286 PROCEEDINGS OF THE AMERICAN ACADEMY. 



him their limits. These were, first, some whitish flecks (sensillse) found 

 dorsally on the ring which contains the ganglion (sensory ring) ; secondly, 

 a pair of pores (nephropores) lying in the furrow just posterior to ring 5, 

 counting the sensory ring as ring 1. 



Gratiolet seems to have assumed that these convenient landmarks 

 indicated respectively the anterior and posterior limits of the mor- 

 phological units or somites. In this assumption nearly all subsequent 

 investigators have followed him uuquestioningly. Vaillant ('70) alone 

 expressed the view that it would be more logical to look for septa, which 

 should, as in the Cha^topoda, mark the limits of the somites. In the 

 case of Pontobdella, Vaillant believed that he had succeeded in finding 

 such septa, but subsequent investigators do not agree with him in this 

 view. 



Gratiolet further observed that toward the ends of the body the 

 number of rings in a somite becomes smaller, and the ganglia are 

 crowded closer together. He suggested that the terminal ganglion at 

 either end of the body probably had resulted from a fusion of originally 

 distinct ganglia, as is the case with the ganglia of certain Mollusca 

 {Helix, etc.). 



Gratiolet thus opened the way for a study of the metamerism of the 

 leech in two directions : first, to determine the number of primitive 

 somites present in the body of the leech ; and secondly, to determine the 

 natural limits of these somites. The first problem indicated has received 

 a satisfactory solution through the splendid work of Whitman ('92) on 

 the Rhynchobdellidre, and of Bristol ('99) on the Gnathobdellidaj. 



From studies of my own, the results of which are in process of pub- 

 lication, I can confirm the conclusion of these investigators, that the 

 body of the leech contains thirty-four distinct somites. 



The second problem, namely, the determination of the true limits of 

 the somite, has, as already indicated, received practically no attention. 

 To a discussion of this problem and some considerations growing out of 

 it, I now invite attention. 



II. Limits of the Somite in the RHTNCHOBDELLiD.iE. 



1. Neuromeres as a Criterion of Somite Limits. 



In determining the number of somites found in the body of the leech, 

 it has been found necessary to rely solely on a study of the nervous 

 system. It would seem natural, therefore, to look in the same direction 

 for a solution of the problem conceiniug the limits of the somite. 



