496 ADAM SEDGWICK. 



earlier stages (Part I, fig. 24 a — d). These latter spaces 

 are comparable with blastocoele spaces of other embryos^ so 

 called because they present the relations of the segmentation 

 cavity of the earlier stages. Such a name is^ however, obviously 

 out of place here, inasmuch as the segmentation cavity is never 

 present. 



In the next stage (f) the dorsal division of the somite has 

 entirely separated from the ventral, so that the two parts may 

 be considered separately. 



The ventral division of the third somite presents the 

 same parts as in the previous stage. These were, it will be 

 remembered, (1) what I may call a vesicular internal part, ex- 

 tending to the hind end of the appendage and forwards as an 

 anterior diverticulum, and opening into (2) a tubular part, 

 projecting ventrally and opening to the exterior. 



In Stage r the vesicular internal part (PI. XXXVI, figs. 36, 

 37, /. s. V. 3) has not only lost its connection with the dorsal 

 (median) part of the somite, but its peripheral part — viz. that in 

 the appendage — has become largely obliterated by the increase in 

 the thickness of its mesodermal walls and by the growth of the 

 slime-gland rudiment. At the same time the tubular part has 

 become longer and more twisted (figs. 37, 38, l.st.^), and its 

 external opening covered up by the lips, which have met on the 

 ventral surface (Part I, fig. 36). The tubular part, therefore, 

 no longer opens freely but into the posterior part of the buccal 

 chamber. 



I will now describe more in detail the structure of the two 

 parts. 



The inner walls of the internal vesicle retain (figs. 36, 37) 

 the character presented by the walls of the anterior diverti- 

 culum of the previous stage (a. v. PI. XXXIV, fig. 23 b) ; i. e. 

 the nuclei are relatively far apart, and separated by a slightly 

 staining protoplasm. The outer wall, on the other hand, is 

 reduced to a thin layer. 



The tubular portion I shall now call the nephridium. Its 

 opening into the vesicle, shown in fig. 38, is a well-defined 

 structure, which I shall call the funnel of the nephridium. 



