admiralty Photographic and Instrument Research Laboratory and were made by the Construction Department, H.M. Dockyard, Chatham.

For the better reception of foreign time signals, 98-ft. Adastra masts have been erected on Wotton Common. One of the masts, which received damage during a heavy gale, has been repaired.

The roofs of the two small transit huts at Abinger were damaged by the winter gales. The roof of one hut received further damage after repairs had been completed and is now awaiting repair.

Modifications are in progress to the electrical supply installation at Abinger. A new cable has been laid across the courtyard between the diesel-generator house and the main switchboard. The supply cable is overloaded and the installation of a new cable is under consideration.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the National Maritime Museum – Harrison's Time Machines Nos. 2 and 4, and the copy by Larcum Kendall.

Harrison's Time Machines Nos.1 and 3 are at present undergoing or awaiting repair at the Chronometer Depot at Bradford-on-Avon.

The Bamberg Broken Transit Instrument, on loan from the Royal Observatory, Edinburgh, has been completely overhauled, preparatory to being brought into regular use for time determinations. In the past difficulty has been experienced from wear of the pivots of the small transit instruments; the effects of pivot errors on the observed transit times cannot be allowed for and the determinations of level error with a striding level are impaired. The pivots have therefore been heavily plated with chromium and then ground and lapped in the workshop to a high degree of precision. An impersonal micrometer, formerly belonging to the Cooke transit, damaged by enemy action, has been fitted, while a second altitude circle and reading microscope have been added. The system of field illumination, designed by Dr. Atkinson, has also been fitted.

II. Astronomical Observations

Photographic Zenith Tube. - The construction of this instrument by Sir Howard Grubb Parsons & Co. is still in progress. Tests on the mechanisms for constraining and driving the carriage have been carried out at the works of Messrs. Grubb Parsons in conjunction with the Observatory. Preliminary tests were made in July 1946 on a small scale model; the results were encouraging but could only be regarded as tentative firstly on account of the difference in scale and secondly because the tests were at that time inevitably made under static conditions which are unrepresentative both dynamically and in duration. These tests were however useful as an indication of the requirements to be met in final tests and in revealing certain features of the system. For example, the spring constraint differs from the ball constraint in being susceptible of adjustment, and the method, since adopted, of adjusting for minimum rotation of the carriage was suggested by the results of the preliminary tests. The carrying out of final tests has had to await the completion of certain components of the instrument itself and is therefore still incomplete. An important modification of the photographic timing-system has been made. In the original design the optic axis was horizontal so that the coordinates on the timing plate were in the right ascension and vertical directions. Knowledge of the vertical coordinate is, however, required only to comparatively low accuracy whereas the declination coordinate is critical. The system has therefore been redesigned so that, the optic axis being vertical, the coordinates on the plate are in right ascension and declination; this change results also in simplification of mechanical design of the constraint and motion of the carriage of the timing-plate. Unfortunately the erratics of motion of the stellar-plate carriage are liable to be greater in the vertical than in the horizontal direction owing to asphericity of the balls on which this carriage rolls. The effect of such erratics is to produce variations of scale value on the timing-plate which introduce irregularities in the time scale (though not, if the centring is correct, in the declination component). This cause of error may however be eliminated by the introduction of a small auxiliary carriage which slides, instead of rolling, on an optically flat surface, and is connected to the main carriage by a spring linkage which provides flexibility in the vertical direction. In the completed design two such timing systems are to be provided on an east-west diameter on either side of the annular stellar carriage. By this means additional accuracy in the time scale is to be expected, while combination of the measurements in declination of the records obtained by the two systems secures a complete check of the motion of the stellar carriage under the actual conditions of operation. This provides a solution of the problem of replacing the earlier visual and long drawn-out static tests by a dynamic test of adequate precision and relevance suitable for use during the initial investigations of performance; it can also be applied in the ultimate use of the instrument, whenever it is considered desirable, to derive, from the records made primarily for time determination, the latent additional information they contain concerning the erratics of motion in declination.

Tests have also been made on the stability of the mounting of the stellar plate holder. In the main these have given satisfactory results, but have indicated certain weaknesses for the elimination of which further attention is required.

Further consideration has been given to the details of the mechanism for reversing the rotary. One of the minor improvements contemplated is the introduction of air-buffers to avoid any trace of bumping against the fixed stops.

In addition to the experimental tests carried out at Messrs. Grubb Parsons, auxiliary experimental work has been done in the Optics Laboratory at the Observatory. For example, preliminary tests were made of the optical system of the photographic timing method with respect to distortion of the projector and selection of the most suitable type of photographic plate. We are indebted to Mr. Wilfred Taylor of Messrs. Cooke Troughton & Simms for assistance in connection with the selection of the most suitable projector and for carrying out independent tests for this purpose. The construction of the auxiliary unit which is required to actuate the Arditron lamp has been completed. This was designed so that. a frequency of 10 flashes per second can be employed when this is required in work on the Horizontal Transit Instrument. It is also suitable for use in experimental work on the P.Z.T., but for exclusive use on the P.Z.T. where the higher frequency is unnecessary a modified form of unit would suffice. Experimental work is also in progress at the Observatory relating to the design of the system of supporting and adjusting the mercury basin, and to the design of basin best suited to give the immunity from tremors and ease of cleaning.

In the Report for 1945 mention was made of the introduction of an optical method of setting the height of the mercury surface. By this method the surface is set to a specific height with respect to that region of the telescope tube which is at the same level. Constancy of scale value is thus preserved only provided the tube length is invariable. It was intended therefore to measure the temperature of the tube and to vary the position of the fiducial point at which the optical settings were made so that in spite of expansion of the tube the optical criterion corresponded to a fixed scale value. The method has recently been modified to avoid the necessity of making the temperature measurements and resetting the fiducial point. An invar pendulum is suspended from the top of the tube and at the side so that its lower end is a little above, and to the side of the mercury basin. A plane mirror which has its reflecting side downwards is attached to the bottom end of the pendulum, and forms part of the optical system. This system is such that settings made on a fixed fiducial point correspond to constant difference in height between the mercury surface and the plane mirror and therefore approximately to constant scale-value.

Some preliminary investigation has been made of the departure from true level, and of the stability of level, of a shallow pool of mercury. The question is of importance in the use of instruments other than the P.Z.T. There appears to be but little direct experimental evidence available, and none at all to the order of accuracy required: namely· 0”.01. The method which seems to give the best promise of securing adequate sensitivity is electrical rather than optical.

A brief description of the optical method of levelling the stellar plate has been prepared for publication, since this was considered to have possible application for other purposes.

The special Measuring Machine which was designed for measuring the stellar plates of the P.Z.T. has been completed by the Cambridge Instrument. Company and received by the Observatory, where preliminary tests have been carried out. In connection with the measurement of such plates it is hoped shortly to carry out some experiments in the Optics Laboratory to determine whether greater precision and ease of setting than is possible visually can be obtained by the use of a photometric device recently developed by Messrs. Ferranti Ltd. and exhibited by them at the Physical Society's Exhibition in April 1947.

III. Time Service.

The installation of the twelve quartz clocks and their associated equipment was completed in October by the engineers of the Post Office Radio Branch. From February 6th, when the rhythmic signal transmitting motor was transferred to the new control room, the whole of the quartz clock equipment has been operated from the battery supply. This has considerably reduced the number of failures due to interruption in the electricity mains supply. The twelve clocks are divided into four groups, classified under the letters B to E; the clocks in each group being numbered from one to three. Two remaining groups, A and F, with phonic motors on loan from the Observatory, are at present running on test at the Post Office Radio Branch Laboratories, prior to their installation at Greenwich.

It has been the general experience with standards of this type that during the initial period of operation the crystal is subject to a rapid and unpredictable frequency drift; the performance of the Abinger clocks during the year has shown some improvement in the stability of the rates. The clocks that have performed most satisfactorily are B1, B2, C3, D3 and E1 and for this reason these have been used to operate the phonic motors, and have been employed for prediction purposes in connection with the transmission of the time signals. B2 has normally controlled the motor used for the emission of the rhythmic time signals, and has provided the source of standard frequency for the experimental 2 Mc/s transmissions. During the period April 16 to July 4, when B2 was subject to some variations of rate, these functions were performed by the clock D3.

The performance of only five clocks can be directly determined, as there are only five sets of frequency dividers, each of which operates a phonic motor. The performance of the remaining seven crystals can be judged only from figures given by the beat counters. These counters have not proved satisfactory, especially when pairs of clocks with large relative rates are involved. Similar difficulties have been experienced elsewhere, and this matter has received attention at the Post Office Radio Branch Laboratories. A new form of beat indicator has been designed, and will be fitted in the Greenwich installation. The present beat counters at Abinger will be replaced by the improved pattern as soon as supplies become available. As an interim measure, in order to keep a check on the Abinger clocks, a simple frequency comparator, for use in conjunction with a decimal counter chronometer, has been constructed in the electronics laboratory and is now in regular use. Although by this means frequencies may be compared with an accuracy of the order of one part in 10¹¹, only instantaneous relative values are determined and these may differ slightly from the mean daily rates.

The reserve phasable phonic motor, which is provided with rhythmic contacts of the drum pattern, has been fitted with two sets of additional cam-type contacts of a new design, which are being tested prior to their incorporation in the new signal transmitting motors now nearing completion at the factory of Messrs. Muirhead & Co. These contacts, designed to operate at intervals of one minute or longer, were tested at a speed of 61 operations per minute in order to compare the magnitude of the contact variations with those observed in the present type of drum contacts. It was found that the mean scatter was reduced from 0.10 milliseconds to 0.06 milliseconds, which was not sufficient improvement to warrant their use at the higher speeds in the new motors, as they are noisier in operation, more costly, and subject to greater wear than the drum type.

The installation in the new control room is nearing completion, but progress has been show, largely owing to the fact that the staff engaged on this work is also responsible for the maintenance of the operational equipment. With the necessarily complex instruments now employed for precision time-keeping, efficient maintenance and adjustment has become an important section of the work of the department, and great attention is being paid to the practical design of new items of equipment to facilitate testing and rapid replacement of defective units. All essential parts of the system will, where possible, be duplicated, in order to permit observing staff, who may not necessarily be fully acquainted with all the technical details of the operational equipment, to switch over from a faulty unit to a reserve.

Five electronic send relays of an improved design, which allow for convenient maintenance without removal of the chassis from the rack, have been built and installed. By means of a switching panel, the send relays can be connected at will to the four outgoing lines, and provision is also made for switching the Post Office lines to the various test positions necessary for the location of faults by the Post Office engineers.

Several modifications have been made to the decimal counter chronometers, which have been remounted on standard 6’ 6” racks together with associated units. Provision has been made for the selection of one of four standard count frequencies, and a device has been fitted which enables the counter to start and stop on consecutive pulses on the same channel. Seconds impulse dials for checking comparison times have been provided. In addition to the plug and jack selection panel of the type previously used for the connection of clock and radio signals to the counter, sets of push button switches have been provided, which are connected to the circuits normally required during the daily routine.

Four National H.R.O. and two Hallicrafter SX. 28 radio receivers have been modified for time signal measurement and special attention has been given to the screening of leads to and from the sets in order to minimise interference arising from other apparatus in the department.

During the period May-July 1946, Small Transit B was in use at Abinger for astronomical observations for the determination of clock errors. The instrument was then removed to Greenwich, its place being taken by Transit D. The numbers of observations secured at Abinger during the two periods were 40 and 73, making a total for the year of 113. During the period August 1946 to April 1947, 99 observations were taken at Greenwich with Transit B, these observations being combined with those secured at Abinger into a system upon which the performance of the various clocks has been assessed.

At the request of the Director of the Commonwealth Observatory, Mount Stromlo, Canberra, a special programme of observation was undertaken to obtain an improved value of the longitude of the Mount Stromlo Observatory. This involved the erection at Abinger of a rhombic aerial in order to receive special radio time signals emitted by the transmitter, VHK₁₄,at Belconnen, near Canberra. Assistance in the erection of this aerial was given by 90 Group, Air Ministry, whose cooperation was much appreciated. Good reception of the time signal emitted at 19.00h U.T. (10.350 Mc/s) was maintained, in general, over the period August to December, after which the programme was discontinued. The England-to-Australia radio link was effected by means of the rhythmic time signals emitted from Rugby (GKU₃) on a frequency of 12.455 Mc/s. Reception at Mount Stromlo of these signals was not at first satisfactory, but was greatly improved when, with the assistance of the Engineer-in-Chief, Radio Branch of the G.P.O., arrangements were made for the transfer of GKU₃ from a directional to a non-directional aerial. A discussion of the results shewed that no correction to the adopted longitude of Abinger was necessary, but that the previously accepted longitude of Mount Stromlo placed the Observatory 0s.103 too far to the East.

The normal programme of time signal transmission and reception has been continued at Abinger. There has been some increase in the number of time signals received. These now include transmission from France, Russia (Moscow and Tashkent) and America, as well as the long wave and short wave British time signals. In addition daily comparisons have been made, when reception conditions permitted, between the Abinger clocks and the seconds pulses superposed on .the WWV standard frequency emissions of the Washington Bureau of Standards.

Damage caused by icing to the aerial at the Rugby radio station prevented transmission of the GBR 10.00 and 18.00 signals between February 8 and April 13, 1947. During this period the emission of the signal was transferred to the station GBZ at Criggion, which operates on a frequency of 15.2 kc./s. Although in this way continuity of the signal was assured with an adequate accuracy for navigational needs, the very slow and unsteady build-up of the radiated signals, combined with variations in land-line lag and signal strength at the receiver, made some loss of precision inevitable.

On April 18th at 11.00 U.T. a special time signal consisting of ten minutes of mean time seconds dots was sent out from the Observatory and radiated by the British Broadcasting Corporation for the purpose of timing the He1igoland demolition seismic experiments. At the request of Prof. Lindblad, Director of Stockholm Observatory, arrangements are being made to transmit a time signal consisting of a continuous series of rhythmic dots from 11.05 to 16.20 U.T. during the eclipse on May 20, 1947. These signals will be radiated from Rugby GBR on 16 kc./s and from two suitably directed short wave transmitters which have been allocated for the purpose by the cooperation of the Engineer-in-Chief, Post Office.

At Greenwich a Dent regulator controlled by Shortt 67 has been providing a reserve service of six-dots time signals for the B.B.C. and hourly signals for the Post Office. Comparisons between this clock and those at Abinger have been made daily over a land-line. The batteries, charging. equipment, and control board for the six quartz clocks have been installed and work is in progress in preparation for the delivery of the clocks themselves.

The staff of the electronics laboratory has cooperated in the design of equipment for the control room, and in the construction of some of the units. The experimental transmitter, which was described last year, was in use until April 23 for the radiation of the 2 Mc/s standard frequency transmissions. An aerial matching unit has been designed and made, and a considerable increase in radiated power has been effected. A 300 watt transmitter has now been adapted for this purpose, and has been in use since April 24. In the design of the 100 kc./s to 2 Mc/s converter great attention has been given to securing a high phase stability. Reception is reported to be entirely satisfactory at the Post Office Research Laboratories and at the British Broadcasting Corporation Receiving Station, and precision frequency comparisons are made daily. A voice announcement has replaced the morse call sign previously employed and at the end of each transmission, a provisional correction to the radiated frequency is given. Comprehensive switching, modulating and monitoring equipment for semi-automatic operation of the transmitter has been designed, constructed and tested, and is now in course of installation.

The long wave receiver, fully described in the 1945 report, has been further modified. A new radio-frequency unit, with higher gain to permit effective operation on GBZ, which gives a lower signal strength than GBR, has been designed and is now under construction. Other modifications have been made to make the receiver self-checking in its adjustments, and more compact.

The possibility of using frequency dividers of the pulse counting type down to very low frequencies, thus dispensing with phonic motors, is being investigated. There are many applications in which the erratics inherent in phonic motors cannot be tolerated, and it is expected that this divider may find uses apart from the long wave receiver testing equipment for which it was primarily intended. For this testing equipment special linear time base circuits have been designed and tested and a portion of the cathode ray tube equipment has been constructed.

1948

The Report here presented refers to the period from 1947 May 1 to 1948 April 30 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

(a) Greenwich. … The old Upper Computing Room has been made into a Control Room for the six quartz clocks; it has been provided with a new ceiling and re-decorated, as has the old Wireless Room.

(b) Abinger. A new sliding shutter has been fitted to the roof of the Bamberg transit hut. A number of dangerous trees in the immediate vicinity of the buildings or bordering on public footpaths have been felled. The lavatory facilities have been extended, and minor works of repair and maintenance carried out. Fluorescent lighting has been installed throughout the offices, laboratories, workshop and control room.

(C) Herstmonceux. ... A start has also been made with the Solar Building ... It is planned to mount a Shortt master pendulum on the main pier; although this will not be thermostatically controlled, it will serve to carry the time between radio signals, and will in fact be the nearest approach to a precision standard at Herstmonceux for some years.

II. Instruments, Apparatus and Library.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock

To the National Maritime Museum – Harrison's Time Machines Nos. 2 and 4, and the copy of No.4 by Larcum Kendall.

Harrison's Time Machines Nos.1 and 3 are still at the Chronometer Depot at Bradford-on-Avon, where they will undergo necessary repairs as soon as other work permits.

Bamberg Transit

The Bamberg broken transit, which had been very thoroughly reconditioned in the Observatory workshop, was installed at Abinger in July. The pivots had been chromium-plated and then lapped by hand with extensive Mikrotast measurements. Their final figure departed from the circular by a very few hundredths of a micron only. The actual operational figure, including the effects of oil and all errors of reading, is such that a whole night’s levels frequently fall well within an overall range of a second of arc, and so far as can be judged by appearance the chromium is standing up well to wear. The final scatter of a night’s observations is about two thirds of that previously obtained with Small Transit D.

Small Transits

Small Transit D itself had meanwhile deteriorated; it was withdrawn from use in January, since it could no longer make a sufficient contribution to justify the labour involved. Transit C is at present being modified at Greenwich to replace it; the pivots have been plated and figured, the new illumination-system has been installed, and a weight-relieving gear is being added. Transit B, which is at present in use at Greenwich with stainless steel pivots, is continuing to give good service, but all these non-broken instruments suffer from the drawback that the striding-level must continually be removed and replaced, with a risk of affecting its zero, and with the serious disadvantage that the level can only be placed on the pivot in certain positions, so that unequal wear results. A scheme for hanging the level by long invar rods, so that it can stay permanently in place, is under consideration, and some work has also been done on an optical levelling method involving a mirror mounted on a reversible pendulum as a standard vertical plane.

Quartz Clocks

The six quartz crystals, groups A and F, were installed at Greenwich early in the summer of 1947 and have steadied down so satisfactorily that they are already among the best in the Observatory. (This is no doubt partly due to the fact that they had previously been run for extended periods in the Post Office laboratories). However, the heat developed by their driving and auxiliary circuits is such that the apparatus is running dangerously near the temperature at which breakdown may be expected. Arrangements are being made to ventilate the cubicles to a variable extent, so as to permit control at a lower temperature. The temperature is also seriously dependent, at present, on the mains voltage, and a voltage-regulator is to be installed.

Control Room

At Abinger, the new control room has been completed sufficiently to bring it into full operation. The work has taken three years, and though a considerable number of unforeseen changes in requirements have occurred since it was first planned it seems well adapted to present needs, and the general work of signal-reception and clock-comparison has been greatly facilitated. The experience gained will be of great value in planning the arrangements at Herstmonceux.

The new signal-transmitting motors have been received from Messrs. Muirhead, but much experimental and constructional work will have to be completed before they can be installed. As other more urgent work, such as modifications to the crystal thermostat control circuits, still remains to be completed, it is unlikely that the motors will be brought into use before the end of this year. Trouble has been experienced recently with the moving-coil chronographs. They have given long and valuable service, but the stock of replacement coils has been exhausted and no new source of supply has been found. The moving-coil units have accordingly been replaced with modified and rewound polarized relays, which have a very light moving unit. They are giving a performance which compares favourably with that previously obtained.

A contract has been placed for the supply of two Belin chronographs similar to those used at the Bureau International de l’Heure. These chronographs are specially designed for the recording of mean-time and rhythmic time-signals, and should meet the long-felt need for a permanent record to supplement the precision measurements made possible with the decimal counter chronometers. A R201 triple diversity receiver has been made available on loan by the War Office in order to investigate its suitability for the reception of distant radio time-signals. The tests are delayed, however, pending completion of negotiations for the erection of suitable aerials.

Further experiments on reception of long-wave time-signals have been carried out and, as a result, a new prototype receiver has been installed and compared with the previous model. From the experience gained, and as a result of further experimental work in the electronics laboratory, a new design has been prepared and four receivers are now under construction. Two will be used at Abinger, and two at Greenwich in replacement of the Admiralty pattern long-wave receiver now temporarily employed.

The old control room has now been completely cleared of equipment with the exception of the pendulum clocks, and will be used to house the quartz clock obtained from the Physikalische Technische Reichsanstalt. An examination of this clock and of its short-term performance has been completed at the Post Office Radio Branch Laboratories, and is summarised in Radio Branch Report 1610. It is hoped to set the clock in operation at Abinger, together with new divider equipment and the original German phonic motor, and to determine its stability over long periods.

Photographic Zenith Tube

Much further experimental work has been carried out for this instrument. In several cases the employment by the makers, in trial work, of gears and materials that were unsuitable had in fact the valuable result of drawing attention to difficulties that might also arise, to a much smaller but perhaps still significant extent, with first-class equipment. As a consequence it now seems likely that the differential roller will be preceded by a fairly high-quality screw, and will in itself give a speed-reduction considerably smaller than had been contemplated. The carriage will probably not run on mechanical flats but will be suspended, and the azimuth will be determinable without traversing the carriage over a longer (and thus less-worn) range than that of normal operation. If the trials along these lines prove successful, all precision parts of the movement will be controlled entirely by elastic links, with no parts subject to machinal wear. Present indications are that the scatter in positional accuracy in the relevant directions will be only a very few tenths of a micron at most. It is hoped that these tests will be completed shortly, so that the final stages of construct ion can be proceeded with.

