DESCRIPTIONS OF GENERA AND SPECIES 231 



The distinctively Tubificid characters appear to be (i) the presence of two kinds 

 of setae, capilliform and uncinate, and the frequent modification of the last to form 

 pectinate setae; (2) a long spermiducal gland, with a protrusible penis and a prostate; 

 (3) a supra-intestinal, in addition to a dorsal, vessel ; (4) the contractility of at least 

 one pair of the perivisceral trunks ; (5) absence of penial setae. 



I should, therefore, regard the genera Tubifex, Hemitubifex, Spirosperma, Psammo- 

 ryctes, and Lophochaeta, as occupying a central position in the family. So far as can be 

 seen at present, llyodrilus is the simplest form of Tubificid, connecting this family 

 with the Naidomorpha. It may be, therefore, placed at the base of the genealogical tree 

 which indicates the probable relationships of the different genera. Simplification of 

 structure is doubtless often due to degeneration ; but the resemblance which llyodrilus 

 shows to the Naidomorpha is not entirely due to the fact that it, like them, is a worm 

 of simple organization. The simplification of the male efierent apparatus, absence of 

 prostate and penis, might, indeed, be regarded from this point of view; but the 

 connexion of the intestinal blood-plexus with the ventral vessel, the network formed 

 by the nephridial tube immediately behind the funnel, and the presence of penial setae, 

 are characters shared with the Naidomoi-pha which are by no means suggestive of 

 degeneration; neither is the integumental blood-plexus. 



Limnodrilus and Clitellio are so far specialized genera, in that they have lost the 

 capilliform setae ; the latter genus also has lost the distinctively Tubificid prostates. 

 I should regard these forms as an offshoot, which has, however, progressed but a little 

 way from one or other of the more typical genera. An attempt to fix the position of 

 Telmatodrilus is more difficult. It shows certain resemblances to the very aberrant 

 genus Branchiura, in the possession of several contractile perivisceral trunks, and in 

 the integumental network. The last-mentioned character, as well as the absence of any 

 specially dilated hearts, connects Telmatodrilus also with llyodrilus. Perhaps these 

 characters indicate that Telmatodrilus has originated from the Tubificid stem a little 

 before the acquirement of all the characteristic features of the group. Branchiura 

 also must be looked upon (in my opinion) as a, comparatively speaking, primitive 

 type. It has not lost the integumental network ; the perivisceral loops of five or six 

 segments are contractile, the spermiducal gland is like that of llyodrilus, but the 

 relations of the gland to the vas deferens indicate a specialization, as do also the 

 branchiae of the hinder segments. The same arguments apply to Hesperodrilus. 



Bothrioneuron is also a much-specialized type, as is shown by the absence of 

 spermathecae, and by the peculiar form of the prostates. 



The mutual relations of the different genera, as I believe them to be, are indicated 

 by the accompanying phylogenetic scheme. 



