FIG. 26a. Planetary Nebula N.G.C. 7009. Composite drawing by Curtis from photographs made with the Crossley Reflector of the Lick Observatory. With the slit of a 3-prism spectrograph placed on the longer axis, the bright nebular lines were found to be inclined to the direction of the slit, owing to the rotation of the nebula, as shown (exaggerated) in the upper part of this figure, and as reproduced from the spectrogram in Fig. 266.

masses, in accordance with the simple laws of physics, and the planets and their satel

lites of our system, as they exist to-day, are the result.

I see no reason to question that a spiral nebula could originate in this manner; the close passage of two massive stars could, in my opinion, produce an effect resembling a spiral nebula, quite in accordance with Moulton's test calculations on the subject. Some of the spirals have possibly been formed in this way; but that the tens of thousands of spirals have actually been produced in this manner is another question, and one which, in my opinion, is open to grave doubt. The distribution of the spirals seems to me to negative the idea. If the close approaches of pairs of stars are producing the spirals we should expect the spirals to occur and to exist preeminently in and near the Milky Way structure, for that is where the stars are; and that is pre-. cisely where we do not find the spirals.

I think it is more probable that our stellar system as a whole is a spiral nebula, or has analogies to a spiral nebula, and that our solar system has been formed from an insignificant detail of spiral structure, than that our sun and its system of planets and moons should be the evolved product of an entire spiral nebula. Of course we have not

the proof of this; but the chances would appear to be strong that the ancestor of our solar system was a mass, nebulous or otherwise, comparatively diminutive in size.

There seems to be no reason to doubt the value of Herschel's opinion that a planetary nebula will develop into a star. Laplace's hypothesis that our solar system has

FIG. 27. Visual Image (right end) and Part of Visual Spectrum (left end) of the Planetary Nebula N.G.C., 418.

[The horizontal line in the left half of the figure is a part of the continuous spectrum of the stellar nucleus of the nebula at the right end. Of the three white circles in the spectrum the left one is the hydrogen Beta image of the nebula 14 sec. of arc in diameter, the middle circle is the "second nebulium" green image of the nebula 9 sec. of arc in diameter; and the right one is the "first nebulium" green image of the nebula, 11 sec. of arc in diameter.]

developed from a small rotating nebula is still exceedingly valuable, but it has been so buffeted by the winds and waves of criticism that many of the details have had to be thrown overboard. I think we should give ourselves some assurance that certain of the planetary nebulæ may develop into systems bearing many resemblances to our solar system. Campbell and Moore have been able in the past year and a half to prove, by means of the spectrograph (see Figs. 26a and 26b), that these bodies are in rotation-just as we should expect them to be from their more or less symmetrical forms-around axes passing through their centers; these axes, in general, at right angles to their longest dimensions. The rotation of our sun and the revolution of our planets about the sun, all in one and the same direction and very nearly in the prin

cipal plane of the system, afford a close analogy.

The rotational velocities of the gases composing the principal rings in the planetary nebulæ are comparable with the orbital speeds of our great planets, Jupiter, Saturn, Uranus and Neptune; and the rotational speeds of the gases are slower and slower as we go out from the principal rings, which is true of the orbital speeds of the planets of our system.

It appears that there is very little material between the stellar nuclei and the principal rings of the planetaries. The material which we should normally expect to find there has apparently been drawn into the central stars. Very little material is left in that space to condense into planets, just as in our solar system the four inner planets, Mercury, Venus, earth and Mars, and the many asteroids, are of almost negligible mass. Jupiter and the other three outer planets contain 225 times as much matter as the earth and the other three inner planets and the more than 800 asteroids.

If we assume that the rotational speeds of the particles composing the rings and other outer structure of the planetary nebulæ are controlled by Newton's law of gravitation, we have means of estimating the masses of their central nuclei, or stars. The indications are extremely strong that the planetary nebulæ are in general at least as massive as our solar system: many times as massive in some cases, possibly less massive in others. In them we seem to have enough materials to develop into systems comparable in mass with our solar system.

More than twenty years ago I observed that the different gases composing several of the nebulæ are not uniformly distributed throughout the nebular structure. The hydrogen in some of the nebula that were

critically examined was found to extend out farther than do the other chemical elements. Such was the case in the great Orion nebula, in the Trifid nebula, and in one of the small planetaries, N. G. C., 418. Viewed in the telescope, the latter was observed to be a disc of greenish-blue light, about fourteen seconds of arc in diameter, with a well-defined star near the center of the disc. Viewed in the spectroscope, the spectrum of the central star was a continuous line of light, and the composite disc was resolved into three separate discs (see Fig. 27); the largest one of hydrogen, fourteen seconds of arc in diameter; a smaller one, corresponding to the first green line of the element nebulium, eleven seconds of arc in diameter; and a still smaller one, corresponding to the second line of nebulium, nine seconds of arc in diameter. The nebulium did not extend out so far from the central star as the hydrogen. Wolf and Burns applied photographic methods to a similar study of the Ring Nebula in Lyra, in 1908 and 1910, respectively, and found differences both in the sizes and in the detailed structure of the spectral rings. Wright has in the past two years carried the development of the subject much further, by photographic methods applied to the principal planetary nebulæ (see Fig. 28). He finds, for example, that the distribution of the helium in the structure of the planetary nebulæ always favors the cen

tral nucleus or star more than the hydrogen and nebulium distribution do. In some of the planetaries, the helium has apparently been drawn entirely into the central nucleus. In one of the planetaries the hydrogen globes seem to persist brilliantly after the nebulium images have become reduced in size, or have become exceedingly faint. We can scarcely doubt that in these phenomena we are witnessing certain stages of progress in the gradual evolution of the planetary nebulæ into the stars which we see at their centers, or, possibly, into systems of stars and planets. The materials, for the most part, seem to have been drawn -possibly are still moving-into the central stars, into suns that are forming; and a very little of the materials shown by observation to be revolving around the central suns may ultimately be left to form planets of the systems. Let us recall that in our own solar system 99 6/7 per cent. of the materials are in the sun, and that only one seventh of one per cent. of the materials exists outside of the sun, in the eight major planets and their moons.

The different sizes of the elliptic images of different chemical elements in the spectrum of a planetary nebula give some basis. for the speculative thought that the chemical composition of the large outer planets of our solar system may be quite different from that of the small inner planets.

While a strong case can be made out for

the evolution of planetary nebulæ into stars at their centers, with possible planets revolving around them, we must not conclude that all stars have been formed from planetary nebulæ. There are reasons for rejecting that view.

1. Amongst the many millions of stars whose images have been examined in the telescopes or on photographic plates, fewer than 150 planetary nebulæ have been found. Unless the planetary-nebula stage of existence is lived very rapidly, the numbers are too few to play a controlling part in the

development of stars in general at any point in stellar evolution.

2. The average speeds of the planetary nebulæ and of the different classes of stars are now fairly well defined. The average speed of the planetary nebulæ is about seven times that of the extremely blue stars, which are the only ones, we shall see later, that we need consider as the immediate descendants of the nebulæ. There are indeed individual stars which are traveling as rapidly as the individual planetary nebulæ, but on the average the discrepancy