The electrical method of studying the stability of level of a mercury surface, which was mentioned in the last Report, has been developed and tested. Two stainless steel disks are mounted close above the mercury, and the two condensers so formed are placed in two arms of a bridge circuit. The sensitivity is limited by the fact that with small gaps the voltage must be kept low, or else the mercury surface is seriously distorted by electrostatic attraction. With increased circuit-stability and lower voltage the method appears promising, but further study has been held up by staff losses. In the course of this work, an opportunity arose of testing out the method by which it is proposed to level the stellar plate of the Zenith Tube; it was found to be satisfactory for the purpose. A description of the method has been published.

Some tests were made of the new Furt microphotometer developed by Messrs. Ferranti, to study its usefulness as a pure position-measuring instrument for stellar images etc. Direct visual methods were found to be slightly better for ideal test-objects, but it appears possible that if somewhat greater electrical stability can be attained the microphotometer might be superior, especially for photographic images and irregular objects generally.

III Astronomical and Radio Observations

Time Service

The following observations for clock error have been obtained during the periods stated:-

	Abinger, Small Transit D (to Jan. 20)	73
	Abinger, Bamberg (from July 13)	188
	Greenwich, Small Transit B (year)	270
	Total	531

The large number of observations (especially at Greenwich) has been made possible only by having two duties on one night whenever practicable. This, together with the fact that very few observers were engaged (at Greenwich only two) has considerably steadied the personal equations, and a noticeable improvement in overall accuracy has resulted, in addition to improvements of instrumental origin. At the same time, the greatly increased number of fairly steady quartz· clocks now available to the Observatory has made the drawing of reliable clock-error curves less difficult. Intercomparisons between the 12 clocks at Abinger, the 6 at Greenwich, and some others at the Post Office Radio Branch and the N.P.L., are made by differencing the times of reception of the GBR 1000 signals, and the most suitable crystals out of a total of nearly 30 can thus be readily identified and used for prediction and control of signals, and as primary standards.

Quite apart however from this direct contribution to the increased accuracy and reliability of the service, the use of clocks at four establishments possesses many other advantages. By a statistical analysis of the reception times of the GBR time-signals on clocks at each establishment, it is possible to estimate the magnitude of the signal erratics and errors of radio-reception. By applying the final published corrections derived for the GBR time-signals to a clock at an establishment which has not directly contributed to their assessment, an independent estimate of the uniformity of the adopted time-system may be deduced. A recent analysis of the results for 1946 has shown that the mean deviation of the 1000 GBR signals from a smooth curve was reduced during the year from 0.9 to 0.4 milliseconds. This is further reduced to 0.3 milliseconds if corrections are applied for known variations in the lag of the land-line from Abinger to the transmitter at Rugby. The mean absolute value of the error of radio-reception, measurement, and other local errors was reduced in the same period from 0.3 to 0.2 milliseconds. At the end of 1946, the corresponding radio-reception erratics for the Post Office Laboratories and the National Physical Laboratory were 0.3 and 0.2 milliseconds respectively. The values of the erratics were also examined for a period at the end of 1947, and were found to be too small to be detected by the simple methods of analysis previously employed; they are estimated to be of the order of one-tenth of a millisecond.

The long-term accuracy of the best crystals is now sufficient to show up effects due to the earth’s polar motion. If the earth’s rate of rotation as a whole were constant, and if a perfect clock were available, polar motion would introduce spurious clock-error at any observatory that was not on the equator, since evidently all meridians must pivot about their mid-points when the poles move. As long as available clocks could not show this effect, it was immaterial whether the zero meridian was regarded as pivoting about Greenwich or about its mid-point on the equator, and in the former case (the actual one) the longitude of Greenwich itself could be permanently zero by definition. It has now become inexpedient to continue this convention; the mean longitude is still zero, but the actual value must be regarded as fluctuating. If an up-to-date knowledge of the polar movement were available, allowance for the displacement of the meridian would permit a more certain assessment of clock performance; the determination of changes in the quadratic term is much mere difficult if these are overlaid by a smooth systematic variation which may exceed 20 milliseconds each way in the course of the year. The international latitude results are not available soon enough for this purpose; but the Washington results, obtained from the Photographic Zenith Tube, are reduced very promptly and are of high quality. For some considerable time now the results have been sent to Greenwich by air-mail as soon as each group has been reduced at the Naval Observatory. This welcome cooperation is of great value; the longitude of Washington (+77°) is sufficiently near 90° for this component to be practically sufficient by itself, and the corrections so obtained have been used for assessing frequency-drift since 1947 July. For the purpose of computing the final Greenwich Mean Time the effect of polar motion is re-introduced; the time is thus that of the actual (variable) meridian, and any arbitrariness in the adopted polar motion is almost completely eliminated. However, since many users of precision time-signals are more interested in their physical uniformity than in their astronomical significance, a column has been added to the published “Times or Reception” giving the polar correction explicitly. This procedure admittedly provides only an approximate correction for the effects of polar motion, but it is believed that the residual uncorrected effect is too small for any crystal to detect at present. If the polar motion could be computed in advance and included in the Almanac, it is probable that only the uniform time would be employed, and would of Course be called Greenwich Mean Time; unfortunately, this is not the case, and the question of nomenclature is somewhat difficult.

The Bureau de l’Heure has now recommenced publication of the regular tables and graphs relating the time finally adopted at fourteen observatories to the mean of them all. The results for 1946, which have recently appeared, fully confirm the view that the final Greenwich times were very smooth and also very closely correct; the extreme range of Greenwich on the mean is from +15 to -37 milliseconds. These curves are not corrected for polar motion; using the results for the International Latitude Variation Service which have new become available for 1946, the corrections to be applied to the time determinations of each of the 14 observatories can be computed and allowed for. Only the Greenwich and Washington curves are smooth enough to be materially improved by this action; the curve “Greenwich minus Mean” becomes a steady linear drift from 0 to -20 milliseconds throughout the year, and the curve “Greenwich minus Washington” shows a steady value of +40 ms in the first quarter of 1946, followed by a drift from +40 to +10 ms in the second quarter and a steady value of +10 ms for the remaining six months. The Greenwich drift might in principle be due to errors by the other 13 observatories, but this seems unlikely since the Washington drift on the mean is in the opposite direction; the GreenwichWashington difference, and especially its change of 30 ms in three months, cannot be explained in this way, and appears considerably larger than any possible reception-lags. The importance of these comparisons will evidently increase as instrumental accuracy improves; meanwhile there can be no doubt that Greenwich and Washington show the smoothest curves at present, with Greenwich somewhat the better of the two.

Of the Abinger and Greenwich clocks, C3 and D3 (Abinger) and A1, A2 and Fl (Greenwich) are the best at present; C3 and D3 have in fact superseded the Post Office Group IV for long-term prediction, the latter having deteriorated. Group IV will be withdrawn shortly for modification, and it is fortunate that the Greenwich groups have settled down so rapidly. Up till February, the Greenwich-Abinger comparisons involved use of a land-line; since that date, the installation of decimal counters at Greenwich has made possible a direct use of GBR in addition.

Routine transmissions and receptions have continued. The long-wave and shortwave 1000 and 1800 transmissions are controlled from Abinger; the hourly signals to the Post Office and the “six-pips” to the B.B.C. are available from both stations. The six-pips service is also relayed to the Post Office for distribution to Service users. British long-wave and short-wave signals, and signals from America, France and Russia are received at Abinger and also, partially, at Greenwich. The WWV signals (seconds contacts superimposed on a standard frequency) which are transmitted continuously, are received at both stations, but have not yet been included in the “Times of Reception”, mainly because they are received at variable times. A series of time comparisons with WWV made every few minutes over some hours indicates that the accuracy of measurement, including effects due to changes in the transmissionconditions, corresponds to a mean error of ±0.3 ms and there seems little doubt that the reception could be reduced to some standard time with sufficient accuracy.

In August it was decided to try transmitting the 1800 signal to Rugby on a direct current line, while keeping the 1000 signal on the voice-frequency channel hitherto used. Comparison showed that the lag of the D.C. line is smaller, and much more constant from day to day, than that of the V.F. channel; the maximum variation in five months was about 6 ms as against 12 ms for V.F. From 1948 January therefore, both transmissions have gone by the D.C. line; the short-period variations occurring in the ten minutes between test signal and signal proper are now about ±0.5 ms.

Since 1944 it has been the rule to limit deliberate changes in the rate of the signal to one millisecond per day, so as to cause the minimum inconvenience to users whose main concern was the interval (i.e. frequency) rather than the absolute time. If, as a result of reassessment of clock error, particularly after a period of poor observing weather or unpredictable changes in clock rate, it appeared that the signals were in error, readjustment of the time signals was made by jumps of 40 milliseconds (or multiples of 40 milliseconds) in preference to deliberate gross adjustments of rate. Such jumps, when required, were always introduced for the GBR 1000 signal on a Wednesday.

The limitation of jumps to 40 milliseconds has proved to be a serious handicap; it has meant in practice that signals have frequently been transmitted systematically in error over a period of some weeks. As from May 1, therefore, corrections will be applied as required, but only on Wednesdays, to the GBR radio time-signals in jumps of 20 milliseconds (or multiples of 20 milliseconds). At the same time, the deliberate adjustments in rate will be normally limited to 0.6 milliseconds per day.

The 2 Mc/s daily frequency transmission from Abinger, originally introduced to provide a convenient means of making frequency comparisons between the clocks used for the control of the time service, has proved to be of much wider interest. Details were therefore published in the technical press in November, and the facility is now in very general use. From October, an estimate of the transmitted frequency has been based on daily comparisons with the Abinger standards, which are combined numerically into a “mean crystal”. Comparisons are effected by means of a frequency comparator of very high accuracy, which is a slightly modified version of a Post· Office design. A special receiver for comparing the frequency of the Droitwich 200 kc/s transmission with any of the Abinger standards has also been designed and constructed. It incorporates a new type of synchronized oscillator, developed in the laboratory, which appears to be adaptable to a wide range of uses. A short report is being prepared for publication. From 1948 March 2, daily comparisons of the frequencies of C3 and D3 with that of the Droitwich 200 kc/s transmitter have been made and the results communicated to interested establishments. The whole question of the future programme of standard frequency transmissions in this country is now under consideration, and the Royal Observatory is co-operating with the other departments concerned in the formulation of definite proposals for a service which will also carry super imposed time signals.

As no detailed information was available on the interference which may be caused to radio-reception at Herstmonceux by the Pevensey radar station (not at present in operation) tests were carried out in the neighbourhood of the similar station at Bawdsey. Interference on frequencies used for radio time-signals was less than had been expected, and the tests showed that it would be possible to select suitable reception sites in an area behind and within two miles of the Bawdsey station. The results were not altogether conclusive, as owing to different aerial characteristics, transmitter power, and local conditions, the level of interference at Herstmonceux may be different from that at Bawdsey. A further series of tests is therefore planned when the Pevensey station is in full operation.

Tests have also been made on reception-conditions in the Rugby district; the proposed standard frequency service, as well as a possible reserve Time Station, may ultimately be established in this area, and it was desired to determine the degree of interference which might arise from the high power transmitters there. Tests were made at a site only two miles from the Post Office Radio Station, and in the direction of the British Broadcasting Corporation station at Droitwich. It was proved that reception of the normal range of foreign radio time signals could be obtained with complete freedom from significant interference. As a part of the same field programme, simultaneous film records of the GBR and Rugby short-wave time-signals applied to a cathode ray oscillograph were taken at the Rugby site and at Abinger, in order to obtain some indication of the variations which might be expected in the relative effective travel time of short-wave and long-wave signals. When the short-wave time-signals were first introduced, it was realised that accurate monitoring of the transmissions might not be possible at Abinger, which is within the skip area of the transmitters. It is for this reason that, in the final published corrections, the reception times of the British short wave time-signals are given in units of hundredths of a second only. The work of reading up the film records has been long and laborious, but is now almost completed.

While the equipment was at the Rugby site, film records were obtained of the GBR time signal dots on an oscillograph without the interposition of a radio receiver, for comparisons with the build-up of the signals at the output of the long wave receivers at Abinger. Further tests on similar lines, planned in the light of the experience gained, are contemplated.

The total amount of computation, and also of installation and maintenance work, has steadily increased in the Time Department. Some simplifications of method have proved possible, but the work of intercomparison is so extensive that the load on the staff remains heavy. Nevertheless, some arrears remaining from the war period have been made up, and the distribution of the “Times of Reception” has been accelerated. At the date of the last Report these figures were some 18 months in arrears, but results have now been completed for 1947 December. Some progress has also been made with copy for the “Greenwich Observations”.

XI GENERAL

Mr. Smith and Mr. Pope visited Paris in November to see the work of the Bureau de 1’Heure, the Laboratoire National de Radfoelectricite, and other scientific institutions. Visits were also paid to various firms engaged on the manufacture of frequency-standard equipment, and to Etablissement Belin, from which firm two chronographs have since been ordered.

1949

The Report here presented refers to the period from 1948 May 1 to 1949 April 30 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings

(a) Greenwich

… The old Cookson hut has been fitted up to accommodate a petrol-driven A.C. generator which has been obtained in order to have a standby power-source for the quartz clock installation.

(b) Abinger.

The premises at “Cornerways”, Abinger Hammer, which were requisitioned during the war for use as offices and which, since the war, have been used as a staff hostel were vacated at the end of 1948. Negotiations for the tenancy of “Feldemore”, Holmbury St. Mary, referred to in the previous report, were completed. The hostel was transferred there in January, 1949. The house provides also some office and storage accommodation for the Time Department and accommodation for seven married members of the staff.

Several of the halyards on the masts supporting the aerials on Wotton Common were broken during heavy gales. The masts were lowered for fitting new halyards and subsequently re-erected.

II. Instruments, Apparatus, and Library.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock.

To the National Maritime Museum: – Harrison's Time Machines Nos. 2 and 4, and the copy of No.4 by Larcum Kendall.

The removal of the Chronometer Department to Herstmonceux and pressure of work have prevented any further progress in the repair of Harrison's Time Machines Nos.1 and 3.

Photographic Zenith Tube.

The experimental work has now been concluded. A number of comparative records of different forms of drive was measured in the laboratory at Abinger. These showed that the best results were obtained by the use of the differential roller with reduced magnification preceded by a screw. It has therefore been decided to adopt this form of drive. But in the general design of the rotary, provision has been made for easy transfer to an alternative form if unforeseen difficulties in actual operation should be encountered.

It has been decided to make all the carriage constraints elastic on the lines suggested in the previous report.

Certain simplifications of design have been made. The four chronographic records pertaining to a single star will be superposed, and the chronographic plate will only be moved between the observations of successive stars. This eliminates the need for fiducial marks on the plates, and thus simplifies the optical design of the chronograph and reduces the work of measurement. It also eliminates error due to misorientation of the measuring wire and facilitates the checking of N.-S. displacements of the carriage. In order to make this superposition feasible it is necessary to control the relative phases of the four series of traverses. Coltrol [control] is achieved by adapting the time-sequence so that the traverse and rest periods of the carriage are specific factors of the flashing interval of the strobotron lamp, and by selection of the phase relation between the sequence and the flashes. It is also arranged that the complete record of a star is broken up into groups of four

for ease of identification and that the records of corresponding traverses are adjacent; the previous need for a mechanical method of securing this condition is thus eliminated. The optical method of levelling the stellar plate and squaring on the objective has been replaced by a simpler and better system. The use of the suspended silica rod by which, in conjunction with the optical method of setting, it had been proposed to maintain constancy of separation between stellar plate and mercury surface has been provisionally discarded because it only controls one of the factors upon which scale value depends. But since the relative importance of this factor is at present unknown, the possibility of restoring the rod in future, should the need be shown to arise, has been provided for.

Certain other features of the general design have been dealt with. Means of obtaining the absolute value of distance between the photographic plate and the mercury surface-in order to determine the displacement of the stellar plate from the gaussian plane have been added by the provision of a micrometric measuring-rod. Reduction of chronographic magnification has been made in order to accommodate the requirements of the new carriage constraints. The carriage clamp, which needed modification to remedy defects and to suit the new constraints, has been redesigned and improved. The rotary clamp, which had previously only been considered in principle, has now received detailed treatment. A satisfactory design for the stellar plate-holder, which needed modification to suit the new type of rigid spider mounting, has been completed. Attention has also been given to the design of the stellar shutter, and the relative merits of a number of types have been investigated. This feature of the instrument can, however, receive separate treatment, since it is external and therefore has no direct bearing on the main design. Further work has been done in the laboratory at Abinger on the errors of measurement of the stellar photographic records.

Some consideration has also been given to the question of the best type of housing for the Photographic Zenith Tube, and the general lines to be followed have been agreed.

The present position may be summarized by saying that the main features of the instrument have now all been decided; that the corresponding basic designs have been made, that the detailing of designs is in progress, and that some parts of the instrument have already been constructed.

Small Transits.

Small Transit C, which was being modified last year, is now nearly ready for use. The reversing gear has been arranged to serve also as a counterpoise, so that the bearing area could be reduced to the narrow ring where the figure is really good. Small pressure-rollers have also been added in the micrometer, converting it approximately to a kinematic mounting; so long as there is play on the guides, there is in fact a serious possibility of a permanent error which does not go out on reversal. The instrument has not yet been fitted with the proposed hanging level, and will now be taken into use without waiting for this.

Clocks and Radio Equipment.

Some tests have been carried out on a new type of crystal oscillator, in which the amplitude of oscillation is greatly reduced. A suitable crystal for these experiments was presented to the Observatory by Standard Telephones and Cables Limited. This assistance is much appreciated; thanks are also due to Mr. Jacobsen of the same firm for his advice and help. A further crystal has since been made available by the Post Office Radio Branch in order that experimental work may be continued while the performance of the other crystal and drive circuit are being examined. The preliminary tests were very promising, and further work was commenced on an improved form of oven control circuit. Pressure of other work has necessitated the temporary stoppage of these investigations, but it is hoped that it will be possible to resume them before long.

The German quartz clock, PTR VIII, is now being installed in one of the clock rooms at Abinger. The temperature control circuits, frequency dividers, and power amplifier have been completely rebuilt. Work is in progress on the power supply circuits. Only the oscillator, buffer amplifier and motor are being retained unaltered.

Two tape chronographs of an improved and more compact design, primarily intended for field work, have been constructed. The Belin chronographs have been received and, after some modifications, one has been installed in the Control Room to provide a permanent record of all radio time signals. The other has been placed in the cellar and will be used for the continuous and simultaneous recording of the time differences between six quartz clocks. At present it is being used for tests, made at the request of the Hydrographer, on the uniformity of rate of contact chronometers.

The design of the receiver units of the long wave reception and measuring equipment has been standardized, and three receivers have been completed. Two are

now in use for the routine work at Abinger. The Observatory is indebted to Standard Telephone and Cables Limited for their generosity in supplying the quartz crystals used in the selective stages of these receivers. A report has been prepared on the receiver units, and has been distributed on a limited scale. Design work on the ancillary equipment is approaching completion and special attention is being paid to the need to provide for easy maintenance and simplicity of operation. The R201 receiver loaned by the War Office has been returned owing to the difficulty experienced in obtaining suitable sites for the three aerials.

Some modifications and improvements have been made to the design of the Droitwich receiver, used to derive a 100 kc/s output controlled by the carrier frequency of the Droitwich transmitter. Two complete receivers have been constructed, one being in regular use in the Control Room, and the other serving as a reserve.

The experience acquired in the laboratory in the design of receivers working on 16 kc/s has been applied in the construction of a sudden phase anomaly recorder for use by the Solar Department, based on an experimental model developed at the Cavendish Laboratory Cambridge. The instrument has been tested at both Abinger and Herstmonceux and is working very satisfactorily.

III. Astronomical and Raidio Observations.

Solar Eclipse. 1948 November 1.

Four 50-foot collapsible radio-masts, which were overhauled and repaired at Abinger, carried the aerials for the reception of the time-signals, and formed the largest single item of the apparatus transported from England. Petrol driven generators and batteries were overhauled and tested at Abinger, where the whole of the equipment was packed in a specially prepared naval radio container for shipment. The two observers went out and back by air.

Special time signals were sent out during the eclipse from Abinger via the Rugby radio station, with the co-operation of the Post Office.

Time Service.

Astronomical observations for the determination of time have been made throughout the year at Abinger using the Bamberg transit, and at Greenwich using Transit B. The number of observers has been limited in order to ensure that each is adequately represented in the results obtained over any period of a few weeks. Whenever possible two groups of observations are taken at each station on the same evening, which are separately reduced. At Abinger, a total of 320 observations were secured on 174 nights. At Greenwich, both observers had to discontinue observing in 1948 April for medical reasons; several temporary observers used the small transit instrument B during the summer, but their observations were not used for the determination of clock performance. By September, conditions had improved and, since October, the Greenwich small transit results have been used again. Of the total of 143 observations on 106 nights, 91 observations secured on 67 nights were used in the operation of the time service.

Up to the installation of phase-integrating rotary beat counters at Greenwich and Abinger in 1947, the only clocks for which satisfactory results could be obtained were those controlling phonic motors through the limited number of divider chains; intercomparisons of the remaining clocks were effected by impulse beat counters which were subject to frequent mechanical failure. The new counters have made continuous long-term intercomparison of all crystals possible, and the number of quartz clocks which can be considered for the operation of the time service has thus increased from eight to twenty-four. For a period, nearly thirty were available, but on June 21st the Post Office Group IV at Dollis Hill was taken out of service for major overhaul, leaving only the five ring-type crystals in operation. Not all of the available crystals have a performance which is sufficiently good to justify their inclusion, and several can be ruled out on inspection. Nevertheless the increase in the amount of usable data has involved a corresponding increase in the computing work. Various experiments have been made in lay-out and presentation to ensure that the maximum of information about the clocks should be displayed in a convenient form with the minimum of delay.

Attention has also been given to the problem of developing improved methods in the prediction of clock corrections for the control of time signals. The observations for time determination have steadily become more numerous, and possibly more accurate, than those obtained during the war years, while the performance of the best clocks has shown a slight, but definite, deterioration; it is evident that the limit to the accuracy of the time service provided is still set not only by the scatter of’ the time-observations but also by the quality of the clocks available.

