Airy’s 1843 address to the Board of Visitors arguing the case for a new altazimuth instrument

The text below was published as part of the 1844 Report of the Astronomer Royal
Click here to read it in its original format.



Read at the Special Meeting of the Visitors on 1843, November 10.


THE most important object in the institution and maintenance of the Royal Observatory has always been the Observations of the Moon. In this term I include the determination of the places of fixed stars which are necessary for ascertaining the instrumental errors applicable to the instrumental observations of the Moon. These, as regards the objects of the institution, were merely auxiliaries: the history of the circumstances which led the Government of the day to supply the funds for the construction of the Observatory shews that, but for the demands of accurate Lunar Determinations as aids to navigation, the erection of a National Observatory would never have been thought of. And this object has been steadily kept in view when others (necessary as fundamental auxiliaries) were passed by. Thus, during the latter part of Bradley's time, and Bliss's time (which two periods are the least efficient in the modern history of the Observatory), and during the latter part of Maskelyne's presidency (when, for years together, there is scarcely a single observation of the declination of a star), the Observations of the Moon were kept up with the utmost regularity. And the effect of this regularity, as regards its peculiar object, has been most honourable to the institution. The existing Theories and Tables of the Moon are founded entirely upon the Greenwich Observations; the Observatory of Greenwich has been looked to as that from which alone adequate observations can be expected, and from which they will not be expected in vain: and it is not perhaps venturing too much to predict that, unless some gross dereliction of duty by the managers of the Observatory should occur; the Lunar Tables will always be founded on Greenwich Observations.

With this impression, it has long been to me a matter of consideration whether means should not be taken for rendering the series of Observations of the Moon more complete than it can be made by the means at present recognised in our observatories. The interruptions of the regularity of the series are, at present, very numerous; and the uniform suppression of observations during the Moon's passage through entire portions of her orbit of great extent is in reference to theory, extremely injurious. On examining the printed observations during seven years in which I have directed the Royal Observatory, I find that there are but two observations of the Moon within three hours of the Sun, one of which is imperfect. And the whole number of observations within six hours of the Sun is small as compared with the remainder. Thus one-fourth of the Moon's course is absolutely lost, and one-half is very imperfectly observed. It is needless for me to point out the influence of this deficiency upon the Variation, and several theoretical terms whose argument is related to the Variation. But, besides this, the loss of observations (sometimes from clouds beginning to obscure the Moon when it has actually entered the field of the meridional instrument) in those parts of the revolution which may be considered observable, is a most serious misfortune. For the Sun or a Planet, it is of little importance that the observations of one or several days are lost: for the Moon, it is of great importance, because the relation of sequence in its rapidly-changing Errors of Tables is thus entirely lost to the view.

If, now, we had the means of observing the Moon in any part of the sky, the deficiencies to which I have alluded would be, in a great measure, supplied. In fine weather, the Moon might perhaps be always observed within an hour and a half of the Sun, and (when Cancer is near the meridian at the time of observation) probably nearer. In other parts of the lunation, the Moon might be observed almost every night: for there are very few nights when the Moon is above the horizon in which she cannot be seen pretty well at some time. The number of complete observations made at present is under one hundred in a year: I should expect that, with proper extra- meridional means, the number of nights of efficient observation might be made to exceed two hundred.

After careful consideration of the ways in which these observations might be made, I have come to the conclusion that there is but one instrument with which observations could certainly be made possessing the required accuracy, and that with it the observations certainly could be made. This instrument is the Altitude and Azimuth Instrument. No form, however, in which I have seen this instrument appears to me to be sufficiently firm for this purpose (in which the azimuths are as important as the altitudes). I should propose to construct a new instrument, with circles of three feet diameter, in general form resembling the Palermo or Dublin circle, but framed on the same principles of massiveness and strength and with the same exclusion of adjusting power, which I have adopted in the Ordnance Zenith Sector.

