Under a sky more transparent than that of England the harvest was yet richer and the general summary published in 1864 gave the positions of 5,079 of these objects. Very few nebulæ found later worthy of interest escaped the eyes of the Herschels.

No thought was taken at that time as to what could be the origin of these curious objects. Their vague aspect gave little faith in their permanency. At one time the hope was held that they might rapidly change before our eyes. Laplace, after meditating upon the spherical or flattened figures of the planets, upon the existence of the ring system of Saturn, upon the close coincidence of the planes of the equators and the orbits of the planets, became convinced that the sun and the planets must have once been parts of the same large, very diffuse cloud. We might then expect the history of the solar system to repeat itself among the many other nebulous clouds in the realms of space. What would be more natural than to see among the nebulæ successive stages in this evolution from such clouds, the material of suns and planets of the future. And, accordingly, he devised that celebrated hypothesis which has since been the cause of so many polemics.

For the convenience of reasoning, Laplace gave to his primitive cloud a figure of revolution, a general rotation about an axis and a density decreasing regularly from the center outward. Upon all these points the great mathematician showed no spirit of intolerance, and would have willingly consented to improvements. But it was much later that objections were raised.

The assiduous observers of nebulæ found that these objects were mostly of a much less simple structure. This was shown first by the principal nebula of Orion, which was selected because of its extent and brightness. Within the same limits where Huygens drew a uniformly bright surface, astronomers provided with better telescopes found strong contrasts of light and shade, filaments and entangled jets, indications of physical connection between this cosmic cloud and numerous stars. All these points are revealed in the beautiful drawings left by J. Herschel, De Vico, W. Bond, Lassell, G. Bond, and Lord Rosse. The divergencies, often striking, may be interpreted through the marvellous plates taken by Prof. Ritchey at the Yerkes Observatory and at Mount Wilson. The same features are not shown by the various artists and by the chemical processes. Even photographic plates have their "personalities" as well as artists. However, we have the right to hope that the plates are more impartial in the features which they reproduce. The long exposures employed often destroy the details easily recognized by the eye in the central and brighter parts. But for the reproduction of the faint and more extended portions the superiority of the plates is unquestioned.

To sum up, the great nebula of Orion is a very complicated object, very rebellious against graphical representation by which means we had hoped to show by a comparison of drawings what changes may have taken place during the course of a century. The early drawings have in this respect very little value and the elaborate discussion which Holden based upon the sketches of Bond has not in general been found convincing.

This nebula departs too får from a globular form or rather from a figure of rotation to be taken as giving support to the Laplacian hypothesis. No one could trace in it a prelude to the formation of a narrow and regular ring surrounding a larger central body. Several annular nebulæ were noted by W. Herschel, but among them not one had a nucleus of any importance.

If we must find in the sidereal universe a picture of what took place in our system, then we would have greater hope of finding it among the planetary nebulæ. In the smaller telescopes they appear as small round, somewhat brilliant, diffused spots, but in stronger instruments like bright stars embedded in dense atmospheres. But such systems were too small and too distant to tell us much of the details of their structure before spectroscopic methods were developed.

Such was the condition of affairs when Lord Rosse, in 1850, showed the existence of a distinct series of nebulæ, having besides the central nucleus several successive envelopes. But these envelopes, instead of being separate and concentric, as the advocates of Laplace's hypothesis would have expected, were spiral in form. They showed streamers, growing progressively larger, at first in the direction of the radius, then curved around all in the same sense. No theory had predicted such an appearance.

The instrument used by Lord Rosse and made under his direction was a gigantic telescope, 6 feet in aperture, a size not since surpassed despite many courageous attempts. Judging from drawings, it could have been used only near the meridian. Nor was sufficient protection provided against the weather, either for the observer or the mirror. The necessary access to the upper part of the tube was possible only by the use of heavy and complicated machinery. Such a piece of apparatus required the assiduous and careful maneuvering of several assistants. Official astronomers, with strict limitations and limited means, could obtain such cooperation only with great trouble and for very little time. Is it necessary to seek further for the reason why the great instruments of Lord Rosse and the Herschels, despite their great services, had such a short career and were used only by their makers?

The object which first seemed to offer to Lord Rosse an unusual character is numbered 51 in Messier's catalogue. It is to-day considered the most typical and the most curious of the spiral nebulæ. If we examine how Lord Rosse drew it in 1850, we will find that the rays do not come out from the nucleus in all directions but normally and only from two diametrically opposite regions. The curvature, pronounced at the start, decreases later but irregularly. One of the spirals departing further from the center terminates in a secondary bright nucleus. The principal spiral continues its path undisturbed and completes at least a turn and a half before fading away. The appearance of these structures, so fine, so geometrical, so prolonged, gives the impression of a rapid whirling movement.

Long afterwards, in 1878, Lord Rosse returned to this same object. The general appearance remained the same, but the number of filaments, their fineness and regularity of curvature seemed much decreased. After mature examination, it appeared that the early appearance had been judged too geometrical just as seems to be the case with the canals of Mars. It looked as if now the principal spiral expands into the secondary nucleus.

