F. W. Very— A Cosmic Cycle. 191 



mode of attack which is full of promise. Still more powerful 

 are the new magnetic methods. The difference of frequency 

 in the Zeeman components of a Fraunhofer line is constant for 

 all the lines of a group, even through a series of homologous 

 elements, but differs in different groups of lines for the same 

 element. 



The problem of the composition of the elementary atom has 

 been approached by the study of 



(a) spectral series, 



(b) enhanced lines, 



(c) the Zeeman effect. 



By these tokens we may conclude that the atoms are resolva- 

 ble. It may be beyond our power to actually separate and sort 

 the various component motions, but they may now be rationally 

 conceived. A host of facts has been presented for mathemati- 

 cal analysis, and a problem has been set before us of even 

 greater interest than that of the celestial motions, which here- 

 tofore has been the noblest of all. 



Relative Abundance and Distribution of the Elements. 



If the atoms are indeed in process of formation in space by 

 absorption of radiation, there must be a general tendency to an 

 average composition throughout space, and a corresponding 

 general resemblance in the aggregates of matter resulting from 

 concentration. There is, on the whole, a considerable simi- 

 larity in stellar spectra. The divergences from a common 

 type are not greater than might be anticipated from rearrange- 

 ment of materials and thermal changes, occurring subsequently 

 to the first concentration, with a small residual change which 

 furnishes an argument in favor of the present hypothesis. 

 Gaseous nebulae cannot be admitted into the argument, since 

 the relative simplicity of their spectra is due to the absence of 

 conditions suitable for exhibiting the radiations of their more 

 complex ingredients, and not to actual simple composition. 

 The rapid transformation of spectra of temporary stars to the 

 nebular type proves this. 



In opposition to the idea of an average composition of matter 

 in all parts of space must be placed the very different densities 

 of the planets ; but since they have probably been formed from 

 our sun long after the original aggregation, and when hetero- 

 geneity had already developed in the parent mass, this fact has 

 no bearing on the argument. We might say the same of the 

 differences in the composition of meteorites, if we were as con- 

 fident of their origin as we are of that of the planets. On the 

 whole, the secondary origin of comets and their meteoric 

 products seems to be indicated. 



