496 Cresswell Shearer 



certain parts of the mesendoderm this Separation is more niarked 

 than in others, especially in the regions of the endoderm lateral to 

 the primitive streak and in that region towards what will be later the 

 pre-orai lobe. Nevertheless the process is a general oue in which 

 there is invaginated a mesendoderm from which the mesoderni 

 separates off over the entire surface of the endoderm. 



De Selys Longchamps in his latest work (25), is also of the 

 opinion that this process is a general one. He says "le mésoblaste 

 de l'embryon prend son origine uniquement dans des cellules se 

 dctachant isolément de l'endoblaste ... il y a formation, aiix dépens 

 du mésendoblaste , d'un mésenchyme primaire et persistance du 

 blastocèle embryonnaire" (pag. 12). 



It remains for me to consider several minor points in the origin 

 of the mesoderra. Caldwell has put forward the view that the 

 uephridial pouches give off cells which form the mesodermal liniug 

 to the posterior body cavity. In pl. 32 fig. 37 it will be seen that 

 there are mesodermal cells already in this body cavity before the 

 nephridial pouches areformed and while the pouches are too rudimentary 

 to take any part in the formation of these cells. Ikeda has noticed 

 that sometimes mesodermal cells seem as if partially detached from 

 the ends of the pouches and about to separate off, but as I hope 

 to show later these cells are really giving rise to the tube of the 

 nephridial canal and take no part in the formation of mesoderm. 

 They take on a filiform shape with numerous processes, as can be 

 Seen partially in fig. 8. But this shape is soon lost and the cells 

 are transformed into the cuboidal cells of the nephridial canal. 

 Contrarr to Caldwell's opinion I have shown the pre-orai body 

 cavity, if we are to consider it as such, is from tbe first un- 

 paired in origin, and not paired as it would be if derived as he 

 believed from the anterior diverticula. In returning to Caldwell's 

 former vicw that the anal pit marks the old terminal position of 

 the primitive streak, I know that this is contrary to what Ikeda 

 has shown to be the case in Pìi. ijimai. He has shown that in this 

 species there is no genetic connection between the primitive streak and 

 the anal pit, as the primitive streak has disappeared some time from 

 the external surface of the larva before the anal pit has appeared. 

 This is not the case, however, with the larva of Ph. kippocrcpia 

 and that from Faro, where the ectoblastic Clements grow somewhat 

 more slowly. The primitive streak is still marked somewhat faintly 

 on the external surface as a slight depression of the ectoderm 