Again, looking at the same object as photographed by Keeler at the Lick Observatory, it is evident that the second drawing of Lord Rosse is the more faithful. But other important details are brought to light. The junction of the two principal spirals with the main nucleus is no longer radial but tangential. By their evident discontinuity we are led to strongly doubt that they can be considered as trajectories. Various points of the two spirals are the origin of independent rays, each curved in the same sense as the main spiral but with entirely different initial directions. At the starting points of the secondary rays we always find a star, or if we look closer, a group of stars. Upon a plate of the same nebula, taken by Dr. Isaac Roberts, 180 condensations were counted on the lines of the spirals.

It is evidently well in the presence of such immensely vast objects, so different from any that we have at hand for experiments, to build as much as possible on firm structural groundwork, neglecting no evidence concerning their form, their structure, or distribution in space. Thus armed, we may approach with less danger their life history and seek to know how these strange organisms are born and how they grow.

First, what can be stated as to the distribution of the spiral nebulæ, for instance with regard to that most natural plane of reference, the mean plane of the milky way?

If we consider nebula irrespective of class, we can state on this score a well-defined law. These objects show, as to their direction

from the earth, and doubtless also as to their absolute position in space, a marked antipathy to the plane of the milky way, the galactic plane.

This fact was noticed long ago by the philosopher and sociologist, Herbert Spencer. It is shown by the often-published figure constructed by Proctor. The principal catalogued nebulæ are indicated by so many points. The white spot near the south pole corresponds to the Magellanic Clouds, a small region where nebulæ and clusters abound. A place of similar nature, though less important, lies in the northern hemisphere close to the milky way. Apart from these two exceptions, the milky way traverses, throughout nearly all its whole extent, regions poor in nebula which cluster chiefly near the north pole of the milky way.

We

But is this law of distribution the same for the spiral nebulæ ? For some years it was generally admitted that it was not, that the spirals were irregularly distributed as regards the milky way. might therefore treat them as strangers and keeping in mind their circumvolutions, bifurcations, gaps and the fact that they inclose so many stars and clusters of stars, consider each one as an independent milky way.

To-day that conclusion does not seem so assured since Keeler has pointed out that many of the faint nebulæ, showing to the naked eye no trace of a central nucleus or spiral structure, reveal on longexposure photographs both these characteristics. Now we are beginning to ask whether the greater number of nebulæ are not spiral and whether statistics, including all of them, would not show that the great majority of these objects are related to the milky way. A photographic exploration of the entire sky with a powerful instrument is necessary to solve this problem.

Apart from their structure, which too often escapes us, is there no other easily determinable characteristic which may serve to classify the nebula? Could we not, for example, group them, as we have the stars, according to the richness of their spectra in absorption lines?

Huggens, trying to do this, noted that they readily fall into two classes. One shows a spectrum composed of bright lines like that of a gas made luminous electrically. These are often called the green nebulæ because the greater part of their light is concentrated in a bright green line in their spectrum which has never been identified with any known terrestrial element. Provisionally it is considered as an indication of an unknown element which has been named nebulium. Of the four lines to which the spectrum of a nebula of this class is usually limited, the third in order of intensity is the only one upon whose origin we are agreed. It belongs to the spectrum of hydrogen.

A moderate dispersion may be used with this class without weakening the lines of the spectrum too much. Keeler showed that the brightest line does not occupy exactly the same position in all the green nebulæ. Naturally, these small differences are interpreted as a sign of radial velocities. The 14 nebulæ for which satisfactory results have been obtained give for the radial component figures ranging from 18 to 64 kilometers per second. There is a predominance of negative values, evidently not because the green nebulæ show a tendency to approach us, but because the greater part of them which may.be easily observed are situated nearer the constellation Hercules toward which our sun is moving, carrying us along with him. Contrary to what is true of the nebulæ in general, the majority of the green nebulæ lie in the milky way. The existence of these gaseous bodies, owing their light to a more or less extended mass of gas, has been considered as furnishing the experimental basis formerly lacking for the Laplacian hypothesis.

Interesting as these results are, we will not dwell upon them as they take us away from our subject. Indeed, of all the nebula whose spiral structure is beyond doubt, not one belongs in the class just described. Not one is adapted to the determination of its radial velocity. All of them, as well as the great majority of the faint nebula without definite form, shine with a white light which the prism transforms into an apparently continuous spectrum. This spectrum is too faint for the detection of absorption bands. However, there is some justice in calling it purely stellar. The white nebulæ owe the greater part of their light to the stars which are clustered within them. As to the great nebula of Andromeda, which is the brightest of the spiral nebulæ, we may add that the majority of the stars of its central portion are of the solar type.

The contribution of the spectroscope to the study of the spiral nebulæ is on the whole somewhat restricted. The services rendered by photography are, on the other hand, inestimable. The great part taken by this method of study dates from the invention of the sensitive bromo-gelatin plates. The green or gaseous nebulæ, whose light more strongly affects the photographic plate, brought the first success. The photographs of Paul and Prosper Henry, of Isaac Roberts, and of Keeler early showed evidence of a physical relationship between the stars and the nebulæ, even in the case of the gaseous nebule. This connection is yet closer in the spiral nebulæ, of which we will now speak exclusively.

About the year 1900 they were looked upon as rare and scattered objects. Keeler undertook to form a collection of the most remarkable nebulous objects and was led to the two following unexpected conclusions: First, many nebulæ formerly classed as globular, annular, 85360°-SM 1912-11