142 Dn. 



J. STEPHENSON ON THE MORPHOLOGY, CLASSIFICATION, 



the accessory spermatliecal glands in Draioida schunkarai 

 (Michaelsen, 16) ; similarly there are associated yet indepexident 

 glands in the geuei-a Hoplochceiella and Erythrceodrilus, belonging 

 to another family, the Megascolecidje. 



(6) The Origin of the Moniligastridce. 



If the testis sacs represent segments, they must have existed 

 as such in the ancestors of the present genera. Let us take 

 Desmogaster, as the most archaic of existing genera of Monili- 

 gastridpe, and expand the testis sacs into segments ; "we get a 

 form which may be represented as in the accompanying figure 

 (text- fig. 15). 



Text-figure 15, 



VM 



XI 



db:^ 



XII 



Desmogaster, with tlie testis sacs expanded to appear as segments. 



Now let US take the remai-kable genus Syngenodrilus, described 

 by Smith and Green (20), and separated by them as a distinct 

 subfamily. The actual arrangement of the sacs and septa is 

 approximately represented in the subjoined figure (text-fig. 16), 

 which is a diagrammatic copy of the section illustrated in the 

 original paper. 



Text -figure 16. 



v.s. o.s 



XII XIII 



XIV XV 



VII YUl IX X / XI 



pr 



Syngenodj-ilus. The external segmentation is indicated. Fr., prostate. 



Note the three pairs of prostate gia.nds, which, however, are none 

 of them in close association with the male dvicts or pores ; it is 

 quite possible, therefore, that these prostates of Syngenodrilus 

 have nothing to do with the structures that go under the same 

 name in the other genera of the family. If we expand this 

 fioure, so as to sepai'ate the testis sacs (which are close up against 



