THE THIRD SOMITE. 81 



ventral part does not break up into spaces, nor does it become 

 traversed by muscles and connective tissue, but persists through 

 life as a vesicle with an epithelial lining. The lateral sinus 

 has nothing to do with the coelom, but comes from the space 

 marked b. lat. in my sections. The funnel of the nephridium 

 opens always into the lateral part of the somite, of which, 

 indeed, it is a part ; and does not become blindly closed. The 

 funnels of the other nephridia do not open into the space 

 of the feet, but into the lateral division of their proper 

 somites exactly as do the salivary glands. 



Kennel further maintains that the funnel only of the adult 

 nephridium is mesodermal. I cannot accept this ; it is an 

 altogether fanciful view. It may be true, but it is a point quite 

 impossible to settle by sections. With regard to it, I have 

 only to say that the ectodermal ingrowth at the opening of the 

 nephridum is extremely inconspicuous, and that at the early 

 stage, immediately before and after the establishment of the 

 external opening, an ectodermal part such as Kennel describes 

 can only be made out with difficulty. 



The dorsal divison of the third somite separates from 

 the ventral in an embryo slightly older than that from which 

 series fig. 23, Plate VII was taken, viz. one in which the cerebral 

 grooves were slightly more advanced than in fig. 22, but not so 

 much developed as in fig. 33, Plate VIII. After its separation, 

 it becomes much reduced in size (fig. 36, 6?.5. 3),then still smaller 

 (Plate IX, fig. 45, d. s.), and finally vanishes (fig. 46, d. s.). 



This completes what I have to say about the third somite 

 up till Stage f, by which time the adult condition of the parts 

 is practically attained. The remaining somites conform, on 

 the whole, to the type described. They do, however, present 

 certain diflferences, of which, perhaps, the most important are 

 found in the posterior part of the body. In the posterior 

 somites the dorsal divisions do not become obliterated, but 

 persist and give rise to the generative glands (Plates VIII, 

 IX, figs. 41, 43, d. s.=gen. o.). It will be more convenient, 

 however, to defer the detailed consideration of this and other 

 peculiar features of the remaining somites until after the 



