ACQUIRED CHARACTERS 205 



It will be observed that Weismann's view is based 

 upon a rigidly mechanical conception of organisms and 

 of their propagation. The plausibility of his argument 

 lies to some extent in our inability to imagine any me- 

 chanical process by which a change in a functional 

 part of the organism can so affect the germ-cells as to 

 cause them to reproduce that change in the same part 

 of the organisms of offspring. In the absence of com- 

 plete knowledge of the process of propagation, and of 

 the laws that determine heredity, the basis of his argu- 

 ment is not free from elements of insecurity, and his 

 exclusion of non-mechanical or superphysical factors 

 cannot be proved to be warranted. In short, his posi- 

 tion is to a considerable extent based upon speculative 

 premises, and is largely supported by an appeal to our 

 inability to furnish an exact and demonstrable descrip- 

 tion of the process of transmitting acquired characters. 



Over against his contentions must be placed the cir- 

 cumstance that many modern biologists, in spite of 

 their acquaintance with Weismann's arguments, are 

 led by observed facts — especially in the fields of 

 paleontology, domestic breeding, and medical science 

 — to the conclusion that acquired characters are in 

 many instances certainly transmitted. The issue is 

 one between a partially speculative theory of cellular 

 processes and the inferences ordinarily drawn by 

 paleontologists, professional breeders, and physicians 

 from the facts which come within their observation.^ 



1 See V. L. Kellogg, Darwinism To-day, pp. 271, 272. He gives 

 a bibliography of the subject on p. 305. Herbert Spencer's argu- 



