246 W. R. Coe — Development of the 



cells which apparently correspond to " posterior entoblast " cells in 

 the annelid embryo. A further discussion of the origin of the 

 mesoderm will be given below. 



The process of invagination is somewhat different than in M. cceca. 

 The pear-shaped endoderm cells near the middle of the lower surface 

 become pressed further into the segmentation cavity so that a slight 

 ventral depression is left (Fig. 4, PI. xxxni). This marks the begin- 

 ning of the invagination. The individual cells slip away from the 

 outer surface, but remain closely in contact. As those in the center 

 pass inward they draw in their irregular, tapering, lower extremities 

 and arrange themselves about a central cavity — the archenteron. 

 This cavity is very narrow and communicates with the exterior by a 

 small, circular mouth (Fig. 4). As the process goes on it becomes 

 more like the typical invagination described above for M. ccBca. 



The intestinal canal is early differentiated into two regions as in 

 the two species above described, and is likewise provided with a simi- 

 lar intestinal valve (Fig. 6, PI. xxxiii). The esophagus remains very 

 narrow for several days, but later becomes much enlarged. It never 

 develops, however, to the great extent characteristic for the two 

 other species. Buccal folds are present as thickenings of the pos- 

 terior esophagal Avails. 



The muscular system develops as in the other species, and is 

 arranged in much the same way. The lateral bands to the lappets 

 consist of more numerous fibers which are not so regularly arranged. 

 The two bundles running from the apical plate to the anterior border 

 of the esophagus are well marked in this species. 



During the formation of the gastrula the ectoderm becomes very 

 thin, and the apical plate forms as in 31. coeca. There is usuallj' but 

 a single flagellum. This is nearly as long as the body. 



The mesenchyme cells are several times more numerous and larger 

 than in the other species, but have the same general arrangement 

 (Figs. 3-6, PI. xxxiii). Some of those which find their way into the 

 lappets later acquire a reddish tinge. There is also a slightly red- 

 dish tinge to the marginal cells of the lappets. 



This may, perhaps, be looked upon as an indication that this pilid- 

 ium possibly develops into the Pilidium gyrans of Miiller. The 

 time of the year when both may be found probably agrees. More- 

 over the geographical distribution of P. gyrans corresponds more or 

 less closely with that of C. margmatus. The great abundance of 

 C. niarginatus, and the enormous number of eggs produced, makes 

 the chances of obtaining pilidia of this particular species greater 

 than is the case with almost any other species. 