It had been anticipated that the quartz clocks would prove of such long-term stability that it would be necessary to remove the effect of polar variation from the time-observations in order to assess the clock performance adequately. This is done as a matter of principle but while the comparison of finally adopted time between Washington and Greenwich is thereby materially improved none of the quartz clocks available to the time service is capable of clearly showing the effect as an apparent deviation from uniformity of run and prediction is not appreciably facilitated by the correction. Instead of predicting clock corrections from long runs, it has been found better to take a shorter term view of the clock performance, giving much more weight to the recent observations. A long-term linear and parabolic ephemeris is still cleared from the rresults , for each clock selected as suitable; but after that, attention is focused on the last fifty observations only, and these are fitted to a linear ephemeris by least squares. It is this linear ephemeris which is used for predict ion; it is re-computed as soon as five new observations are available, the oldest five being then dropped. The influence of the quadratic term actually used in the first ephemeris is thus slight, and the second stage can be put on a pure routine basis; in difficult cases, however, the highest grade of personal judgment has still to be relied on, and the long-term behaviour must be specially reviewed. The indications from the selected clocks are then coordinated to obtain the best estimate of clock correction. Since the adoption of this procedure, the number of occasions on which it has been necessary to introduce a “jump” into the signal has been much reduced.

A new service for the benefit of those departments and establishments which maintain accurate frequency standards is the systematic estimation of absolute frequency. A monthly circular giving the estimated frequencies has been prepared and distributed since 1948 April, and a similar weekly service was instituted in 1948 June. The information has proved useful to the recipients, in spite of some uncertainty in the estimate of frequency from the current observations. The definitive reception times of the radio time signals are of necessity published three or four months in arrear, which is too late to be of much use to those who wish to know the current performance of their frequency standards.

A preliminary study of the frequency estimation service over the comparatively short period since its introduction shows that the monthly estimate, made from ten to fourteen days after the end of the month treated, has been within ±1 part in a hundred million for 81% of the period and within ±½ part in a hundred million for 70% of the period. Comparable figures for the weekly estimate were:- within one part, 53%; within half part, 23%. The maximum error has never reached two parts in a hundred million. These assessments of errors in the estimates are based on the clock error curves used in the preparation of the bulletin of reception times, and this at once raises the question of the possibility of further modifying these curves. Examination of the charts shows that in general the adopted curves could not suffer an adjustment of more than 5 milliseconds at any point without failing to be in agreement with the observations, but there are occasions when a strong suspicion has been aroused that the observations over some four or six weeks have systematically differed from the general run over several months before and after. In such cases, unless the observations are smoothed very drastically, there is no alternative but to draw a curve showing an identical local deviation on all clocks. If the deviations are attributed to a cause which affects the observations in a systematic manner for some weeks and they are ignored in drawing a smooth clock-curve, the adjustment to the adopted line might amount to 20 milliseconds. The rates or frequencies defined by the finally adopted curves would be but little affected by such adjustments, provided that the curves remained fairly smooth; it would be very hard to justify any change as large as one half part in a hundred million in the frequencies defined from the adopted curves. This process of defining an absolute frequency estimate assumes that the period of the earth’s rotation has remained constant over the period of several months for which the curves are drawn, an assumption which may not be strictly correct.

With the increase in routine computing, a welcome expansion in office space has resulted from the acquisition in February of the library at “Feldemore” for long-term computing work.

The routine reception and measurement of radio time-signals has been continued, some fifteen different signals being normally measured each day. All intercomparisons of clocks, and between clocks and radio time signals, have been made with the decimal counter chronometers. The standards of accuracy in measurement have been maintained, and, in the case of the GBR long-wave radio time-signals still further improved.

The Droitwich carrier frequency is measured daily at 10.30, and the corrections published in the monthly Time Service Circulars. Since the beat comparator is used not only for these measures, but also for routine intercomparisons between the standards, a duplicate instrument has been constructed and installed in accordance with the policy of providing a reserve for each instrument used in the regular operation of the service.

The service of time signals controlled from the Observatory has been maintained unchanged, except that from May 24 to 29 inclusive, while GBR was out of action for servicing, the long-wave time-signals were radiated from GBZ. The new phonic motors for the control of the time signals have been tested and one is in part operation. The driving amplifiers have been completed and work is in progress on other related equipment. A proposal is under consideration for the radiation from Rugby of mean-time signals in addition to the present rhythmic signals. Provision has been made on the phonic motors for the control of mean-time signals.

The 2 megacycle standard frequency transmission from Abinger has been continued and is proving of increasing value to other departments and establishments and to commercial firms. Owing to delays in concluding arrangements for the loan of a higher power transmitter and for the erection of an improved aerial system the original transmitter is still in use. The ancillary equipment, constructed within the Observatory, and originally intended for use only during a limited experimental period, has stood up well to daily use for some three and a half years; the whole equipment now requires complete rebuilding and it is hoped at the same time to make some modifications as a result of the experience gained.

Throughout the year the clocks C3 and D3 have been the only Abinger clocks used for long-term predictions. Each had one failure of temperature control during the year, D3 on May 10 and C3 on November 3. The clocks Cl and Dl have at times run uniformly for a few months and have been plotted for short periods. The E group was disturbed in June when the temperature control units were replaced by new units employing two thyratrons in parallel, to reduce the risk of faults caused by thyratron failure. This brings them into line with the equipment at Greenwich. None of the oscillators in this group has yet settled to a steady run. The B group has shown promise of improvement, but has not yet proved sufficiently steady to warrant plotting. In particular, the violent changes previously noted in B3, which had originally operated at Edinburgh as Al, appear to have ceased , and it is hoped that this oscillator will return to the excellent performance which it gave at Edinburgh. In spite of these slight improvements, the Abinger clocks remain inferior in quality to those at Greenwich and Dollis Hill; the Post Office Radio Branch has recently been approached with a view to further action being taken on this matter.

At Greenwich, the clocks Al and Fl have continued to give the best performance as long-period standards, chiefly because they seem less affected by mains disturbances. There was a complete mains failure at Greenwich on January 1 which affected all the clocks; these two showed a rapid recovery, and were not seriously impaired as long-period standards. Of the remaining clocks, F2, which has always been subject to large erratic changes of frequency, developed a fault in the maintaining amplifier and has been removed for inspection. The frequency of A3 continues to increase very rapidly, a characteristic which has been attributed to an unduly protracted ageing process.

The clocks in service at Dollis Hill have been the five ring-crystal oscillators, groups 5 and 9. Of these, 9A and 5B have proved, like Al and Fl, to have a long-term superiority due to a smaller susceptibility to disturbance. 5C and 9C have given good results between the occasions of disturbance, July 20 and March 5. 9B suffers from a leak in the crystal holder, and has always been liable to large erratic changes, comparable with those formerly experienced on B3. The quartz clock B at the National Physical Laboratory has suffered several serious disturbances during the year and is not yet considered worth plotting. Comparing the clocks at the various stations, it appears that the Greenwich and Dollis Hill clocks are giving a similar standard of performance, while the Abinger clocks are of decidedly lower standard. Omitting the violently erratic F2 and 9B, the best Abinger clocks are comparable with the worst remaining clocks at the other two stations.

Reference was made in last year’s report to tests carried out at Rugby and Abinger to determine the order of magnitude of the errors which may arise in monitoring the short-wave time signals at Abinger. As only simple equipment was used, and the duration of the trials was very severely restricted, the data obtained are somewhat limited. For the sample signals examined, reception at Abinger was not subject to effects arising from an abnormally long travel time, and there was no indication of day-to-day anomalies, though without much more extended tests it would be unwise to conclude that such anomalies may not occur. Under normal conditions the differential lag on long and short waves appears to be small, and more elaborate equipment would be required to determine the actual difference with precision. The results already obtained afford some justification for publishing the times of reception of the British short-wave time signals to three decimal places of a second instead of two places as hitherto, thus bringing them into line with the times published for other radio time signals. The published figures will be applicable to the majority of receptions though there is a risk that anomalous travel path may cause appreciable error in the measurements of occasional signals.

As the result of a suggestion by Mr. Ratcliffe of the Cavendish Laboratory, a programme has been arranged in co-operation with the U.S. Naval Observatory for the determination of the velocity of propagation of very long radio waves across the Atlantic. Conflicting results have been derived from the measurements, of limited accuracy, which have been made in the past. Using precise methods it is hoped to determine the relative times of arrival of the radio signals at each station with an accuracy of the order of a few tenths of a millisecond. An accurate determination of the speed of travel is not only of value in time service and longitude work, but it should also make it possible to discriminate between various theoretical assumptions on the mechanism of propagation. Measurements over a period of six weeks should provide an accurate determination of the velocity, but it is planned to extend the measurements over twelve months in order to decide whether the velocity has the same value under varying ionospheric conditions. By the co-operation of the Director of U.S. Naval Communications, the Navy transmitter at Annapolis, NSS, will transmit time signals on a frequency of 18 kc/s within five minutes of the normal times of transmission of the British radio time signals from Rugby, GBR, on a frequency of 16 kc/so One of the standard long-wave receivers, which has been slightly modified for the purpose, will be used at Abinger for the reception of the signals. A new frame aerial is at present under construction and will be used ‘for the routine receptions of the Rugby signals, thus setting free the rotatable frame for the special measurements. The signals. will be recorded by photographing a cathode ray oscilloscope trace with a 35 mm. film camera. The cathode ray and and timing equipment is now under construction in the laboratory and a camera based on modifications of a standard REB4 unit is being made in the Workshop. For processing the film, modifications are being made to the automatic developing equipment and a film drier has been constructed. A continuous check will be maintained on the performance of the radio and measuring equipment, using the test oscillator at present employed for calibration of the long-wave receivers; it produces pulses similar in shape to those emitted by the Rugby transmitter, and the lag of the receiver can be measured. To ensure that the equipment used on both sides of the Atlantic will be identical in all significant respects, full details of all the equipment designed at Abinger have been forwarded to the U.S. Naval Observatory, and also to the Department of Engineering of the Pennsylvania State College, where additional records will be obtained.

Work has been resumed on the preparation of copy for the volumes of ‘Greenwich Observations’ from 1937 onwards. The copy for the period 1937-1940 is nearing completion.

XI. General

Mr. H. M. Smith attended the conference of the Societe Chronometrique de France in Paris in April, and presented a paper on Radio Time Signals.

1950

The Report here presented refers to the period from 1949 May 1 to 1950 April 30 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

(b) Abinger.

A new septic tank has been supplied. At “Feldemore”, where the water supply had not proved satisfactory, a new system has been installed.

II. Instruments, Apparatus, and Library.

The following instruments are on loan to the Royal Observatory: –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope: – clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock.

To the National Maritime Museum: – Harrison's Time Machines Nos. 2 and 4, and the copy of No.4 by Larcum Kendall.

The removal of the Chronometer Department to Herstmonceux and pressure of work have prevented any further progress in the repair of Harrison's Time Machines Nos.1 and 3.

Photographic Zenith Tube.

In the previous report it was stated that all the main features of the instrument had been decided. The current year has been mainly occupied in the considerable amount of detailed work required to translate the agreed general principles into the specific designs which must precede actual construction. This has included a general review and tidying up of previous proposals and decisions as well as the raising and solving of problems of detail that had not hitherto received consideration. Construction of the instrument has followed hard on the heels of the completion of various sections of the detailed drawings, and has made substantial progress. In the course of the detailed design work the opportunity has been taken of making a few minor modifications where improvement or simplification resulted. The following special points may be mentioned·-

1. When the details of the time-sequence of operations came up for final review, it appeared from further consideration of the various factors involved that the balance of advantage lay with a four cycle sequence, and not with a five cycle sequence. The drawback to continued successive application of a four cycle sequence is that, the initial orientation of the rotary being the same for all the stars of the programme, congestion of the stellar images can occur. This difficulty can be obviated by breaking the series part way through the programme at a point where a suitable time-gap between stars occurs.

2. A simple optical system has been added by means of which the chronographic record can be calibrated in terms of displacement of the stellar plate carriage. It is thus possible to determine the uniformity of magnification of the chronograph over its range, and also to determine from the chronographic record the absolute value of the speed of the carriage and thus to set the speed to the correct value.

3. Provision has been added for independent oil damping of the natural period oscillations of the suspended stellar plate carriage, since it was found after assembly that the damping by the link bar was insufficient.

4. The design of the rotary reversal mechanism has been modified. While the basic principle of the cable system has been retained, an entirely new mechanism for operating it by solenoids has been designed.

5. This in turn has led to a simplification of the rotary clamping mechanism, which is now withdrawn by a gravity operated mechanism applied to the reversing cables, so that simultaniety of action of the clamps is no longer required.

6. The mechanical design of the instrument has been dimensionally adapted at various points so that a specially devised system of safety devices can, if found desirable, be inserted.

7. The stellar shutter, which is to be external to the instrument, has been redesigned. The obstructing members are made segments of a cylinder which rotates about a horizontal axis. This form is simple in mechanical design and operation, while the obstruction of the aperture is symmetrical. An improved method of imparting motion of the desired rapidity has been devised. The previously intended operation by solenoids has been replaced. Spring action is used, the initial potential energy being converted into kinetic energy during the first half of the motion, and the kinetic energy reconverted to potential during the second half. Only the small residual supply of energy necessary to make good the frictional losses is supplied electrically. Trials of this arrangement with a full scale model have proved successful.

8. The details of the auxiliary electrical control and indicating systems have been worked out and completed.

9. Provisional designs have been prepared for the building to house the instrument and also for the separate building to house auxiliary equipment.

During the meeting of the British Association in Newcastle in September, 1949, a number of the components of the instrument which had at that date been completed were assembled for demonstration at the works of Messrs. Grubb Parsons. At a meeting of Section A a brief address was given describing the main principles and features of the instrument.

Small Transit.

Work on Small Transit C was completed during the year. On final re-assembly, it was found that the micrometer which had been used with this telescope was of such dimensions that a practicable focus-setting was quite out of reach, although it had been focussed and used successfully at Abinger before reconditioning; the explanation was found to be that the objective had been used backwards, in which position it gives over half an inch longer working distance and surprisingly good images. In order to use the lens correctly with this micrometer, it has been necessary to shorten the telescope-tube, and this has now been done. The instrument has been fitted with the “full-aperture” illumination system, designed by Dr. Atkinson, which has been found satisfactory in the Bamberg Broken Transit. A rheostat-control has been fitted whose position and direction of operation always appear the same to the observer, whether the instrument is reversed or not and whether it is pointed north or south.

III. Astronomical and Radio Observations.

Time Service.

Throughout the year the Time Service has been based upon astronomical observations made at Abinger, using the Bamberg Broken Transit instrument, and at Greenwich, using Small Transit B. At Abinger, 250 observations were secured on 144 nights, and at Greenwich, 187 on 134 nights, making a total of 437 observations during the twelve months covered by this report. Owing to various staff changes, some experienced observers are no longer available, and new observers are being trained to take their places. Transit C, whose over-haul, has now been completed, will be installed at Greenwich, in place of Transit B, which will be transferred to Abinger, where it will be used for additional time determinations and for training.

For the operation of the Greenwich Time Service, the transit observations are referred to clocks of the quartz crystal type which exhibit a stability of rate considerably superior to that of the best pendulum clocks. Intercomparisons between the quartz clocks themselves indicate the extent to which the transit observations should be smoothed in assessing clock performance. For the past few years it has been suspected that some systematic and quasi-systematic errors were present in the observations, which showed persistent deviations extending over periods of several months. Sufficient information having accumulated to permit a thorough investigation to be carried out, the presence of an annual fluctuation in the period of rotation of the Earth was clearly shown, and its magnitude was reliably determined. The effect is a persistent one, though somewhat variable in phase from year to year. In the course of this investigation the material was examined for evidence of any rather sudden changes in the rate of rotation. No large changes were discernible, and it is doubtful if small changes would be detectable with transit observations of the present standard of accuracy.

The annual variation in the Earth’s rotation is such that the Earth is slow in the spring and fast in the autumn, relative to uniform time, by about 0.06 sec. The amplitude appears to be sensibly constant from year to year. Corrections for this effect are now being made not only in the retrospective assessment of clock performance, but also in the prediction of clock error and rate for the current operation of the time service. Corrections for polar motion are also applied, based on the provisional results of latitude variation observations at the U.S. Naval Observatory, Washington, which are supplied weekly by courtesy of the Superintendent of the U.S. Naval Observatory. The radiated time signals and time signal reception times are based on Greenwich Mean Time, as at present defined; but frequency estimation and assessment of clock performance are referred to a time system freed from the effects of polar variation and annual fluctuation.

Radio time signals from France, Russia, and U.S.A. have been regularly observed throughout the year, and from Australia and Germany since October. Reception of the American and French time signals at 20h U.T. was discontinued in September. Daily frequency comparisons have been made between the Observatory standards and the transmissions from Abinger (GMT, 2 mc/s) and Droitwich (200kc/s). The rhythmic time signals, radiated from the long-wave Rugby transmitter G.B.R., and associated shortwave transmitters, have been controlled from Abinger. These signals are derived from the new phonic motors described in the last report, which are now in full operational service. Two of the motors are running at Abinger, arranged as main and reserve transmitting motors, with fully automatic change-over from main to reserve in the event of a stoppage or failure of the main motor. The six-dots time service to the B.B.C. and other organizations, and the hourly signals used for the control of the G.P.O. Speaking Clock, have been controlled from these motors since October. A third motor of this type is running at Greenwich, and has provided a stand-by service since April. The transfer of the G.P.O. and B.B.C. signals from Dent regulators, under the control of Shortt clocks, to phonic motors, under the control of quartz clocks, concluded the history of the long and useful service rendered in the operation of the Time Service by the Shortt Free Pendulum clocks.

Because of the delay in concluding arrangements for the introduction of a mean time signal, the pre-signal series of dots was changed in September, as an interim measure, from rhythmic to mean time spacing. This series is radiated for four minutes, and has proved of great convenience to establishments where rhythmic contacts are not available.

The routine preparation and distribution of the various Time Department publications have been continued. The Time Service Bulletin, which contains reception times of all radio time signals received at the Observatory, has been published for each month of 1949. Quarterly supplements, containing reception times of the WWV time signals of the U.S. National Bureau of Standards, have also been distributed. The reception times of the British short-wave time signals, given to three decimal places of a second, have been forwarded to the Bureau International de l’Heure. The Time Service Circulars, which have been issued monthly, now contain frequency estimates based upon a uniform time system, and provide valuable information for the current assessment of the performance of frequency standards. The distribution of the Time Service Weekly Frequency Estimate has been continued, and the circulation slightly extended. Some progress has been made in the preparation of material for the volumes of Greenwich Observations. The material for the years 1937, 1938 and 1939 has been prepared, and an amended lay-out has been adopted for the period since the transfer to Abinger in 1940.

By the continued co-operation of the Post Office Radio Branch, the National Physical Laboratory, the B.B.C. Receiving Station, Tatsfield, and the Research Division, Marconi’s Wireless Telegraph Co. Ltd., regular information is received concerning quartz clocks and frequency standards at Dollis Hill, Banbury, Teddington, Tatsfield, and Great Baddow.

Of the clocks at Abinger, Bl and B2 continue in service as main and reserve standards for the control of outgoing signals. B3 has been temporarily removed. In the C and D Groups, the best performances have been shown by C3, Dl and D3. The E group has been disappointing, and the three oscillators have been withdrawn from service. The clocks at Greenwich, with the exception of F2, have run well in the periods between the somewhat frequent stoppages and interruptions; Fl has become the first Observatory clock to be used as a long-term primary standard. In addition, Al, A2 and F3 have each been used for short-period work. A3 continues to exhibit an excessive ageing drift. F2 crystal was removed for examination, and a minute crack was eventually found in one corner; it has accordingly been discarded. The German quartz clock, PTR VIII, has been installed, but after a number of stoppages it has been found that gradual ageing of components had occurred in the original buffer-amplifying stages; pressure of other work has temporarily stopped work on the clock.

The Dollis Hill standard 9A has shown a very high degree of frequency stability throughout the year, the maximum deviations from a linear frequency drift being ±3 parts in 10⁹, 9B has been withdrawn from service. 9C and 5C have maintained their former mediocre performance, while 5B, although not attaining the excellence of 9A, has nevertheless provided useful information for the major part of the year. Two new standards, EA and EB, which employ ring-crystals suspended by threads, have been brought into operation. The phenomenon of ageing is noticeably diminished with this form of crystal, and these clocks are already of great value in the operation of the Time Service. Information relating to the Post Office standards at Banbury is now communicated to the Observatory. Of the four oscillators originally available, Green has since been withdrawn from service; Red has been fitted with a suspended ring-crystal in place of the GT-cut plate previously employed; Blue, a mains-operated oscillator, is used as a subsidiary standard; Black has performed well until recent months, when it deteriorated, possibly owing to unavoidable mechanical disturbances.

At the National Physical Laboratory, a number of changes have been taking place as a result of the installation of new standards. Information relating to clock B is received regularly, but several erratic changes are apparent, and the clock is not utilised in the Time Service. The results are of great value, however, in the study of accurate reception of long-wave time signals, since the errors of reception at Teddington are remarkably small.

The standards at Tatsfield and Great Baddow are used primarily as frequency standards and not as clocks, and are employed at the Observatory to provide a completely independent check on the various services maintained. The communication of the monthly results from these stations is much appreciated. At Tatsfield, the GBR time signals are used to establish absolute frequency, which is then compared with the GMT frequency transmissions and with the American time and frequency transmissions from ,WWV. The weekly and monthly frequency estimates, supplied by the Time Department, are applied, and so these data receive an independent check at Tatsfield. The work at Great Baddow is largely concerned with the development of accurate quartz-crystal oscillators; the results communicated to the Time Department permit the other standards available to the Time Service to be compared with the high-precision 5mc/s. oscillators of the Marconi Company. Though the oscillators are at present running in the laboratory without room-temperature control, the standard of performance has been comparable with that of many of the best of the 100 kc/s. crystals running under better controlled conditions.

