CASTLE. — METAMERISM OF THE HIRUDINEA. 295 



can judge from secoud-hand knowledge of the subject, are entirely 

 accurate. 



A comparison of Figures 1 and 4 shows that abbreviation has been 

 more extensive in the head end of Nephelis than in that of Glossiphonia. 

 The process also does not progress anteriorly with such even and regular 

 gradations as in the case of Glossiphonia. Nevertheless, abbreviation 

 takes place in both cases by practically the same steps. First the 

 anterior end of the somite is affected, then the posterior end. 



XXIV. 



XXV. 



Fig. 5. Posterior end of the body of Nephelis lateralis, dorsal view. Somite 

 limits according to the view of Bristol are shown at the left of the figure 

 (XXIV.'-XXVII.') ; according to the writer's view, at the right (XXIV.-XXVII.). 

 After Bristol ('99). 



Figure 5 gives a dorsal view of the abbreviated somites at the 

 posterior end of Nephelis. Somite XXIV. (right half of figure) is the 

 last unabbreviated somite and shows the typical condition, — a sensory 

 ring in the middle of the somite preceded as well as followed by two 

 non-sensory rings. In somite XXV. (ri<rht side of figure), the sensilla? 

 indicating the fixed third ring from which to reckon, rin^s 4 and 5 are 

 seen to be partially united. In somite XXVI., rings 1 and 2 are com- 

 pletely fused together and imperfectly uuited with the sensory ring, 3; 

 rings 4 and 5 are likewise fused together, but retain their distinctness 



