12 C. M. CHILD. 



appear rather early in development and often far from their 

 final position, which they attain by migration. The evidence 

 from observation in these cases also is merely evidence foi early 

 differentiation or specification of germ cells, not for continuity 

 of undifferentiated germ plasm. If the germ cells are independent 

 of the soma why should the extensive migrations occur? They 

 behave as if they were physiologically integral parts of the 

 organism, rather than mere parasites upon it, for certainly their 

 migration cannot be regarded as autonomous. Perhaps it should 

 also be pointed out that the evidence for the origin and migration 

 of the primitive germ cells in these forms rests entirely upon 

 the study of fixed material. So far as I am aware, the migrations 

 have not been observed in the living embryos. Does not at least 

 the possibility exist that cells of this character may appear as 

 temporary differentiations or stages in various parts of the 

 embryo, or perhaps that cells with a low rate of metabolism 

 possess these characteristics? But without laying loo much 

 weight upon these last suggestions, we are, I think, justified in 

 maintaining that the early appearance of primitive germ cells in 

 vertebrate embryos, when taken as a fact of observation, con- 

 stitutes evidence for early embryonic differentiation or specifica- 

 tion, rather than support for the germ plasm hypothesis. 



On the other hand, there is a very large number of forms, both 

 plants and animals, in which the germ cells do not appear until 

 development is completed and then in many cases only period- 

 ically. Moreover, in many of these cases they arise from tissues 

 which are visibly differentiated and constitute functional parts of 

 the organism. The coelomic epithelium of the polychaetes, for 

 example, is certainly not, if appearances have any significance, 

 an undifferentiated tissue, yet at certain seasons or tinder certain 

 conditions the germ cells arise from it. It is of course easy to 

 assume chat the nuclei of these cells always contain undiffer- 

 entiated germ plasm, but this is merely forsaking fact for hy- 

 pothesis. 



The parenchyma of the turbellaria and the cestodes is likewise 

 not morphologically an undifferentiated tissue, yet it gives rise 

 to germ cells at a certain stage of development and, if my observa- 

 tions are correct, even the more highly differentiated muscle 