The operation of the quartz clock installation at Abinger has shown the need for the greatest degree of independence between individual oscillators. The present arrangement, whereby oscillators are mounted on racks in groups of three, is being discarded in favour of individual mounting on brick and concrete piers. As these changes are made, the opportunity is taken to rebuild the maintaining amplifiers, to modify the temperature control unit, and to substitute an improved form of crystal oven. Three piers have been constructed in the now vacated E group cellar, and extensive re-wiring has been carried out. This cellar will accommodate one crystal transferred from D group, the B3 adjustable oscillator, and a new suspended ring- crystal which has just completed initial tests at Dollis Hill. The other three cellars will be modified in succession. Work on the development of improved maintaining amplifiers and temperature control units has made some progress during the year, and the possibility of employing magnetic amplifiers is being investigated.

A new service of standard frequency transmission has been introduced on a restricted scale. The service is sponsored by the Department of Scientific and Industrial Research, and operated by the Post Office. The transmissions from Rugby MSF at 1030 to 1045 each day on a frequency of 60 kc/so are measured at Abinger, and the results communicated daily to the Post Office Radio Branch. To facilitate these comparisons, a new standard frequency receiver is being built, in which the incoming frequency is converted to 100 kc/s. for convenient comparison with the local standards. Transmissions are also made on frequencies of 5 mc/s. and 10 mc/s., and for the reception of these and other standard frequency transmissions, a new form of standard frequency receiver has been developed and the prototype is nearing completion.

The equipment designed for accurate measurement of the reception times of signals radiated by the Rugby and Annapolis transmitters was completed in time for the proposed series of trans-Atlantic measurements for the determination, in cooperation with the U.S. Naval Observatory, of the velocity of radio propagation. In order that the equipment used on both sides of the Atlantic should be identical, an additional receiver was constructed and sent to America. A new form of photographic registration was developed, and was found to give very clear records. A film projection unit was also constructed to facilitate measurement of the records. Unfortunately the reception of the Rugby signals at Washington proved unsatisfactory, and the programme has been postponed.

Various ancillary items of equipment have been completed for the long-wave receivers, and one of the ‘frame aerials, which was found to have deteriorated, was completely dismantled and rebuilt. Some new chronograph units have been made, and will be used in the observing huts at Abinger. A series of tests have also been carried out on receivers of standard Admiralty pattern in order to determine their suitability for the reception of time signals by survey vessels.

XI. General

At the invitation of the Societe Chronometrique de Suisse, Mr. H. M. Smith attended the conference of the Society at Geneva in August, and read a paper.

1951

I. Grounds and Buildings

(b) Abinger. A new generator house has been erected to house the emergency power supply equipment. A bathroom has been added to the Caretaker’s residence.

II. Instruments, Apparatus and Library.

The following instruments are on loan to the Royal Observatory –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory:–

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock.

To the National Maritime Museum: – Harrison's Time Machines Nos. 2 and 4, and the copy of No.4 by Larcum Kendall.

To the South Bank Exhibition of the Festival of Britain: – The Airy Sidereal Standard Clock [Dent 1906] and one Time Signal Transmitting motor.

Further progress has been made on the repair of Harrison's Time Machines No.1. Mr. D. W. Fletcher, Superintending Civil Engineer, H.M. Dockyard, Chatham, is keenly interested in this horological work and is voluntarily undertaking a substantial part of it. No further progress has yet been made on No. 3.

Photographic Zenith Tube

The bulk of the constructional work on this instrument has been completed. In the near future a preliminary complete assembly will be made so that various tests of performance can be carried out before the components are finally cleaned and painted. One point of uncertainty which can only be settled by test in situ is whether or not the stellar-plate carriage must be clamped during reversal of the rotary; it is hoped that this will be unnecessary, but provision has been made for introducing clamping if required. Certain features of the work carried out during the current year may be briefly summarized.

The control and indicating console, the chronographic system, and the carriage drive mechanism have been completed. Trial of the clutching system showed the necessity of providing a larger impulse than had been anticipated in order to keep the time-uncertainty of operation within the tolerance. A new mercury floating basin has been constructed of similar design to that of the new U.S.A. instrument, and tests at Abinger have confirmed the efficacy of floatation as an anti-vibration device. The mounting of the tilted focussing plate has been designed so that it is adapted for use with the standard plate holder. The plate itself is located" in defined geometrical relation to the normally used horizontal plate by being supported on two knife edges.

The auxiliary framework which surrounds the telescope and supports the synchronous motors, the reversing mechanism and other components, was originally planned to be supported on four pillars let into the concrete pier. With this arrangement problems of the relative positioning of parts arose. It was decided, after inspection of a temporary partial assembly of the instrument, that it would be preferable to enlarge and strengthen the sole plate and to support the pillars from it. The whole instrument thus becomes self-contained. A rearrangement of the essential features of the building has also been made. The relatively small circular platform for the operator has been enlarged into a complete floor, with a trap door for access to the mercury basin and with sections removable for cleaning purposes. The shutter and the small derrick for lifting components are to be mounted on girders supported by the end walls of the building. Detailed designs for the support of the derrick and shutter have been prepared.

Further attention has been given to the question of measuring machines for the P.Z.T. The results of tests made in the Laboratory at Abinger and of tests made subsequently by the N.P.L. show that the machine which it had been hoped to use for measurements of the stellar plate is not suitable either in design or construction to give the accuracy desired, though it will probably suffice for the measurement of the chronographic plates in which optical premagnification is used, and where orientation is not critical; it will thus meet a requirement which was not foreseen until the photographic chronograph had been developed.

The design is under consideration of a special machine for the stellar measurements, which is to satisfy the following requirements:-

The machine is to be capable of measurement in the x-co-ordinate only, measurements in the y-direction being derived from measurements made in the x-direction after the plate has been transposed by rotation through a nominal angle of 90°. Translational motion of the plate (without measurement) in the y-direction is to be provided. An advantage of such a design is that the error of angle between the ways need not be explicitly determined since it appears always in conjunction with the arbitrary orientation of the plate. This error is not only not easy to determine with precision but is liable to variation which would necessitate frequent redetermination. The critical quantity is the angle of transposition, and it is proposed to determine this for each plate by autocollimation on a 90° stainless steel prism mounted on a turn-table which carries the plate. The angle of the prism, which should in fact be very stable, can be pre-measured or checked at any time on the bench. As the means of measurement a calibrated scale with a comparator system employing two microscopes is to be used, instead of a screw.

If it is decided to use a visual method of observation, consideration has been given to the possibility of employing superposed fields of view and securing exact superposition of image and a scale division by the use of a single optical micrometer as an alternative to the use of separate fields and two micrometers. The possibility of replacing visual observation by a physical microphotometric criterion is, however, under consideration. Special attention will be given to the mechanical constraints to secure the necessary coplanarity and collinearity of essential components.

Small Transit C

Work on this instrument was described in the previous report. On trial, it was found that the new illumination system was not bright enough, and it was necessary to remodel the source of light and make other changes. This work has been completed. Before bringing the instrument into use, various ways of eliminating the present procedure for obtaining the level were considered; the continual placing of a striding-level onto the pivots wears pits in them, besides apparently disturbing the level itself, and a permanently-hung level was clearly preferable if it could be arranged. It was finally achieved by mounting the level rigidly on the horizontal axis of the telescope itself, so that it always reverses with the instrument; it follows that it cannot be read unless the telescope is pointing near the zenith, but there appears to be no disadvantage in this. The instrument was mounted in the old Small Transit pavilion, which had been overhauled and re-wired where necessary, and the level-determinations show a definite improvement on the old method, besides being quicker to obtain; parallel and simultaneous observations with the small transit instrument in the Altazimuth pavilion were on the point of starting at the end of the report year.

III. Astronomical Observations

Time Service

Throughout the year the Time Service has been based upon astronomical observations made at Abinger, using the Bamberg Broken Transit instrument, and at Greenwich, using Small Transit B.

At Greenwich 320 observations were secured on 166 nights, and at Abinger, 249 on 123 nights, making a total of 569 observations during the twelve months covered by this report. Transit C, referred to in last year’s report, has been set up in the Small Transit Pavilion at Greenwich, and is undergoing final adjustments. It has not yet therefore been possible to release Transit B for return to Abinger, which station has been without a second instrument since January 1948. The levels used with the Bamberg and Small Transit instruments were recalibrated by an optical method in the laboratory at Abinger.

The adopted longitude of the Small Transit mounting in the Altazimuth Dome at Greenwich was changed from 0s.150 E to 0s.145 E on July 1 as a result of a special local survey.

In order to provide as uniform a time system as possible for the assessment of performance of the standard quartz clocks, corrections have been applied to the observations to minimize the effects of the annual fluctuation of the Earth’s rotation and of the polar motion. The effect on time determinations of the polar motion has been estimated, as in previous years, from the current latitude variation observations of the U.S. Naval Observatory, made with the photographic zenith tubes at Washington and at Richmond, Florida, the results being forwarded weekly by air-mail.

The astronomical observations, corrected for polar motion and annual fluctuation, are referred to quartz clocks at Abinger, Greenwich and elsewhere. The silk-suspended Z-cut quartz ring oscillators introduced at the Post Office Research Station, Dollis Hill, in 1948 and 1949 have continued to show a marked superiority over the GT-cut quartz plate type. Two quartz ring oscillators are now in use at Abinger and have already shown themselves to be better than any previous Abinger clocks. The first of these, which was installed in May, had practically no ageing effect and rapidly settled to a frequency drift that was indistinguishable from zero, thus greatly enhancing the value of the clock. By the end of 1950, after barely six months of operation, the clock was brought into use as a primary long-period standard. The second quartz-ring was installed in September, and had settled to a practically constant drift in frequency by the end of 1950; it is now in use as the reserve Abinger standard.

The modification of the clock cellars at Abinger from rack-mountings to single pier-mountings is now more than half completed. One cellar is in full operational use, with two quartz-ring oscillators and one adjustable-frequency GT-plate oscillator. A second cellar is in partial use, with one adjustable-frequency GT-plate oscillator and two oscillators temporarily fitted with ordinary GT-plate crystals, until the next two quartz-rings are delivered. The third cellar is in process of modification, and should be in use before the end of 1951. The remaining cellar contains the last rack-mounted GT-plate oscillators, and will be modified to single-pier mounting in due course for the purpose of housing oscillators of experimental design during tests. The three adjustable-frequency oscillators, from which are selected the transmitting standards, were previously kept as a complete group; there was a risk of disturbance of them all whenever adjustments were made on anyone of them. In future, they will be dispersed among the three operational cellars.

The service of International Time Signals and of Six Pips Time Signals broadcast by the B.B.C., together with the Hourly Signals used by the Post Office for the control of the Speaking Clock, has been maintained from the phonic motor transmitters at Abinger.

Since October 1 a five-minute series of dots, each of approximately one-tenth of a second in duration, and at intervals of one second, have been radiated at 0955–1000 and 1755–1800 G.M. T. daily. These signals are convenient for checking the performance of high-precision quartz-crystal controlled standards of time and frequency. The rhythmic series formerly sent at these times is still required by users for checking chronometers by the method of coincidences, particularly for navigation and for field survey work. The rhythmic signals are therefore radiated at 1001–1006 and 1801–1806 G.M.T. daily. The transmission of the mean time signal before the hour and rhythmic signal after the hour conforms with international usage.

The service of radio time signals controlled by the Royal Greenwich Observatory is designed to provide reasonable world-coverage, and is accordingly transmitted on long waves from Rugby (GBR, 16 kc/s) and on associated short-wave transmitters selected to meet summer and winter conditions. Both the mean time and rhythmic series are radiated in this way, thus providing users in most parts of the world with a choice of suitable signals on at least one wave-length. The preliminary signals for determination of line-1ag continue to be radiated at 0946–0950 and 1746–1750. Arrangements are being made for the long-wave transmitter GBZ to act as reserve in the event of the non-availability of the GBR transmitter.

The reserve service of Hourly Signals to the Post Office and Six-dots Signals to the B.B.C. was maintained from Greenwich throughout the year.

By the continued co-operation of the Post Office Radio Branch, the National Physical Laboratory, the B.B.C. Receiving Station, and the Research Division, Marconi’s Wireless Telegraph Co. Ltd., regular information is received concerning quartz clocks and frequency standards at Dollis Hill, Banbury, Rugby, Teddington, Tatsfield and Great Baddow.

The clocks at Dollis Hill have continued to serve as long-period standards for use in the Time Service. Interest in the older quartz-ring oscillators has declined somewhat with the advent of the silk-suspended rings described last year, though they still have great value as long-period standards. At Banbury, the usefulness of the clocks has been marred by a long series of troubles in the power supply arrangements. The standards at Rugby were installed for the control of the new service of standard frequency transmissions from MSF, and include one silk-suspended ring of very high stability. The necessity for occasional re-adjustment of the frequency of this oscillator impairs its value as a long-period standard clock, but it provides a powerful control on frequency variations at all stations receiving the transmissions.

At the National Physical Laboratory a new standard fitted with a silk-suspended quartz-ring was brought into use in May, and has displaced the GT-plate oscillator as the primary quartz standard. It has further increased the value of the Teddington reception of the GBR time signal as a completely independent indication of the erratics of the transmitted signal. Similar independent checks upon the Time Service are provided by the regular comparisons made with the local standards at Tatsfield and at Great Baddow, as described in the previous report. The German quartz clock

PTR VIII has been set going at Abinger, and regular comparisons with the other Abinger clocks have been carried out since June 27. The performance appears comparable with that of a normal GT-plate oscillator.

Radio time signals from Australia, France, Germany, Russia, and the U.S.A. have been regularly received throughout the year. Daily frequency comparisons are made between the Observatory standards and the transmissions from Rugby (MSF, 60 kc/s) and Droitwich (200 kc/s).

The routine preparation and distribution of the various Time Department publications have been continued. The Time Service Bulletin containing reception times of all radio time signals received at the Observatory has been published for each month of 1950, together with the quarterly supplements giving reception times for the WWV time signals of the U.S. Bureau of Standards. The Time Service Circular containing frequency estimates referred to a uniform time system has been published about the middle of each month for the preceding month. The Time Service Monthly Frequency Estimate, prepared at the same time,’ gives direct assessment of the predicted, current and revised frequencies of selected quartz crystal standards, and is sent to the Post Office Laboratories at Dollis Hill and Banbury. The Time Service Weekly Frequency Estimate was discontinued on September 30, as its primary purpose was to supplement the GMT frequency transmissions from Abinger (2 Mc/s), which ceased on that date consequent upon the inauguration of the MSF transmissions. The material for the volumes of Greenwich Observations has been prepared up to the end of 1949.

In the Electronics Laboratory a considerable amount of design and experimental work has been carried out on new types of crystal drive circuits working with extremely low amplitude of crystal oscillation. Three separate forms of circuit have been subjected to careful test and analysis, and the most promising design is now being constructed in prototype form. Some delay is being experienced in the delivery of special high-stability components, but it is hoped that the experimental oscillator will be in operation by the end of the year. Similar delays have arisen in the development of an improved type of quartz crystal oven: the oven control circuit, which is operated from battery supplies, has already been tested and found satisfactory.

To facilitate the careful checking of the performance of the quartz crystal drive circuits, an accurate phase-meter has been constructed, which is capable of measuring phase variations of less than one-hundredth of a radian. It has proved of the greatest value in the checking and adjustment of the rebuilt amplifiers of Post Office design. A new form of regenerative divider has been developed, and the prototype is performing satisfactorily. It is proposed to replace all the dividers at Abinger and Greenwich with the new pattern. A special receiver has been built and installed at Abinger for the comparison of the standard frequency transmissions from Rugby (MSF, 60 kC/s) with the Observatory standards. Messrs. Standard Telephones and Cables, Ltd. kindly gave a 60 kc/s quartz crystal filter for use in the new receiver.

A new type of electronic pulse divider, converting from 100 kc/s to 1 pulse per mean time second, has been developed, and is undergoing reliability tests. The same units can be arranged to delay a pulse by any desired interval within the range 0s.0000 to 0s.9999, in steps of 0s.0001, the delay time being selected by means of four decade switches. It is also possible to arrange the units to give pulses at intervals of one sidereal second. Another use for the pulse divider is as a new form of decimal counter chronometer, which besides having certain advantages over the type now in use in the Department is also considerably more compact. A complete decimal counter chronometer of this type has been built for the Control Room at Greenwich.

The programme of trans-Atlantic measurements of radio propagation velocity, referred to in the report last year, was finally abandoned, owing to the impossibility of securing an adequate standard of reception of the Rugby radio signals at the U.S. Naval Observatory. The special long-wave receiver, which had been constructed at Abinger, was returned to this country, and has now been overhauled and prepared for installation at Greenwich. The programme for the completion of the Greenwich control equipment is thus making slow but steady progress.

The special standard frequency receiver prototype was completed and put into operation, but tests showed that interference between the WWV transmissions on 5 Mc/s and 10 Mc/s and those on the same frequencies from Rugby made accurate measurement difficult.

VI. Chronometers.

The Shortt Free Pendulum No. 40 with its precision Slave Clock has been mounted in a temperature controlled clock room under the West courtyard to provide a more accurate time standard. A mean solar dial is now operated in the Solar dome and an hourly signal is transmitted to the Solar Office to afford a reliable time base on the recorder for the Sudden Phase Anomaly Indicator.

In addition to the normal repair work the Cooke regulator from the Solar Dome, the Shelton regulator from the Abinger Magnetic Station and the Free Pendulum and Slave Clock (Shortt 40) were overhauled in the Chronometer Workshop. The Dent Sidereal Clock with chronometer escapement, formerly used for many years as the sidereal standard at the Royal Observatory, was thoroughly overhauled at Herstmonceux in April. The clock was subsequently installed at the South Bank site of the Festival of Britain, where it forms part of the exhibit showing the development of time standards.

1952

The Report here presented refers to the period from 1951 May 1 to 1952 April 30 and exhibits the state of the Observatory on the last named day.

1. Grounds and Buildings

(a) Greenwich. As mentioned in last year’s report, the Wren building has been taken over by the Ministry of Works, and will be treated as an Historic Monument.

(b) Abinger. The walls of the Time Department Building and of the Laboratories have been treated with a water-proofing solution. All unpainted exterior woodwork at Feldemore has been treated with wood preservative.

(c) Herstmonceux. Preliminary plans of the building to house the Time Department, the Nautical Almanac Office, the Workshops and to provide storage accommodation have been prepared by the Architect as a basis for discussion.

II. Instruments, Apparatus and Library.

The following instruments are on loan to the Royal Observatory –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock.

To the National Maritime Museum: – Harrison's Time Machines Nos. 2 and 4, and the copy of No.4 by Larcum Kendall.

Work has been completed on Harrison's Time Machines No.1. This work was undertaken without charge by Mr. D. W. Fletcher, lately Superintending Civil Engineer, H.M. Dockyard, Chatham, has been carried out with great skill. When Lieut. Commander R. T. Gould restored this timepiece, he left the temperature compensation inoperative; Mr. Fletcher has succeeded in bringing it into action and has made various improvements. No further progress has yet been made on No. 3.

Harrison's Time Machine No.4 and Larcum Kendal copy, shown as on loan to the National Maritime Museum, have been examined and cleaned in the Watch Repair Shop at Herstmonceux, These instruments, together with Harrison's No. 1, will be loaned to the special exhibition 'The British Clockmakers' Heritage' to be held at the Science Museum, 1952 May to September, after which they will be returned to the National Maritime Museum.

Photographic Zenith Tube.

The instrument, together with the new sole plate and supporting pillars, has been assembled in the works of Sir Howard Grubb Parsons & Company for tests of performance to be made. One result of these tests was that clamping of the stellar plate carriage during reversal is not necessary.

Some trouble was at first experienced with the reversing mechanism, but this was overcome in due course. Another trouble was vibration of the stellar plate carriage as a whole. This was traced to the synchronous motors, and was much enhanced when, as frequently occurred at this stage, one of the motors settled into an asynchronous type of running. This trouble was cured, and the amplitude of vibration was much diminished by additional bracing of the supporting pillars. The motors, clutches, and the rotary reversing solenoids have been enclosed and air will be drawn past them to minimize any temperature effects.

At various stages of the tests chronographic records were taken both for single runs and for complete cycles which, together with observations made with the reversal autocollimator, provide checks on the mechanical performance. Measurements of these photographic records were made both in the laboratory at Abinger and on the Zeiss machine at the National Physical Laboratory. The results were generally satisfactory; in the meantime the instrument was dismantled for the carrying out of various finishing processes.

A minor alteration to the rotary, which had appeared for some time to be desirable, has been effected, providing easy access to the rotary stops so that they can be readily removed for servicing. Before this alteration such access would have been possible only after extensive dismantling of the instrument.

The stellar shutter has been completed. Tests show that its speed of operation and degree of repetition fulfil the requirements laid down. Construction of the tilted focussing plate and of the auxiliary equipment for controlling the height of the mercury basin have also been completed.

The control and indicating console has been delivered to Herstmonceux: delivery of the instrument itself, now contingent on the finishing processes, should not be long delayed.

Measuring Machine for P.Z.T. stellar plates.

Considerable further work has been done on the design of this machine, and to assist development and discussion a full scale model has been constructed. The optical system has been so arranged that settings can be made either by means of a microphotometric criterion or visually. Construction of the auxiliary electrical equipment required for the microphotometric method has been completed; when delivery of the necessary photocells has been made it is proposed to carry out an adequate bench-test of the method before proceeding to actual construction. The microphotometric method has the following advantages:

(a) Relative ease of observation and absence of fatigue and consequently, other things being equal, probable superiority in speed of operation.

(b) An appreciable superiority in precision, especially for faint or ill-defined images.

(c) Superiority in sensitivity. Since measurement is possible at lower contrast between image and background than with a visual setting, it is possible that the measurable range of stellar magnitude may be extended.

Although the use of a microphotometric method, based on the scanning method due to Dr. Furth, had previously been considered, the decision to adopt a microphotometric criterion was made after seeing the Double-Trace Scanning Microphotometer shown by Dr. Rymer of Reading at the Physical Society Exhibition of Scientific Instruments in 1951: its essential features appeared to be suitable to the present purpose. It has the advantages of being less cumbrous, more flexible, and superior in the method of setting. On a cathode ray tube two traces are formed, one of which is the mirror image of the other. The criterion of setting is coincidence of these traces, which is preferable to the setting of a single trace to a fixed fiducial line. Shortly afterwards Dr. Rymer very kindly demonstrated the instrument at Reading, discussed its application to our purpose, and provided details of his circuits.

