No. 554] FITNESS OF ENVIRONMENT 107 



accordingly there must always be exchange of matter and 

 energy with the environment. 



Obviously these few conclusions can make no claim to 

 completeness. Fully to describe life, the discovery of 

 many other fundamental characteristics is necessary, in- 

 cluding such as are related to inheritance, variation, evo- 

 lution, consciousness and a host of other things. But in 

 the formation and logical development of such ideas 

 there is danger of fallacy at every step, and, since the 

 present list will suffice for the present purpose, further 

 considerations of this sort are best dispensed with. This 

 subject should not be put aside, however, without clear 

 emphasis that the postulates which have been adopted 

 above are extremely meager. The only motives for aban- 

 doning further search are the economy and the security 

 which are thereby insured and the very great difficulty of 

 extending the list. 



The Envieonment 



Even at the earliest period in the evolution of a typical 

 star there appears to be a progressive variation in the 

 chemical composition from center to periphery. Theo- 

 retically it seems inevitable that the heaviest elements 

 should be concentrated in the interior and that those of 

 lowest atomic weight should be present in the greatest 

 amount near the surface. Actually, spectroscopic investi- 

 gation fully confirms this view. Thus the spectra of 

 typical hot stars show that hydrogen is an inevitable con- 

 stituent of their superficial parts. Indeed the universal 

 presence of hydrogen under such circumstances is un- 

 doubtedly one of the most clearly established facts of 

 stellar astronomy. As stars cool and become red the 

 spectral changes quite as unmistakably point to the pres- 

 ence of carbon. Accordingly wo possess the best of evi- 

 dence and the best of reasons for the belief that large 

 Quantities of hydrogen and carbon must exist at or near 

 the surface when a crust forms upon a cooling star. 



The nature of the chemical combinations into which 

 tbese elements at first enter is perhaps open to some 



