TRANSFORMISM 267 



organization. Such observations, and experiments bearing upon 

 the same problem, have been made again and again and with 

 apparent success. Yet there has been constant controversy as 

 to what these experiments and observations demonstrate. With* 

 out doubt it has been fashionable in biological investigation, since 

 the time of Weismann and all through the period of modern 

 genetical studies, to distrust such views as were almost dogmatic- 

 ally held before the time of Weismann. And the experiments 

 and observations expected to demonstrate Lamarckianism are 

 laborious and difficult to verify and so some doubt always attaches 

 to their results. But what big experiments have been made do 

 seem to point strongly to the conclusion that small adaptations, 

 or even apparently indifferent bodily changes made throughout 

 many generations, at last become changes included in the develop- 

 mental organization. That means that individual acquirements 

 become congenital changes, or there is transformism. 



Clearly we have, at present, no generally accepted hypothesis 

 of transformism. (i) Contemporary investigation is almost 

 entirely experimental and the tendency is to apply its results to 

 the study of the evolutionary career. Certainly we can actually 

 observe transformism in progress in all cases where plants and 

 animals are domesticated and bred as " stocks " that have 

 utilitarian value. Certainly every Mendelian experiment demon- 

 strates transformism. But it would be foolish to argue that results 

 obtained by human, experimental control can be extended to the 

 past, pre-human phase of life on the earth unless we insert into 

 that past some agency comparable in effect with human, experimental 

 control. And even then all Mendelian results that are applicable 

 to the study of the evolutionary process must postulate the 

 occurrence of mutations, that is, real novelties of character, and 

 this formidable problem is untouched by any contemporary 

 experimental work. (2) The Natural- Selection hypothesis, logic- 

 ally powerful as it is, nevertheless depends on some estimate 

 as to the existence, in wild nature, of inheritable novelties of 

 character, or mutations, that may be the material for selection. 

 Therefore, to render that hypothesis verifiable we must have some 

 notions as to the numerical values of occurrence of these mutations, 

 and some fairly precise notions as to the rate at which evolutionary 

 changes have proceeded. And again we merely take for granted 

 the occurrences of these mutations — without any notion as to how 