There is one position, within the precincts of the Royal Observatory, which appears to me well adapted for such an instrument. It is on the south side of the Chronometer Room and South Dome, and between the two windows of the former. Another position might be obtained by dismantling the Advanced Building and erecting a dome, &c., on its site. In either case it would be necessary to build a pier for the instrument and walls for the dome, to the height of two low stories. The former position would command a free horizon throughout the Moon's path, excepting that in the north-west (a part of little importance, because the air is thickened by the smoke of London) the erections at the top of the Octagon Room rise to the elevation of eight degrees above the horizon. The access to this place would be either through or by the side of the South Dome. In the latter position, the horizon would be somewhat less interrupted. I should propose to cover the instrument with a drum-dome, and should keep it at such a height as not to interrupt the view from the South Dome.

Upon all details connected with the construction of the building, the construction and mounting of the instrument, the method of observing with the instrument, and the method of computing the results, I am fully prepared to explain my views.

I think it probable, that the use of this instrument would in some shape occupy the time of one observer. Its running expense might thus be estimated at £150 per annum. Of the first expenses I can give no accurate estimate; but I should suppose that they would be £700 or £800. To give an idea of the relation of these sums to the importance of the object aimed at, I may state that I consider the Lunar Observations to be justly chargeable at present with one-third of the whole expenses of the Observatory, or with £1000 per annum; and that I have always considered one complete lunar observation to be worth £10.

I now beg leave to submit these suggestions to the consideration of the Board of Visitors, and to request their judgment as to the propriety of making a representation to the Board of Admiralty, with the view of carrying into effect the proposed scheme.



Royal Observatory, Greenwich, 1843, Nov. 4.




Read at the Special Meeting of the Visitors on 1843, November 10.


WITHIN the circular wall which supports the Dome, I propose to build a cruciform pier on a separate foundation, having no connection with the circular wall from the ground, but approaching near to it at the extremities of its cross-arms. Instead of a cruciform pier, a three-rayed pier may be used: but as the same description, with small alterations, will apply to both, I shall confine my description to the cruciform shape. This cruciform pier I propose to raise to within a small distance (from one to two feet) of the floor of the dome, so that it will not be touched by the joists of the floor, which are to be supported entirely by the circular wall. The internal diameter of the circular wall needs not to exceed 10 feet (and may probably be less): the length of the arms of the cross may be about 9½ feet. Upon the top of this cross I propose to plant an iron cross (or three iron rays, if the pier be three-rayed), the length of whose arms exceeds by a small quantity the length of the arms of the pier-cross, but so little that it does not touch the circular wall (its length may be 9¾ feet): the extremities of the four (or three) arms of this cross to be connected by four (or three) sides, welded in the same piece: thus the iron work would form a square (or triangle), whose sides are about 7 feet, and which has diagonals in the same piece of metal. Upon each side of the square, &c., I propose to erect an iron triangle, connected by screw-bolts to the angles of the square, and rising about 10 feet above the square. The four points forming the four vertices of these triangles would be four angles of a square to be attached to them, of the same dimensions as the lower square; but lying in such a position that its angles correspond to the sides of the lower square, and vice versa. Of this upper square I propose that the sides be in one piece, but that the crossed diagonals be in another piece, and that they rest in forks at the angles of the square, their ends being cut with a screw-thread so that they can by means of nuts be drawn endways. This power of screw-motion is for the purpose of adjusting the place of the intersection or center. Every drawing-strain upon one of the arms will imply a bending of the transversal arm. At the intersection or center will be the Y for the upper pivot of the azimuthal motion, into which I' the pivot will be forced by a long piece of iron or steel screwed upon it, whose spring will insure perfect bearing. In this manner I conceive that most perfect firmness would be given to the upper pivot of the azimuth motion. No adjustment nicer than that depending on the screws, at the end of the cross is to be provided for it.

