EASTMAN: FISHES FROM UPPER EOCENE OF MONTE BOLCA. 317 



from Trygon relates to the more distal position of the caudal spine, which arises 

 at a distance behind the pelvic arch about equal to the maximum width of the disc. 

 Jaekel in his memoir above referred to rightly holds that this does not constitute 

 a valid differential character, for examination of a number of specimens shows that 

 the relative position of the caudal sting is about the same in all. 



Two examples of this Eocene ray are contained in the Bayet Collection of the 

 Carnegie Museum, one small and preserved in counterpart, the other a beautiful 

 specimen, larger than the type, and showing many structural details in great per- 

 fection. The small, evidently immature individual, is cataloged as No. 4521 

 +4521a, and the larger adult specimen bears the Catalog No. 4304. An illus- 

 tration of the latter is shown in Plate XLIII. In this the various cartilages of the 

 head, especially those about the mouth and scapular arch, and of those forming the 

 axes of the pectoral fins, are clearly visible, and a number of small teeth, of the 

 characteristic Trygon-type, are also seen to be attached to the palato-quadrate 

 cartilage. An impression is preserved of the body-walls of the trunk on either side 

 of the vertebral columns as far as the point of insertion of the caudal spine. The 

 latter displaj^s a median dorsal groove, bears a double series of strong posterior 

 denticles, and has a total length of about 9 cm. 



Subclass TELEOSTEI. 

 Order Solenichthyes. 



This ordinal term, first proposed by Dr. C. T. Regan for the Centriscoids only, 

 and afterwards (in 1909) extended so far as to include the Aulostomids and Lopho- 

 branchs, marks the present-day conception of the evolutionary history of that 

 group of physoclistous fishes with abdominal ventrals, of which the sticklebacks 

 form the well-known ground- type. 



The Gasterosteids and their immediate allies were first united by Cope in 1887 

 under the comprehensive designation of Hemibranchii, and the limits of this sub- 

 order were enlarged by A. Smith Woodward in 1901 to include also the Lopho- 

 branchs of Cuvier, pursuant to the view of Knerand Steindachner (1863) that these 

 are only extremely specialized sticklebacks with tufted gills. For this same associ- 

 ation of Lophobranchs and Hemibranchs the new name of Thoracostei was proposed 

 by Swinnerton in 1902, and that of Pthinobranchii was suggested for it by 0. P. 

 Hay at about the same time. Boulenger, in 1904, having attempted to show that 

 the Lamprididse are related to the Hemibranchs, defined the enlarged suborder 

 which he called " Catosteomi " as consisting of the forms just named, together 



