from tin; Marl-Hlate of Midilerhlye. 61 



a space betwouii tlicm l<ss than lialt" their diainetcr. They arc 

 in the fonn of irre^uhirly rounded bosses, with tlie surface 

 smooth and gh>ssy, and the margins sinuous and i)ro(hiced into 

 points. Sometimes, however, they arc nmeli clonj^atcd, and 

 are frequently very irregular in shape, with the marginal pnj- 

 longations much produced, variously formed, sharp or obtuse. 

 Others have one margin comparatively smcjoth, the points 

 being confined to the opposite side. Some are quite devoid of 

 all such irregularities, the margins being smooth or oidy a 

 little sinuous at one side ; these are rounder and larger than the 

 others (fig. 8). Another form (fig. 7), not by any means un- 

 common, is irregularly stellate, with the rays ridged and some- 

 times a little bifurcated. 



From the fineness of the cutting-margin in the so-called 

 ClimaxoduSj it was inferred in the former paj)er, so frequently 

 referred to, that the food must have been composed of some soft 

 material. We are disposed to draw the same conclusion from 

 the structure of the teeth oi'Jaua.ssa hituminosa. The scoop-like 

 cutting-margin is certainly much used, for it is almost always 

 greatly worn in a regular nuinner ; only in one instance have 

 we seen it a little broken. It would be an efficient instrument 

 in cutting vegetable substances, and these might afterwards 

 require the aid of the crushing-disk. 



In corroboration of this view of the food, we may quote a 

 passage from IMlinster, who says, of his Byzenos latipinnatus^ 

 that '^ the intestine seems to have been very full when the 

 fish died. It is more elevated than the other parts of the body. 

 On some places one sees in the interior a black earthy mass 

 in which small pointed bodies appear, Avhich are like small 

 pieces of shiny coal." May not this " black earthy mass " 

 and " pieces of shiny coal " be carbonized vegetable matter, 

 the food of the fish ? 



It is unnecessary here to dilate on the affinities of Janassa, 

 as the subject is discussed in the former paper, already quoted. 

 We may remark, however, that the full investigation of the 

 Permian s])ecies has only the more confirmed our opinion of 

 its close alliance with tlie Coal-measure form (the so-called 

 Climaxodus linrju(t'formis)^ and of a certain relationship of 

 both to Myliobatcs and Zijgohates. 



We may also state that Janassa is more closely related to 

 Petalodus than was at first thought ; for we now find that the 

 latter genus is provided with both symmetrical and oblique 

 teeth ; so that it is quite probable that they may be found to 

 be an-anged in much the same manner as those of Janassa^ 

 especially as the former have been found in vertical series, as 

 previously stated. 



