292 CONCLUSION 



wholly different order of certitude from that upon which the 

 co-operative alliance between Heredity and Natural Selection 

 depends. On the one hand we have the facts directly observed 

 and recorded in Chapters IX-XIII ; while on the other Weismann 

 himself says,' " It is true that we have never directly observed the 

 process of natural selection," and he then goes on to illustrate the 

 kind of evidence that exists for it, and its cogency, by adding, 

 " neither has anybody directly observed the mode in which the 

 heat of the animal body is generated by the processes of combus- 

 tion going on in the blood and in the tissues, nevertheless this is 

 believed as a certainty." We, see, therefore, that for the establish- 

 ment of the doctrine of Animal Heat as well as that of Natural 

 Selection, although the amount of direct observation has been 

 altogether inadequate to bring about conviction, the lacunae — as in 

 scientific investigations generally — have been filled by the exercise 

 of reason. Yet the tendency with my critics has been to demand 

 nothing but observation, " continuous and unbroken " " and to 

 allow no place whatever for reason, even to supply the smallest 

 links in the chain of evidence. That this is no exaggerated state- 

 ment any one may see for himself who will carefully peruse 

 Appendix H of my " Studies in Heterogenesis," and will contrast 

 the very cogent evidence which 1 adduced as to the reality of the 

 transformation of Hydatina eggs into great Otostomata, with the 

 incredulity with which this evidence was received, not only in this 

 country but abroad. 



So far, indeed, from there being any a priori objections to per 

 saltunt development or Heterogenesis, which, as I claim to have 

 shown, occurs so frequently, the fact of its occurrence ought to be 

 considered thoroughly harmonious with all that we know con- 

 cerning simpler forms of matter, and with the constitution of 

 living matter itself. 



As Herbert Spencer said,3 it is a cardinal fact " that proteids admit 

 of multitudinous transformations ; and it seems not improbable that, 

 in protoplasm, various isomeric proteids are mingled. If so, we 

 must conclude that protoplasm admits of almost infinite variations 

 in nature." But the ' ids ' of Weismann as well as the ' physio- 

 logical units ' of Herbert Spencer are essentially protoplasmic in 

 nature and, therefore, of great molecular complexity. These are 



' " Studies in the Theory of Descent," Trans., p. 643. 



' "Studies in Heterogenesis," p. 130. 



3 " Principles of Biology," vol. i., 1898, p. 67. 



