l60 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 52 



A few forms still living exemplify the manner in which the extreme 

 modification of the last has been attained ; these forms, it is true, 

 are not in the direct line of descent, but they are not very far off. 

 The common Dog-fish of the New England coast (Squalus acan- 

 thias) has a slender tail, but there is a regular gradation from the 

 preanal region, or trunk, into the postanal, or tail, and the pectorals 

 have the slender bases characteristic of the Sharks generally. The 

 Guitar-fishes (Rhinobatidce) still have the regular gradation of the 

 trunk into the tail, but the pectorals have a broad basis of union with 

 the body and head, and a narrow disk is thus formed. In the ordi- 

 nary rays {Raiidce) the tail has become disproportionately slender 

 and the disk wider and more sharply differentiated ; in the Sting- 

 rays (Dasybatidce) the tail has almost entirely lost its muscular de- 

 velopment, but the disk is much like that of an ordinary ray. The 

 tail of the Sting ray is essentially like that of the Devil-fish, but in the 

 Devil-fish the disk has become extended sideways into acutely angu- 

 lated and wing-like fins. The homologies of the respective parts are 

 thus evident. In the course of evolution, more and more resort has 

 been had to the pectoral fins for progression and the tail correspond- 

 ingly disused; the culmination has been reached in the Devil-fishes, 

 which progress by wing-like flapping of their pectorals and the tail 

 is carried inert behind. 



The tranformation of shark-like forms into the ray-like type must 

 have commenced early in Mesozoic times, for well-developed repre- 

 sentatives of the Dasybatids and Myliobatids were living in the Cre- 

 taceous epoch and were abundant in the Eocene. It has been be- 

 lieved that no fossil remains of Devil-fishes have been found, or 

 rather identified. If this had been a fact, it might have been partly 

 explained by the pelagic habitat of the species and partly by the 

 reduction of teeth and spines, the parts most likely to be pre- 

 served. There is, indeed, one record of an extinct form which, 

 however, only takes us one stage back in the geological series. 

 The record is of a supracaudal tubercle from the "phosphate beds" 

 of South Carolina, which are supposed to be of post-Pliocene age ; 

 the tubercle has been considered by Joseph Leidy to represent an 

 extinct species closely related to the living Devil-fish of the same 

 State and has received from him the name Ccratoptera unios; it was 

 described and figured in 1877 in the Journal of the Academy of Nat- 

 ural Sciences of Philadelphia (2nd ser., viii, 248-9, pi. 34, figs, i, 2). 



The individual development of the fishes is to a large extent par- 

 allel with the evolution of the type from the shark-like form to the 

 ray-like one. 



The Devil-fishes form a familv of rav-like Selachians to which the 



