1904.] OSTEOLOGY OF CLUPEOID FISHES. 491 



of the forms studied there ai'e two hypohyals on each side, and 

 the lower of the two is lai'ger than the upper. The first pharyngo- 

 branchial is ossified in Chlrocentras, but remains cartilaginous in 

 the other geneiu ; the ossified ligament known as the spicular 

 bone is pi'esent in all except Chlrocentras. In Encjraidis and 

 Coilia the second hypobranchials are fused with the sides of the 

 second basibranchial, and in Chanos the third hypobranchials are 

 fused with the sides of the third basibranchial, but in none of the 

 three cases ai-e the sutures obliterated. 



Comments ok the 8kull of the Clupeoib Fishes. 



Of the eleven genera the skulls of which have been described 

 in the pi'evious pages the gi'eatest interest centres arovind Chanos 

 and Chirocentrus. As i-egards the others, the craniological 

 charactei's ai'e such as would justify the placing of Engraulis and 

 Coilia in one family, the Engraulidae, arid Clupea, Fellona, 

 Pellonula, Pristigaster. Hijperlophas, Chatoi'ssus, and Dnssamieria 

 in another, the Olupeida?. 



Both Engraulis and Coilia have a lai-ge gape, a backwardly 

 thi-ust quadrate Ijone, and a large and pi-ominent mesethmoid. 

 The skull of Coilia difibrs fi-om that of Engraulis in several 

 i-espects, but these are all of such a nature as might be explained 

 by high specialisation ; such chaiuctei-s, foi- instance, are the 

 reduction of the temporal foramen, the obliteration of the pre- 

 epiotic fossa, the absence of the auditory fenestra, the absence of 

 paired posterioi- wings of the parasphenoid and the closure of the 

 eye-muscle Oanal, the absence of the opisthotic limb of the post- 

 tempoi-al, and the loss of the posterior of the Uvo heads by which 

 the hyomandibulai- ai'ticulates with the ci'anium. The j^aiudoxical 

 extension of the maxilla, behind the mandibulai- articulation is 

 foreshadowed in some species of Engraalis, e. g. E. mystax and 

 E. setirostris. 



The skull of Chatoessas conforms with the Clupeoid type, in 

 spite of cei'tain aberi-mt featiu'es, such as the absence of projecting 

 wings fi-oin the back of the pai-a sphenoid, the intimate articulation 

 of the epiotic limb of the post-tempoi'al with the cranium, the 

 small size of the mouth, the bounding of the month above by the 

 pi-emaxilla alone, the absence of teeth, and the loss of one of the 

 .suiiiiaxilke. The characters of the skull do not warrant the 

 sepai-ation of Chatoessus from the Clupeida? to constitute a distinct 

 family, although possibly on other grounds the action of Gill 

 (Smithsonian Miscell. Coll. No. 247, 1872, p. 17) and Jordan and 

 Gilbert (Bull. U.S. Nat. Mus. No. 16, 1882, pp. 262-274) may 

 pi'ove to be justifiable. 



Chirocentrus agrees so closely in the structul'e of its skuU with 

 the Clupeidse, that appeal must be made to other organs of the 

 body for evidence to support the views of those who would make 

 of it a distinct family, the Chirocentridaj (e. g., Valenciennes, 

 Hist. Nat. Poiss. xix. 1846, pp. 150-168 ; Kner, Reise der 



