212 Publications of the Some Pioneer

Home , Frank Schlesinger, James Bradley, Lewis Morris Rutherfurd, Meridian circle

212 PUBLICATIONS OF THE

SOME PIONEER OBSERVERS1

By Frank Schlesinger

In choosing a subject upon which to speak to you this evening, I have had to bear in mind that, although this is a meeting of the Astronomical Society of the Pacific, not many of my audience are astronomers, and I am therefore debarred from speaking on too technical a matter. Under these circumstances I have thought that a historical subject, and one that has been somewhat neglected by the, formal historians of our science, may be of interest. I propose to outline, very briefly of course, the history of the advances that have been made in the accuracy of astronomical measurements. To do this within an hour, I must confine myself to the measurement of the relative places of objects not very close together, neglecting not only measurements other than of angles, but also such as can be carried out, for example, by the filar micrometer and the interferometer; these form a somewhat distinct chapter and would be well worth your consideration in an evening by themselves. It is clear to you, I hope, in how restricted a sense I am using the word observer ; Galileo, Herschel, and Barnard were great observers in another sense and they were great pioneers. But of their kind of observing I am not to speak to you tonight. My pioneers are five in number ; they are Hipparchus in the second century b.c., Tycho in the sixteenth century, Bradley in the eighteenth, Bessel in the first half of the nineteenth century and Rüther fur d in the second half. Newton was undoubtedly the greatest astronomer of all time; Hipparchus, in the opinion of many a competent judge, stands next to him. Historians in the field of literature have often lamented the fact that we know so little of the life of Shakespeare, but our knowledge of him is full and voluminous as compared with what we know of Hipparchus, whose position in science is of the same order of brilliance as that of Shake-

1 An address delivered at a special meeting of the Astronomical Society of the Pacific, at New Haven, Connecticut, on April 11, 1929, at which the Bruce Gold Medal of the Society was presented to Dr. Schlesinger.—Editors.

speare in literature. All that we know with certainty of Hip- par chus is that he was making observations from the island of Rhodes around the year 150 b.c. We do not know when or where he was born, nor when or where he died. Concerning even his work we know practically nothing at first hand. With the exception of one unimportant fragment of his own, our knowledge of what he did comes from the writings of an astronomer who lived three centuries later. Of his many contributions to the science, it is only to his success as an observer that we can now refer. Gnomons of various forms, the tall vertical obelisk among them, were the earliest astronomical instruments. By their aid the ancients became acquainted with the Sun's apparent motions from hour to hour and from season to season. The only other instrument to be used extensively before the time of Hipparchus was the crude cross-staff, two sticks or strips of metal at right angles to each other, the one containing two sights like those on a modern rifle; the other engraved with a scale of tangents, and movable through the center of the first ; this device can be made to yield the angular separation of any two objects. Hipparchus used, and probably invented, some forms of the armillary sphere and of the astrolabe, an arrangement of three or more circles, capable of giving in the hands of a very skilful observer differ- ences in celestial latitude and celestial longitude between two objects. With instruments of this type Hipparchus was able to set down the positions of the planets and the stars with an average error that was in the neighborhood of four minutes of arc, or the equivalent of about one-eighth the angular diameter of the Sun or the Moon. The rough character of these observations is, I think, surprising at first sight, especially when we consider that they were not excelled until seventeen centuries after Hipparchus' time. Surely it ought to be possible to point to a star with an error much less than one-eighth the diameter of the Sun or Moon. I have in my hand a target at which a marksman has fired ten shots from a distance of fifty feet. The ten holes are merged together and on the average they deviate from their common center by less than a tenth of an inch, or about thirty seconds of arc. This

