538 PROCEEDINGS OF THE ACADEMY OF [July, 



cut off from, though in contact with, a fairly distinct group of cells 

 bordering the surface of the body. It thus appears that the cells 

 enclosed by the ring of larger cells become transformed into a sense 

 organ, perhaps the dorso-terminal sense organ known to occur in 

 many solenogastres. 



In the latest stages the test becomes considerably reduced in size 

 and the trunk gains proportionately in prominence. This increase 

 in the extent of the trunk appears to be wholly due to the division 

 of definite ectoderm trunk cells and not to any products supplied 

 by the test. Measurements show conclusively that the test cells 

 gradually shrink in bulk, probably due to the absorption of their 

 nutritive products, and karyokinetic spindles indicate activity on 

 the part of the trunk ectoderm. To what extent this proceeds it is 

 impossible to state. In early stages there is one, possibly two cells 

 situated immediately beneath the cells (fig. 1, C, b) that I believe 

 furnish the material for the cerebral ganglia. In later stages there 

 are indications that this deeper seated cell has undergone a few divis- 

 ions, and the resulting products occupy the space (fig. 1, D) between 

 the alimentary canal, cerebral ganglia and test cells of the head 

 region. They may possibly represent mesoblastic products, but 

 their relatively large size (for the sake of clearness, they are smaller 

 in the diagram than in reality) and their position suggests that they 

 may supply the material for the head epidermis as the test recedes. 



Turning now to the development of the species described by 

 Pruvot, we find that especially in Proneomenia aglaophenioe the early 

 development follows essentially the same path as in Halomenia. 

 In the other species, Myzomenia banyulensis, there is a decided 

 difference in the size of the cells during the early cleavages, but in 

 both cases the close of segmentation finds the larvse constructed 

 upon the same plan. The test is evidently of greater size than in 

 Halomenia and more completely envelops the remaining cells so 

 that they are hidden in lateral view, but the arrangement of the 

 cells is evidently quite similar to that shown in diagram C. The 

 enclosed elements are supposed to be endodermal in character, and 

 the region (depressed in the species studied by Pruvot) bordered 

 by the test is termed the blastopore. I believe both of these state- 

 ments are incorrect as I shall now attempt to demonstrate. 



Generally speaking, the velum of the trochozoa forms only an 

 insignificant portion of the ectoderm. In the solenogastres it has 

 expanded to such an extent that it comes in contact with the cells 

 destined to form the cerebral ganglia, and posteriorly it forms a 



