416 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



and persist," 1 Weismann soon had to abandon his amphimixis 

 hypothesis (already repudiated long since by Darwin). Gradually 

 he came to the hypotheses of " germinal selection," or struggle for 

 food between the determinants of the germ plasm, as a probable 

 cause of inherited modifications, and " parallel induction." In these 

 two hypotheses he thus acknowledged that the germ cells are modi- 

 fied by external causes, so as to reproduce in the offspring the somatic, 

 or body changes produced in the parent by the environment. Only 

 in his second h} T pothesis he suggested that the germ cells are influ- 

 enced directly by the external agencies — not through the modifica- 

 tions produced by the environment in the organs and tissues of 

 the body. It hardly need be said that most biologists received this 

 last suggestion not as a new workirfg hypothesis but as a veiled con- 

 cession of Weismann to his opponents. In fact, the hypothesis was 

 not a generalization born from the study of changes going on in 

 germ cells under the action of external agencies; it was advocated 

 only as an hypothetical explanation for the facts that contradicted 

 the previous hypotheses of Weismann. But till now, " we are told 

 by the specialists who have studied the subject," it is impossible to 

 ascertain in one single concrete case of inheritance how the modifica- 

 tion was produced in the germ cells — through the body cells or 

 independently of them. 2 



Some biologists saw in " parallel induction " an interesting new 

 line of research, and they followed it. But Darwin, who already 

 knew this hypothesis long before Weismann resorted to it, pointed 

 out with full right, in Variation, that although a simultaneous modi- 

 fication in some definite direction of the body cells and the germ 

 cells takes place in certain special cases, this can not be a general 

 cause of the hereditary transmission of variations. Like Amphi- 

 mixis, this hypothesis does not account for the inherited adaptive 

 variations, the necessity of which for the evolution of new species 

 Darwin already saw in 1868, and we still better see now. 



In short, Weismann's attempt to combine the pre-Darwinian con- 

 ception of innate predetermined variations with the Darwinian 

 principle of natural selection has failed; and an attentive reader 

 of his last work, Vortriige zur Descendenztheorie, especially the 

 pages 258-315 of the second volume, will himself see how little there 

 remained from that attempt. By his criticisms of some facts, which 



1 Essays, i, 100. 



2 Cf. L. Plate, Selektionsprinzip, 4th edition, 1913, pp. 441-442. The same view, as it 

 was pointed out by Prof. Hartog, is held by E. B. Wilson, the author of a standard work 

 on the cell : " Whether the variations [he writes] first arise in the idioplasm [the germ 

 plasm] of the germ cells, or whether they may arise in the body cells, and then be 

 reflected back upon the idioplasm, is a question to which the study of the cell has thus 

 far given no certain answer" (The Cell in Development and Inheritance, 2d edition, 1900, 

 p. 43.'!, quoted by Marcus Hartog in his work. Problems of Life and Reproduction, London, 

 Murray, 1913, p. 198, chapter on the inheritance of acquired characters). 



