222 PROCEEDINGS OP THE ACADEMY OF 



creased in the abdominal and caudal portions." This generalization as ap- 

 plicable to the representatives of Acanthopterygian families generally, can be 

 considered in connection with the predominance of the true Malacoptery- 

 gian fishes in northern waters fishes in which the increase in the number of 

 vertebra? is a normal feature. 



In the following list, no revised deceptions of the species are offered, as all 

 have been since admitted. The object is chiefly to sustain the generic nomen- 

 clature formerly proposed. From this must be excepted the genera Oxychei- 

 linus and Crassilabrits Swainson, which were based on false characters, the 

 figures of the types having been defective. It is possible that those types are 

 generically distinct, but their true characters require to be yet given. The 

 name Malacocentrus must be also suppressed, Dr. Bleeker having slightly pre- 

 viously given to the same genus the name Novaculichthys. The distribution of 

 the genera among subfamilies is admitted with much hesitation. 



Subfamily CHCEROPINJE (Gthr.) 



Genus HARPE (Lacepede) Gill. 



This genus was originally founded by Lace"pede for the reception of the 

 species since generally known as the Cossyphus bodianus C. V. ; it was charac- 

 terized by the dentition, allusion being made to the presence of the enlarged 

 teeth at the front of the jaws as well as behind, and the intervening small 

 teeth erroneously described as compressed and triangular, and by the falci- 

 form production of the dorsal, anal and ventral fins, as well as the extension of 

 the outer rays of the caudal. To it was also erroneously attributed a com- 

 pressed and triangular barbel on each side near the commissure of the lips, a 

 mistake probably due to some defect in the drawing which afforded Lacepede 

 his knowledge of the genus. It was only known through a design of the na- 

 turalist Plumier. As the genus and species are recognizable from the descrip- 

 tion of Lacepede, the name should have been retained for that group of which 

 his species is the type, even if Cuviers subsequent name Cossyphus had not 

 been pre-occupied. 



The genus Cossyphus as finally limited by Bleeker and Giinther is scarcely 

 natural as such, five according to the author's views being embraced in it. Giin- 

 ther, in his revised list, has admitted seventeen species, besides two as doubt- 

 ful ; these may be distributed among the genera, indicated in the analytical 

 table given below. 



The genus Harpe contains five known species, H. rufus (= Cossyphus bodi- 

 anus C. V.); H. pulchellus (C. p. Poey) ; E. eclancheri (G. ex Vol.) and the two 

 species of Lower California. 



I. Posterior canine tooth developed. Scales 1. 1. 30 34 



(-3G). 

 a. Dorsal and anal fins never produced into falciform 

 lobes (old world). 



1. Limbs of preoperculum scaly ; head oblong, 



snout produced in front Lepidaplois.* 



2. Limbs of preoperculum scaly ; head high, snout 



decurved from forehead, and rather ab- 

 breviated Euhypsocara.f 



3. Limbs of preoperculum naked ; snout convex.... Gymnopropoma.J 

 (3. Dorsal and anal fins produced into very long falci- 

 form lobes in adult, (new world) Harpe. 



II. Posterior canine teeth obsolete. Scales of lateral line 



about 39 Achoerodus.$ 



* Type. Cossyphus axillaris C. Y. f Cossyphus anthioides Gthr, 



I Cossyphus I'ihmulatus C. V. Cossyphus Gouldii Rich. 



[Sept. 



