26 SIRENOID AND CROSSOPTERYGIAN GANOIDS. 



and small teeth consolidated and enlarged by addition of fresh substance into those 

 of the adult? We have as yet no direct information on these points, but it is 

 highly probable that there was no replacement. If they were hard from the first, and 

 replaced as they became too small, we should expect to find abundance of the small shed 

 teeth. This is not the case, and nearly all are of what may be considered the adult size ; 

 besides, it is easy to see that great temporary inconvenience would result from the sudden 

 loss of all the grinders, while replacement one by one would lead to distortion, and 

 almost equally cripple the masticating power. If, on the other hand, the teeth were hard 

 from the first, and enlarged by successive additions, this could only take place along the 

 outer edge if the function of the teeth were maintained during its growth. Upon this 

 supposition the angle, as the first formed part, should be greatly worn, while lines of 

 growth should be apparent upon the surface. Neither of these indications can be traced 

 either in the recent or fossil teeth. It is probable that the tooth was at first compara- 

 tively soft and vascular, that it grew, replaced the loss from wear, and changed its figure 

 as requisite, becoming fully calcified and rigid only in the adult. When this point was 

 reached, further increase could take place only by the addition of fresh layers to the base. 



Such a dentition is plainly adapted to a herbivorous diet, and this conclusion is 

 established by observation of the recent species. 1 The arrangement resembles in its general 

 features that of a Ruminant mammal, except that the incisors are in the upper instead of 

 the lower jaw, a point of difference which may be due to the circumstance that the 

 Sirenoid approaches the plants upon which it feeds from below, the Ruminant from 

 above. 



When in use the first mandibular denticle is received into the first palatal furrow. 

 The chief contact of the two sets of teeth is apparently between the tips of the ridges of 

 the lower and the corresponding hollows of the upper. Hence the palatal denticles are 

 usually less worn than those of the mandible, while the first mandibular denticle is 

 commonly much worn and the corresponding hollow deep and extensive. Evidence of 

 wear is often to be found also in both sets upon the angle and the internal margin 

 posterior to the angle. 



The British species of Geratodus range from the Keuper to the Oolite. C. Icevissimus 

 (a MS. species of unknown authority) comes from the Upper Keuper sandstone of Ripple, 

 Worcestershire, and C. Phillipsii, Ag., from the Stonesfield Slate. It is unfortunate 

 that the latter specimen, like some others upon which specific determination partly rests, 

 can no longer be found. I see no reason to doubt the express statement of Agassiz 3 that 

 such a fossil did exist, confirmed as it is by Phillips himself. 3 Among fossils purchased 

 by the British Museum from the Mantell collection is a Ceratodus tooth imbedded in an 



1 In the stomach of a recent Ceratodus I find much vegetable matter, and also some bivalve molluscs 

 (Cyrena). 



2 'Poissons Fossiles,' t. iii, pp. 135, 176. 



3 'Manual of Geology,' table on p. 324 (1855). 