accuracy was attained by means of two ordinary sights, very much like those used by Hipparchus, and separated by about the same distance. We are led by such considerations as this to infer that the accuracy of Hipparchus' observations depended to only a very minor extent upon skill in pointing. The errors in graduating his circles were a more important source of inac- curacy. We must remember that these graduations had all to be made by hand and without any magnifying device whatsoever. On a circle thirty inches in diameter, a minute of arc corresponds to about 0.004 inch, and it is therefore not surprising that Hipparchus' observations were not more accurate, nor that they held the record for accuracy for so many centuries. Considering the state of the mechanic arts of the times, it was obvious that the only way in which the accuracy of astronomical measurements could then be increased would be by using larger graduated circles. But this was impractical with the types of instruments that were in use, and with the way they had to be used. The armillary sphere, for example, had to have its circles adjusted to certain planes, and these planes were con- stantly changing their apparent positions on account of the rotation of the Earth. For these reasons no considerable improvement over the accuracy attained by Hipparchus was possible until a new type of instrument was invented; one that remained stationary and that could therefore be made large enough to contain graduated circles of much greater radius. It was Tycho Brahe (1546-1601) who made this important invention. Tycho was the eldest son of a Danish nobleman, and therefore had the choice of a very different career from a scientific one. But his tastes ran so strongly in this direction that against the wishes of his father, and for the most part secretly, he studied mathematics and astronomy at the two uni- versities he attended. In the end he expended the whole of his considerable fortune, inherited from his father and from an uncle, in the construction of great astronomical instruments. These were housed, as all of us know, in palatial establishments on the island of Hveen and later at Prague. Tycho's position as one of the great leaders of our science is conceded by all ; and yet, in my opinion, his contributions have been underrated.

This may be on account of his theory of cosmogony, known as the Tychonic System, which placed the Earth at the center of the motions of the Sun and of the Moon, but made all the other planets revolve about the Sun. This theory has ever since been universally characterized by historians as "unfortunate/' "retro- grade," and the like. But no theory ever fitted all the facts, as they were known at the time, more perfectly than did this one. I wish I had time this evening to defend the Tychonic System from this point of view, but I must not wander too far from my subject. Tycho's most striking instrument was the mural quadrant, a quarter circle no less than nine feet in radius fixed on a north- and-south wall. At its center was a stationary sight and on its circumference a movable one. As an object crossed the meridian the sights were trained upon it, the time of meridian crossing noted, and the position of the movable sight on the circle recorded, thus furnishing information from which the right ascension and the declination could easily be deduced. On the circumference of such a circle one minute of arc corresponds to .032 inch, and this Tycho could subdivide into six parts, each one of which corresponded to 10 seconds. The average error of his b^st observations was about 40 seconds, a very great improvement over anything that had been done before. There is much evidence to show that Tycho had very defi- nite objects in mind in making all his observations, and there can be little doubt that had he been spared (he died at fifty- five years) he would have examined his planetary observations from the same point of view as actually did his friend and pupil, Kepler, after Tycho's death. Upon these observations Kepler built his three laws of planetary motion, and upon these in turn Newton erected his proof of the law of universal gravi- tation. Less than ten years after Tycho's death came the invention of the telescope and its application to astronomy. This enabled pointings to be made with extreme nicety, but for the same reason that we have just been emphasizing, it did not bring about any immediate or considerable improvement in accuracy, nor could it do so until better methods of graduating and read-

