606 APPENDIX D2. 



the anterior segments alone or along the whole length of the 

 body." So that in these last cases there are, in all the segments, 

 parts from which arise generative products. The fact that these 

 parts are not definite ovaries and testes is irrelevant. Ovaries 

 and testes are developed generative structures, and in the order 

 of evolution are preceded by undeveloped ones ; and the fact that 

 these undeveloped ones are found in little-developed members of 

 the type conforms perfectly to the hypothesis. [I may remark in 

 passing that here is a good illustration of that process of evolu- 

 tion which, in the above speculation of Prof. Korschelt, is sup- 

 posed to be inverted : many dispersed, similar, and indefinite 

 parts, are integrated into a few localized and definite parts.] 



In continuation the critic above quoted says : — " My position 

 is that the repetition of segments in an Annelid is a phenomenon 

 of the same nature as the repetition of hairs in a Mammal or of 

 scutes in a Reptile ", and he proceeds to give instances of repe- 

 titions of organs in other types, as of the reproductive struc- 

 tures and excretory system in the young Dog-fish or of the 

 ovaries in Amphioxus. These examples do not seem to me 

 relevant. No parallelism exists between the repetition of a par- 

 ticular organ in an animal, and the repetition of an entire cluster 

 of organs constituting a physiological whole. The repetitions of 

 the ovaries in Amphioxus and of the excretory system in a young 

 Dog-fish, occur without threatening to divide into similar parts 

 the entire organism. But the segmental repetitions in an annu- 

 lose creature implicate the structures at large, and would, if 

 pushed a little further, result in separate creatures. The segment 

 of a low Annelid contains alimentary, vascular, nervous, excre- 

 tory, reproductive, sensory and locomotive organs — all the organs 

 required for carrying on life, save certain organs of external 

 relation which its position excludes. When there is shown some 

 vertebrate animal, or proto-vertebrate animal, that is divisible 

 into parts each of which is in great measure physiologically inde- 

 pendent, I shall feel obliged to abandon my position. 



While this appendix is in hand I have received from another 

 expert, whose view is in general agreement with my own, a letter 

 containing the following passage : — 



" You will see that Dohrn's theory was the antithesis of your own view 

 of vertebrate structure, namely that the vertebra? were formed by the seg- 

 mentation, from mechanical causes of a body originally simple. This view 

 of yours has been confirmed by later researches, which have shown that 

 the most primitive forms allied to the Vertebrates, possessing the essential 

 organs, viz., gill-slits, notochord, and dorsal nerve cord, are not segmented 

 animals, like Annelids and Crustacea, but simple animals, having at most 

 three regions, not exactly corresponding to segments. These primitive 

 unsegmented forms are Ascidian tadpoles, Balanoglossus, and certain other 

 primitive forms. The embryology of Vertebrates also proves that they are 



