394 Ralph S. Lillie 



characteristic relations of tliis cpithelium to the blood-vessels. As 

 explained above, the formation of the scries of dorsal vasciilar pro- 

 cesses, and also of the vesicle of the ventral lip, is in ali probability 

 primarily dependent upon the increase in epithelial surface. I bave 

 not however been able directly to study the early condition of these 

 characteristic structures, which do not arise until a late stag:e of 

 development and bave not yet begim their appearance in the most 

 advanced larvae at ray disposai. 



Terminal Vesicle. 



The terminal vesicle — as might bave been expected from the 

 similarity of its adult structure to that of the glandulär region — 

 is formed as a dilferentiation of the most posterior portion of the 

 primitive nepliridium. There is no ectodermal invagination and 

 from the resemblance in structure between the two regions it would 

 appear a priori probable that the vesicle, like the tubulär portion 

 of the nephridium, is chiefly if not eutirely of mesoblastic origin. 

 A complete fusion, however, takes place at an early stage between 

 the ectoderm and the posterior end of the nephridium, and it is 

 from this region of fusion, in which the limits of ectoderm and 

 mesoderm become indistinguishable, that the vesicle is differentiated. 

 In ali probability, therefore, its distai portion is derived from a 

 region which originally was of ectoblastic origin; it seems however 

 impossible to decide the extent to which each germ-layer has taken 

 part in its formation. In point of fact the vesicle is differentiated 

 as a whole, without regard to limits of germ-layers, from the ex- 

 treme posterior region where the nephridium and the ectoderm 

 become iudistinguishably fused (see Plates 24 and 25, Figs. 34, 

 37—42). 



Fig. :U, Piate 24 represents the early nephridia of somites V and 

 VI in a larva of 14 somites. The nephridium is fused with the ecto- 

 derm slightly behind and internai to the ventral hooked setae at a 

 point wliere the ne|)hropore afterwards opens. At the point of 

 fusion, ectoderm and mesoderm appear simply to become continuous 

 with each other without any visible line of demarcation. The por- 

 tion of the nephridium immediately adjoining the fusion remains 

 similar to the rest of the organ until a relatively late stage 

 cf. Plates 24 and 25, Figs. 37 — 12). Fig. 40 represents a cross section 

 thruugli the jMisterior portion of the first nephridium of a larva of 

 the full nuiiiber of somites and a length of abdut 1,75 mm. The 