ing circles were devised. This is the explanation of fact that has puzzled so many commentators: that Helvelius (1611- 1687), for example, was able to make observations with ordinary sights upon instruments like Tycho's nearly or quite as accurate as those of his contemporaries who used the telescope. We come now to the long period in astronomy during which the transit circle was the principal instrument in every observ- atory. In its essential features this is very like Tycho's mural quadrant except that the telescope is substituted for his sights. The history of this instrument is a fascinating one. Since its invention by Roemer (1644-1710) every generation of astronomers, including our own, has made considerable improvements in its construction or the manner of using it; and now I think it can be said that no piece of purely scientific apparatus em- bodies so much thought as does the modern transit circle. It was in the hands of James Bradley (1692-1762) that the transit circle first gave results of the order of accuracy that modern astronomy demands. Bradley was graduated, at Oxford in 1714 and took orders, as did many others in that day whose real inclinations were in other directions. Through influential friends he secured two livings, but as one of his biographers remarks with, I think, unconscious humor, his duties in connection with these appointments do not seem to have interferred with his work. Later he was appointed Savilian Professor at Oxford and still later (1742) he succeeded Halley as Astrono- mer Royal. He held both these posts until his death in 1762. In 1718 Halley had discovered that the stars are not abso- lutely fixed with respect to each other, but that they have proper motions. It is curious that this discovery came so late. It was made in Bradley's lifetime and only ten years before he discovered and correctly interpreted that much more subtle phe- nomenon, the aberration of light. Among the stars observed by Hipparchus and Ptolemy so many centuries earlier, there are eleven whose motions exceed one second per annum. The slowest of these moved 27 minutes across the sky (nearly equal to the angular diameter of the Moon) in the interval between the observations of Ptolemy and those of Tycho. The bright star Arcturus moved a degree in the same interval, and μ Cas-

siopeiae (the most rapid star in Hipparchus' list) a degree and a half. These quantities are much greater than could be as- cribed to errors of observation ; nevertheless it was reserved for Halley, more than a century after Tycho's death, to recognize that real changes were concerned. Bradley was the first to see the importance of these motions ;,as early as 1748 he states that the motions of the stars may in part be due to a perspective re- flection (as it were) of the motion of our own Sun among the stars. He thought that to determine the latter motion observations extending over very many years would be required. He could not foresee that the man who would first determine the direction of the solar motion, William Herschel, was already on the scene at the time he made this remark. At any rate, like a true and loyal observer he set about making the observations that would form the basis for this investigation by some future thinker, whoever he might be and whenever he might appear. This is the origin of Bradley's truly wonderful catalogue of the positions of more than 3000 stars, a catalogue that Bessel well named "Fundamenta Astronomicae." The average error of a single observation by Bradley is 2"2 in right ascension and 1^7 in declination; as each star was observed on the average five times, the average error of a catalogue place is 1" or less. No one had come anywhere near such precision before, and no one surpassed it or even equaled it till more than half a century after Bradley's death. Though nearly two centuries have gone by since these observations were made, they are still used in the determination of proper motions ; "and until comparatively recently, the proper motions of only those stars that had been observed by Bradley could be accurately derived. In addition to his observations with the meridian circle, Bradley made many observations with several zenith sectors, and it was from these that he discovered both the aberration and the nutation. This form of instrument is a long hanging telescope that can be used only for observing zenith distances of objects that pass nearly overhead in their apparent diurnal motion. The zenith distance of such objects can be determined by noting how far from the vertical (as indicated by a delicate plumb line) it is necessary to tilt the telescope in order to bring

è the object into the center of the eyepiece. Bradley's observations with these instruments are even more astonishingly accurate than those made with the meridian circle. The variation of latitude was discovered by Küstner in 1888, and since then this interesting motion of the Earth has been well observed ; but our knowledge of it is not confined to these recent years. In par- ticular, Chandler showed in 1891 that Bradley's observations are delicate enough to reveal and define this variation as far back as 1726, more than a century and a half before it was established by Küstner. In describing this work of Bradley's, Chandler, who seldom indulged in superlatives, is carried away by a veritable storm of admiration; he speaks of Bradley's "wonderful sagacity," his "immortal work," and its "wonderful accuracy," and says that going back to it after one hundred and sixty years is more like advancing into an era of more re- fined observing than that from which we pass. A remarkable fact in connection with Bradley's career is that he invented no new instrument nor any striking modification of one. His success was due to the completeness with which he studied and understood his tools, and the ingenious devices he employed for eliminating or avoiding the errors to which they were liable. If we analyze the causes of his success we conclude that he was so good an observer because he was so good a mathematician, a remark that applies to other observers as well, among them the two pioneers that we have already considered, Hipparchus and Tycho. The next striking improvement in observing was made by Bessel of Königsberg, who was born twenty-two years after Bradley died. Though Bessel never attended a university, he became not only a great observer but a great mathematician, and was entirely self-taught in these sciences. As in the case of Bradley, he did not devise the instrument that he made famous. The credit for the invention of the heliometer is shared by Bouguer and Dollond. In this instrument, as most of you are aware, the object glass is split into two halves, and measurements are made by noting how far the two must be moved apart to make the image of one object formed by one of the halves coincide with the image of a second object formed by