In considering the design of the measuring machine, based on this principle, it was clear that the system of employing superposed fields of view, referred to in last year's report, was ill adapted to the microphotometric method, and was therefore abandoned; and that the proposed use of a ruled glass scale as the basis of measurement was readily adaptable to a compensating optical system which eliminates the effect of any residual drift in the position of the source of light.

An experimental optical system was set up in the Optical Laboratory to test various points in practice, including the micrometer system which is of the rotated plate type.

One feature of the mechanical design which, from the operational point of view, is a great improvement on the previous instrument is the use of a turn-table to rotate the plate holder and fiducial optical square through 90°. Previously the square plate holder had to be lifted by hand clear of the machine and replaced after rotation.

The turn-table is located on the Y-carriage and moves between two adjustable stops which are provided for easy and rapid rotation of the nominal 90°. The actual angle of rotation is measured for each stellar plate with high precision on the autocollimator.

Another proposal tending to ease and safety of operation is the provision of a bench jig which, by using a simple optical projection method, will enable the plate to be lined up before its holder is located on the turn-table.

At various stages during the development of the design much help and encouragement has been given by discussions with the National Physical Laboratory. The construction of the model has been facilitated by the recent installation of a small lathe in the laboratory.

III. Astronomical and Radio Observations.

Time Service.

Throughout the year the Time Service has been based upon astronomical observations made at Greenwich using Small Transit B and at Abinger using the Bamberg Broken Transit instrument.

At Greenwich 201 observations were secured on 160 nights and at Abinger 308 on 135 nights, making a total of 509 observations during the twelve months covered by this report. Transit C was included in the equipment sent to the Sudan for the observation of the eclipse of February 25, so that the position of one of the observing stations could be accurately determined. It was dispatched on December 4 and has not yet been received back. Routine re-calibration of the Bamberg level has been carried out twice during the year and no change in scale value has been detected.

The analysis of the performance of the clocks has been continued and has provided strong evidence that the annual fluctuation in the rate of rotation of the Earth is subject to considerable changes in phase and amplitude. The values obtained for 1950–51 are significantly smaller than those given by many previous concordant analyses. It is consequently not practicable to apply corrections month by month to the time determined astronomically so as to produce a uniform time system. The best quartz clocks must be accepted as defining a time system that is approximately uniform while the astronomical determinations of time over two or three years are used to determine for each clock the error at a given epoch and the rate. If the annual fluctuation were constant its effects could be eliminated by comparing monthly mean errors of each clock, separated by an interval of twelve months; the true annual rates of each clock would then be determined. This technique is used as a control over the annual fluctuation but is only applicable to clocks of the highest precision with uniform runs extending over two or three years. Determinations of the annual fluctuation over a twelve-month period are now made every three months. As in previous years, the effect on time determination of the polar motion has been estimated from current latitude variation observations of the U.S. Naval Observatory. Comparison of the provisional corrections so determined with those based on the international latitude variation data, which are not available until much later, has shown satisfactory agreement.

At Abinger the quartz ring oscillator E5 has continued to show a high standard of stability. E6 is also operating satisfactorily, with a .steady decrease in the rather large frequency drift. Two new ring crystal oscillators were installed in August as C5 and C6. The performance of C5 has been fair but that of C6, which had been equipped with long-life valves, was disappointing; minor modifications were made to the circuits of C6 in March and a considerably higher standard of short-term stability has resulted: it is too early to decide whether the long-term stability will show a corresponding improvement. If it does, similar modifications will be made to the circuits of C5. The modifications in B cellar were completed in August and the oscillators were set going with GT-plate crystals. It is proposed to replace the crystals in B5 and B6 with ring crystals when they are available from the Post Office Radio Branch Laboratories.

A three-phase 45 KVA diesel-driven AC generator has been installed and is fitted with automatic starting equipment which brings it into operation in the event of failure of the mains electricity supply. The 25 volt battery which operates a motor generator providing a 50 cycle AC supply for the quartz crystal thermostats in the event of mains failure has been replated.

At Greenwich two brief interruptions occurred in the HT supplies to the clocks, but no change of rate of the oscillators occurred. In November oscillator A2 was stopped while changing the main amplifier and buffer valves, causing a change of rate. In November the rate of A3 oscillator was re-adjusted and in December there was a violent change of rate of oscillator F3 for no apparent cause.

Twelve new regenerative dividers of the type described in last year’s report have been made commercially to the pattern of the prototype made in the Electronics Laboratory. The three regenerative dividers in the Greenwich installation have now been replaced by the new type, and two further dividers have been added: work on the replacement of the dividers at Abinger is in hand. Two long-wave receivers and two decimal counter chronometers have been set up at Greenwich. One of the decimal counter chronometers is of the type developed in the Electronics Laboratory; the other is of the commercial type previously employed, modified to incorporate gate and trigger units of the new pattern. A transmitting motor, together with ancillary circuits for the transmission of the six-pips and the hourly signals, has been installed and a number of send relays with a control panel and line-meter panel are being installed. All this equipment has been made up at Abinger in complete rack units for ease of erection and wiring, and has been designed to facilitate its removal to a new site if and when required.

By the continued co-operation of the Post Office Radio Branch, the National Physical Laboratory, the B.B.C. Receiving Station, and the Research Division, Marconi’s Wireless Telegraph Company Ltd., information is received concerning quartz clocks and frequency standards at Dollis Hill, Rugby, Teddington, Tatsfield, and Great Baddow.

The clocks at Dollis Hill have continued to provide information of great value in the operation of the Time Service. The ring crystal oscillator 9A maintains its high standard of performance and other ring crystal oscillators continue to serve as long-term standards. The new ring crystal oscillator, Ql3, at the National Physical Laboratory is running satisfactorily and now provides useful data for the Time Service.

The International Time signals, the six-pips time signals, and the hourly signals used by the Post Office for the control of the Speaking Clock, have been sent out regularly from the phonic motor transmitters at Abinger. One of the Rhythmic Phaseable Phonic motors has been converted to provide seconds contacts. The reserve service of hourly and six-pips signals has been maintained from Greenwich.

The preparation and distribution of the various Time Department publications have been continued. The Time Service Bulletin containing reception times of all radio time signals received at the Observatory has been published for each month of 1951. Three supplements to the Bulletin are also published. Supplement A contains the G.M.T. reception times of the WWV time signals of the U.S. National Bureau of Standards. Supplement B gives the G.M.T. reception times of the Rugby GBR radio time signals and corrections for the effects of polar variation and annual fluctuation, thus permitting a direct comparison between the GBR signals and a provisional uniform time system. Supplement C gives the reception times of the MSF seconds pulses which are also referred to a provisional uniform time system. The Time Service Circular, containing frequency estimates referred to a currently estimated uniform time system, are published about the middle of each month for the preceding month. The Time Service Monthly Frequency Estimate, prepared at the same time, gives estimates of the predicted, current and revised frequencies of selected quartz crystal standards and is sent to the Post Office and the National Physical Laboratory. Some progress has been made in the proof reading for the volumes of Greenwich Observations whose publication had been delayed by the War.

The Maintenance and Installation Section at Abinger undertook some of the preparatory work for the Solar Eclipse of February 25. Six radio receivers were modified and re-aligned, special recording units were constructed and tested, and rhombic aerials made up. Three chronographs were constructed, and two others fitted with DC motors, in the workshop at Abinger.

A millisecond printing chronograph constructed by Messrs. Edouard Belin of Paris was delivered to the department for test.

At Greenwich the Time Ball was dropped to mark the opening of the Festival of Britain on May 3rd and then daily at twelve noon until work on the Octagon Room prevented its further use.

Details of the observations made at Abinger with the Bamberg Transit Instrument were transferred to punched cards for analysis by Hollerith equipment. As the equipment at the Nautical Almanac Office was in full use and not available for this work, a member of the Observatory staff carried out the machine operations with the Hollerith equipment of the Mathematics Division of the National Physical Laboratory, through the courtesy of the Superintendent of the Division. This Hollerith analysis has been supplemented by various investigations by conventional methods undertaken at both Greenwich and Abinger and the results are now being analysed and collated.

The higher standard of accuracy now demanded has made it necessary to employ rigorous computing procedure in cases where simpler graphical methods previously sufficed, and also to carry out the computing with an increased number of significant figures. Greater use is therefore being made of calculating machines. A Brunsviga and a Britannic machine have been supplied, while a National Accounting machine is held on loan from the Nautical Almanac Office in order that its value to the Time Department may be ascertained. It has already been employed for the analysis over a period of twelve months of the six-figure beat-counter readings at Abinger, for which it has proved particularly suitable.

The prototype crystal oscillator G1 designed in the Electronics Laboratory, referred to in last year’s report, has been in operation since November. It incorporates the GT crystal previously presented to the department by Messrs. Standard Telephones and Cables Ltd., which has since been resealed by the firm. The results obtained so far have been very encouraging. The temperature control circuits and thermostat are operating satisfactorily to within the very close limits demanded by the design, and the oscillator which maintains the crystal in oscillation at a low amplitude shows a very high standard of short-term stability. A second experimental oscillator G2 has been constructed and is in process of adjustment and installation.

Preliminary experiments have been carried out with a view to designing and constructing more accurate and convenient equipment for frequency .comparison and some constructional work is in hand. Further experience with the pulse dividers developed in the Electronics Laboratory has provided adequate confirmation of their accuracy and reliability. The installation of pulse dividers has now become a necessary operational requirement, as the scatter of the seconds pulses derived from the phonic motors limits the precision of clock intercomparisons and of the measurement of radio signals such as the seconds pulses superimposed on the MSF transmissions.

A number of minor items of equipment have been designed and constructed in the Electronics Laboratory. A reserve 60 kc/s standard frequency receiver has been installed, and a prototype for the seconds pulses of the MSF transmissions has been completed and is now on test at Greenwich. A noise receiver, constructed by Messrs. Morley and Dukes under the supervision of the Cavendish Laboratory, Cambridge, was tested in the laboratory and installed in the Solar Department at Herstmonceux. A test set for this receiver has also been made. The phase-anomaly received originally constructed in the laboratory and then converted to operate as a noise receiver has since been re-converted for the recording of phase anomalies and has been re-installed in the Solar Department. A chronograph drive unit was constructed for the Chronometer Department, and a low-frequency oscillator for the Dye coil magnetometer of the Magnetic Department. Considerable assistance has been given to the Optics Laboratory in the design of the electronic equipment for the P.Z.T. Photo-electric Measuring Machine.

The equipment of the laboratory has been augmented by the addition of a Cossor double-beam industrial oscilloscope complete with a camera attachment, two Furzehill oscilloscopes, a Marconi signal generator and a Marconi low frequency circuit magnification meter.

1953

The Report here presented refers to the period from 1952 May 1 to 1953 April 30 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings

(a) Greenwich. The Time Ball was brought into operation again on October 26th and has been dropped at 1300 daily since then, with occasional exceptions due to dangerously high winds, or to the need for minor adjustments.

The main office of the Time Department was transferred to the old Wireless Room in October, at which time the Control Room, which had been serving both functions, was repainted. Internal painting has also been carried out in the Astrographic dome, and on the staircase and landing leading to the new Time Department office and to the Sheepshanks dome.

(b) Abinger. Considerable inconvenience to the operational work of the Time Service has been caused by the large variations in the voltage of the mains electricity supply. As a result of discussions with the Electrical Engineering Manager, Chatham, and the Supply Authority, a static balancer has been installed as a temporary measure, pending modifications in the supply arrangements.

(c) Herstmonceux. Preliminary plans of the buildings to house the Time Department and Nautical Almanac Office, the Observatory Workshop, the Chronometer Repair Shop, and to provide miscellaneous storage accommodation and a works pound, have been considered. Some modifications to these plans were required and revised plans are in course of preparation. It is hoped that it will prove possible to accept a tender for construction by the spring of 1954.

II. Instruments, Apparatus, and Library.

The following instruments are on loan to the Royal Observatory –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory: –

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – A syncronome slave clock.

To the National Maritime Museum: – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No.4 by Larcum Kendall.

Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No.4 by Larcum Kendall. were loaned to the Science Museum for the special exhibition 'The British Clockmakers' Heritage', which was organised by the Antiquarian Section of the British Horological Institute to be held at the Science Museum, and open from the 23rd May to the 14th September, 1952.

An unprecedentedly large proportion of the apparatus sent out to the Sudan eclipse stations received damage. Castings have had to be welded, or entire fresh patterns and castings made, for two coelostats, both of which had been borrowed, and for the reversing gear of Small Transit C, which had been taken to the En Nahud station to determine its longitude; there was a good deal of minor repair work to be done on this transit in other respects also.

Transit C has been erected in the old Small Transit pavilion, on the “Bradley” meridian. This instrument employs the micrometer originally made in the Observatory Workshop for the “personal equation machine”. It has also been fitted with weightrelieving gear; the bubble from the striding-level, which was in use with Small Transit B, has been mounted directly on the central cube of the instrument, and reverses with it, so that there is now no tendency to wear flats on the upper surfaces of the pivots by placing the striding-level on them. Levels can thus only be taken with the telescope pointing to the zenith; but since the pivot-form has been made very good nothing would be gained by using other positions, and on the old system the readings were subject to a progressive risk of falsification as wear developed. A spare bubble has been similarly fitted to the cube of Small Transit B; comparison revealed no actual discontinuity in the readings, due to the change, but the new arrangement is quicker and much safer.

In October, a tape-chronograph was installed for use with the Airy Transit Circle. For some time previously this instrument had been sharing a chronograph with the Small Transit, which was inconvenient. The old Airy barrel chronograph, with the original conical pendulum drive, is still kept in use; the tape record is ordinarily used for measuring, as it has the longer scale, but the sheets of the barrel chronograph provide a much more convenient form of file, in case any subsequent reference to the chronograph records is required.

Photographic Zenith Tube

The various finishing processes referred to in the previous report have now been completed. Some portions of the instrument have been delivered to Herstmonceux and all will be received before the pavilion to house the instrument has been erected.

Measuring Machine for P.Z.T. Stellar Plates

The photocells required for the microphotometric method of setting have recently been delivered and it is thus hoped that investigation of the method will shortly be resumed.

Further consideration of the design has suggested the desirability of certain modifications of the previous proposals. The mechanical lay-out is much simplified if a single carriage is used as it is in the Zeiss measuring machine; the elimination of the Y-traverse carriage, the guiding ways of which needed to be of very high accuracy, considerably facilitates the mechanical construction. In order to employ the microphotometric method of setting it is necessary to retain the system of measuring only in one co-ordinate and of transposing through 90° to obtain the other measurements. In the system now proposed, however, the angle of transposition is not critical since the optically polished square, which was previously intended for an autocollimation measurement of transposition, now becomes the co-ordinate frame of reference and errors of orientation are only of the second order. To facilitate the transposition it is proposed to retain the turn-table to carry the reference square and to mount it on the single carriage.

Since the pattern of star images to be measured is not random but is nominally rectangular, it is considered preferable to control the motion of the carriage by a pair of rectangular guide frames (the constraints of which are not critical) rather than by a mechanical linkage of the type used in the Zeiss machine. This system has the further advantage that the measuring scale can be carried by the X-guide frame. Consequently the scale suffers only minute displacements relative to its supports and a simple elastic mounting can therefore be provided. A further proposal which has been under consideration is to replace the divided glass scale, as the basis of measurement, by a Moiré fringe micrometer of the type at present in course of development by the National Physical Laboratory. This has the primary advantage of providing the highest precision required, while avoiding the necessity of calibrating the scale. Moreover, since the scale "divisions" would be very small (of the order of lµ), the range of the optical micrometer, which it is proposed to retain for settings on the stellar plate so as to avoid the need of exact adjustment of the carriage, would be very considerably reduced.

III. Astronomical and Radio Observations

Time Service

Throughout the year the Time Service has been based upon astronomical observations made at Greenwich using Small Transit B and at Abinger using the Bamberg Broken Transit instrument.

At Greenwich 188 observations were secured on 149 nights and at Abinger 232 on 131 nights, making a total of 420 observations during the twelve months covered by this report. At Greenwich, the adjustable bubble, which had been used on Small Transit C, was mounted on the tube of Transit B on June 19. The bubble from the old striding level has been mounted on instrument C. Parallel tests between the two instruments are being made. It is intended to withdraw Transit B and to give it a thorough reconditioning, including fitting it with precision-worked chromium-plated pivots, as soon as it can safely be spared. At Abinger, observations are being made in the early morning as well as in the evening, as a control over periodic errors in the system of right ascensions.

The time determinations made at Greenwich and Abinger are reduced to the standard Greenwich meridian by the application of the appropriate adopted longitudes. It has been found that there are significant variations between the time systems defined by the two instruments. From an examination of the residuals of observations from the finally-adopted time, “observer corrections” are derived which take into account the “personal equation” of the observer concerned as well as the mean difference between his observing station and the adopted mean of the two stations, called the “station correction”. The determined “observer corrections” are, not strictly constant; for current use the adopted values are based on an examination of a series of quarterly means.

As in previous years, the astronomical observations have been corrected for the effects of polar motion, for which the current latitude variation data, communicated regularly by the U.S. Naval Observatory, are used. These provisional estimates are subsequently amended, when the polar motion data from the International Latitude Service become available. The observations, after correction for polar variation, are referred to the mean of five or six selected quartz clocks and analysed to derive values of the seasonal fluctuation in the Earth’s rotation. The values have proved to be in substantial agreement with those obtained in 1950–51, which were smaller than those derived in earlier analyses, though a slight tendency for the amplitude to show a small and persistent increase has been noticed in the last few quarters.

An investigation has been made of the effects of systematic star-place errors on time determination, based on the Normal System N30 of Dr. H. R. Morgan, who kindly supplied data in advance of publication. When the Greenwich and Abinger time observations are referred to the N30 system, the deduced annual fluctuation for current epochs is reduced from about ± 30 ms. to about ± 20 ms., and is in close agreement with the annual fluctuation derived from the American P.Z.T. observations, for which the star places are tied to the FK3 places in the equatorial zone, which the N30 catalogue indicates are substantially free from periodic errors. The Canberra time observations, when corrected for the FK3-N30 periodic difference in R.A. at the Canberra zenith, are also in satisfactory agreement. The results have been reported in a paper read to the Royal Astronomical Society.

The Time Service Bulletin has hitherto been based on the G.M.T. (or U.T.) determined at the Royal Greenwich Observatory, and the time signals transmitted by the Observatory have been regulated in conformity with extrapolated G.M.T. A high precision in time intervals is required for the standardization of frequencies. In recent years, the Time Service Bulletins have given the corrections required to remove the inequalities from the adopted G.M.T. and so to provide an approximately uniform time system.

The effect on astronomically determined time of the polar motion depends on the position of the observatory; that of errors of a periodic nature in the adopted right ascension system depends on the declinations of the stars observed which, in turn, depends on the latitude. The astronomical determinations of time at two different observatories are consequently not directly comparable.

From 1953 July 1, the transmitted time signals and the Time Service Bulletin of the Royal Greenwich Observatory will no longer be based on G.M.T. but on a provisional uniform time system. The deviation of G.M.T. (as previously used) from the adopted time system will be included in the Bulletin for those users desiring to maintain continuity with the old system. The adopted co-ordinates of the pole will be included in a supplement to the Bulletin for the convenience of astronomical users. Continuity in the adopted time systems and in the adopted co-ordinates of the pole will be preserved by a suitable smoothing process.

The transition in the regulation of the transmitted time signals from G.M.T. to uniform time will be effected during June 1953, by means of an adjustment in the trend or rate of the signals. As the deviation of G.M.T. from uniform time will be approximately zero at this time, there will be no need to make an adjustment in the time of the signals.

Experience throughout the year has further confirmed the superiority of the oscillators employing ring crystals, and standards of this type have been used exclusively for prediction and for the preparation of final corrections. At Abinger, oscillator E5, having a silk-supported ring crystal, was stopped on November 25 after running without interruption for 2 years. The crystal was transferred from a Post Office type 14 oven to a type 17 oven and connected to a modified maintaining amplifier. After careful re-alignment it was put back into service on December 9. Oscillators B5 and B6, having GT-cut crystals, have been taken out of service. The ring crystal oscillator E6 was removed from service on April 6 for temperature adjustment and re-alignment. Various minor modifications are also being made to it. The equipment has been cleared from D cellar in order to provide accommodation for new clocks of the pattern developed in the electronics laboratory.

Two oscillators, to be mounted on brick piers at Greenwich, are being built to the same specification as those at Abinger; each will contain a GT-plate crystal removed from the Greenwich A group. The buffer valve of the maintaining amplifier of clock Fl failed on September 26; as the oscillator had been running for more than five years, advantage was taken of this opportunity to change all the valves. Oscillators Al and A2 have given a reasonable performance throughout the year, but A3 and F3 have suffered a number of changes of rate.

The Observatory has continued to enjoy close co-operation with the Post Office Radio Experimental and Development Laboratory and the National Physical Laboratory and information on the quartz clocks at these establishments has been communicated regularly and employed in the control of the time service. Information has also been received on the frequency standards at the B.B.C. Receiving Station at Tatsfield and the Research Division, Marconi’s Wireless Telegraph Co. Ltd. A general account of the quartz clocks employed in the Greenwich time service is being published in ‘Monthly Notices’.

The installation of the new regenerative frequency dividers has been completed. Five pulse dividers operating from 1000 pulses to one pulse per second were brought into use at Abinger on January 1. These pulse dividers have replaced phonic motors for clock inter-comparisons and for mean time radio time signal reception, but the motors are still used for time signal transmission, rhythmic signal reception and sidereal time. The dividers are phaseable in steps of 0.1 second to facilitate the reception of radio signals when a short interval between signal and clock is desirable; the mean deviation of the output pulses is of the order of one microsecond compared with 0.1 milliseconds for the phonic motors.

