476 WILLIAM K. GREGORY 



regarded as homologous with the infraclavicles seen in Ganoids 

 and all the resemblances to the Acipenseroidei are doubtless 

 analogical, not homological. 



In the armored forms an elaborate system of tuberculated 

 bony plates protects the head and shoulders (in a manner analo- 

 gous to the plates of Dinichthyids among Arthrodires) , and 

 is supported posteriorly by the coalesced neural arches and by 

 the stout shoulder girdle. The anterior fin ray, in the pectoral 

 and dorsal fins, forms a great bony spine, which is erected and 

 locked (in the dorsals) by an ingenious modification of the 

 underlying neural spines, or (in the pectorals) of the basals. 



The adaptive radiation of structure and habit among the iooo 

 species of Siluroids is extraordinarily great, and may indicate 

 ■a great antiquity for the group; but many seemingly annec- 

 tant forms still exist, and Eigenmann 1 traces all the higher 

 subfamilies back to the American Diplomysteidae, which retain 

 dentigerous maxillaries forming part of the border of the mouth. 2 

 Fossil forms are rare. Rhineastes from the Wasatch or Lower 

 Eocene of North America is probably related to the Pimelodinas, 

 from which, says Eigenmann, 3 "the present North American 

 forms are, not unlikely, lineal descents." The gigantic Leopard 

 Catfish Bagarius yarreli of India and Java is represented in the 

 Pliocene of the Siwalik Hills in Northern India. 



Although some of the genera belonging to the least specialized 

 subfamilies Pimelodinae and Tachisurinse are marine, the ma- 

 jority of Catfishes shun competition with the higher Teleosts by 

 living in muddy instead of clear water, a fact which may have 

 determined the survival of the group. In correlation with this 

 habit: (i) the eyes are often comparatively small; (2) the direction 

 in which food lies is detected by the barbels or even by the skin, 

 both of which in the naked forms are sensitively gustatory as 

 well as tactile in function (Herrick) ; (3) in many genera the 



1 " Revision of the South American Nematognathi or Catfishes," Calif. 

 Acad, of Science, Occasional Papers, Vol. I, 1890. 



2 The single genus Diplomystes should not be confused with the Clupeoid 

 genus Diplomystus . 



s " A Catalogue of the Fresh-Water Fishes of South America," Proc. U' 

 S. Nat. Mus., Vol. XIV, 1891, p. 281. 



