FISHES OF FAMILY BLENNIIDAE 3 



Gosline added another character of which he was not quite so certain: 

 anteriormost anal pterygiophores are attached to the first haemal 

 spine. 



All of the above characters, except the presence of a basisphenoid 

 (absent as noted above) and the absence of a subocular shelf (see 

 discussion below on circumorbital bones) are descriptive of the 

 blenniids I have examined. 



Gosline closed his discussion of the Blennioidae (Blenniicae in 

 discussion below) by stating that "the relationship of the tropical 

 blennies to any other [perciform] fish group is by no means clear." 

 My study affords no better conclusion. 



Hubbs (1952) included three families of living fishes in the Blen- 

 niicae: Tripterygiidae, Clinidae, and Blenniidae. I (1955) questioned 

 Hubbs' basis for separating the Clinidae and Blenniidae, redefined 

 the two famiUes, and transferred the subfamily Chaenopsinae from 

 the Blenniidae, in which Hubbs (1953) had placed it, to the Clinidae. 

 Not all my information was accurate, but I still find my general con- 

 clusions to have been warranted. Whether the Chaenopsinae deserve 

 to be recognized as a family separate from the Clinidae, as Stephens 

 (1963) proposed, will not be treated here. 



I (1966) discussed problems of the interfamilial classification of the 

 famihes in the Blenniicae and gave characters further separating the 

 Blenniidae from the other included families. My statements in that 

 paper to the effect that all the members of the superfamily might be 

 combined into a single family now appear unwarranted. 



Norman (1943) gave the last and, actually, only comprehensive 

 infrafamilial classification of the Blenniidae, including a synopsis of 

 the genera and principal species. He recognized three subfamilies: 

 Ophioblenniinae, Blenniinae, and Salariinae. The Ophioblenniinae 

 have been shown by various workers to be larval stages of various 

 salariin genera, and one genus {Somersia Beebe and Tee-Van) included 

 in the Ophioblenniinae by Norman has been shown by me (1959a) to be 

 the juvenile of a species of Clinidae — Labrisomus nuchipinnis (Quoy 

 and Gaimard). Norman's Blenniinae includes my Nemophidinae and 

 part of my Blenniinae. Norman's Salariinae I have relegated to the 

 rank of a tribe, which forms the remainder of my Blenniinae. The 

 number of genera now available in the Blenniidae greatly exceeds the 

 number available during Norman's day. A synopsis is therefore given 

 below (without indication of generic synonymies). I have not included 

 a list of the nominal species in each genus as such will not be possible 

 until the genera and species are better known. 



The results of my study are presented in the following order: 

 first, a classification of the family Blenniidae to the tribe level, 

 including a list under each tribe or subfamily of the included nominal 



