1896.] NRW YORK ACADEMY OF SCIENCES. 63 



from those of other species. The number of its primary mar- 

 ginal denticles is greater. In this respect, however, an accu- 

 rate comparison cannot be made, since all of the specimens of 

 ■which have been described in other species of Edestus are pos- 

 sibl^^ imperfect — i. e., lacking in proximal or terminal elements. 

 Of those described, the spine of U. daoisii, as figured by Mr. 

 Henry Woodward, is probably the most perfect, having fourteen 

 primary marginal denticles. The greatest number possessed by 

 the other species, E. minor, E. heinriehsii, and E. giganteus 

 does not, as far as the writer is aware, exceed seven. In the 

 ■case of E. vorax, no greater number than four appears to be 

 known, judging from the figure of Prof Newberry * which was 

 doubtless taken from the best specimen in the collection of Dr. 

 Leidy. In the latter case it must, however, be noted that the 

 spine bears every evidence of liaving been not widely unlike that 

 of E. heinrichsii. Further differences of E. lecuntei include : 

 the tapering character of the spine, the antero-posterior differ- 

 ences in the primary denticles, the possible absence of sec- 

 ondary denticles, and finally' the obliteration of the segmental 

 elements of the spine-base. In all of these regards this spine is 

 widely unlike all other members of its genus. 



An interesting comparison of the characters of five species of 

 Edestus may be made by reference to the accompanying figures, 

 PI. V, Figs. 2-6. For in these the spines have been reduced to the 

 same size, represented in relatively the same aspect, and ar- 

 ranged with a view of suggesting the lines of their relationship. 

 The writer has thus placed the form, E. heinrichsii, at the top, as 

 presenting the simplest type of Edestus, and has ranked the re- 

 maining forms in the following order: E. giganteus, E. minor, 

 E. davisii, E. lecontei, the last as the most highly specialized. f 



A review of the structural features of these forms has been at- 

 tempted in the table on the following page. 



The morphological relations of the spines of Edestus. 



In the foregoing pages the writer has assumed that Edestus 

 is to be regarded as an elasmobranchian spine. As far as the 

 jibove comparison is concerned, however, this determination need 

 not be strictly adhered to, for from the morphology of verte- 

 brate dermal structures the arrangement of the forms represented 

 by the five species might still hold good — i. e., the position E. 

 heinrichsii as the most generalized form, and of E. lecontei as 



* J. S. Newberry. "On the Structure and Relations of Edestus, with a Description of 

 a Gigantic New Species." Annals ^f. Y. Acad, of Sci. Vol. IV, No. 4, 1888. Pp. 10, pis. 3. 



ti?. vorax has appeared too fragmentary to be admitted in this comparison: its 

 structures would probably place it between the types E. giganteus and E. minor. 