In view of the occasional slipping experienced with the rotary beat counters, the power to the driving motors has been increased, which has cured the trouble. At the same time the phase relationship was made uniform throughout the six groups of counters to avoid discontinuities in the last figure during the change over from one group to another. Two additional groups of beat counters are in course of construction in the workshop to deal with the requirements of the additional experimental clocks. An electrically-operated camera has been installed to photograph the rotary beat counter dials. The shutter is at present hand-controlled but provision is being made for automatic control from the time signal transmitting motors. Routine photographs are taken daily on roll film, which is developed weekly. A second magazine has been adapted to allow single exposures to be taken on bromide paper when required.

The gate circuits of the decimal counters at Abinger have been replaced by the new type which, as reported last year, had been fitted to a similar counter at Greenwich. The modification has led to higher accuracy and reliability and makes the ancillary trigger unit unnecessary.

A Belin printing chronograph has been tried out and tested for accuracy and a report issued. At the request of the Hydrographer several methods of determining the error of a chronometer by using radio time signals have been investigated and a report is being prepared.

The receiver for the seconds pulses of MSF on a frequency of 60 kc/s, referred to in last year’s report, was brought into regular use at Abinger on December 1. A significant increase in measurement accuracy has been achieved. It was originally intended as a prototype, but as changes in the MSF transmissions are pending, further construction has been postponed.

The international time signals, the six-pips time signals, and the hourly signals used by the Post Office for the control of the Speaking Clock, have been sent out from the phonic motor transmitters at Abinger. The reserve service of hourly and six-pips signals has been maintained from Greenwich, and has been called on a few times when faults occurred on the Abinger lines. The circuits of the send-relays, which were referred to in last year’s report as under construction at Abinger for installation at Greenwich, have been modified and a new and improved type of valve incorporated. The relays are now complete, together with control and monitoring panels, but final wiring of the rack is held up pending delivery of a special multicore cable.

For an experimental period of four weeks from May 19, the GBR time signals from Abinger were sent from London to Rugby on one channel of a three-channel high-speed VF line. A reduced scatter resulted and a request was therefore made to the Post Office for the permanent allocation of this channel, which was brought into regular use on September 26; the physical line previously used is retained as a reserve. Transmissions of Greenwich time signals from Rugby on both long and short waves have been monitored at the Observatory although, owing to reception conditions, satisfactory short wave reception cannot always be achieved. Regular measurements are made of the carrier frequencies of MSF (60 kc/s) and Droitwich (200 kc/s) together with the reception times of the seconds pulses superimposed on the MSF transmissions. Radio time signals from Australia, France, Germany, Russia, and the U.S.A. have been received throughout the year. Since November signals have also been received from Argentina. Occasional measurements have been made of Canadian and Japanese radio time signals.

Publication of the Time Service Bulletin, together with its supplements, has been continued. A number of requests have been received from abroad for the complete series of the Bulletin. The Time Service Circulars and Monthly Frequency Estimates have been published regularly on or about the 15th of each month. Copy for the volumes of Greenwich Observations has been completed up to 1949 and the material for 1950 and 1951 is approaching completion.

The second experimental oscillator, G2, which was mentioned in last year’s report, has been in operation since July. It incorporates a soldered wire-supported GT-cut crystal, which had previously been used in a conventional circuit as oscillator E2. Its performance had never attained the standard of accuracy which had been expected either in long-term or short-term stability. In the new circuit a high standard of short-term stability has been achieved, the scatter of the daily measures of frequency being reduced from about ± 1 part in 10⁸ to about ± 1 part in 10¹⁰ over periods which may extend over more than a month. The long-term performance cannot yet be assessed. A further experimental oscillator G3, which incorporates the clamped GT-cut crystal previously employed in the operational clock B5, has been constructed and is undergoing test. The first experimental oscillator G1 was stopped in December and is now being rebuilt in the light of experience gained with G2 and G3. A fourth standard is under construction so that experience may be gained in the use of a ring crystal in the new circuits; a slightly sub-standard ring crystal is being made available for the preliminary tests by courtesy of the Post Office Radio Experimental and Development Laboratory.

For the study of the short-term stability of the new experimental oscillators and of the operational ring crystal oscillators, frequency measuring equipment of high accuracy is essential. A direct-reading frequency comparator is therefore being constructed, with which it is hoped to achieve a comparison accuracy at 100 kC/s of ± 2 parts in 10¹¹. A new phase comparator has also been developed and constructed, which is capable of measuring phase changes of less than 0.001 radians at 100 kc/s.

Preliminary work has been carried out in the electronics and optics laboratories on methods of recording star transits by automatic photoelectric records. Ancillary equipment for the photomultiplier tubes has been constructed and tests have been made. The work is continuing.

1954

The Report here presented refers to the period from 1953 May 1 to 1954 April 30 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

(a) Greenwich.

A new boiler has been installed for heating the Record Rooms and Time Department Office.

The Time Ball has been dropped regularly, exceptions being made (on account of high winds) on January 15, March 1, and March 2.

In the Meridian Group, the Control Building for the Photographic Zenith Tube has been completed.

The foundations for the Photographic Zenith Tube Pavilion are under construction and the manufacture of the pavilion is well in hand at the works of the contractors. The plans for the pavilion to house the Bamberg Small Transit are nearing completion. It had been expected that both of these buildings would have been completed during the year, but there have been further delays. It is hoped that both buildings will be erected by the end of July 1954. The pavilion which housed the Cooke R.T.C. at Greenwich is to be modified to house the Melbourne R.T.C., together with its two collimators, at Herstmonceux. The drawings of the modifications are in hand.

The details of the design of the building to house the Nautical Almanac Office, Time Department, workshops and associated buildings have been under discussion. The Consulting Architect has not made the progress expected on the working drawings and the commencement of the work on site has accordingly been delayed. It is hoped, nevertheless, that an early start will be made on this building.

II. Instruments, Apparatus, and Library

The following instruments are on loan to the Royal Observatory –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Greenwich Observatory: –

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – Shortt clock No.16 with slave.

To the National Maritime Museum: – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No.4 by Larcum Kendall.

III. Astronomical and Radio Observations.

Time Service.

At Abinger the Bamberg broken transit instrument was in use throughout the year, 204 observations being obtained on 121 nights. At Greenwich, Transit B was withdrawn from service on September 16, on the completion of the parallel tests mentioned in the previous report. Transit C was in use in the Courtyard dome until July 18; it was then transferred to the Altazimuth dome, but was returned to the Courtyard dome on January 1. Transit B was transferred from the Altazimuth dome to the Courtyard dome on July 18. Transit B was adopted as the standard Greenwich instrument until June 30, and Transit C from July 1. The number of observations at Greenwich used for controlling the clocks was 196, obtained on 152 nights. In addition, for the parallel comparison between Transits Band C there were 69 observations on 69 nights, between 1953 March 25 and September 15.

The pivots of Transit B have been chromium plated, and the instrument is being re-conditioned in the workshop at Greenwich. The figure of the pivots of the Bamberg instrument, which had been in continuous operation for some six and a half years, was examined. Wear had occurred at the parts of the pivot which are in contact with the supporting Y’s and the pads of the hanging level, but the departure from circularity nowhere exceeded 1 micron and has no significant effect on the determined time.

The application of observer corrections to the time observations was discontinued on July 1, with the exception of a significant personal correction retained for one observer (AC). It was found that the personal equations of the observers, with this one exception, are much smaller than the fluctuating difference between the two stations, Abinger and Greenwich.

The series of morning observations undertaken at Abinger was discontinued on May 1; no significant differences of a systematic nature were detected between the morning and evening observations. A similar series was commenced at Greenwich on January 31, after Transit C had been finally mounted in the Courtyard dome.

The determinations of time have been corrected for the effects of polar motion, based on current latitude observations communicated regularly by the D.S. Naval Observatory. After this correction, they have been compared with several selected quartz clocks and analysed to derive the apparent annual fluctuation in the rotation of the Earth. The fluctuation is in close agreement with those observed in the past two or three years.

During the summer of 1953, the time determinations indicated an increasing systematic departure of the standard clocks from the performance predicted on the basis of the observations of the previous 6 to 12 months. Clock inter-comparisons gave no reason to suspect that the clocks had suffered any change, and the divergence was attributed to a change in the rate of rotation of the Earth. The magnitude and form of the change could not be precisely evaluated because of the considerable scatter in the time determinations and because of some uncertainty about the rate of frequency drift of some of the clocks. The change was represented at first as a simple decrease of 1.3 milliseconds in the length of the day, occurring in 1953 March, but subsequent evidence suggests that it was more complex, and may have occurred as early as 1952 December.

Clocks.

The Time Service has been based on clocks employing silk-supported ring crystals, together with an experimental clock employing a GT-cut plate crystal, which has been included because of its excellent performance. Two new ring crystals have been obtained from the Post Office Radio Branch, and installed as B5 and B6. The performance of B6 has been good, but B5 showed variations which correlated with changes of atmospheric pressure, and it has been returned for examination. The oven

temperature of oscillator C6 was readjusted in August; oscillator C5 was taken out of service in March for overhaul and modification. Oscillators E5 and E6 have run throughout the year. Additional buffer amplifiers are being fitted in the 100 kc/s outputs of the ring crystal oscillators to minimize the possibility of variations in the output load causing changes in frequency. Oscillators B4, C4 and E4, of the GT-cut plate type, have been employed as working standards.

Three oscillators of the pattern developed by the Electronics Laboratory have been installed in D cellar. Oscillator D6 has been provided with a four-point supported GT-cut plate crystal: D4 and D5 are awaiting suitable crystals.

At Greenwich, two oscillators, similar in design to B4, C4 and E4, have been installed in a newly prepared clock room. They have been designated H1 and H3, and incorporate the GT-cut plate crystals previously used in oscillators Al and A3. Their power supplies are obtained direct from the A.C. mains, but an alternative supply is provided by battery-driven rotary converters, which automatically start up when a mains failure occurs.

The Observatory has continued to receive information on the Post Office quartz clocks at Dollis Hill and Banbury, those of the National Physical Laboratory at Teddington, of the B.B.C. Receiving Station at Tatsfield, and of the Research Division, Marconi’s Wireless Telegraph Company Ltd.

Experimental Clocks.

The second experimental oscillator G2 which, as reported last year, attained a high standard of short-term stability, has now been in operation for over 18 months. During the year 1953 its rate approximated closely to an exponential curve: when this is allowed for, the performance of the clock becomes comparable with that of the ring-crystal oscillators. The experimental oscillator G3 has continued in operation, but its performance is erratic. The rebuilding of oscillator G1 was completed, and oscillator G4, which contains a slightly sub-standard ring crystal, has commenced its first test run. These clocks have not as yet attained the high standard of performance of G2.

Some of the experimental clocks have been equipped with special long-life valves. For the best results, it is essential that the low-tension supply should be held within closer limits of voltage than are needed for conventional valves. The L.T. supply to the experimental group has accordingly been fitted with a voltage stabilized rectifier. Similar long-life valves are being fitted in some of the operational clocks, and constant voltage devices are being obtained for installation in the operational power supplies.

Work has proceeded on the development of a high-precision frequency comparator. The first experimental model gave an accuracy of a few parts in 10¹⁰, and construction is in progress on a new design which is expected to give still higher accuracy.

Control Room Equipment.

The rotary beat counters in the Control Room at Abinger have been augmented by the installation of two additional groups, enabling the experimental oscillators to be investigated in the same way as the operational clocks. The beat counter dials are photographed automatically at the same time each day.

The 100 kc/s to 1 kc/s regenerative dividers referred to in last year’s Report have performed satisfactorily. The circuit has been modified to permit adjustment in steps of 0.01 and 0.001 seconds as well as in steps of 0.1 seconds. A prototype pulse divider employing Dekatron counting tubes has been designed and constructed and is undergoing tests. This type of divider has the advantage of consuming only one-third of the power required by the thermionic dividers. Two 1 kc/s phaseable transformers are under test: they have been designed to permit phase adjustment of the 1 kc/s input to the pulse dividers, thereby enabling the output pulses to be set to an accuracy of 0.0001 seconds. A contract has been placed for two 1 kc/s phaseable transformers, to be driven continuously by suitable gearing from aphonic motor and to provide an output of 1,000 cycles per sidereal second.

Cold-cathode trigger tubes possess advantages for many time service applications, and various experiments have been carried out. An automatic sender for the morse preamble which precedes the radio time signal has been constructed using these tubes, and a time signal switching clock is being built.

The precision of the signal pulses obtained from the drum contacts of the phonic motors has been affected by the transfer of material from the positive drum to the negative wiper arm, causing craters in the leading edges of the rotating contact segments. The contact drums have been insulated from their spindles so that the polarities can be reversed.

With the co-operation of the Post Office, tests have been made of various types of circuits for the control of the Rugby transmitters by direct land-line from Abinger. The system at present in use consists of a 3-channel V.F. line from London to Rugby, keyed at London by direct current pulses from Abinger. Promising results have been obtained with a private telephone line from Abinger to Rugby, the signal taking the form of keyed pulses of 1 kc/s tone. By this system, which is essentially a single channel wide-band V.F. circuit, the erratics have been reduced by one half.

At Greenwich a signal forming unit similar to that at Abinger has been installed. The phonic motor circuits are being modified to work with this unit, and the various selector contacts are being aligned. The installation of electronic send relays has been completed. Serious variations in mains supply voltage have given cause for concern, and discussions are in progress with Electrical Engineering Manager, Chatham, regarding the provision of a constant voltage device to ensure adequate stability in the supplies to essential equipment.

Time Signals.

The international time signals, the six-pips time signals, and the hourly signals used by the Post Office for the control of the Speaking Clock, have been controlled by the phonic motor transmitters at Abinger. The reserve service of hourly and six-pips signals available from Greenwich has twice been called upon during the period under review. At the request of the Radio Research Station, Slough, several special time signals were transmitted: these consisted of 20-millisecond pulses at intervals of one-tenth of a second, the pulses being lengthened to 40 milliseconds at the exact seconds, and to 60 milliseconds at the exact minutes.

The transmissions of the international time signals from Rugby on both long and short waves have been monitored at the Observatory, but propagation conditions sometimes cause the reception of the short-wave time signals to be uncertain. Routine measurements are made of the carrier frequencies of MSF (60 kc/s) and Droitwich (200 kc/s), together with the reception times of the seconds pulses superimposed on the MSF transmissions on 60 kc/s and 2½Mc/s. Radio time signals have been received throughout the year from Argentina, Australia, Canada, France, Germany, Russia, and the United States.

The Time Service Circular and the Monthly Frequency Estimates have been distributed at the middle of each month. The Time Service Bulletin, with its various supplements, has been published in quarterly instalments. Proof-reading of copy for the Time Service Section of the volume of Greenwich Observations for the years 1941, 1942, and 1943 has been completed or is in progress.

Miscellaneous.

Experimental work on the automatic recording of star transits with photoelectric equipment has been continued. A number of transits have been observed; it appears that the system now being investigated may form the basis of a practicable instrument.

The optical and mechanical design of the photoelectric measuring machine for the P.Z.T. stellar plates has been in abeyance owing to the prolonged illness of the officer concerned. The electronic equipment has been constructed and preliminary tests have been made with a temporary bench arrangement in the laboratory. The results give some promise that the desired accuracy may prove capable of achievement.

A prototype unit for the cathode ray tube display of chronometer and watch ticks has been constructed, and a quartz crystal controlled drive circuit is under construction for the Belin drum chronograph.

Measurements have been made to determine the level of the radio interference emanating from overhead power lines at Herstmonceux. As a result, the S.E. Electricity Board has been asked to re-route one of its proposed new lines so that it will be at a safe distance from the site selected for the time signal reception aerials.

Shortt 49 free pendulum and slave have ‘been dismantled and transferred to the Chronometer Department at Herstmonceux.

1955

The Report here presented refers to the period from 1954 May 1 to 1955 March 31 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

(b) Abinger.
The Admiralty Civilian Hostel at Feldemore, Holmbury St. Mary, which in the past nine years has provided accommodation for many of the staff employed at Abinger, was closed on February 13. Some of the rooms which have been set free are in temporary use as a flat: the remaining rooms will be used for stores and other official purposes.

c) Herstmonceux.

The small pavilion to house the Bamberg Small Transit and the pavilion for the Photographic Zenith Tube have been erected. A few modifications to the latter building are required before final acceptance.

The foundations for the West Building, which will house the Nautical Almanac Office, the Time Department and the workshops and which will also provide miscellaneous storage accommodation, have been commenced and are in an advanced stage of construction. The erection of the steel framework of this building has been commenced.

II. Instruments, Apparatus, and Library.

The following instruments are on loan to the Royal Observatory:-

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Greenwich Observatory: –

To the Royal Observatory, Cape of Good Hope: – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge: – Shortt clock No.16 with slave.

To the National Maritime Museum: – Harrison's Time Machines Nos. 1, 2, 3 and 4, and the copy of No.4 by Larcum Kendall.

Photographic Zenith Tube

In November representatives of the makers commenced the installation of this instrument and of the control console in the separate buildings. The installation has reached an. advanced stage. The movement of the plate carriage is at present controlled from the frequency of a tuning fork vibrating at a nominal 1000 c/mean time second. The speed of the plate carriage has been found not to be precisely that appropriate to the diurnal motion of a zenith star, but can be corrected by changing the frequency of the fork to 1000 c/sidereal second. The tuning fork has therefore been returned to the makers for this change to be made. The installation will be completed on its return. It is intended, after the quartz clocks have been moved to Herstmonceux, to use the amplified output of 1000 c/sidereal second for driving the plate carriage during the stellar exposures.

III. Astronomical and Radio Observations.

Time Service.

At Abinger the Bamberg broken transit instrument was in use for the whole of the eleven-month period covered by this Report, 180 observations being obtained on 103 nights. The contact length and the level scale value received the usual routine check in October, and the values adopted were confirmed. At Greenwich during the same period Transit C was employed in the Courtyard Dome. Altogether, 234 observations were obtained: 103 of these were made on 79 mornings, and 131 on 127 evenings. The observations at Abinger and Greenwich, totalling 414, were combined for the purposes of time determination.

It was noticed in May that the effective length of the micrometer contact of Transit C differed from the value adopted from previous measurements. There was no independent evidence to indicate precisely when the change had occurred. As the instrument had been in operational service for less than a year, it was decided to recompute the observations on the assumption that the new value of the contact length had been applicable throughout. This micrometer was replaced on June 11 by the Cooke micrometer, which had previously been used on Transit B and which had been reconditioned and the drive altered so as to double the speed of manual turning. The contact length was redetermined in June and November and was found to have the same value as used throughout its 25 years of service. Since the introduction of this micrometer the mean error of the time determinations at Greenwich has been appreciably reduced.

The determinations of time have been corrected for the effects of polar motion, based on current latitude observations communicated regularly by the U.S. Naval Observatory. Various instrumental changes have occurred, but overlapping series of observations have preserved the continuity of the polar motion adopted. After this correction, the Greenwich and Abinger time observations have been compared with several selected quartz clocks and analysed to derive the apparent annual fluctuation in the rotation of the Earth. There has been no significant change in the amp1itude and phase of the observed fluctuation.

Clocks.

The Time Service continues to be based on clocks at various stations: during the period under review, the clocks at Abinger have provided the majority of the long-term standards employed. The best performances at Abinger have been those of C6 and E5, and valuable supplementary information has been obtained from B5, B6, C5 and E6, but the last named still exhibits a large frequency drift and small but disappointing departures from uniformity. All of these oscillators are provided with ring crystals. The subsidiary standard clocks employing GT-cut plate crystals, which are used to control the time signal transmitting equipment, are proving inadequate for modern requirements, being insufficiently uniform in their performance. B4 is the only one of the three sub-standards which can be satisfactorily employed; fortunately, however, the frequency of one of the primary ring-crystal standards, B6, is fairly near nominal, and the rate of this clock has been sufficiently low for it to be employed not only as a primary standard but also to control time-signal transmitting equipment. Consideration is being given to various alternative ways of resolving the difficulties of the present unsatisfactory position.

The clocks at Greenwich, which are of the GT-cut plate type, although adequate for the purpose of transferring the Greenwich time observations to the primary ring-crystal clocks, have not themselves attained the high standard of uniformity requisite in a primary standard clock. The clocks A2, F1 and F3 have run without interruption: H1 and H3 were stopped on November 21 by a fault on a common power supply. In order to reduce the large variations in the mains supply voltage at Greenwich, a Ferranti Voltage Stabiliser was supplied and installed on February 11 by the Electrical Engineering Manager, Chatham.

The Observatory has continued to receive information on the Post Office quartz clocks at Dollis Hill and Banbury. The information relating to the clocks at the National Physical Laboratory is now received by private telephone line via Dollis Hill.

Experimental Clocks.

Reference was made in the previous Report to the installation in D cellar at Abinger of three oscillators of the pattern developed in the Electronics Laboratory. These clocks are at present used as experimental clocks and not as operational standards. The quartz crystal resonator (of GT-cut plate type) which previously gave an excellent performance in the experimental oscillator G2 was transferred in the summer to oscillator D4 where, after an initial ageing period, it has given a performance comparable with that of primary standards incorporating ring crystals. D5 is operating temporarily with a GT-cut plate crystal, which has been supplied by the Post Office for special low-temperature tests which have been arranged in collaboration with the Clarendon Laboratory at Oxford. A similar crystal in oscillator D6 has been running for some time, but with disappointing results: it is planned in the near future to replace this crystal by the ring crystal now incorporated in oscillator E6.

As an alternative to the employment of a condenser in series with a crystal to bring the frequency of the oscillator conveniently near to its nominal value, the use of a continuous phase-shifting device has been considered. An experimental model incorporating a motor-driven phase-shift capacitor has been constructed and is now undergoing tests. Equipment for the precise measurement of the temperatures of the thermostatically controlled ovens has been received after considerable delay, and has been used to check the temperature variations in the new-type quartz crystal oven. Results obtained so far indicate that the variations do not exceed 0.001 degree in the course of one month. Investigations are being made into the possible advantages of employing transistors in quartz clock circuits. A departmental report on the design, construction, and installation of the laboratory pattern quartz clocks has been prepared and is in course of publication.

Control Room Equipment.

