300 Mr. G. A. Bouleugei- on the 



number; and tliere is no example of tlielr having again 

 increased in number except concurrently with a multiplication 

 of their supports. 



The Zeidffi or Cyttidse, comprising the genersi Zeus, Zenion, 

 and Cj/ttus, to which I would add the little-known Grammico~ 

 lepi's, described by Poey and by Shufeldt * from a single 

 specimen, have been placed by some authors near the Chasto- 

 dontidge, by others with the Scombriformes or with the 

 Beryciformes. Tliey form a perfectly natural family, which 

 may be defined as follows : — 



Acanthopterygians without subocular shelf or suborbital 

 stay for the preeoperculum, with double basis cranii t, well 

 developed entopterygoid, and strongly protractile prsemax- 

 illaries. Two nostrils on each side. Grill-membranes free 

 from isthmus ; 7 or 8 branchiostegal rays ; gills 3^ ; pseudo- 

 branchiai well developed. Lower pharyngeal bones separated. 

 Vertebrse 30 to 46, the anterior with sessile ribs, the posterior 

 prtecaudals "with long neural spines bent forwards and with 

 transverse ])rocesses directed downwards, forming haamul 

 arches and bearing the ribs at their extremity : epipleurals 

 much reduced or absent j hypural large, without the basal 

 spine present in most Perciformes and all Scombriformes and 

 Percecoces^ bearing fewer than 20 rays. Dorsal and anal fins 

 elongate, the former with a distinct spinous portion, the 

 latter with 1 to 4 spines detached from the soft portion. 

 Pectoral fin supported by four pterygials, of which three are 

 in contact with the perforated scapular bone ; posttemporal 

 forked and solidly attached to the skull. Ventral fin with 1 

 spine and 6 to 8 soft rays. 



The family which I propose to name Amphistiidse, with 

 the single genus Amphistium, Ag. (^Macrostoma, Ag.), from 

 the Upper Eocene, agrees with the Zeidte in all characters 

 that can be ascertained on the preserved remains, except that 

 the vertebrae are fewer (24), the spines of the vertical fins 

 are reduced to a few adnate to and continuous with the series 

 of soft rays, and the scales are more normal and imbricate. 



Now, as already pointed out by Agassiz, these Amphistiids, 

 provided they be possessed of the last half-gill absent in the 

 Zeidge, and this is a character which unfortunately cannot be 

 ascertained on the fossils, appear to realize in every respect 

 the prototype of the Pleuronectidae before they had assumed 



* Journ. of Morphol. ii. 1888, p. 271. 



+ I am unable to confirm the statement made by Starks (Proc. U.S. 

 Nat. Mus. yxj. 1898, p. 470), that the basispheuoid is absent and that the 

 parietals unite in front of the supraoccipital. 



