AMERICAN FORMS. 125 



years a vast number of such remains were obtained, through 

 the energy of the Eev. P. B. Brodie and Mr. Beccles. All 

 these specimens were obtained from a single bed, and 

 many of them indicated forms more or less closely allied 

 to those from Stonesfield and Stuttgart. It was thus 

 shown, once for all, that mammalian life must have been 

 locally abundant throughout the Jurassic period. This 

 conclusion was subsequently amplified by the discovery in 

 the upper Jurassic rocks of North America of a whole 



K 40. Lower Jaw of Triconodon j twice natural size. 

 (After Marsh.) 



host of small mammals very closely allied to those from 

 Dorsetshire, a large number of which have been described 

 by Prof. O. C. Marsh, some of whose figures are here 

 reproduced. Many of these small Jurassic mammals 

 (Figs. 39 and 40) were evidently carnivorous, and such 

 carnivorous forms exhibited two distinct types of dentition. 

 In one of them (Fig. 39) there was a numerous series of 

 cheek-teeth behind the tusk, or canine (a), each of which 

 carried three cusps arranged in a triangle ; while in the 

 other type (Fig. 40) the cheek-teeth were fewer in number, 

 and had the three cusps on their crowns ranged in the same 

 line. From this peculiarity the animal to which the second 

 type of jaw belonged was appropriately named Triconodow. 

 The first type corresponds to the amphithere of the Stones- 

 field slate, while the second is more like the phascolothere 

 of the same formation. 



In Fig. 40 it will be observed that there is a pecu- 

 liar groove (g) running along the inside of the jaw, and 