The prototype Dekatron divider has given a satisfactory performance for over twelve months, and six more dividers are approaching completion. Two are already working in conjunction with the new cold-cathode trigger-tube switching clock, a 60-61 frequency converter, and magslip phasing devices in an electronic installation capable of providing all the time signals transmitted by the Department. The switching clock, which is based on a Post Office idea, is designed to perform all the functions of the existing phonic motor transmitters. The rhythmic impulses are derived from a 1 kc/s source by means of an electronic 61/60 multiplication stage followed by a 1000-1 Dekatron converter. This has been operating reliably since December. The two solar-sidereal transformers, mentioned in last year’s Report, have been delivered and tested, and one is now in operation with a 20-1 frequency divider supplying 50 cycles per sidereal second to drive a synchronous clock dial in the Bamberg hut. This is an experimental installation for providing experience in the operation of the time distribution system that is being planned for Herstmonceux.

A new model of the Cintel decimal counter chronometer, which measures time intervals of up to one second with an accuracy of one microsecond, is being modified to suit Time Department requirements. It is the first of three new counters that are being prepared for installation at Herstmonceux.

In continuance of the tests on the private telephone line mentioned in last year’s Report, two keyed 1 kc/s tone sending devices have been constructed and were brought into operation on December 1 for the control of the Rugby radio time signals. With the co-operation of the Post Office, a number of special test transmissions have been made with a view to ensuring that the time signal erratics are kept as small as possible.

The four rectifiers charging the quartz clock L.T. batteries have been replaced by Westat constant potential rectifiers to prevent large fluctuations in valve heater voltages, which are detrimental to the long-life valves with which some of the standards are equipped.

At Greenwich, the rewiring of the phonic motor circuits has been completed to work with the new signal forming unit. The full operation of the equipment awaits modification to phonic motor transmitter No. 3, which is now in hand at Abinger. The drive circuit for the Gate Clock has been replaced by one of an improved design

Time Signals.

The international time signals, the six-pips time signals, and the hourly signals used by the Post Office for the control of the Speaking Clock, have been controlled by the phonic motor transmitters at Abinger. The reserve service of hourly and six-pips signals has been available from Greenwich. At the request of another department, a series of special time signals has been transmitted for four hours each week-day since February 14 from the Post Office radio station at Criggion.

At Greenwich, the Time Ball was dropped daily throughout the period under review, except on three days when high wind or snow on the mast prevented the ball from being raised.

The transmissions of the international time signals from Rugby on both long and short waves have been monitored at the Observatory. Routine measurements are made of the carrier frequencies of MSF (60 kc/s) and Droitwich (200 kc/s) , together with the reception times of the seconds pulses superimposed on the MSF transmissions on 60 kc/s and 2½ MC/s. Radio time signals have been received throughout the year from Argentina, Australia, Canada, France, Germany, U.S.A., and U.S.S.R.

The Time Service Circular and the Monthly Frequency Estimate have been distributed at the middle of each month. The Time Service Bulletin, with its various supplements, has been published in quarterly instalments. Preparation of copy for press has proceeded: 1944 is ready for vari-type correction, and 1946 has been corrected. The tabular material for 1952 and 1953 has been completed.

Miscellaneous.

For the convenience of observers of the total solar eclipse of 1954 June 30, special time signals were transmitted between 0900 and 1500 by the Post Office long-wave transmitter at Rugby and by associated short-wave transmitters. In order that the signals might serve also for a programme of radio observations, the seconds dots were lengthened from the normal duration of 0.1 second to become dashes 0.3 second in length. On the ten days preceding the eclipse, rehearsal time signals were transmitted between 0945 and 1030. In view of the difficulty experienced in obtaining reliable measurements at Abinger of the Rugby short-wave transmissions, the co-operation of the Fernmeldetechnisches Zentralamt at Darmstadt and of the Geodätisches Institut at Potsdam was invited. Results obtained at Abinger, Darmstadt and Potsdam have been described and tabulated in a departmental report. The assistance of the two German establishments is gratefully acknowledged.

A Belin drum chronograph has been transferred from Abinger to Herstmonceux for the rating of chronometers and watches. A portable mains-driven quartz crystal frequency standard has been built to control the speed of the drum, and a five-way amplifier with one microphone input permits simultaneous recording of four contact chronometers and one chronometer or watch without contacts.

A prototype model of an improved receiver for the measurement of the seconds pulses of the 60 kc/s standard frequency transmission has been constructed and is undergoing tests. Electronic equipment is under construction for use in the proposed tests at the Clarendon Laboratory on the operation of a crystal resonator at low temperatures.

The problem of making precise frequency comparisons at widely separated points by means of very long wave radio transmissions has recently been under consideration by the United Kingdom Study Group VII of the C.C.I.R. Various methods have been proposed, one of which is being further developed in the Electronics Laboratory.

As a preliminary to the International Geophysical Year, a programme of international co-operation has been organized under the chairmanship of Professor Boella with a view to the better determination of the propagation times of radio time signals. On selected days the co-operating stations are asked to make precise measurements every three hours of the times of reception of the seconds pulses superimposed on local and distant standard frequency transmissions. The Royal Observatory has taken part in the programme on the days so far selected, three in December and two in March, using a visual cathode-ray tube technique specially developed for the purpose. Tests have been arranged to determine the suitability of the method for more general applications in the reception of conventional time signals.

During the past twelve months experience has been gained in the use of a new method, which avoids any personal bias, of extrapolation of clock performance for the purposes of “steering” the transmitted time signals. After some disappointments arising from adverse circumstances in the early part of the period, the method has more recently shown considerable promise, and the trial period has been extended. In anticipation of the delivery of an impersonal prismatic astrolabe, a pilot computation for the compilation of the observing list and factor table was undertaken.

An increasing proportion of the Time Department’s work has been devoted to the formulation of detailed plans of the new installation in the West Building at Heretmonceux. Final decisions on these matters have been delayed as long as possible in order that the adopted designs should incorporate the most up-to-date techniques. The whole pattern of the Department’s work is at present in a state of rapid change, which makes it extremely difficult to foresee and cater for the probable needs of a few years hence. The situation is further complicated by the special responsibilities which will fall upon the Time Department during the International Geophysical Year.

XI. Genreral.

… visitors included members of the Court of Assistants of the Worshipful Company of Clockmakers, …

Mr. H. M. Smith attended the International Congress of Chronometry in Paris in October, and presented two papers, one by himself and Mr. J. D. Pope on “L’appareil actuel du service horaire de Greenwich”, and the other by himself and Mr. R. H. Tucker on “Calibrage des etalons de frequence, quartz et atomiques”. These papers will be published in the transactions of the Congress. Mr. Smith afterwards visited the Paris Observatory to study the impersonal prismatic astrolabe.

Consideration has been given to the participation of the Observatory in the general programmes of observation to be undertaken during the International Geophysical Year, 1957 July to 1958 December inclusive. Since many geophysical phenomena are closely related to the various aspects of solar activity, it is essential that the Sun should be kept as nearly as possible under continuous observation during this period. The normal hours of observation of the Sun will therefore be extended, as at other solar observatories, and additional observations will be made during the alert periods, when special activity is to be expected. In connexion with the world programme of longitude and latitude determinations, the photographic zenith tube will be used for the observations at Herstmonceux, and it is planned to obtain an impersonal prismatic astrolabe for the observations at Greenwich. With each of these instruments time and the variation of latitude can be determined free from personal equations, so that Greenwich and Herstmonceux will be tied into the world network with high accuracy. The receptions of radio time signals will be considerably increased. The seasonal changes in the rate of rotation of the Earth and the movements of the pole depend upon the general circulation of the atmosphere. The information on these subjects that will be derived from observations at many stations and at various altitudes during the International Geophysical Year should enable a closer evaluation to be made of the relationships between these phenomena. The times of reception of radio time signals will provide information about the velocities of travel of radio waves of long and short wavelengths and of the dependence of these velocities upon ionospheric conditions. It is hoped to extend the cosmic ray observations to include the recording of the neutron flux. Approval has been given for the building of the magnetic observatory at Hartland, and the buildings are expected to be completed in time for equipment to be installed and observations to be commenced before the beginning of the International Geophysical Year. The necessary overlap in observations between Abinger and Hartland will therefore be made during that period, so adding to the available magnetic data. The Abinger magnetic observatory will be closed after the completion of these observations.

1956

The Report here presented refers to the period from 1955 April 1 to 1956 March 31 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings

(b) Herstmonceux.

The main fabric of the West Building has now reached an advanced stage of construction.

The pavilions for the Photographic Zenith Tube and the Cooke Reversible Transit Circle were completed during the year.

II. Instruments, Apparatus, and Library

The following instruments are on loan to the Royal Observatory –

A small transit instrument, of the broken type, with stand, by Bamberg, from the Royal Observatory, Edinburgh.

The following instruments have been lent by the Royal Observatory.

To the Royal Observatory, Cape of Good Hope – Clocks, Dent 1916 and Dent 2013.

To the Cavendish Laboratory, Cambridge – Shortt clock No.l6, with slave.

To the National Maritime Museum – Harrison’s Time Machines, Nos. 1, 2, 3 and 4 and the copy of No.4 by Larcum Kendall.

To the Jeremiah Horrocks Observatory – Clock Dent 2009.

Photographic Zenith Tube.

The erection of the P.Z.T. by the makers was completed during the year. The final stages, involving the adjustment in azimuth of the plate carriage motion and obtaining satisfactory images on the stellar and chronograph plates, occupied several months. A few experimental plates for time and latitude were obtained in July and September. Azimuth observations were continued in October. The instrument was brought into regular use for time and latitude determinations on 23 November.

Several instrumental faults revealed themselves, which delayed the observational work on a number of occasions. Steps are being taken to bring the instrument into a state capable of giving reliable performance.

III. Astronomical and Radio Observations.

Time Determinations.

At Greenwich, astronomical time determinations were made throughout the year with Transit C in the Courtyard Dome, 99 observations being obtained on 96 mornings and 160 on 157 evenings. The reduction in the mean error of the observations, referred to in last year's report, has been maintained. At Abinger, the Bamberg broken transit instrument was removed from service on 4 November for replacement of the wiring which had become faulty on account of condensation. The opportunity was taken to give the instrument a meohanical overhaul. When observations were resumed, there appeared to be a discrepancy of some 100 milliseconds. In spite of various efforts to locate the cause, it was not until 17 February that backlash between the lead-screw and the wire carriage was detected by means of a displacement indicator gauge. On further dismantling the micrometer it was found that the return spring was oversize for its housing. A new spring was made and inserted, but was apparently of insufficient strength. A third spring was made and inserted on 23 March. Subsequent observations indicate that the discrepancy persists, and the investigation continues. In all, 232 observations were taken on 138 nights, of which only 146 on 89 nights were used for time determination. The combined total for the two stations of observations used for time determination was thus 405.

Corrections for polar variation have been derived from current latitude observations communicated regularly by the U.S. Naval Observatory. There is satisfactory agreement between these provisional corrections and those adopted by the Bureau International de l'Heure for use in current timekeeping work. In accordance with the recommendations of the International Astronomical Union, the B.I.H. values have been employed from 1 January, but the U.S.N.D. values are still reduced as a general check. Similarly, there is satisfactory agreement between the annual fluctuation in the rate of rotation of the earth derived from the Greenwich and Abinger observations and the seasonal fluctuation adopted by the B.H.I. for use in 1956. In these circumstances it will be possible, for the first part of 1956 at any rate, to identify observed astronomical time with G.M.T. in 1955 and with U.T.O. in 1956, without introducing any discrepancy between the provisional uniform time system (P.U.T.) of 1955 and the analogous system (U.T.2.) of 1956.

Clocks.

Quartz clocks at several co-operating stations have been used in the operation of the Time Service, but the principal long-period standards have again been found among the Abinger clocks employing ring crystals. The increasing demand for short-period accuracy in the “working” clocks (which are adjusted: to be very near nominal frequency) have been met by the provision of an adjustable condenser in the B6 oscillator. This arrangement successfully disposes of the difficulties mentioned in the previous report, and provision is being made for the attachment of variable condensers if required in all the quartz clocks to be installed at Herstmonceux.

The quartz clocks at Greenwich have continued undisturbed with no marked changes of rate. There was a power failure on 10 January lasting nearly 5 hours due to contractors in the Park cutting the electricity supply cable. The emergency generator was brought into operation and there was little effect on the rates of the clocks.

Time Signals.

The International time signals, the six-pips time signals, and the hourly signals to the Post Office have been controlled by the phonic motor transmitters at Abinger. A reserve service of six-pips and hourly signals has been available from Greenwich. Routine measurements are made of the time signals transmitted on long- and short-waves from Rugby; the carrier frequencies of MSF (60 kc/s) and Droitwich (200 kc/s), the seconds pulses superimposed on the MSF, WWV and IBF standard frequency transmissions; and radio time signals from Argentina, Australia, Canada, Czechoslovakia, France, Germany, U.S.A., and U.S.S.R.

At Greenwich the Time Ball was dropped daily except when it was too windy or when there was snow on the mast.

The Observatory again participated in the special programme of international time signals measurements carried out on two days in every third month. A visual method of measurement, developed in collaboration with the Electronics Laboratory, has continued to give satisfactory results, which have been communicated to M. Boella. Improved equipment for permanent use is being designed.

Control Room Equipment.

The telephone line between Dollis Hill and Abinger has been adapted to carry, in addition to the normal telephone service, a frequency of 2.5 kc/s derived from one of the Post Office Standards. At Abinger, this frequency is multiplied up to 100 kc/s and compared with the Abinger standards by means of rotary beat counters. A Post Office telephone line has been rented so that accurate clock pulses can be sent from Abinger to Herstmonceux to operate the Photographic Zenith Tube timing circuits.

A new type of long-wave receiver for use on the 16 kc/s Rugby time signal has been constructed and is now in full use at Abinger. This improved model has the advantage of small size and simplicity and makes use of modern components. Two similar receivers will be required for use at Herstmonceux.

Electronics Laboratory.

Experimental work on operating a quartz crystal oscillator with the resonator at the temperature of liquid hydrogen has been carried out at Oxford with the help of the Clarendon Laboratory. The experience gained has suggested modification in both the low temperature and the electronic measuring equipment. Results so far are promising, the Q factor of the resonator for, instance, increased by a factor of five. It is intended to continue these experiments.

On a request by Study Group 7 CCIR to investigate the possibility of high accuracy frequency comparison by means of the 16 and 60 kc/s controlled transmissions from Rugby, experimental equipment for continuous recording of the phase of these transmissions has been built. An accuracy of a few parts in 10¹⁰ has been achieved. Certain improvements in the system are currently being investigated.

Because of its accuracy and simplicity this method of frequency checking has been chosen for the next series of tests at Oxford.

A transistor amplifier, replacing part of the oven control circuit of a quartz crystal oscillator, has been constructed and has operated successfully for six months. No change in accuracy of temperature control, compared with a thermionic valve amplifier, could be detected (less than ±0.º001 C). Since then a stabilized magnetic amplifier has been developed in the laboratory, making it possible, under certain circumstances, to dispense with thermionic valves altogether in the control unit, with the added advantage of drastically reduced power consumption.

A new method of adjusting and checking the phase of the maintaining amplifier of a quartz crystal oscillator without stopping it has been developed. One of the new maintaining amplifiers equipped with this device showed phase variations of less than 0.º17 over a period of four months, corresponding to frequency changes of the order of one part in 10¹¹ or less.

The programme of work on quartz crystal oscillators has now been concluded. A report intended for publication is in course of preparation.

VI. Chronometers.

The movements of Standard Clocks used in the Chronometer Depots at Plymouth, Portsmouth and Gibraltar were overhauled during the year and the Clock Dent 2009 belonging to the Observatory was reconditioned prior to being issued on loan to the Jeremiah Horrocks Observatory.

1957

The Report here presented refers to the period from 1956 April 1 to 1957 March 31 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

(b) Herstmonceux.

The West Block has not been handed over to the Royal Greenwich Observatory, but by special arrangement quartz clocks have been installed. Three oscillators are now running

III. Astronomical Observations.

Time Determinations.

Astronomical observations for time determination have been made throughout the year with Small Transit C in the Courtyard Dome at Greenwich. Data from the Photographic Zenith Tube at Herstmonceux have not so far been utilised in the Time Service.

The internationally-adopted values for the annual fluctuation in the rotation of the Earth, and for the polar variation, have been applied to the time observations in the derivation of the uniform time system, U.T.2.

Clocks.

The ring-crystal clocks at Abinger have continued. to form the basis for the control of the time signals, and supplementary information has been made available concerning the ring-crystal clocks at the Post Office and the National Physical Laboratory. Results of regular comparisons between a ring crystal clock and the caesium resonator at the National Physical Laboratory have also been forwarded to the Observatory: the establishment of U.T.2. has, however, been based solely on the comparisons between the astronomical observations and the quartz clocks. During the year under review, the frequency defined by the caesium resonator and by the time system U.T.2. has been in agreement to well within one part in 10⁸.

Three quartz clocks were set up in cellars in the new West Building at Herstmonceux early in 1957 and are now running. These have been designated H.11, H.12, and H.13, and employ the ring-crystals previously at Abinger in clocks B.6, G.4, and B.5, respectively.

Time Signals.

The schedule of time signals has been maintained as previously, except that the international time signals have been radiated since October from the reserve transmitter Criggion GBZ instead of from Rugby GBR owing to extensive work on the GBR aerial system. Difficulties previously experienced with the control of GBZ from Abinger, which gave rise to an enhanced scatter of the time signal dots, have been substantially reduced by the co-operation of the Post Office in the provision of a telephone line, similar to that used for the control of GBR, enabling the Observatory to transmit the dots in the form of pulses of 1,000 c.p.s. tone.

The various Observatory publications containing corrections to the times of transmission of radio time signals have been prepared and distributed as usual.

1958

The Report here presented refers to the period from 1957 April 1 to 1958 March 31 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

Abinger.
The Observatory Establishment was vacated and handed over to Admiralty Works Department on December 31.

Herstmonceux.
The West Block has been handed over to the Royal Greenwich Observatory. Departments have moved in and work is proceeding satisfactorily.

III. Astronomical Observations.

Time Determinations.

Observations made with the Small Reversible Transit C were used until October 16 in the establishment of the provisional time system employed for the adjustment of the clocks controlling the radio time signal transmissions. It was moved from the Courtyard Dome at Greenwich following an observation on April 23 at 03.9 U.T. and transferred to Herstmonceux, where the first observation was obtained on April 23 at 20.0 U.T.

From October 16 the P.Z.T. observations were used exclusively in current work, and retrospectively from July for the establishment of the definitive time system given in the Time Service Bulletin.

Clocks.

The last of the quartz clocks was removed from Abinger in August; six clocks are now running at Herstmonceux, and one more is installed and under test. The three additional operation clocks set going since the last report are:- H.14, H.15, H.16, incorporating the ring crystals used in the Abinger clocks E6, D5 and D6 respectively. The crystal in clock H.14 developed a fault, and was replaced by that formerly used in Abinger clock C5.

With the co-operation of the Engineer-in-Chief, Post Office and the Director of the National Physical Laboratory a direct link has been installed between Herstmonceux and the National Physical Laboratory, Teddington, by means of which quartz clocks at the National Physical Laboratory and at the Post Office Radio Experimental and Development Branch are continuously displayed on rotary beat counters at Herstmonceux. Daily comparisons are made. Since the National Physical Laboratory make regular comparisons between their quartz clocks and their caesium standard, it is possible to compare the Herstmonceux quartz clocks, and therefore ultimately astronomical time given by the P.Z.T., with the caesium standard. The results of these comparisons are being studied.

Time Signals.

Reception of radio time signals was transferred during the year from Abinger to Herstmonceux as follows:- GBZ. Abinger till July 21; Herstmonceux from July 22. British Short-waves. Abinger till July 31; Herstmonceux from August 1. MSF and all foreign time signals. Abinger till August 14; Herstmonceux from August 15.

Daily reception of GBZ commenced experimentally at Herstmonceux May 28, and from July 1 the 1000 GBZ signals were transmitted from Herstmonceux, but the Abinger reception of the pre-signal test series was used for the adjustment of the time of the transmitted signals. The associated short-wave transmissions at 1000 and all the 1800 signals were controlled from Abinger. This continued until July 22 when all signals were emitted from Herstmonceux. Operational work at Abinger ceased on August 15.

Throughout the whole of the period under review, the Rugby GBR transmitter has been undergoing overhaul, and the 1000 and 1800 signals have been radiated by the reserve transmitter GBZ at Criggion.

Approval has been given to a change in the schedule of the 1000 and 1800 radio time signals. The rhythmic series, with which the Rugby signals were inaugurated in 1927, is to be discontinued after 1958 July 31. The cessation of signals of this type has been recommended by the International Astronomical Union, and will also facilitate various planned improvements in the Time Service.

Control Room.

The new time service installation at Herstmonceux is now virtually complete in all essential features. Advantage has been taken of the opportunity to redesign units and incorporate circuit improvements. All measurements are now made in terms of seconds pulses derived electronically from the quartz oscillators themselves, instead of by means of mechanical contacts on phonic motors, and with a consequent increase in precision. All outgoing time signals are now similarly derived, but the phonic motors operate a safety circuit to minimize the risk of an erroneous signal being emitted. Sidereal time, required for recording the observations in the Meridian Group, is provided by means of a solar/sidereal transformer at 1 kc/s followed by electronic equipment for the derivation of seconds pulses.

Electronics Department.

Ancillary equipment for use with the Hilger and Watts digitizer, and timing equipment for P.Z.T. contacts have been constructed and installed.

A special chronograph drive unit for time signal reception on board ship was constructed for the Admiralty and is now undergoing trials on board H.M.S. Vidal. Three further units are nearly completed.

1959

The Report here presented refers to the period from 1958 April 1 to 1959 March 31 and exhibits the state of the Observatory on the last named day.

II. Instruments.

Astrographic 13-inch Refractor.

Provision has been made for the rapid exchage, twice each month, of the breach end for the Markowitz Moon camea on loan from the U.S. Naval Observatory. The latter is used fors some ten days in each lunation.

III. Astronomical Observations.

Time Determinations.