the other half. You will readily see that the two semi-objectives must be good ones to make this instrument accurate. It was Fraunhofer, Bessel's close contemporary, close neighbor, and close friend, who first succeeded in making a serviceable heliometer. BesseFs great triumph with this instrument in measuring for the first time the parallax of a star is too well known to justify description this evening. By the time Bessel made these observations (around 1836) the accuracy of the results obtained with meridian circles had improved somewhat since the time of Bradley, so that two or three of the best of these instruments now gave 0^9 as the average error in right ascension and 0^6 in declination. Bes- sel's heliometer measures were greatly superior, yielding 0^2 as the average error. The next great improvement in accuracy came from an entirely unexpected quarter. In 1839, Daguerre first made photography a practical art. Arago seems to have been the first to see the advantages of the new method in delineating celestial objects. At his suggestion, Daguerre himself as well as Foucault and Fizeau had by 1845 secured promising photographs of the r Sun. During the eclipse of 1851 Busch obtained an excellent picture of the corona, and soon after de la Rue and especially Rutherfurd photographed the Moon with a success that was not surpassed- for nearly half a century. But it is a far cry from such applications of photography as these to the precise determination of celestial angles. This subject has been one of long and sometimes slow growth in which many investigators have taken part ; but among the pioneers in this subject one figure stands out head and shoulders above his colleagues. Lewis Morris Rutherfurd was born near New York City in 1816. He studied for the law and practiced it for a few years. But in early manhood he became deeply interested in astronomy, and soon forsook other occupations in its favor, a decision that he could make all the more easily because of his private means, for he was one of the wealthiest men in New York. In the garden of his fine house at Second Avenue and Eleventh Street (then one of the handsomest parts of New York City, but now sadly crowded into ugliness by the merciless

march of the city's progress), he erected a well-equipped ob- servatory, and here he worked industriously for many years. Shortly before his death in 1892 he gave all this apparatus and other astronomical equipment to Columbia University. Rutherfurd has written astonishingly little of his astronomical work, and that little appears in out-of-the-way places. This is the reason that so few historians of our subject, especially those abroad, know how prominent a part he played in this important development ; and it is also the reason why so long a time elapsed before astronomers fully recognized the value of photography as a means of measuring. Had Rutherfurd taken the trouble to make his work impressively and widely known, this application of photography might well have been in general use a quarter of a century earlier than was actually the case. His first experiments, in 1858, were made with an ordinary 11-inch refractor. He then tried the insertion of correcting lenses between the objective and the plates; next he went to a reflecting telescope, and finally, in 1863 and 1864, he constructed, in his home and with his own hands, the first objective (llj^-inch aperture) to be corrected for the violet and the blue region of the spectrum instead of the green and yellow. A few years later he made a 13-inch objective which could be used for both visual and photographic observations, a third lens being placed over the other two when the latter kind of work was contemplated. In these years he constructed, likewise in his own home, the first measuring engine for photographic plates. He took all the precautions necessary to center the plates and to orient them properly ; and he showed by ingenious experiments that there is no serious distortion of the film during the photographic processes. These questions were not touched upon again by any other astronomer until the eighteen-eighties, when experiments were set on foot in preparation for the work of the great Astrographic Catalogue. Rutherfurd went on to accu- mulate an extremely valuable collection of photographs of clus- ters and other interesting objects, and most of these were measured under his direction. The reduction of most of these measures was carried out many years later, chiefly under Pro- fessor Jacoby's direction at Columbia University.

