294 PROCEEDINGS OP THE AMERICAN ACADEMY. 



ring of the somite (ann. n.), a structure typically found in annulus 2 

 (see somite VIII,, Fig. 4). 



Somite VI. clearly consists of three annuli, of which we can identify 

 3 and 2 by the dorsal seusillipe and the anterior nerve ring respectively. 

 In place of annuli 4 and 5 we have in somite VI. a single annulus, in 

 which, however, Bristol found no nerve ring. 



Somite V. consists, apparently, of two annuli. But what the distri- 

 bution of nerve V.'' shows to be undoubtedly the anterior nerve ring of 

 this somite lies just within the limits of the next anterior annulus (IV.', 

 left half of Fig. 4). Our first impulse would be to include that annulus 

 in somite V.; but a careful study of the distribution of nerves I.^-IV." 

 shows that the annulus in question contains important structures (includ- 

 ing the eye and associated sense organs) belonging unquestionably to 

 somite IV. It is possible to suppose that in this case the most anterior 

 part of somite V. has fused with somite IV. It seems to me, however, 

 more reasonable to explain the condition as illustrating a general 

 tendency in the head region for nerves to be cari'ied forward of the 

 somites to which typically they would be limited. For example, nerve 

 v." (Fig. 4) is carried forward ventrally as far as the last annulus of 

 somite II.; nerve IV." runs into the anterior annulus of somite II. ; 

 nerves II." and III." are carried forward into somite I. In Glossiphonia 

 (Fig. 1) also a similar tendency can be recognized in the distribution of 

 nerves III."-V." The nerve ring found in the posterior portion of the 

 annulus assigned by Bristol to somite IV. , (Fig. 4, left half) cannot 

 reasonably be considered the posterior nerve ring of somite IV., because 

 it is connected exclusively with nerve V." (See Bristol's PI. Yll., 

 Fig. IG.) Moreover the posterior nerve ring has disappeared in somites 

 V. and VI., we should therefore expect to find it wanting also in the 

 more abbreviated somite IV Finally, unless this nerve ring does 

 belong to somite V., that somite contains no nerve ring at all, and somite 

 IV., which is much more extensively al)breviated, still retains a nerve 

 ring. This seems very improbable, for we find that metameric structures 

 omitted from an abbreviated somite, do not appear in other somites still 

 more strongly abbreviated. 



In determining the external limits of somites I. -IV., Bristol has been 

 guided by the position of the metameric sense organs and the distribution 

 of the metameric nerves I. "-IV.", which represent the nerves given off 

 from one side of primitive ganglia I.-IV. respectively, completely fused 

 into single trunks. The somite limits indicated by him in the case of 

 these four metameres are, accordingly, neuromere limits, and, so far as I 