The Time Service was based on observations made with the Photographic Zenith Tube. The observations were corrected for the effects of Polar Variation and Annual Fluctuation, using the international figures published by the Bureau International de l’Heure.

Moon Camera.

After minor modifications had been made in the Observatory Workshop, the dual-rate Moon camera was mounted on the 13-inch Astrographic instrument in June. As a result of preliminary trials the procedure has been adopted of using the Moon camera during the second and third quarters of the lunation, omitting a few days around full Moon. The exact periods are selected on each occasion both in regard to the position of the Moon and the availability of a suitable star field. Considerable difficulties were experienced inititally in securing satisfactory plates, but during the winter months conditions proved more favourable and some two dozen pairs of plates havs so far been selected for measurement.

In addition two consignements of plates taken at the Government Observatory, Perth, Western Austrailia, have been received in accordance with the International Geophysical Year programme whereby the Royal Observarory acts as a measuring station.

Clocks.

Two more quartz clocks have been installed during the year, H17 and H18, incorporating the ring crystals used in the Abinger clocks, E5 and C6 respectively.

The direct line link with the National Physical Laboratory and Post Office laboratories has been maintained throughout the year, but with some interruptions due to line faults.

The Post Office clock EB serves as a valuable reference standard, having now been in continuous operation for nearly ten years.

The frequency of the National Physical Laboratory clock Q26 is measured every few days in terms of the caesium standard and the results are communicated to the Royal Observatory. This arrangement permits comparisons between the atomic frequency and the frequency defined by U. T. 2 derived from the Herstmonceux P. Z. T. observations.

Signals.

The last rhythmic time signal controlled from the Royal Greenwich Observatory was broadcast at 1800 on June 30.

The reserve transmitter GBZ at Criggion was used for the transmission of the time signals until 1000 on November 8.

The main transmitter GBR at Rugby was restored to service at 1800 on November 3, thus providing a short overlap.

During January the control of the B.B.C. six pips signals and the P.O. hourly signals was changed from the Direct Current signalling system employed hitherto. All external signals are now controlled from the Royal Observatory by means of pulses of 1000 c. p. s , current transmitted by telephone lines.

With the resumption in November of the MSF 60 kc/s standard frequency transmissions from 1430, the apparatus formerly employed at Abinger was brought back into use for the comparison of the transmitted carrier frequency with the standards at the Royal Observatory.

Electronics Department.

An Electronic Automatic Punching Chronograph was built and installed in the Transit Circle Pavilion and has been in constant use since 1958 June. It was based on the prototype instrument installed in 1957.

Modifications to the electronic and electrical equipment at the P.Z.T. have reduced the time spent on maintenance, improving reliability. This work is continuing.

The Flare Patrol Receivers of the Solar Department were overhauled and installed in the Control Room of the Time Department. The associated recorders were moved to the Solar Dome.

Electronic equipment for an integrating exposure meter to be used with the Yapp 36-inch reflector has been completed and awaits tests and installation.

A photoelectric photometer for work at the Radcliffe Observatory was constructed and has been in use during the winter.

Electronic equipment for running nine watch rate recorders was installed in the Chronometer Workshop.

A number of alterations and additions to the Time Service equipment were made in order to improve the service and its reliability. Two further quartz crystal oscillators were installed.

The digitiser fitted to the Mann measuring machine, used for measuring Moon camera plates, was modified to drive a Hollerith 029 Card Punch.

Workshop.

Two Shortt F.P. clocks and a Dallmeyer Equatorial Mount have been sent to the Cape.

1960

The Report here presented refers to the period from 1959 April 1 to 1960 March 31 and exhibits the state of the Observatory on the last named day.

I. Grounds and Buildings.

The original “small Transit” pavilion has been successfully modified to accommodate the Danjon Astrolabe.

III. Instruments and Machines.

Sidereal clock dials, controlled by the quartz oscillators in the Time Department and driven by battery-operated transistor amplifiers, have been installed in all domes at the Equatorial and Meridian groups. All domes and pavilions now have access to sidereal and solar seconds-pulses, and to sidereal and solar 50-cycle and I000-cycle A.C., controlled directly from the Time Department.

Agreement has been reached between the Royal Greenwich Observatory, the National Physical Laboratory, and the Post Office for the co-ordination of the radio time signals transmitted by Rugby GBR and its associated short-wave transmitters with the MSF standard frequency transmissions. Both series of signals will now be controlled from the same equipment at Rugby and will of course still be monitored by the Observatory; as a result, the land-line to Rugby is no longer required. The Royal Greenwich Observatory retains the responsibility for the time signals, and the National Physical Laboratory for the frequencies: maintenance and operation of the equipment will be the responsibility of the Post Office. Agreement has also been reached with the U.S. authorities concerned for the extension of this co-ordination and the adoption of common standards of time and frequency in the control of the transmissions from the U.K., the U.S.A., the Canal Zone and Hawaii. No change is proposed in the publication of final results. It is anticipated that the new arrangements will be of great practical assistance to many users of the services.

IV. Astronomical Observations.

225 plates were obtained with the Photographic Zenith Tube. The average number of stars per plate is 11.1, with four images of each. All reductions are up-to-date.

The latitude and time results have been sent regularly to the Bureau de l’Heure in Paris and the Office of the International Latitude Service in Turin. The latitude results have been incorporated in the Service International Rapide (S.I.R.).

The time results have been used in the establishment of the Observatory Time Service.

Time Department.

The time service was based on observations made with the Herstmonceux P.Z.T., which were corrected for the effects of polar variation and the seasonal fluctuation in the rate of rotation of the Earth, using the internationally-adopted figures supplied by the Bureau International de l’Heure.

The quartz crystal clocks have continued in uninterrupted operation throughout the year. The direct link with the National Physical Laboratory has been maintained with some interruptions due to line and other faults. The results of comparisons made at the National Physical Laboratory between the caesium standard and one of the quartz clocks there are communicated regularly and these, together with the line link, have made it possible to compare the atomic frequency with that defined by U.T.2. as determined from the Herstmonceux P.Z.T. observations. Throughout the year there was no apparent secular change of the U.T.Z time system with respect to the caesium standard.

The atomic standard does not define a continuous time scale but it may be used for the periodical assessment of the frequencies of selected quartz clocks; thus with the aid of the quartz clocks a time scale may be defined, and this has been done independently on the two clocks H12 and EB. Results from the two clocks are in good agreement, but, over a period of four years, the discrepancy between the two clocks has at times amounted to two milliseconds.

The Observatory is indebted to the Director, National Physical Laboratory and the Engineer-in-Chief, Post Office, for their continued co-operation in making available information of value to the time service.

The reception and measurement of radio time signals has continued, and the reception times of the signals in terms of U.T.2. have been published in the Royal Greenwich Observatory Bulletins. Bulletins covering the period 1956 January to 1959 June have been issued: the Bulletin for 1959 July - September is with the Stationery Office.

54 pairs of moon-camera plates have been selected for measurement from plates exposed at Herstmonceux on 62 nights during the year. In addition nine further consignments of plates have been received from the Government Observatory, Perth, Western Austrailia, making a total of 66 pairs of plates received. Of these 39 pairs have been measured, five pairs have been rejected as unsuitable for measurement, and three pairs have suffered damage in transit. Puched cards giving the measures of 51 pairs of plates from Herstmonceux and Perth have been forwarded to u.S. Naval Observatory, where the reductions are being carried out.

1961 June 10

The Report here presented refers to the period from 1960 April 1 to 1961 March 31 and exhibits the state of the Observatory on the last named day.

I. Equipment, buildings and Grounds

Herstmonceux

The electrical control circuits of the P.Z.T. have been completely redesigned and simplified. Automatic operation of the instrument is now possible.

V. Solar System

No plates have been taken with the dual rate Lunar Camera [Moon Camera] at the Cape since 1959 December 31. During the year, 1o4 of the 136 plates received from Tokyo were measured at the Cape and the results forwarded to the U.S. Naval Observatory. The programme at Herstmonceux was discontinued on 1961 January 11. In addition to 104 pairs of plates secured at Herstmonceux, 87 pairs received from Perth have been measured.

VI. Positional Astronomy

During the year 172 plates have been obtained with the P. Z. T. at Herstmonceux.

The Danjon astrolabe at Herstmonceux was used to observe 130, 2-hour groups during the year with an average of 26 stars each. The HEC 1201 computer of the Nautical Almanac Office has been used for reduction of the observations.

VIII. Time Service

By agreement between the Royal Greenwich Observatory, the National Physical Laboratory and the Post Office, arrangements have been made for the unifying of the MSF standard frequency transmissions and the radio time signals emitted by GBR and its associated H. F. transmitters, with effect from 1960 April l. Under the new arrangement the measurement and control of the radiated frequencies is the responsibility of the Director, National Physical Laboratory, and the measurement and control of the time signals is the responsibility of the Astronomer Royal. This unification has eliminated the need for land - lines between the Observatory and the transmitter; all measurement and checking is carried out by means of the radio link.

A further agreement between the United Kingdom authorities and the organisations concerned in the United States, namely the Naval Observatory, the National Bureau of Standards and the Naval Research Laboratory, has extended the coordination on an international basis. Astronomical observations made with the Photographic Zenith Tubes at Washington and Richmond, Virginia are combined with those made at Herstmonceux for the determination of time and rate. From 1961 January l the time signals from GBR, MSF, NBA, WWV and WWVH have been synchronised. Other observatories, including those in Switzerland, Canada and South Africa, have announced their intention to participate in the coordination plan. Each observatory will, as hitherto, publish independent corrections based on their own observations alone.

The U. T.2 time system of the Observatory is established by plotting the P.Z.T. observations, incorporating the corrections adopted by the Bureau International de l'Heure, in groups of weight 50 on several quartz clocks and fitting a smooth curve to the plotted points. The clocks employed have been selected from those available at Herstmonceux, at Dollis Hill (Post Office Laboratories) and at Teddington (N. P.L.). Of the Herstmonceux clocks, H16 suffered an accidental stoppage in September; the rate of H13 is unsteady; H15 behaved erratically and was stopped for investigation, The Post Office clock EB, which operates beat counters at the Observatory, has now completed nearly twelve years of continuous operation, and has proved of great value in the analysis of observations covering a number of years. The frequency of the N. P. L. clock Q13 is measured in terms of the caesium resonance and, since Q13 operates beat counter at Herstmonceux, the calibration may be transferred to other clocks. The scale of atomic time, referred to in last year's report, has been continued. Throughout 1959 and 1960 this atomic scale, Csl ( EB ), has been in agreement with the independent atomic scale, Al, of the U.S. Naval Observatory to within 2 milliseconds. Comparisons during this period between Csl ( EB ) and U. T.2 have shown no significant secular change.

X. Chronometer repair service

The large Harrison timekeeper, loaned by the Royal Astronomical Society, was overhauled and erected in the entrance hall of the West Building. The Harrison No. 2 timekeeper was overhauled and returned to the National Maritime Museum in December; the overhaul of Harrison No. 3 is at present in hand.

1962 June 2

The Report here presented refers to the period from 1961 April 1 to 1962 March 31 and exhibits the state of the Observatory on the last-named day.

III. Instrument and Machines

(a) Herstmonceux

Ring crystal No. 69 was sent to the G. P. O. for investigation; after a complete renovation it was installed as oscillator H19 in 1961 September. Oscillator H13 was taken out of service in 1961 September and its ring - crystal and oven are being used in an experimental, prototype, transistorized oscillator.

Test equipment has been installed for the MSF 60 kc / s standard - frequency and time - signal receivers.

A new phase - recording and time - signal receiver NBA (18 kc / s) has been designed and constructed. A similar receiver for GBR (16 kc / s) is under construction as well as a phase shifter to permit one of the local frequency standards to be continuously adjusted to be within 1 part in 10¹⁰ of a nominal 100 kc / s.

H. Time Service

During the year 190 P. Z. T. plates were obtained with an average of 11 stars per plate. The time and latitude results are reduced on the I.C.T. 1201 computer and communicated weekly to the Bureau de l'Heure and the Office of the International Latitude Service (now the International Polar Motion Service). Work on a Bulletin giving the collected P. Z. T. time and latitude results for the years 1960–61, and the results for 1958–59 in a revised form, is in hand.

The P.Z.T. working catalogue has been modified for the 7-year period 1965–71 on account of precession. Observation will shortly begin of the additional stars required, in order to determine their places relative to the current groups.

The Danjon astrolabe has been in regular use throughout the year; 244 groups were observed representing 7000 individual observations. Using the I.C.T. 1201 computer the equations of condition for the 11000 observations made since 1959 September have now been obtained. The results of 125 “standard” groups have been computed and preliminary star corrections obtained to refer any individual star to the standard group system. The reduction of all observations with these preliminary corrections will shortly be completed. A preliminary comparison with the results of the Paris astrolabe indicates that the results are of comparable accuracy.

The scale value of the astrolabe screw, deduced from 10,000 observations, shows a slight dependence on temperature, but the effect on the results is negligible. Investigation of 460 values of v_o (micrometer reading for which the effective zenith distance of observation equals the angle of the prism ) does not show the simple dependence on temperature found for the Neuchâtel astrolabe.

A new weekly series of circulars was commenced jointly with the U.S. Naval Observatory in August. These circulars give, for the current week, emission times of the co - ordinated signals in preliminary U.T. 1 and U. T.2 time systems, and in terms of an atomic time scale, A1. Provisional co-ordinates of the pole are also given in order that corrections due to the effects of polar motion may be applied for any observing site. These U.T.1 and U.T.2 time systems are computed by the Royal Greenwich Observatory from P. Z. T. observations made at Herstmonceux, Washington and Richmond. The atomic time scale is computed in the U.S. and based upon caesium standards in the U.K., U.S.A., France, Switzerland and Canada.

No change has been made in the U.T.2 time system published in the Bulletins, which is based solely upon P.Z.T. observations made at Herstmonceux. Increasing use has been made of the caesium standard at the National Physical Laboratory in the computation of a clock whose rate may be relied upon to be uniform. The full potentialities of the atomic time scale have been, however, difficult to realize in practice, owing to the technical and operational difficulties inherent in comparing standards at different sites; the present arrangements represent a compromise. Despite the wholehearted co-operation of the engineers in the Post Office Lines Branch there have been interruptions, due either to the line or the terminal equipment, and the quality of the linkage between the best quartz clocks and the caesium standard has been barely adequate for the stringent requirements of the time service.

The difference in the emission-times of the GBR and WWV signals, working in the “Co-ordinated” system adopted last year, has been kept below two milliseconds, but in principle this could be improved on.

The signals in both countries are emitted on the basis of a rate adopted for the whole year, together with a jump of 50 ms on a pre-arranged date when the accumulated error due to changes in the Earth's rotation makes this seem advisable. The rate for 1961 was set at 150 parts in 1010 below that corresponding to the nominal caesium second, and the rate for 1962 is 130 parts in 10¹⁰ below it. One jump was made in August.

On December 1 the twice-daily time-signal transmissions from GBR were replaced by four transmissions at 0300, 0900, 1500, and 2100 G.M.T. The supplementary short-wave transmissions are still radiated twice daily, but at 0900 and 2100, The increase in the number of signals was originally requested by the survey authorities, but is now essential for the full benefits to be derived from the international co-ordination of radio time signals.

The steadiness of the GBR signals is such that it is sufficient now to monitor only those at 0900 and 1500 on Mondays to Fridays; weekend and evening duties have been eliminated. The number of foreign time-signals which are regularly received has also been restricted. The beat-counters which monitor the various clocks are automatically photographed twice a day.

VII. Nautical Almanac Office

Machines.

The I.C.T. 1201 computer was used … for the reduction of data from the R.T.C., P.Z.T. and Astrolabe (579 hours).

1963 June 15

The Report here presented refers to the period from 1962 April 1 to 1963 March 31 and exhibits the state of the Royal Greenwich Observatory and of the Cape Observatory on the last-named day.

II.Grounds and Buildings

(a) Herstmonceux

The removal of the Chronometer Office from the Castle to the rating and storage rooms prepared in the West Building was completed in May.

Exceptionally heavy snowfalls in January brought down 15 out of the 40 antennae of the "umbrella" aerial near the West Building and thereby caused gaps in the reception of certain time signals.

III. Instruments and Machines

(a) Herstmonceux

The photographic zenith tube has been completely overhauled for the first time since its installation in 1955. All the control circuits have been renewed and the console has been modified to simplify future maintenance. The photographic chronograph has been replaced to determine whether the somewhat defective performance of the contact system is due to faulty contact operation or to some irregularity in the motion of the plate carriage.

A new maintaining amplifier for oscillator H 11 has been constructed and installed.

New phase-recording equipment for GBR 16 kc/s, made and brought into service for the Telstar experiment (see Section IV H ), proved suitable for making time measurements to an accuracy of a microsecond.

(b) Cape

The pendulum clock installation has now been completely replaced by two quartz crystal oscillators previously in use at Greenwich and Abinger. The associated equipment has been designed and constructed at the Cape. The time service has been operated from these oscillators since June, and on January 1 the former ONOGO shipping signal was replaced by one having the standard English form.

IV. Astronomical Observations and Reductions

H. Time Service

During the first quarter of the year, 56 plates were obtained with the P.Z.T. and the results communicated weekly to the Bureau de l'Heure and the office of the International Polar Motion Service. From July 23 advantage was taken of the availability of the astrolabe for current time determination to overhaul the P.Z.T. (see Section III). Regular observations were resumed on February 11.

With the Danjon astrolabe, 249 groups were observed during the year. Reduction of the accumulated observations was brought up to date in May. Corrections have been derived to the adopted zenith distances of each star at transit, and group corrections have been determined in time and latitude. No significant closing error in time has been detected, but the closing error of 0”.25 in latitude can be compensated by an empirical correction of -0".06 sin (ʘ-α) to the observed latitude. The difference between the astrolabe (standard observer) and the P.Z.T. for the year 1961 was astrolabe 0^s.245 west and 1”.49 south of the P.Z.T.: the measured intervals on the ground are 0^s.240 and 2”.00. The apparent discrepancy is mainly due to the different systems of star places used for the two instruments.

Since July the astrolabe has been used instead of the P.Z.T. to define the Herstmonceux system of time and latitude. Empirical corrections of -0^S.005 and +1”.49 have been applied to the astrolabe results to preserve the continuity of the system.

Co-ordination between the time and frequency services of the United Kingdom and the United States of America has continued. The difference between the emission times of the radio time signals from GBR and WWV has not exceeded three milliseconds. The carrier frequencies of the co - ordinated transmissions have been offset from values based upon 9 192 631 770 c / s as the nominal caesium frequency by an amount, agreed internationally, which brings the time signals into general agreement with U.T.2. The offset for the calendar year 1963 is -130 parts in 10¹⁰, as for 1962.

In August a programme was devised at the National Physical Laboratory for the utilization of the artificial satellite “Telstar” in an attempt to ascertain the difference between the time systems of the U.K. and the U.S.A. The experiment involved the simultaneous transmission in both directions over the Telstar circuit of precise time signals from quartz clocks installed at Goonhilly U.K ) and Andover (U.S.A.). The very high frequencies used, 4 and 6 Gc s, are unaffected by ionospheric conditions. This experiment established that at the end of August the U.S. signal WWV was emitted 1.6 milliseconds in advance of the U.K. signal MSF (and GBR).

The caesium standard is now of fundamental importance in time determination and dissemination. Astronomical time, which is required wherever measurements are made from the earth, is not uniform; but for many purposes a uniform time scale of great accuracy is required and this can be obtained by integration from quartz clocks calibrated in terms of the caesium standard. As a frequency standard it is used to calibrate the carrier frequencies of radio time signals. If the signals on both sides of the Atlantic are to agree within one millisecond, and this is the minimum requirement today, then the carrier frequencies must be accurate to better than one part in 10¹⁰. A persistent error in frequency of one part in 10¹⁰ throughout the year leads to an error of three milliseconds in time. Of the present quartz clocks available only H12 and EB approach modern requirements.

VI. Chronometer Service

… In addition, two astronomical regulators by George Graham and Robert Molyneux were overhauled in the chronometer workshop, in addition to the Harrison No. 4 timekeeper mentioned in Section I.

1964 June 6

This Report has been drawn up in a different form from those of previous years, in an attempt to present a more readable account of the year's work separate from (whilst not excluding) detailed information that may be found of rather specialized interest.

The narrative Section (pp. 5 - 14) is intended to be read as a synoptic account of research activity at Herstmonceux and the Cape during the year. It makes no reference to routine work which, important though it is, changes little in nature from year to year. An abbreviated account of this latter work, and such details of the former as are unsuited to a narrative report, are contained in the Appendix (pp. 15 - 24).

Appendix

Instruments and Machines

(a)Herstmonceux

An experimental photoelectric chronograph for timing lunar occultations to 0^s.01 has been constructed, using a two-hole light-switching technique to overcome the variable degree of scattered light encountered. Static recordings show the equipment to be practicable but adverse weather conditions have prevented adequate tests on the astrographic refractor.

Two new Sulzer quartz crystal oscillators already tested at the U.S. Naval Observatory were installed towards the end of March. The lenticular AT-cut crystals oscillate at a nominal 2.5 Mc/s in transistorized driving circuits. An output frequency of 100 kc/s allows direct comparison with the existing ring oscillators. The initial performance of the new oscillators is highly satisfactory.

Time Service

Observations with the P.Z.T. were interrupted in April to study the performance of the plate-carriage driving mechanism. After removal of some constriction of the horizontal constraints, the motion of the carriage was found to be satisfactorily uniform and rectilinear. The contact system was also improved. Since routine observations re-started at the end of July, 89 plates have been secured.

Between April and September 108 groups were observed with the astrolabe, bringing the total for its 4 - year service to 760. Corrections to the adopted R.A. and Dec. have been obtained for the 171 stars observed at both rising and setting transits across the almucantar.

Forthcoming issues of the Royal Observatory Bulletins will describe both the P.Z.T. and the astrolabe, and the methods of reduction used with each instrument.

Co-ordination of the time signals emitted by GBR in the U.K. and by WWV in the U.S. has been very satisfactory. The difference in emission times has been between two and three milliseconds throughout the year, the variation being well within one millisecond.

Except where indicated, all text and images are the copyright of Graham Dolan