THE I.VADEQi'ACV OF THE CEIJ.-THEORY. 121 



inside the embrvonic tissues as in the ejjo-s of fresh-water 

 fishes, but actually outside the tissues on the inner face of the 

 embryt), near its posterior end. Its whole ventral and lateral 

 boundar}' is formed, not by archenteric cells, but by a periblastic 

 layer often as thin as the wall of a soap-bubble, and completely 

 free from all nuclei. It does not even arise as a single cavity, 

 but as numerous minute cavities that look like a cluster of 

 granules. These expand, flow together gradually, and finally 

 form one bubble-like vesicle projecting almost wholly into the 

 transparent yolk. Having attained a maximum size, its slightly 

 concave roof becomes more and more deeply hollowed out, and 

 thus it comes to inclose more and more the cavity, while the 

 latter gradually shrinks in size, and finally vanishes as the true 

 cell-walls close up. Such is briefly the history of this floor-less 

 form-reminiscence of what is a more substantial rudiment in 

 many other embryos. 



This remarkable reproduction of a form-phase that is to last 

 only for a few hours and then pass away without leaving a 

 visible trace of its existence, cannot be explained as due to cell- 

 formation nor as the result of individual action or interaction 

 on the part of the cells. The embryonic mass acts rather as a 

 ;/////, tending always to assume the form peculiar to the state of 

 development reached by its " essential architectonic elements " 

 (Briicke) — elements that are no less real because, like the atom 

 and molecule, they are too minute to be seen by the aid of our 

 present microscopes. 



That cells as such do not participate in this formative act, is 

 shown by the mode of development of the vesicle and by the 

 absence of cells in its ventral and lateral walls. This fact, the 

 absence of cells, has actually been urged recently against the 

 identity of the structure with Kupffer's vesicle, — -an error 

 which one is likely to fall into only while under the delusion 

 that acellular walls cannot be homologous with cellular walls. 



The evidence furnished by Kupffer's vesicle will doubtless 

 lose much of its force with those w^ho have not had an oppor- 

 tunity to study the subject sufficiently to form an independent 

 opinion about it. To some who are better acquainted with the 

 structure, its meaning may still appear to be somewhat prob- 



