460 KINSHIP AND ADAPTATION 



systematic importance seem to have nothing whatever to 

 do with utility. 



Yet, we know that many organs do serve marvelousl}^ 

 well the needs of the organism. There is no reason, however, 

 why their adapti\-e features may not have arisen through 

 mutations, even without selection, and we have seen that 

 initial stages in tlie flevelopment of many an adaptation are 

 of so little use that selection could not reasonably be sup- 

 posed to act on them. At the same time it is of course not 

 impossible in other cases that natural selection may operate 

 under certain conditions now and then occurring. Variations 

 of the mutative sort would then serve especially well as steps 

 in the process of species-making, because of the way in which 

 they are inherited, while fluctuating variations might also 

 sometimes contril)ute to the result, provided incompatible 

 features did not arise as mutations. For the most part, 

 however, selection ma.y now be supposed to play only a 

 subordinate role in organic evolution, its effects sho^^ang 

 chief!}' in the maintenance of a certain standard of perfection 

 in an established type. A plant grows where it can. and it 

 can grow at all only by having the chance, and Ijeing fit to 

 take advantage of it. When we have said this we have ex- 

 pressed about all that it is necessary to admit of the doctrine 

 of natural selection. We must rememljer also that selection 

 has at best but a negative value; it cannot originate any- 

 thing, it can onlj' favor certain individuals by weeding out 

 others. As to mutations, the reader has doubtless already be- 

 come aware of their strilcing likeness to "special creations." 



If it could be shown that acquired cliaracters may be 

 passed over from one mutation to another, we might suppose 

 that a direct influence of the environment is instrumental in 

 originating species. We know that it does control individual 

 peculiarities often in a striking way, and may not improbably 

 account for the constant a]5pearance of features sometimes 

 attributed to other causes. The great difficulty often is to 

 decide which of several possil.de causes may have brought 

 about a given result; and only long continued, careful experi- 

 ments can give a satisfactory answer. So far as we may 

 judge from such extended observations as those of Dr. Stur- 



