TEOCHOPHOEE OF HYDROIDES UXCIXATUS (eUPOMATUS). 569 



seen in the seo-mentation-cavity. Their origin I have not 

 succeeded in observing-. They are shown in Text-fio-. 18, 

 Imr. and l)id. I believe they arise from cells of the third 

 quartette, but as I have not followed the cell-lineage care- 

 fully, I am by no means certain of their exact origin. Thev 

 sink into the cleavage-cavity daring gastrulation, and take 

 up a bilateral position on either side of the blastopore, as 

 shown in Text-fig. 18. They immediately divide, giving rise 

 to some irregular small cells that apply themselves closely to 

 the wall of the stomodteum, and later form larval muscles. 

 One large cell on either side gives rise to a string of cells, 

 which enter into close relation with the coelomesoblast. Fi'om 

 their mode of origin and their subsequent behaviour I think 

 there can be no doubt that they represent the larval or 

 ectornesoblast of Podarke and Thalassema. In addition 

 to these cells, some mesenchyme cells are also constantly seen 

 in slightly later stages (figs. 9 and 10) in the apical region 

 under the "cross-cells.'' Whether they arise by migration 

 of some of the cells from the stomodamm, or bv the sinkino- 

 in of ectoderm ceils in the apical region, which last I think 

 is more likely, I have not determined. As in Podarke and 

 Thalassema and molluscs, therefore, the larval mesoblast 

 can be divided into the median, the portion under the apical 

 organ, and the right and left portion on either side of the 

 blastopore. These cells {Imr. and 1ml. of Text-fig. 18) 

 would correspond with the right and left parts of the ecto- 

 mesoblast of Podarke and Thalassema. 



It is worth repeating the description of these strnctures in 

 these forms. In Thalassema, Torre}^ (41) states, "The 

 most important source of functional mesench^'me, in Thalas- 

 sema, are the three cells 'frouT the third quartette, namel}^, 

 3d^ n n 1, oc^ ^ o p and and 3a,, ., 3- The first two sink into 

 the cleavago-cavity, just before gastrulation, and lie at first 

 close to the coelomesoblast cells. The}^ soon migrate laterally, 

 and bud off simultaneously szuall cells towards the mesoblast 

 cells, dividing like teloblasts, but in the reverse of the 

 ordinary direction. So close is the connection of these cells 



