One hardly knows where, in the history of science, to look for an important movement that had its effective start in so pure and simple an accident as that which led to the building of the great Washington telescope, and went on to the discovery of the satellites of Mars. Very different might have been a chapter of astronomical history, but for the accident of Mr. Cyrus Field, of Atlantic cable fame, having a small dinner party at the Arlington Hotel, Washington, in the winter of 1870. Among the guests were Senators Hamlin and Casserly, Mr. J. E. Hilgard of the Coast Survey, and a young son of Mr. Field, who had spent the day in seeing the sights of Washington. Being called upon for a recital of his experiences, the youth described his visit to the observatory, and expressed his surprise at finding no large telescope. The only instrument they could show him was much smaller and more antiquated than that of Mr. Rutherfurd in New York.

The guests listened to this statement with incredulity, and applied to Mr. Hilgard to know whether the visitor was not mistaken, through a failure to find the great telescope of the observatory. Mr. Hilgard replied that the statement was quite correct, the observatory having been equipped at a time when the construction of great refracting telescopes had not been commenced, and even their possibility was doubted.

It is sometimes said that no man, in passing away, leaves a place which cannot be equally well filled by another. This is doubtless true in all ordinary cases. But scientific research, and scientific affairs generally at the national capital, form an exception to many of the rules drawn from experience in other fields.

Professor Joseph Henry, first secretary of the Smithsonian Institution, was a man of whom it may be said, without any reflection on men of our generation, that he held a place which has never been filled. I do not mean his official place, but his position as the recognized leader and exponent of scientific interests at the national capital. A worldwide reputation as a scientific investigator, exalted character and inspiring presence, broad views of men and things, the love and esteem of all, combined to make him the man to whom all who knew him looked for counsel and guidance in matters affecting the interests of science. Whether any one could since have assumed this position, I will not venture to say; but the fact seems to be that no one has been at the same time able and willing to assume it.

On coming to Washington I soon became very intimate with Professor Henry, and I do not think there was any one here to whom he set forth his personal wishes and convictions respecting the policy of the Smithsonian Institution and its relations to the government more freely than he did to me.

The questions involved in the heading have engaged the attention of astronomers in the past, but does the photographic method contribute evidence more reliable in character than that formerly available?

Let us consider the evidence which has up to the present time been obtained by photography, remembering that the last dozen years covers the whole interval during which it has been accumulated:–

(A) Eleven years ago photographs of the Great Nebula in Andromeda were taken with the 20-inch reflector and exposures of the plates during intervals up to four hours; and upon some of them were depicted stars to the faintness of 17th to 18th magnitude, and nebulosity to an equal degree of faintness. The films of the plates obtainable in those days were less sensitive than those that have been available during the past five years, and during this period photographs of the nebula with exposures up to four hours have been taken with the 20-inch reflector. No extensions of the nebulosity, however, nor increase in the number of the stars, can be seen on the later rapid plates than were depicted upon the earlier slower ones, though the star-images and the nebulosity have greater density on the later plates.

The plates are arranged in classes or groups so as to indicate apparent physical relationship between them, and the Right Ascensions are, as far as practicable, given in the order of time within each group.

The edge next to the printed heading on each plate is the south, and the lower edge the north; the right is the following, and the left the preceding edge.

The scales of the photographs, which are given in the letterpress, are such that by eye alignments of the stars, without the application of measuring instruments, changes which have taken place in their positions or in the structures of the nebulosities, if these changes should not be less than about five seconds of arc in extent, could be detected by comparing corresponding dual plates in this simple manner. The examination and comparison of stars, both as regards their positions and magnitudes, could thus be made in a single day though they should number several thousands on the dual photographs.

Besides this alignment method, measurements by scale and compasses, or by a réseau on glass or other transparent substance, or by a rectangular L-shaped metal rule divided into millimètres on both limbs, or by the superposition of the plates upon each other, are obvious methods available for detecting changes in the position angles and magnitudes of the stars shown on the photographs.

The catalogues of stars are numerous, as will be seen by referring to the list given in Chambers' “Handbook of Descriptive Astronomy,” where 170 are enumerated between that by Hipparchus in the year b.c. 128 and the year a.d. 1876; besides these there are catalogues of nebulæ, and atlases or charts of stars.

In considering these records one is impelled to ask the questions–Have these vast stores of computative and descriptive literature, the product of great energy expended in physical and mechanical operations, and of much thought, been utilized in the advancement of astronomical knowledge to a degree commensurate with the labour and cost of their production?

