II0 PAPERS FROM TORTUGAS LABORATORY vol. xxxiv 



gated dashes below and above the nostrils, and a crozi'er-like stripe with the eye 

 in its crook and the straight stem running obliquely down and back on the 

 cheek; vertical lines of the same color on the preopercle and opercle, two of them 

 extending on the breast before the pectoral; body with many faint vertical lines 

 of light blue; and soft dorsal and caudal barred with them. 



Hypoplectnts vitulinus (Poey), of which I have seen no specimens, seems to 

 differ from typical H. puella only in the absence of blue markings on the head 

 and breast. 



Fish with the elaborate pattern of H. puella may be seen on the reef at Tor- 

 tugas in different shades. These vary from pale amber to dark red-brown, and 

 each is itself variable. Broad bands, most distinct in the darker individuals, may 

 appear in any individual in an instant, and vanish as quickly. One band in this 

 phase passes through the eye and down across the cheek, being bordered an- 

 teriorly by a subocular blue streak; a second band at the nape, extending on the 

 pectoral base, broad above and narrow below; a very broad band extending from 

 4th spine to 7th soft ray, sometimes interrupted at mid-level of side; two other 

 bands, one before end of dorsal base, and the other at base of caudal; the upper 

 part of the last darker and frequently persisting as a distinct saddle when all 

 other traces of bands are lost; snout, interorbital occipital region, and web of 

 dorsal from 3d spine backward sometimes nearly white, interrupting the con- 

 tinuity of the ocular, humeral, and first subdorsal dark bars. 



Referring to group two of the names in synonymy (those with a single inter- 

 rogation point), a slight difference in the color markings on the side of the snout 

 is all that obviously separates H. guttavarius from H. gum mi gutta. Faint band- 

 ing of the dorsal, as opposed to none, distinguishes H. pinnavarius from H. gutta- 

 varius. In fishes so variable in coloration as the vacas, the difference last men- 

 tioned seems quite unsubstantial. I anticipate that these three will be found to 

 be one, which for the present may be called H. gummigutta. 



Hypoplectrus chlorurus, H. nigricans, H. accensus, and H. affinis, of group 

 three (designated with two interrogation points), agree in having no blue mark- 

 ings of the puella type. Their differences are chiefly in shade. Hypoplectrus 

 aberrans differs from them only as the banded phases of H. puella or H. gemma 

 differ from the nonbanded phases. They seem surely one, which until further 

 reduction may be made should stand as H. chlorurus. 



This discussion is intended to supplement that on the genus as a whole. With 

 what has gone before, it indicates an irreducible minimum of three species, or 

 four if H. puella proves to be distinct from H. unicolor. Hypoplectrus gummi- 

 gutta seems a very doubtful species, likely to be merged with H. puella, and H. 

 chlorurus more probably good, though in danger of being merged with H. uni- 

 color, even if it proves distinct from H. puella. 



West Indies to Florida. W. H. L. 



Hypoplectrus gemma Goode and Bean 



This exquisitely colored species was hitherto known from only 1 specimen, 

 recorded as from Garden Key. It has now been observed at various places within 



