of Cllmaxoclus and Janassa. 327 



even, though here and there slight undulations are perceptible. 

 From this comparatively smooth state there is every degree of 

 undulation to the most rugged. In fact, the smoothness is 

 very much owing to wear ; and in such specimens this portion 

 of the crown is generally much reduced in thickness. The 

 form of the grinding division of the crown also varies con- 

 siderably. We have said that the sides arch outwards ; they 

 are, however, not unfrequently quite straight ; and when this 

 is the case, and the anterior ridge is free from undulations, the 

 area assumes the form of an equilateral triangle, with one of 

 the angles directed backwards ; in two or three specimens the 

 area is even wider than long, with the lateral angles more 

 acute than usual. In such individuals the scoop-like cutting- 

 margin occupies half the crown. The root is a wide plate as 

 broad as the tooth, and tapers slightly backwards ; behind, it 

 is rounded, convex above and concave below, and projects 

 backwards on a lower plane, the crown being elevated above 

 its upper surface. 



The second primary or lateral oblique teeth are very in- 

 equilateral, one side being concave, the other convex ; they 

 have only three ridges, with the grooves very wide ; the scoop- 

 like cutting-margin is deep, oblique, and projects laterally on 

 the concave side. 



The largest Petalodontoid or secondary teeth are nearly f inch 

 wide and f inch long ; they are inequilateral and oblique, 

 with one side concave, the other convex ; they are depressed, 

 and the crown is somewhat longer than the root ; the former 

 consists principally of a wide, sharp, hollow, scoop-like cutting- 

 margin, which in fresh specimens is obscurely denticulated ; 

 the grinding-surface is very short, and is represented by only 

 two transverse close-set delicate ridges immediately below 

 the cutting-margin ; the root tapers a little backwards, and is 

 truncate. 



From the character of the teeth above described, it may be 

 inferred that the food of Glimaxodus was composed of some 

 soft material, notwithstanding the rather formidable appear- 

 ance of the grinding- or crushing-surface. The cutting- edge 

 of the scoop-like margin is sharp and thin, and does not seem 

 calculated to seize hard and resistant bodies ; and though it is 

 frequently worn evenly down, its sharpness is maintained, 

 often, apparently, by the wearing of the outside, as though 

 the teeth had been overlapped by those that opposed them. 

 And, moreover, the edge is not broken or chipped, as might 

 be expected if it had rough work to perform, or came into con- 

 tact with bony or shelly bodies. J^l either are the ridges of the 

 crushing-surface broken, but worn regularly, retaining their 

 sharpness, though in a few instances they are miich reduced 



