CASTLE. — METAMERISM OF THE HIRUDINEA. 301 



If, however, a new ring were formed first at only one end of the somite, 

 the probabilities are that it would be at the posterior end. For we have 

 found that ring multiplication occurs rather oftener there than at the 

 anterior end of the somite. Moreover, in somite reduction in Glossi- 

 phonia, we get a stage (Fig. 6, B) intermediate between the one-riuged 

 (Fig. (\, A) and three-ringed (Fig. 6, C) condition of the somite, in 

 which rings 1 and 2 are united into a single broad anterior ring, while 

 ring three is entirely distinct. 



The intermediate condition just described not improbably represents a 

 true phylogenetic stage in the formation of the three-ringed type of somite, 

 for it corresponds exactly with the condition found 

 in a typical body somite of Branchiobdella or of 

 Bdellodrilus (Fig. 8), those curious leech-like 

 annelids, which nevertheless possess certain chae- 

 topod characters. The ganglion (g., Fig. 8) of a 

 somite of Bdellodrilus or Branchiobdella is situ- 

 ated, as we should expect if the views just ex- 

 pressed are correct, in the broader anterior ring 

 of the somite. This ring, according to the 



homology suggested, corresponds with rings 1 Fig- 8. A single body 



and 2 of Glossiphonia (Fig. 6, B and C). somite of Bdellodrilus Ufa- 



T „ x , .. . .... minatus, lateral view, a., 



If, as 1 believe, the common assumption is well , . . 



r anterior ; a., dorsal ; 7., 



grounded, that leeches and chaetopods have been o- a nglion ; p., posterior; 

 derived from a common stock, can we discover v., ventral. After Moore 

 any reason why ring multiplication should take C y 5). 

 place in one group and not in the other ? I think 



we have a sufficient explanation in the fact tint the leech body contains 

 always the same definite number of somites, no matter how large or how 

 small the animal may be. This number is thirty-four, both in the 

 Ilhynchobdellidae and in the Gnathobdellidaa. 



In the chaetopod, on the other hand, the body contains at first a 

 relatively small number of somites, which is increased, as the animal 

 grows in size, through the formation of new somites in some limited 

 region of the body, usually at its posterior end. In some cases the newly 

 formed somites may separate themselves off as a distinct individual, in 

 other cases they serve merely for the elongation of the original indi- 

 vidual to an indefinite extent. 



In the leech, however, there is no provision for increase in the 

 number of somites. A definite number of somites, thirty-four, is laid 

 down early in ontogeny, and never increased. Elongation of the body 



