498 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



included elements. The smaller are usually egg shaped, the larger like bunches 

 of grapes. In the same sacculus the most diverse types are seen side by side. 



In the living object the individual elements of each little group are colorless, 

 homogeneous, and strongly refractive. They stain readily and uniformly and 

 ordinarily show neither nuclei nor chromatin grains. But under special treat- 

 ment most of the elements can be recognized as true cells. 



Roughly in the center the nucleus, including a few chromatin grains, can 

 be differentiated. It is large and vesicular, surrounded by the scant protoplasmic 

 mass of the cell body, which, when the elements are closely crowded, are convex 

 only on the free side, the other sides being flattened against each other and poly- 

 hedral. Very commonly, however, a definite nucleus can not be made out, and 

 there are only to be seen chromatin grains apparently lying freely in the cyto- 

 plasm. The amount of chromatin is variable in the different elements; some- 

 times there is a considerable quantity, and again so little that its demonstration 

 is difficult. 



In a somewhat more advanced stage the cell clusters have become elongated 

 and the constituent cells spherical. Between the larger clusters lie a few smaller 

 irregular groups. There appears to be no definite arrangement of the large 

 groups, although the majority lie with the curved base on the cortical pavement 

 epithelium and the point centrally. Seeliger in a few cases saw clearly the fili- 

 form hyaline process at the pointed end of the cell clusters, though it was shorter 

 than in the adults and never reached the wall of the sacculus. No enveloping 

 membrane was discerned. 



By the isolation of the individual elements of a cluster it is made plain 

 that in spite of the absence of an inclosing membrane these cling rather closely 

 together. They are not all the same in structure. Relatively few can be shown 

 to be complete cells with a cell body and a small nucleus containing a few chro- 

 matin granules. Most of the nuclei contain no chromatin. Within some of the 

 cells is a clear vesicular body, usually placed excentrically, which is perhaps the 

 remains of the nucleus, or perhaps the beginnings of a vacuole. This last type 

 may represent the ultimate development. 



In the oldest larvae of this period five new sacculi are formed between the 

 original five and the oral surface. They do not develop at the same rate, with 

 the result that a single larva shows several different developmental stages in the 

 different radii. The separation of the original cluster of mesenchyme cells into 

 a peripheral and a central part in these sacculi occurs much earlier than in the 

 first five. 



Connective tissue. As in the period preceding the mesenchyme cells form in 

 connection with the elements of the outer layers in the body wall a uniform 

 tissue, in which only the most superficial cells, mostly spindle-shaped and stand- 

 ing more or less at right angles to the surface, can with certainty be traced back 

 to the ectoderm, while the deeper cells, bordering directly upon the hydroccele 

 or the parietal layer of the ccelome, undoubtedly originated from the mesenchyme. 

 In the intermediate layer no separation of the cells according to their origin is 



