460 Mr. C. T. Regan on the 



ray articulated, and there can be little doubt that the foramen 

 between scapula and coracoid is a specialization which has 

 independently arisen within the Gadoid group, and does not 

 indicate affinity with other fislies in which this specialization 

 has also arisen, concurrently with the forward shitting of the 

 ventral fins. 



Since, then, the position of the scapular foramen cannot be 

 used as an absolute character for separating the Gadoids from 

 the Zoarcidas, Brotulida^, &c., other distinguishing features 

 must be found if these families are not to be associated in the 

 same group ; and after examining all the skeletons available, 

 and in several cases making dissections, it appears to me that 

 the following generalizations hold good : — 



In the Gadoids (Gadidse and Macrurid^e) the ventral fins 

 consist of 1-12 soft rays and are below or in front of the 

 pectorals, whilst the pelvic bones are posterior to the clavicular 

 symphysis, to which they are loosely attached by a liga- 

 mentous connexion ; the first two vertebrte have no epi- 

 pleurals, the first epipleural being attached to the first rib. 



In the Blennioids (Blenniidte, Zoarcidje, Brotulidaj, Ophi- 

 diidaj, &c.) the ventrals, when present, consist of less than 5 

 soft rays, sometimes with the addition of a s[)inous ray, and 

 are jugular, the pelvic bones being directly and firmly 

 attached to the clavicular symphysis; whilst the first two 

 vertebrae bear sessile epipleurals. 



It is evident that the Blennioid fishes are modified Acantho- 

 pterygii, but that the Gadoids have originated from some less 

 specialized stock, and that the absence of non-articulated fin- 

 rays, tlie large number of rays in the ventrals, and the lack 

 of direct attachment of the pelvic bones to the clavicles, taken 

 together, must be regarded as primitive features. From their 

 anatomy and appearance I am inclined to think that the 

 Gadoids are not related to the Percesoces, but are derived 

 from some Haplomous stock from which the BerycidjB have 

 also descended, and of which the Stephanoberycida3 may well 

 be the living representatives. They may be distinguished 

 from the Percesoces by the extreme development of the 

 opisthotic, which forms a large part of the lateral wall of the 

 brain-case and extends down to the basioccipital, thus sepa- 

 rating the pro-otic from the exoccipitals. In most Teleostei 

 the exoccipital extends forward below the opisthotic and 

 meets the pro-otic *. 



* Exceptions are the Fierasferidas aud Gobiidae, in which the opisth- 

 otic has the same relation as in the Gadoids (see Emery, ' Fauna und 

 Flora des Golfes von Neapel,' Fierusfer (lySO) ; but in other characters 

 these three groups are widely different. 



