BY J. DOUGLAS OGILBY. 



555 



species could not be generically separated from A. urvillii, reject- 

 , ing in fact Castelnau's genus while allowing his species. 



At the date of this last paper there were therefore five species, 

 two Australian and three American, united together under the 

 common n&meAphritis, namely : — A. urvillii, C.V.; A. undulatus, 

 Jen.; A. porosus, Jen.; A. gobio, Gnth.; and A. bassii {Ca.st.), Ogil. 



In my paper on " Paeiidaphritis bassii, Casteln.," quoted above, 

 the following paragraph will be found on p. 68 : — " In the 

 'Zoological Record' for 1872 Dr. Giinther remarks: ' Aphritis 

 dnvierili. To this species appears to belong Pseudaphritis bassii 



' As I am unable to find any description of the 



former species, I am not in a position to verify or contravene 

 this supposition.'" The same difficulty still i-emains, but it has 

 occurred to me that ' durnerili ' may be a misprint or lapstos 

 calami for ' durvilUi,' by which name Giinther erroneously alludes 

 to Cuvier and Valenciennes' species elsewhere (A.N.H. 1 c.) 



A careful study of the characters of these fishes shows that 

 their association in a single genus is unwarranted, and that not 

 only are Jenyns' two species generically separable from that of 

 Cuvier and Valenciennes, but that Giinther's gobio must be 

 removed from both ; necessitating therefore the division of the 

 heterogeneous Aphritis of the latter author into no less than three 

 distinct genera. 



It now remains only to determine by what names these genera 

 with their accompanying species should be known. 



Tlie first author to detect the generic diiferences between the 

 fish described by Cuvier and Valenciennes and those named by 

 Jenyns was Gill, who, so long ago as 1861, appended to his 

 " Synopsis of the Notothenioids " a note in which he remarks : — 

 " Two species (Aphritis undulatus and A. porosus), referred by 

 Jenyns to the genus Ajjhritis, not only are generically distinct, 

 but belong to a different family, and form a genus nearly related 

 to El' ginus, which will be at an early date described as Eleginops. " * 



* Proc. Acad. Nat. Sc. Philad. 1861, p. 522. 



