during their Orowth and Development. 121 



are always sufficiently great to give rise to the establishment 

 of false species when one has not sufficient materials at com- 

 mand. Epkijjpus argus appears to me, however, to include 

 three species : — the Chinese form, with a small number (20-30) 

 of large spots ; the East-Indian type species, with many spots 

 of moderate size ; and a form from the Sunda Islands with 

 numerous small spots, a pattern which, in young individuals, 

 changes into transverse bands {E. ornatus) . Strictly speaking 

 "we cannot characterize our youngest Ephippus as a " Tholich- 

 thys ; " but nevertheless it has so many points in common 

 with this phase of Chcetodon that we may describe it as being in 

 a "'J'holichthyoid" phase. It somewhat resembles a C/iromis 

 or a Pomacentrum : the body is short, squat, and much com- 

 pressed, the profile of the head nearly vertical, the skin rough 

 and without scales ; the fins are naked ; the pattern consists 

 of dark transverse laands ; the forehead is broad, convex, and 

 protected by two thick, rounded, triangular sJiields, which 

 meet in the median line, but which, posteriorly, embrace 

 between them the apex of a parietal boss ; there is also on 

 each side a temporal boss accompanied by a stout spine, which 

 is the inferior extremity of a triangular suprascapular tubercle; 

 the prolongation of the operculum [i. e. of the prseoperculum 

 and interoperculum) is divided by a notch into two short 

 rounded parts, of which the superior is directed backwards, 

 and the inferior inwards and downwards. 



17. ACANTHUEUS, NaSEUS ; ACEONURUS, KeRIS. 



We now know that Keris and Acro7iurus are respectively 

 only the young forms of Naseus and Acaiithurus. With 

 regard to the development of the Kerides and their transfor- 

 mation into Naseus I may refer to the illustrations which 

 accompany the magnificent work that Dr. Giinther is publish- 

 ing under the title of ' Fische der Siidsee. There are nume- 

 rous analogies between the young individuals of the two 

 genera — the form of the body, which is short, with strongly 

 arched contours, the streaking and partial metallic lustre of 

 the skin, the greater length of the anterior dorsal and anal 

 spines, the different position of the ventrals relatively to the 

 pectorals, &c. My own contributions to the history of the 

 metamorphoses of the genus Acanthurus consist in the indi- 

 cation of the so-called ^'Acronurus^^ form of the two West- 

 Indian species, Acanthurus chirurgus {phlehotomus) and A. 

 cceruleus, and of the still more curious form under which the 

 so-called ^'Acronurus " shows itself in its first phase. In fact 

 I regard as a young example of A. c(sruleus, Schn,, the very 

 marked form of Acronurus represented in pi. v. fig. 4 (of the 



