216 J. MURRAY ON GASTROTRICHA. 



group, and will not attempt to modify the generic arrangement, 

 beyond shifting about some of the species. I am not qualified to 

 deal with the question, but as some little assistance to students 

 1 shall point out some apparent shortcomings of the prevailing 

 classification. 



The three fork-tailed genera, Ichthydium, Chaetonotus and 

 Lepidoderma, are separated on very slight characters, as Stokes 

 (57) recognised in " lumping " them all together. Ichthydium 

 has neither plates nor dorsal bristles ; Lepidoderma has scales, 

 but is supposed to have no bristles ; Chaetonotus has bristles, 

 and may have scales. So if an Ichthydium or a Lepidoderma 

 possesses any dorsal bristles it becomes a Chaetonotus. How 

 many bristles are necessary ? Some so-called Lepidoderma have 

 a very few bristles. L. loricatus, Stokes, has no bristles, while 

 a variety has four near the tail. 



Authors have made the matter worse by entirely disregarding 

 the generic definitions, even those made by themselves. Thus 

 Zelinka's Lepidoderma was instituted first to contain Dujardin's 

 C. squammatus, which was described in these terms, " Revetu 

 en dessus de poils courts, elargis en maniere d'ecailles pointues 

 regulierement imbriquees," and which is thus a true Chaetonotus, 

 following Zelinka's own definition. 



The possession or not of scaly armour is surely itself more 

 important than the presence or not of bristles on the scales, 

 but the character has not been used in classification quite 

 rightly, for the scales are after all only the enlarged bases of 

 the hairs, and there is every gradation from a slightly enlarged 

 insertion to large imbricated scales. 



Authors have further confused matters by professing to identify 

 a,s the species of the earlier authors animals which are quite 

 different from their descriptions and figures. This is pernicious, 

 as the practice nullifies the meaning of language, however 

 precisely used. It may be admitted that the descriptions of 

 Muller and Ehrenberg are insufficient to distinguish their species 

 from the numerous similar species now recognised. But the 

 species must either be dropped as " insufficiently described," or, 

 if we profess to recognise them, it must be in animals possessing 

 at least the characters ascribed to them by their discoverers. 



Ehrenberg has many faults, among which I reckon not 

 least the insufficiency of his descriptions. Frequently these 



