CONTRIBUTIONS TO THE ANATOMY OP THE HIEUDINBA. 427 



determined by series of section through it, the preceding and 

 following annulus, lodges the ganglion. The testis lie partially 

 in this annulus, but by far the greater part of it lies in the 

 second annulus. The nephridial funnel, as I shall show below, 

 lies upon the wall of the testis, and thus has a similar position 

 to that obtaining in Pontobdella. The nephridial pore to 

 the exterior lies in the last annulus of a somite, while in 

 Pontobdella it opens just within the first annulus of the 

 next succeeding somite (fig. 4). 



I have never seen a clitellum in Hirudo; butin Aulostoma, 

 which exactly corresponds in other respects, it commences in 

 the annulus immediately following upon the third pair of 

 nephridial pores, extends over exactly three somites (5, 6, 

 and 7), and the generative pores lie in the second annulus of 

 the second and third of these somites respectively. 



Anteriorly, if we suppose that two annuli are missing (we 

 have seen a similar condition in Pontobdella), and have 

 been fused in the formation of the mouth, we have the 

 first somite presenting three annuli : the second is normal 

 with five annuli. The clitellum is thus seen to extend over 

 somites 5, 6, and 7, the male pore to lie in somite 6, and the 

 female in somite 7 (see fig. 4, g. jo., ^, and g. p., ?). Pos- 

 teriorly the somites are crowded together (Leuckart^s observa- 

 tions upon the fusion of primitively distinct ganglia in this region 

 must be remembered in this connection), but twenty somites may 

 be counted. The anus lies well forward in a region following the 

 nineteenth somite. This region must be regarded as representing 

 in a fused condition at least seven ancestral somites. 



Thus, Pontobdella and Hirudo, which I choose as repre- 

 senting the extreme conditions, Branchellion excepted, with 

 regard to external characters, show such characters to be almost 

 absolutely identical. This seems to me a striking result, 

 the meaning of which I shall refer to later on. 



The other Gnathobdellidse which I have examined in this 

 respect present characters almost identical with those of 

 Hirudo. 