Or, have the astronomers been deterred from undertaking the work of correlation, on a comprehensive scale, because they know that there is a considerable margin of probable error in all these records–particularly in those of earlier date than the middle of the present century–they therefore judge it would be unprofitable to devote their time to making the necessary examinations. They know that when they find differences to exist in the records, it would be uncertain whether they were objective, or were only the result of human errors.

In my own experience I have found differences to exist between photographs and carefully prepared modern charts of stars–differences in the position angles, in distances, in the magnitudes of the stars and in the structure and extent of nebulosities, which were most probably due to errors in charting.

It is a general opinion that the longer the time a sensitive film is exposed, in a photographic instrument, under clear atmospheric conditions, the greater will be the number of stars and the extent of nebulosity imprinted upon the film. But so far as my experience enables me to judge, after twelve years' use of the 20-inch reflector, and more than two years' use of an excellent and specially-made portrait lens combination of 5-inches aperture and 19-inches focus, the limit of photographic effect is reached sometime within ten to twelve hours on clear nights, and with very sensitive films, in the 20-inch reflector. With the 5-inch lens very much longer exposures may be given before the darkening of the films, by atmospheric glare and diffraction effects, reach the same degree of density as in the reflector.

The photographic effect produced by the 5-inch lens with an exposure of two or three hours and upwards is about two stellar magnitudes less than that given by the reflector in the same time and with films of equal sensitiveness. It would, therefore, appear that, given sufficient time, the atmospheric glare would, in both instruments, mask or extinguish the light of faint stars and faint nebulosity, which is provisionally assumed to be equal to that of 18th magnitude stars. When that limit has been reached no fainter light-effect than this would be imprinted on the films; and upon these premises the questions in the following section require consideration.

N.G.C. 7089. G.C. 4678. h 2125.

Sir J. Herschel, in the G.C., describes it as partly resolvable; a globular cluster; bright; very large; gradually, pretty much brighter in the middle; resolvable into stars; stars extremely faint. A drawing of it is given in the Phil. Trans., 1833, pl. XVI., fig. 88, and 1844, pl. XVIII., fig. 88.

Lord Kosse (Obs. of Neb. and Cl., p. 162) describes the cluster, and gives measurements of the position angles, and distances of some of the stars.

The photograph shows the cluster with a large central mass of nebulosity so dense as to obliterate the star-images; but the faint stars surrounding it are arranged in a manner suggestive of their origin from a spiral nebula.

The three clusters depicted on Plate XXVIII., and there are others of a similar character but not yet published, are strongly suggestive of, if they do not indisputably prove, that the same principle of aggregation has been in operation to cause the origin and development of each of these clusters, and I have not been able to detect any clearer evidence of their origin than that of development from spiral nebulæ.

If a reason had to be given in addition to the obvious advantage of this method of publication by printing the photographs with permanent ink it would be afforded by the fact that the records obtained by photography are peculiarly liable to be lost by accidental breakage of the glass negatives. Besides this there is the certainty that after the lapse of a limited number of years the gelatine films will become discoloured; the images will fade, and the faint stars and the faint nebulosities will entirely disappear from view.

I have had within my own experience proofs that the faint stars fade from the films, and will give the following examples:–On the 15th February, 1886, a photograph was taken of the region of the sky with the co-ordinates R.A. 9h. 40m. Dec. North 72°·0 at the centre of the plate; exposure 15m.; area of the plate four square degrees.

Shortly after the photograph was taken I counted 403 star-images on the negative; and on 29th May, 1895, I again counted the stars on the same negative, and found only 272. Therefore, stars to the number of 131 had entirely disappeared from the film in the course of nine and a-quarter years.

Another photograph of identically the same region was taken with an exposure of fifteen minutes on the 22nd March, 1886, and soon after that date I counted 364 stars upon the negative.

My intention, in the pages following, is to convey in brief, and I hope clear form, my views concerning some of the results already obtained by the aid of photography in the elucidation of celestial problems, the complete solution of which cannot for many years yet be obtained; and I may here quote from the preface to the volume, issued in the year 1893, of A Selection of Photographs of Stars, Star-Clusters, and Nebulæ, the following paragraphs, which are applicable also to the present volume.

“It has been my aim, in publishing the photographs and descriptive matter introduced in the following pages, to place data in the hands of astronomers, for the study of astronomical phenomena, which have been obtained by the aid of mechanical, manipulative, and chemical processes of the highest order at present attainable; and that such data should be, as regards the photographs, free from all personal errors.”

“The photographs portray portions of the Starry Heavens in a form at all times available for study, and identically as they appear to an observer aided by a powerful telescope and clear sky for observing.”

