BY E. J. GODDARD. 



55 



According to Whitman's scheme and Castle's scheme respec- 

 tively, we find the metamerism of the anterior extremity to be as 

 follows : — 



It is now regarded by most workers on the group that the 

 somite consisting of a few annuli is more primitive than that 

 which is multianniilate, and with this view I am in accord. This 

 view is then, in efiect, that aunulation of the segment is a 

 secondary character. Consequently when we find two annuli 

 intimately fused we must regard this state of affairs as the result 

 of an incomplete differentiation and not of abbreviation. 



Now we find in L. australis that annuli 5, 6, and 7, 8 are well 

 diiferentiated from each other respectively on the dorsal side, but 

 they are fused ventrally. This shows clearly that 5 and 6 belong 

 to one and tlie same somite, 7 antl 8 to another somite. This 

 being so, it must then be concluded that the sense-papillae repre- 

 sented by the eyes situated on annnlus 8 do not lie on the first 

 annulus of the somite. If Whitman's scheme applied in this 

 case, then we should have th« absurdity of an annulus in one 

 somite originating from that in another somite. In the table 

 given above, I consider the annuli as distributed according to 

 Castle's scheme to be the correct one. On the assumption that, 

 at the extremities, we find the somite developing through the 

 same stages through which the pentannulate somite passed, and 

 that no " fusion " has taken place at all, we find that somite iv. 

 consists of two annuli of which the first is sensory; the second 

 annulus has arisen posterior to this annulus, but, inasmuch as it 



