tion upon them is not altogether unprofitable. The time spent in conjectures and surmises is not wholly wasted; for it is impossible to follow up any of the lines of thought thus opened, without reaching some valuable suggestions, which may pave the way to new discoveries. New truth, in one direction or another, is sure to be reached in the long-run. So, then, we may proceed to theorize in the most barefaced manner, without entirely quitting the legitimate domain of science.
It is plain that the nebular hypothesis would be doubled in importance, and our views of the universe greatly expanded, if it could be shown that an evolution of complex from simple forms of matter accompanied the development of planets from the nebulæ. Evolution could look for no grander triumph. For the evidence to support such a theory, we must depend mainly upon the spectroscope. Let us continue upon our task of finding the intermediate links between the two extremes of planetary growth, and see whether, as we ascend in the line of change, an increased chemical complexity can be observed. Upon this theory, the planets should contain more elements than the sun; the sun more than some of the less advanced among the fixed stars; and these, in turn, should be more highly organized than the nebulæ?. But we must not fail to remember that we are merely speculating, and that the spectroscope, in telling us of the presence of certain substances, does not give us accurate information with regard to the absence of others. In this investigation, we can look to the spectroscope only for hints, not certainties. Difficulties will abound in our path, and, in a paper of this length, we cannot stop to scrutinize them closely. We must bridge many chasms with guesses.
The evidence concerning the constitution of the fixed stars has been furnished chiefly by Secchi and Huggins. The former observer, favored by Italian skies, has done, perhaps, the major portion of the work, and has given us a classification of these bodies. According to Secchi, the stars may be divided into four classes, as follows:
In the first class, which is by far the largest, we find most of the white stars, Sirius, Altair, Vega, Regulus, and Rigel, being especially prominent. These give spectra characterized by the intense development of the four hydrogen lines, which stand out with great distinctness upon a background of the seven primary colors. Lines belonging to some of the metals, particularly to sodium, magnesium, and iron, are also visible, but are exceedingly faint in comparison with those of hydrogen. The distinctness of this element, as compared with the faintness of the metallic lines, is characteristic of the stars of this type. The absence of bands, indicating an absorptive atmosphere, is also noteworthy.
In the second class of stars we find our sun, Arcturus, Aldebaran, Capella, Pollux, Procyon, and many others. Here we have spectra in which the lines of the metals are apparently more numerous, and certainly more distinct, the hydrogen being less conspicuous. In Alde-