In 1898 it was my privilege, as a student at that university, to determine what precision could be realized from Ruther- furd's material if they were measured and reduced by modern methods. On each of eight photographs of Praesepe I measured the positions of 45 stars, and I derived 0^083 as the average error of a single photograph. This quantity was obtained from a consideration of the agreement among the eight photographs themselves. Just recently Dr. Hayn has put these measures to a more severe test, comparing them with six other surveys of Praesepe made with the heliometer and later photographs ; he derived nearly the same average error : (//085 for a single Rutherfurd plate; Some of these plates were secured more than sixty years ago. Let me sum up by putting together the various average errors that apply to the observations we have been considering :

Hipparchus 240^^= Tycho, mural quadrant 30" Bradley, meridian circle 2" Modern meridian circle 0^35 Bessel, heliometer 0'f2 More recent heliometers CK/l Rutherfurd, photographs 0'Ό8 Modern long-focüs photographs 0^025

It is almost certain that the high accuracy now attained by the photographic process will soon in turn give way to some- thing much better. Is it possible to get an inkling now of the direction in which this next improvement will be made? Prob- ably not. In the case of all five of the pioneers to whom we have devoted our hour this evening, the increases in accuracy with which their names are associated were brought about by the invention of new instruments or new methods, though these men themselves were not always the inventors of these instruments and methods. If, however, I were forced to hazard a guess as to the nature of the next step, I should say that it would lie in some modification of the interferometer that would enable us to measure small changes in comparatively large angles. It is difficult to see how this can be done in a practicable

way and how this process of measurement can be anything but very laborious, but doubtless the same or similar pessimistic thoughts were generally entertained at the time any one of the steps we have considered was in the experimental stage. There is one improvement in observing that is easily within our reach, and it is not an inconsiderable one. The task has fallen to me in late years of co-ordinating "a very large number of earlier observations of various kinds. In the process of this work I have been more and more impressed with lack of confidence that observers in general have in their own work. Results that differ from those of predecessors or contemporaries are apt to be regarded with suspicion. It would be possible, but it would be a graceless task, to point out many instances in which lack of courage on the part of an observer and lack of an obstinate confidence in his work has robbed him of the opportunity to make an important discovery. A single chapter in the history of astronomy, namely that concerned with the variation of latitude, offers many such instances. But let me rather terminate my remarks by the pleasanter task of recording a little-known in- stance of the opposite kind, an observer with high courage. Asaph Hall (1829 to 1907) is best known through his discovery of the two satellites of Mars with the Washington refractor. He used the same instrument to carry out a long and valuable series of micrometer measurements, among them sets of observations for the parallaxes of four stars. When he published these results in 1887 he naturally compared them with previous observations on the same objects. He was dismayed to find that his parallaxes were much smaller than the accepted values. Here is what he so says about them : "It will be seen that generally my observations for stellar parallax give results that are smaller than those found by other observers. This is shown most strik- ingly in the case of 61 Cygni, since the parallax of this star has been found by several astronomers to be very nearly +(//5. I regret this discordance, and can only say that I have given as much care as possible to the work, and that the above results appear to be the best that can be derived from the observations." But his regret was needless, for we know now that his results were accurate. For reasons that are now fairly well understood,

it is the earlier ones that are at fault. Here are his four parallaxes (reduced to absolute values) and here opposite each of them is the best information we at present have of the true values, in both cases with probable errors attached :

Hall Modern 40 Erid +0'.'226±'.'020 +0'.'201±'.Ό04 6 Cyg - .018± .008 + .042± .003 Vega + . 137± .006 + . 124± .004 61 Cy + .273± .010 + .300+.003

On the average Hall's values appear to be too small by only 0^012, a quantity not at all surprising in view of the probable errors of his results. A man of less courage might have been induced to delay publishing these results "until they could be studied further," in other words, forever. The history of early parallax determinations indicates very strongly that suppression of this kind must have been frequent. Had Hall's policy been in more general vogue we should have learned much earlier with what sources of error such determinations are beset, and how to eliminate these errors.