564 



SCIENCE 



[N. S. Vol. XXXIX. No. 1007 



get the photographs of their spectra as- 

 sembled in the same observatory, then the 

 classes into which the photographs tend in 

 the long run to be grouped also tend to be 

 such that, at least for some one resulting 

 classification or aggregation of the photo- 

 graphs, the photographs of the spectra of 

 the stars of one of the star drifts are 

 grouped together, not only in the ideas 

 which the astronomers form, but in the 

 physical arrangements towards which cer- 

 tain groups of photographs, of symbols and 

 of statistical tables, persistently tend. 



The principles here involved depend 

 upon the sorts of assimilation which the 

 radiant phenomena of light make possible. 

 For a photograph is a physical expression 

 of a certain tendency whereby the struc- 

 ture of a photographic plate tends to be as- 

 similated to the molecular structure and 

 state of a radiating object — say a star. 

 When the photographs of stellar spectra 

 are grouped in classes, a secondary assimi- 

 lation tends to take place, since similar 

 spectra tend to get either placed or tabu- 

 lated in similar ways. When this second- 

 ary assimilation of the photographs leads 

 to an indirect discovery of the existence of 

 the two star drifts themselves, a tertiary 

 assimilation of the fortunes of those stars 

 whose proper motions are sufficiently simi- 

 lar takes place, and tends to get repre- 

 sented in the knowledge of different as- 

 tronomers. 



The ideas of these various astronomers 

 tend to further assimilation through the 

 means used in scientific communication. 

 The radiation of scientific knowledge con- 

 tinues the natural process which the radia- 

 tion of light and the making of photographs 

 of stellar spectra have already illustrated, 

 and the rule continues to be illustrated that 

 mutual assimilation is one aspect of classi- 

 fication and aggregation, and is a cumula- 

 tive statistical tendency which accompanies 

 them both. 



The insurance companies and the trans- 

 formation of modem civilization through 

 the extension and aggregation of modes 

 and devices whereby insurance is accom- 

 plished, furnish numerous other examples 

 of this law of the fecundity of aggregation. 

 The law, as I have said, holds in general 

 for non-mechanical systems, although, as 

 stellar evolution seems to indicate, it can 

 also hold for mechanical systems. It may 

 hold, in fact, for all natural processes which 

 involve evolution. 



Clerk Maxwell himself believed that the 

 sharp distinctions which separate the dif- 

 ferent classes of elementary atoms, and the 

 different types of molecular structure 

 which determine the spectra of the mole- 

 cules of different elements, are signs that no 

 kinetic theory of the evolution of the chem- 

 ical elements would ever be possible. It is 

 precisely here that the latest advances, on 

 the still so imperfectly defined outlying 

 boundaries of physical and of chemical re- 

 search, give a new significance to the sta- 

 tistical view of nature, by showing that if 

 we take account of sufficiently large aggre- 

 gates of things and of events, a kmetic 

 theory of the evolution of chemical ele- 

 ments becomes a possibility worthy of fu- 

 ture investigation, and certain to receive, 

 in connection with the phenomena of radio- 

 activity, further investigation upon statis- 

 tical lines, whatever be the further for- 

 tunes of the mechanical view of nature, or 

 of this problem about the evolution of the 

 elements. 



Of such speculations one can say that, if 

 ever a theory of the evolution of the chem- 

 ical elements becomes feasible, it will be, in 

 part at least, a statistical theory, and will 

 illustrate in new ways how widespread in 

 material nature is the tendency to that mu- 

 tual assimilation which all the phenomena 

 of radiant energy illustrate, and of which 

 the relatively uniform constitution and dis- 

 tribution of each one of the various chem- 



