NATURAL SCIENCES OF PHILADELPHIA. 263 



Diameter of eye 2^ 



Dorsal Distance from snout 13 



Length of first spine 2^ 



" second spine 4~ 



" spine over anus 4 



Anal Distance from snout 37 



Height at third ray 4J 



" at middle 4h 



Caudal Length 8" 



Pectoral Length 6 



Monograph of the Tridigitate URANOSCOPOIDS. 

 BY THEODORE GILL. 



In the "Synopsis of the Uranoscopoids" recently published in the Pro- 

 ceedings of the Academy of Natural Sciences, the discovery of two new 

 species of the genus Dactyloscopus was noticed, and a promise was made to 

 describe them at another time. The present memoir is accordingly devoted 

 to the description of the three species of that genus now known and of an 

 additional species representing a fourth very distinct group of the same 

 family. 



The discovery of the last mentioned type has necessitated a revision of the 

 characters formerly given to the subfamilies of the Uranoscopoids, the form 

 of the head presenting quite a decided difference. All of the species formerly 

 known were distinguished by the cuboid form of the head, the superior sur- 

 face of which was nearly flat ; the eyes were also entirely superior and next to 

 the sides, and the cleft of the mouth was almost vertical. Such were the 

 characters common to all the species then known. But when engaged on the 

 present monograph, two fishes were found which were readily ascertained to 

 be most nearly allied to the Dactyloscopi, but which were distinguished from 

 them by the remarkable shape of the head. The bones are so modified that 

 instead of producing the normal cuboid form, they cause an elongated conical 

 one, the height and width gradually decreasing toward the snout. The interor- 

 bital area is also very slight on account of the excessive narrowness of the fron- 

 tal bones. Yet, notwithstanding such remarkable differences of form, none can 

 remain doubtful concerning the affinity of the species to the Dactyloscopi. 

 The general form of the body, the squamation, the character of the lateral 

 line, the fringed opercula and lips, the membranous extension of the suboper- 

 culum and interoperculum, the character of the fins and all other essential 

 characters are reproduced in the two forms. The arguments that were 

 formerly adduced in favor of the pertinence of the Dactyloscopi to the same 

 family as the Leptoscopi are equally applicable to the newly discovered form. 

 The relative position of the dorsal and anal fins is even similar to that of the 

 Leptoscopi and different from that characteristic of the Dactyloscopi. Three 

 peculiar modifications of the elongated Uranoscopoid form are now known, 

 and the propriety of referring them to a family distinct from the compara- 

 tively short Uranoscopoids, with more or less mailed heads, is correspondingly 

 increased. But, for the present, all are still retained under the same family. 



The following diagnoses of the subfamilies and their genera are given so as 

 to distinguish them among the other groups. 



Subfamily DACTYLOSCOPINiE Gill. 



Dactyloscopinae Gill, Proceedings of the Academy of Natural Sciences of Phila- 

 delphia, 1859, p. 133. 

 " Gill, op. cit., 1861, p. 116. 



The body is moderately elongated. 



1861.J 



