106 



THE AMERICAN NATURALIST [Vol. XLIV 



This principle is diametrically opposite to the principle 

 laid down in the introduction to this paper, that the na- 

 ture of the relation between facts and theories is such 

 that the facts can not be valued by their compliance with 

 theories. We need theories for inspiration and enlight- 

 enment, we need facts to build upon ; but to discriminate 

 between facts in favor of any mental conception is to 

 place oneself upon the inclined plane of immoderate 

 speculation. 



To sum up: While the evidence shows that as a rule 

 organs originate from similar germ-layers, both in on- 

 togeny and in regeneration, there are also some striking 

 exceptions to the rule. The hypothesis, that the method 

 of regeneration is causally influenced by the course of 

 ontogeny, is, therefore, quite unnecessary as a corollary. 



Witli the elimination of this hypothesis the conception 

 of the atavistic nature of regenerated peculiarities, i. e., 

 the conception of a repetition in. regeneration of phylo- 

 genetic processes, loses its chief logical support. This 

 last theory, however, is also objectionable (1) because of 

 its inherent inconsistency, (2) because it depends upon 

 more or less problematic assumptions. 



With both hypotheses, those of the repetition in re- 

 generation of ontogenetic and of phylogenetic processes, 

 now discredited, it would be venturesome to take sides in 

 unsettled questions of animal morphology upon the 

 ground of evidence deduced from a study of regeneration. 

 But even if the hypothesis were correct, to accept it as a 

 working principle is to put oneself deliberately into the 

 logician's vicious circle— proving theories with facts 

 approved of by the theories. 



Abel, M. Regenerationsvorgange bei den limikolen Oligochaeten. Zeitschr. 



f. wissenschaftl. Zool., Vol. 73, pp. 1-74, 1902. 

 Barfurth, D. Die experimentelle Regeneration iiberscbussiger Gliedmassen- 



theile (Polydactylie) b.-i ■{■■<■ \u:phil.i.-ii. .!/•<•/». /. l^.f.nH-ivna^m., 



Vol. 1, pp. 91-116, 1894. 



