﻿284 PROCEEDINGS OF THE NATIONAL MUSEUM vor.. lOft 



having examined the above-mentioned specimens and compared them 

 with the original description of H. sardina, I conclude that M.C.Z. 

 No. 17868 is the only specimen that can be accepted as the holotype; 

 the rest are paratypes. It agrees with the original description in 

 every respect, except that the ventral-scute count is 27 instead of 25. 

 A count of 25 ventral scutes is very rare in this species (see table 1) 

 and there is the possibility that a typographical error was made in 

 the original publication or that Poey made a mistake in his count. 

 The specimen 125 mm. in length, U.S.N.M. No. 4794, also agrees 

 with the original description in almost every respect but has a scute 

 count of 28. In my opinion the decisive factor in favor of M.C.Z. 

 No. 17868 as being the holotype of H. sardina is that Poey, in his 

 original description, states that "la ligne du ventre est plus prononc^e 

 que celle du dos," which is true of this specimen. The U. S. N. M. 

 specimen, in contrast, has the dorsal outline more pronounced than 

 the ventral (an obvious artifact in preservation). In this species the 

 dorsal and ventral outlines have approximately the same curvature. 



I follow Storey (1938, pp. 28, 41) in placing, with reservations, 

 Harengnla jaguana Poey (1865, p. 189) from Cuba in the synonymy 

 of H. humeralis. This species was originally described from Bahia 

 de Jagua (Cienfuegos Bay), southern Cuba, and is stiU known only 

 from the original description; the types, apparently, have been lost. 

 The status of this little-known species might be clarified by the 

 study of material from the type locality. 



The material (U.S.N.M. No. 38469) recorded from Nassau, 

 Bahamas, by Lee (1889, p. 672) as Clupea sp. is H. humeralis. The 

 material recorded by Breder (1927, pp. 12, 13) as Sardinella 

 macro phthalmus I observed to be a mixture of two species: No. 14, 

 a specimen 126 mm. in length from Koyal Island, Bahamas; No. 15, 

 a specimen 102 mm. in length from Port Francis, Isle of Pines; and two 

 specimens 122 and 139 mm. in length in No. 16, from Glover Reef, 

 are H. humeralis. The remaining seven specimens in No. 16 are H. 

 clupeola. I have not seen No. 17 (1 specimen), from Siguanea Bay, 

 Isle of Pines. 



Together with H. callolepis, H. humeralis is widely separated from 

 all the other American members of the genus by a number of char- 

 acters, as shown in item la of the key, already pointed out by Storey 

 (1938, p. 23), but she did not recognize H. callolepis as a valid species. 



A very distinctive character not hitherto used and common to H. 

 humeralis and H. callolepis is the presence of a row of pointed teeth 

 forming a cutting edge on the inner side of the palatines. This 

 character, and those already referred to, might justify a new subgenus, 

 as suggested by Storey (1938, p. 23), but no subgeneric distinction 

 should be attempted until a thorough study of the Indo-Pacific species 

 is made. 