Upon the center of the brick cross, and covering the center of the lower iron cross, I propose to plant a stone pier. In the top of this I would insert three metallic forks, very solidly made. These forks are to receive three ribs of the lower fixed circle. This circle I propose to be a solid flat cone of iron or bell-metal, 3 feet in diameter, and 6 inches thick in the center, with a conical hole for the bearing of the lower azimuthal pivot, a racked ring for pinion-action, a smooth ring for clamp-action, and a graduated ring at its circumference. All these are to be turned at the same chucking on the lathe. The circle will be laid with its ribs in the forks, and then examined by the artist for horizontality, with a spirit level. As he finds it wrong, he will correct its position by filing the forks. When it is adjusted in a satisfactory way to horizontality, it will never be moved again. Thus the bearing of the lower azimuthal pivot, and the azimuthal graduated circle with its slow-motion circle and clamp circle, are provided for.

For the lower part of the azimuthal frame, I propose to use a strongly-ribbed flat of cast iron or bell-metal (the ribs being on its lower side, and 3 inches or more deep), with a vertical pivot having a hemispherical end to work in the conical hollow of the lower circle. The whole of this should be cast in one piece. The upper surface should be planed. The length about 33 inches and the breadth 21 inches. This basis would carry no fine work, except the clamp-screw, and the slow-motion pinion.

For the two uprights, I propose to use two trough-shaped pieces with their lower ends planed (the opening of the trough being covered as far as the insertion of the horizontal axis permits), each about 4 ft. 6 in. high, the breadth of its trough about 21 inches, and the depth 10½. One of these uprights to carry two microscopes for reading the lower circle; the other to carry two microscopes for reading the lower circle, four for reading the vertical circle, and blocks for carrying the levels transversal to the horizontal axis. Each also to carry blocks for the levels parallel to the axis. Each vertical trough with the parts that I have mentioned to be cast in one piece, as in the Ordnance Zenith Sector. The vertical which does not carry the four microscopes will carry the semicircles in which the clamp-screw and slow-motion pinion act.

The upper connection of the upright troughs will be only a flat piece 9 inches broad, ribbed, and carrying the upper pivot.

By the sides of the upper connecting piece will be two levels parallel to the horizontal axis, resting on blocks which are near to the outside extremities of the vertical supports, and which are attached to those supports. Perhaps, however, they may be more conveniently placed at the bottom of the supports.

Two levels transversal to these and to the horizontal axis will be placed on the outside of the vertical which carries the four microscopes, resting on blocks which are attached as near as possible to these microscopes.

The Y's for the support of the horizontal axis will be cast in the trough pieces, and will be adjusted only by filing.

This completes the description of the azimuthal axis.

The part moveable in altitude is to consist of a double cone (each cone being widened to a flat at its base) with the telescope between the bases of the cones. If possible, one cone (carrying the graduated circle), its pivot, and the telescope, should be cast in one piece, and the rest in another piece. On the part opposite the graduated circle, and near to the eye end of the telescope, would be the clamp-screw and the head of the pinion for slow motion.

The telescope should have crossed wires in the field, six horizontal and six vertical. The plate containing them should be firmly fixed, their adjustment being effected as nearly as possible by the instrument-maker.

This completes the description of the instrument.



The following would be the method of making observations with the instrument.

The observation, as regards the celestial body, would be entirely one of transits; whether the observation were for azimuth or for altitude.

For an observation of azimuth, the azimuth clamp-screw would be pinched, and the observer would at transit over every wire move the telescope and vertical circle by means of the pinion acting on the vertical racked semicircle, so that every transit should be observed over the middle part of each of the vertical wires. The four microscopes of the lower or horizontal circle would be read, and likewise the two levels which are parallel to the horizontal axis.

For an observation of altitude, the clamp-screw of the vertical circle would be pinched, and the observer would at transit over every horizontal wire move the azimuthal frame by means of the pinion acting on the lower racked circle, so that every transit should be observed over the middle part of each of the horizontal wires. The four microscopes of the vertical circle would be read, and also the two levels which are transversal to the horizontal axis.

Whichever observation be made, the instrument should be immediately reversed in azimuth, and a new observation of the same kind should be made. The clock should be immediately compared with the transit-clock, the comparison being made by the coincidence of beats of a half-second solar chronometer.

This completes the observation.



