THE EDWAEDSIA-STAGE OF LEBUUNIA. 287 



probably shares in the formation of the muscular system, for in 

 connection witb it delicate fibrils are recognizable on the column- 

 wall (fig. 21) and mesenteries (fig. 16). 



The limiting cells next the supporting lamella of the column- 

 wall are not arranged in any definite layer, and appear to differ 

 in no respect from those more central. 



Nuclei sometimes occur here more abundantly than else- 

 where, and often in a similar manner along the two faces of the 

 mesenteries. They are probably connected with the formation of 

 the mesogloea, and some seem to become included within it. 



To recapitulate : The archenteron terminates internally in a 

 small central cavity, from which four canals radiate and open 

 into two circumferential sinuses, extending slightly above and 

 below. In the middle archenteric region of the larvse the sinuses 

 subdivide into eight distinct chambers, one for each radial 

 division, and ultimately each of these terminates in two cornua, 

 sixteen in all. Below the archenteric region the sinuses terminate 

 in eight V-shaped canals, each of which is formed, as it were, by 

 the union of two cornua from adjacent mesenterial spaces. 

 Owing to this the lower terminal chambers are mesenterial in 

 position, while the upper must necessarily be intermesenterial. 

 A ciliated limiting layer closely resembling in structure, and in 

 absolute continuity with, the archenteric eudoderm is throughout 

 easily distinguished from the general mass of undifferentiated 

 tissue. This system of spaces I regard as a larval ccBlome or 

 body-cavity arising in connection with an archenteron, but not 

 cut off from it. 



Instead of representing a definite and independent system of 

 larval coelomic diverticula, it may be urged that the spaces are 

 merely the initial clefts in the disorganization of the solid interior, 

 to be completed later. Against this I would point out that the 

 canals and spaces all exhibit the distinct limiting layer, very 

 different from anything to be associated with disintegrating 

 tissue; and that in the intermediate larvse, where the process 

 can be followed, the breaking down of the cells commences just 

 below the internal opening of the archenteron, attacking first 

 the central mass. It is not until very late that any alteration takes 

 place in the more proximal and distal spaces. And further, there 

 is evidence that the greater part of the limiting layer becomes the 

 unilaminar epithelium of the adult cavity. It is not disintegrated. 



22* 



