1904.] NATURAL SCIENCES OF PHILADELPHIA. 717 



also cover a large portion of the lateral surface. Behind the blastopore 

 the descendants of X form a small group of cells, which in fig. 53 have 

 not yet covered in the primary mesoblasts. Mitotic activity in this 

 region has been very slight, as may be gathered from examination of 

 the optical section shown in fig. 54. The effect of active mitosis in 

 the posterior dorsal region is perceived in the thinness of the ectoderm 

 in this region. The X cells posterior to the blastopore appear to ad- 

 vance and fuse with the other ectoderm cells surrounding the blasto- 

 pore. Whether they actually form part of the blastopore I cannot at 

 present say, but judging from their position in fig. 55 it seems probable 

 that they do. The stomodaeum is formed, as has already been stated, 

 at precisely the point where the blastopore closed. It then moves 

 rapidly forward to occupy the position shown in fig. 59. This move- 

 ment is caused by the growth of the cells posterior to the blastopore, 

 which up to this time have been slow in dividing, as well as by new 

 additions from the posterior stem cells. This ventral plate of cells 

 derived from X corresponds in origin and position to that group known 

 among the annelids as the "ventral plate." Examination of the ven- 

 tral side of an embryo at this stage shows numerous mitotic figures, 

 and the rapid growth of this region is testified to by the thinning of 

 the ventral ectoderm and by the dorsal flexure shown in figs. 59 and 60. 

 From this time on, it is evident that the further growth of trunk 

 ectoderm is due entirely to growth of the cells of the X group, for, since 

 the appearance of the segments from in front backward indicates ter- 

 minal growth, it is evident that the descendants of X can alone be 

 concerned. 2d or X, then, contributes at least the larger part of the 

 trunk ectoderm, as is the case in all the polychaetous annelids whose 

 cell lineage has been studied. A larva of four segments is shown in 

 horizontal section in fig. 61. By comparison of this figure, which is 

 drawn to the same scale as the other figures, an adequate idea can be 

 obtained of the relatively enormous extent of the trunk ectoderm. 

 When it is recollected that this latter is formed principally, if not exclu- 

 sively, from 2d, without any addition of food material from the exterior, 

 the colossal size of 2d is very readily comprehended. A very similar 

 case is found in the development of Arenicola. In this form, as 

 already remarked, the first somatoblast is also unusually large, and 

 furnishes material for the first three segments of the adult worm. In 

 Dinophilus direct development has wholly supplanted the larval type, 

 since the material stored up in 2d forms six body segments, the 

 total number found in the adult. 