The following would be the process of computation: –

By the comparisons of clocks, the clock-time of transit is to be reduced to transit-clock-time, to sidereal time, and (for the Moon) to mean solar time. The tabular R.A., N.P.D., and hour angle, of the objects observed (Stars and Moon) are to be prepared, and with these and the astronomical colatitude for the Stars and the geocentric colatitude for the Moon, the tabular azimuth and zenith distance of every object (Stars and Moon's center) are to be computed.

As regards the Moon, this azimuth and zenith distance are at present referred to the geocentric zenith, and are therefore in a state fit for the application of parallax. Therefore, applying first the geocentric semidiameter, the geocentric Z.D. of the limb from the geocentric zenith is found; and, correcting this for parallax, the Greenwich Z.D. of the limb from the geocentric zenith is found, which, by an easy computation, is changed to Greenwich Z.D. of the limb from the astronomical zenith. Also, with the geocentric Z.D. of the center from the geocentric zenith, and the geocentric semidiameter, the geocentric azimuth of the limb is found, and this is easily converted into astronomical azimuth of the limb.

Thus all the tabular places in zenith distance and azimuth, for comparison with observations, are found.

Now for the observed place: –

1. In Altitude. – An approximate zenith point being known, the zenith distance of a star can be expressed from observation in the usual way (from the microscope readings of the vertical circle and level readings). Every single observation of a Star, corrected for refraction, will give a zenith distance comparable with the tabular zenith distance, and will thus give a correction of the zenith point. But it will be prudent to take the correction   from observations in reversed positions, because the zero of the levels will be eliminated, and, when the clock-time is suspicious, it may be corrected from this observation.

The corrected zenith point thus found is to be used for the Moon's limb, and, after correction for refraction, the true zenith distance is found, which will be comparable with the tabular zenith distance.

2. In Azimuth. – Before a zero of azimuth can be found, corrections are requisite  – For the readings of the levels parallel to the horizontal axis-For the inclination of the horizontal axis – And for the error of collimation of the mean of the vertical wires in the telescope. The first is easily computed. The third may be found by observation of a distant object in the horizon, taking the lower circle reading when each of the wires is brought on it Face Right, and, again, when each wire is brought on it Face Left.

The second may be found by observation of a tolerably high Star, Face Right and Face Left, correcting for the first and third, and determining the second so that the difference of azimuth-readings Face Right and Face Left shall be the same as the difference of the tabular azimuths.

Or the second and third may be found together, by observations Face Right and Face Left, of a high Star and a low Star.

Every observation of a Star is then to be corrected for these three elements, and then it will give an azimuthal zero.

By applying this, with corrections for the three elements, to the observation of the Moon, the true azimuth of the Moon will be found.



In this manner there would be found a series of errors of the Moon's place in zenith distance and in azimuth, at times not exactly the same. From these it will be easy to infer the errors at one instant; and then there are two courses open. Either to apply these errors to the Moon's tabular Z.D. and azimuth computed for that instant, and (reversing the computation) to infer from them the R.A. and N.P.D. at that instant, which will be compared with the tabular R.A. and N.P.D. Or, from the errors in Z.D. and azimuth, to find the errors in R.A. and N.P.D. by the use of factors similar to P, Q, R, S, in the Appendix to the Greenwich Observations, 1836. I prefer this latter course.

The duty imposed upon the observer would be, in every fine night to observe at least two -well-determined Stars, and to observe each of these in altitude and in azimuth, Face Right and Face Left: making in all at least eight observations.

And, whenever the Moon is visible, to observe at least one altitude and one azimuth, and, if possible, two altitudes and two azimuths.

And, in order to be certain about the Moon's diameter, it is necessary that he should be prepared beforehand with calculations of the time when the Moon's horns are vertical; and at such times he must observe repeated zenith distances of the Moon, upper limb and lower limb alternately.

It is indispensable that the observer be prepared to watch from Moon-rise to an hour or more after Sun-rise, or from an hour or more before Sun-setting to Moon-setting, as the case may be.



1843, November 4.