378 Annals of the Carnegie Museum. 



In South America and Africa related forms are known from supposed 

 early Tertiary fresh-water strata, and at the present day the genus 

 Diplomystus itself survives in the rivers of Chili and New South Wales. 



The type-species D. dentatus Cope, and the scarcely separable 

 D. analis and D. pectorosns, have the dorsal ridge-scutes broader than 

 long, and finely denticulated at their posterior border. In species 

 having a slenderer and more elongate form of body, like the so-called 

 D. humilis and D. attus of Leidy, the dorsal scutes are narrow and 

 simple, with one pointed projection. This distinction is considered 

 by Jordan to be of generic, or at least subgeneric importance, and the 

 two last-named species are transferred from Diplomystus proper to 

 the closely related genus or subgenus Knightia of Jordan. The type 

 species of " Knightia " is K. eoccena Jordan, = Diplomystus pusillus 

 Cope, = Clupea humilis Leidy. 



A second closely related genus or subgenus, named Ellipes, has 

 recently been established by Jordan upon the evidence of Brazilian 

 Clupeoids from late Cretaceous and supposed early Tertiary horizons. 

 The type species is named E. branneri, a small fish presenting much 

 the same configuration as the type of Diplomystus; and accompanying 

 it in the same horizon is the so-called E. riacensis, which is slenderer 

 and more elongate, and hence approaches D. eoccena (= " Knightia ") 

 in general aspect. In the opinion of Dr. Jordan, D. longicostatus Cope, 

 from the Upper Cretaceous of Brazil, should also be classed under 

 Ellipes, but from this view the present writer dissents. When one 

 has to deal with material that in the main is of fragmentary nature, 

 and when much inconstancy is to be observed among the characters 

 recognized as having specific value, it seems advisable to hold to a 

 broad conception of genera and families, and to avoid drawing dis- 

 tinctions of such narrow margin as to obscure natural relationships 



Explanation of Plates. 



PI. XXIII, Fig. i. Diplomystus goodi, sp. nov. X i/l. 



PI. XXIII, Fig. 2. Diplomystus goodi, sp. nov. X 5/4- 



PI. XXIV, Fig. i. Enchodus, sp. ind. X 4/5- 



PL XXIV, Fig. 2. Diplomystus goodi, sp. nov. X i/i. 



