VERTEBRATES. 121 



a single row of large, flattened or conical tubercles ; lateral 

 surfaces set with numerous small tubercles, arranged in longi- 

 tudinal and sometimes transverse rows; posterior margin 

 rounded or flattened, without hooks, sometimes carrying one 

 or more rows of relatively large tubercles. 



These spines, with our present imperfect knowledge of them, seem to present 

 some anomalous characters of special interest, as they separate them somewhat 

 widely from most of the defensive spines of living placoid fishes, the study of 

 which has thrown so much light on those found fossil. Of these exceptional 

 characters, the most remarkable is their reversed curve. Nearly all of the 

 dorsal defensive fin-rays hitherto known, of living or fossil sharks, are straight 

 or more or less curved backward, the point being turned toward the caudal 

 extremity. They are also planted obliquely in the integuments in such a man- 

 ner that the point is thrown backward even where they are straight, the poste- 

 rior margin of the exposed and ornamented portion being shorter than the 

 anterior. Such is the position of the spines of Spinax, Cestracion, etc., of recent 

 fishes, and was the position of those of Hyhodus, Ctenacanthus, and most other 

 extinct genera, as is proved by the line of contact with the dorsal surface, 

 sloping upward from the convex to the concave border. This is also generally 

 indicated by the curvature, the position of the medullary cavity (nearer the 

 posterior margin), the greater degree of ornamentation along the anterior, the 

 presence of hooks upon the usually sulcated posterior edge or face, etc. In 

 Drepanacanihus, on the contrary, we find the line of contact with the dorsal 

 surface generally plainly marked; sloping upward at an angle of 45° from the 

 concave to the convex margin; the medullary cavity nearest the convex edge, 

 and opening first there; the concave margin most highly ornamented, etc., etc. 

 All. of which characters seem to indicate that while the base was inserted in its 

 normal condition, projecting forward and downwards into the integuments of 

 the back, the exposed portion was curved forward instead of backward. It is 

 true that if the position of the base was reversed, projecting downward and 

 backward instead of forward, the spine being thrown entirely in advance of the 

 base, we should then have, with the normal curvature of the exposed portion, 

 the line of demarcation between the ornamented summit and the plain base, as 

 we now find it, slanting upward from the concave to the convex border. It 

 may not, therefore, be now possible to determine with absolute certainty in 

 which of the two ways these spines were worn. The place and form of the 

 medullary cavity, however, and the ornamentation, seem to indicate, as before 

 stated, that while the base held its normal position the summit was curved for- 

 ward. This structure would require no change in the muscular attachments of 

 the base, by which the spine was elevated and depressed, while any other sup- 



—16 



