FISSIPAROUS SPECIES OF TUBICOLAR ANNELID 339 



only furnishing additional evidence for the justice of that division of 

 the annelids into the Anne/ides proper, characterized by the separation 

 of their sexes — and the ScoUides, characterized by their hermaphro- 

 dism — which was first established by M. Milne-Edwards, and which 

 has been very generally received. 



However, on a careful survey of the whole class of worms, many 

 facts come to light which throw considerable doubt on the propriety 

 of raising unisexuality or hermaphrodism into distinctive characters 

 of large groups. We have hermaphrodite Rotifera, and unisexual 

 Rotifera. The Nemertidce and Microstomum are unisexual, the other 

 Turbellaria hermaphrodite ; there appears to be considerable doubt 

 ■as to the universality of hermaphrodism in the Trematoda even ; and 

 Echinorhynchus, which cannot be placed very far from the Toeniadce 

 and Distomata, is well known to be unisexual, and there is therefore, 

 perhaps, nothing so very anomalous in the discovery of a truly 

 hermaphrodite tubicolar annelid. It is another question how far it 

 need affect the classification to which I have alluded. 



The fluctuation in the terminology of the classification of the 

 annelids, in fact, has proceeded from the very common but always 

 obstructive practice of giving notional instead of trivial names to 

 incomplete groups of animals. Cuvier divided the annelids into 

 •errant, tubicolar, terricolar, &c., deriving his terminology from the 

 habits of those with which naturalists were then acquainted ; but, 

 with the advance of knowledge, it was found that some of the Errantia 

 inhabit tubes, while one main division of the " Terricola " consists of 

 aquatic worms ; and thus these notional terms, instead of aiding the 

 memory as they were intended to do, served simply to originate and 

 propagate erroneous conceptions. There can be no doubt that the 

 ■divisions established by Cuvier are essentially natural, and had he 

 ■devised some happily unintelligible Grecism, instead of the names 

 Avhich he actually adopted, they would have stood, their definitions 

 altering with the progress of knowledge, until this day. 



The divisions proposed by M. Milne-Edwards possess exactly the 

 qualification which is here wanting. Annelides and Scoleides may 

 mean anything, and, as names of groups, may very conveniently 

 remain, even if it should be found necessary to remodel the whole 

 definition which was primarily assigned to them. It appears to me, 

 therefore, that if the statements which follow be confirmed, they will 

 lead, not to an alteration or sub-division of the group of Annelides, 

 but to a widening of its definition so as to include hermaphrodite 

 forms ; or perhaps it would be better to admit that owing to the 

 imperfection of our knowledge, we have not yet a definition of either 



Z 2 



