ASSOCIATION OF SOMATIC AND GERM CELLS IN CESTODES. 313 



In the accompanying figure, however, I have shown what I 

 consider good evidence in support of the former v'ew. 



This shows part of a proglottid of Dipylidium caninum con- 

 taining numerous early testes. One of these contains but two 

 cells, of which one is a developing flame cell (t'} showing only 

 the cone of cilia, while the other is a testis cell. Another shows 



A part of the parenchyma of Dipylidium caninum, showing developing testes (t), 

 two of which contain flame cells (t 1 and t 2 ), camera drawing. X 900. 



several testis cells and one developed flame cell (/ 2 ), which is not 

 yet however connected with any capillary. 



Obviously, I cannot prove that the groups containing t 1 and t 2 

 are developing testes. Their similarity of structure to neigh- 

 boring testes is however sufficiently evident. 



But whether or not these flame cells and the cells of the testes 

 in which they lie are of common parentage, their occurrence 

 within the testes is evidence of the simplicity of cestode develop- 

 ment; for it shows that apparently any cell of one tissue (testis 

 or parenchyma) may develop into a specialized cell of another 

 tissue (flame cell) entirely apart from any other cells of the 

 latter; the apparently determining factor being a stimulus of the 

 physiological environment. 



The question why two cells in close proximity to each other 

 and probably of common descent, should develop on the one 



