Classification of the Blenmoid Fishes. 279 



ankylosis of the ptery^'oid and mesopterygoid, and the 

 jMolongation foiwards ot" the cleithra within the isthmus as 

 a pair of slender processes, with the pelvic bones attached at 

 tjieir extremities. The lower fork of tiio post-temporal is 

 shortened and attached to the opisthotic by a ligament, and 

 the coracoids are in contact. There are 72 vertebra? (20 + 52) : 

 the first five are as in Srotula, except tliat only the first rib 

 is expanded ; the anterior six pairs of parapophyses (on 

 vertebrpe 6-11) are strong and broad, much as in Merluccius, 

 the rest are normal. 



Principal genera : Ophidium, Ofophidiuin, Lepophidium, 

 Gnif/pterus, from tropical and temperate seas, some in- 

 habiting deep water. 



Derej odic/ithi/s, Gilbert, from the North Pacific, has the 

 mouth non-procractile, the body naked, and the gill-openings 

 more restricted than the others ; it may not pertain to this 

 family and may prove to be related to the Zoavcidaj. 



Family 3. Fierasferid*. 



DIfEer externally from tlie Brotulida) in that the anal fin 

 extends further forward and the vent is placed at the tliroat, 

 caufhil and pelvic fins aie absent "^j and the mouth is non- 

 protraciih\ The craniu ii shows many striking resemblances 

 to that o\ B.otula, but differs in that the parietals meet above 

 the supraoccipital, the occiiiital crest is weak, and the ex- 

 occijitals do not take part in its formation, and the enlarged 

 opisth'itic reaches the basicccipital, sharing with that bone 

 and the pro-otic in the formation of the auditory bulla f. 

 The lower fork of the post-temporal is reduced to a little 

 knobj otherwise the pectoral arch is as in Geni/pterus. In 

 Fierasfer acus [fide Emery) the vertebrre number 125 to 144, 

 of winch 17 or 18 are pra-caudal ; in F.dentatus there are 26 

 pisecaudal vertebrje; the first rib is more strongly expanded 

 in the former species than in the hitter. 



Seeing that tlie Fierast'erida? had always been placed near 

 the OpI.idiidte, and that Emery's anatondcal researches con- 

 firmed this view as to their systematic position, it is not 



* I at one time thouglit that a reduced homocercal fin was present in 

 some ^iera^feridfe, as iu the IJrotulidse ; but on looking into the matter 

 I find that whenever a caudal tin appears to be present it is due to 

 regeneration after the end of the tail has been broken off, 



t These features were first described by Emery (Faun. ii. Flora d. 

 Golf. V. Neap. ii. 1880), and I am able to confirm" the accuracy of his 

 account of the head-skeleton, after preparing and examining that of 

 F. acus. 