In the processes employed for obtaining the photographic illustrations contained in this volume the same instruments have been used, and the same care has been exercised in the production of the illustrations of the various objects as in the first volume; but owing to improvements in the manufacture of photographic films, and to the extended data now available beyond that which had been obtained up to the year 1893, when the first volume was published, certain deductions concerning the evolution of stellar systems are now permissible which six years ago would have been justly considered premature.

On photographs which have been taken with long exposures the stellar images are large, so that a considerable amount of uncertainty is involved in bisecting them with accuracy when measuring position angles of and distances between such stars; and the method of making several measurements and resetting the instrument for each one in order to obtain a mean position for the centre is troublesome. But if with a fine needle point the centre of the photographic disc could be accurately punctured, one careful measurement made from such a well-defined centre point would be more reliable than the mean of several approximate measurements.

The following method of puncturing the star discs I have found in practice to give satisfactory results:–

A microscope is taken that has a revolving stage and also a sub-stage for holding and centering an achromatic condenser. By removing the front lens of the condenser and replacing it with a brass cap in which is inserted a fine needle point at its centre, we obtain a fixed point which can by the screw adjustments of the condenser be accurately placed in the optical axis of the microscope. In the ocular is placed, through the ordinary micrometer slit, a finely ruled glass réseau or a spider-line cross with the intersection of the cross, or of any two lines of the réseau at right angles brought into the optical centre of the instrument.

Very sensitive gelatine films, such as are used in photography, when exposed during several hours to the sky in taking stellar photographs, become more or less darkened during development. The darkening is chiefly due to atmospheric glare caused by starlight; and the nebulous circles seen round the bright stars are caused by the glare and by diffraction effects produced by the objectives, or mirrors, of the instruments employed in photographing.

I have made some experiments to enable us to judge to what extent the glare and diffraction affect the finished photographs, a summary of which experiments may be given here. They were made by exposing simultaneously plates in the 20-inch reflector, the 5-inch lens camera, and to the sky in a blackened box, measuring 7-inches square by 12-inches in height, with the open end exposed to the zenith, the exposures respectively being made during precisely equal intervals of time. The plates were selected so as to be equal in sensitiveness, and the development was performed in a similar manner in each coincident trial.

The plates exposed in the box were 6-inches square, and equal areas on each of them were (1) left uncovered; (2) covered with black paper; (3) covered with different thicknesses of polished plate glass. The plates when developed showed the comparative effects of the unobstructed full sky glare as well as the effects of the application of complete and partial covering with plates of glass or with sensitometer figured scales.

The next step to be taken is to verify the accuracy of the above statements regarding the grouping of the stars into lines and curves; assuming that the examiner has, like myself, been convinced of their reality. As a test of this, as well as an example let us examine Plates 2 to 9, upon which the eye readily detects many groups of stars arranged in lines and in curves, each of them containing several stars; similar configurations to these can be seen on the other plates, and if I had chosen to print hundreds of others that are in my cabinets, each covering four square degrees in the sky, similar configurations would be seen upon them.

This persistency of the lines and curves of stars on the plates leaves no room for doubt that they are the effects of physical causes, and cannot be due to coincidence only; and when the photographs of the spiral and other nebulæ are examined a reasonable explanation of the formation of the curves and lines will be made manifest.

It is not my intention to submit elaborate arguments, or mathematical formulæ, in the discussion of the photographic evidence contained in this and in the first volume of my photographs–these will in the future, when a sufficient interval of time has elapsed, occupy the thoughts of the correlators, the measurers, the computers and of the mathematicians–my aim is now to point out the evidence, and the relationships to each other of the several classes of objects that are found depicted, untouched by hand-work, upon the photographs.

A century has elapsed since Laplace suggested that the sun, and planets, might have been evolved out of nebulous matter, but his imagination did not lead him to realize the much larger idea that stellar systems might also have been evolved from matter similar in its constitution.

I now propose to submit a series of photographic copies of my original negatives from which we may obtain strong evidence, if not complete demonstration, of the evolution of stellar systems.

The first part of the series will consist of photographs of rich fields of stars, and of clusters showing various degrees of concentration; these will be followed by a series of the spiral nebulæ some of them symmetrical in form, and others less symmetrical though clearly spiral. Following these again will be a number of nebulæ of circular, annular, and irregular forms, and, lastly, nebulæ consisting of large areas of cloud-like matter having irregular structural characteristics.

The appearances to which I now wish to draw special attention in the examination of these photographs are the numerous curves and lines of stars that are associated together in separate groups. The stars are of nearly equal magnitude; of approximately equal distances apart in each group, and the groups are independent of each other and of the surrounding stars.