362 ACTION OF NATURAL SELECTION 



sible that the arm muscles can have a secretion differ- 

 ent from that of the other muscles of the body. The 

 greater muscular development of the man as a whole, 

 however, may lead to the production of slightly more 

 muscular children than the average. 



On our hypothesis, the heritableness of mutilations 

 and injuries is not admissible. It is almost inconceiv- 

 able that each spot of skin on the body, or each finger, 

 should have a specific secretion, and that an injury to 

 it, by changing its secretion, should so affect the germ- 

 plasm as to produce a similar change in the correspond- 

 ing area of skin or the finger of the offspring. How, 

 then, is it possible to account for the various apparent 

 instances of inherited injuries, such as are quoted by 

 Eimer,* Cope,f and others who believe in the transmis- 

 sibility of such characters? There certainly seem to 

 be a small number of thoroughly well authenticated 

 cases, but the number is so small that we may perhaps 

 attribute them to mere coincidence. The millions of 

 instances of injuries which show no trace of any trans- 

 mission provoke no remark, as it is only what we are led 

 by common experience to expect. Supposing, on the 

 other hand, a child exhibits any birth mark or de- 

 formity bearing some similarity to an injury or mutila- 

 tion in a parent, it is at once hailed as a remarkable 

 case of inheritance of an acquired character. 



There are, however, certain cases of the apparent in- 

 heritance of acquired characters which require more 

 detailed criticism. These are the well-known experi- 

 ments and observations of Brown-Sequard on injuries 



* " Organic Evolution/' p. 173. 



f " Factors of Organic Evolution," p. 431. 



