ON BRITISH FOSSIL MAMMALIA. 7<J 



orders. It is on this account, and because the tooth of the fossil in question 

 differs in the shape of the middle and size of the accessory cusps from that 

 of any known species of Didelphys, that I regard its reference to that genus 

 as premature, and the affinities of the species to which it belongs as needing 

 further evidence before they can be determined beyond the reach of doubt. 

 Mr. Charlesworth, by whom the present fossil was first described and figured, 

 has accurately specified the differences above alluded to in the shape of the 

 crown of the tooth as compared with the false molars of the true Opossums : 

 they are seen in the more equilateral or symmetrical shape of the middle 

 cusp, the greater development of the posterior talon, and the presence of 

 the anterior talon at the base of the middle cusp : the grounds on which his 

 determination of the fossil was founded are not stated. 



The crowns of the spurious molars of the placental Ferae, which present the 

 same general form as the fossil, are thicker from side to side in proportion to 

 their breadth ; the spurious molars of the Dasyurus, Thylacinus, and Pkas- 

 cogale, differ in like manner from the fossil. It is in the marsupial genera 

 Didelphys and Perameles that the false molars present the same laterally 

 compressed shape as in the fossil. Now besides the perfect tooth, the fossil 

 includes the empty sockets of two other teeth ; and the relative position of 

 these sockets places the Perameles out of the pale of comparison. On the 

 hypothesis that the present fossil represents a species of Didelphys, the tooth 

 in situ unquestionably corresponds with the second or middle false molar, 

 right side, lower jaw. This is proved by the size and position of the an- 

 terior alveolus. 



Had the tooth in situ been the one immediately preceding the true molars, 

 the socket anterior to it should have been at least of equal size and in juxta- 

 position to the one containing the tooth. The anterior socket, however, is 

 little more than half the size of the one in which the tooth is lodged ; it is 

 also separated from that socket by an interspace equal to that which separates 

 the first from the second false molar in the Didelphys Virginiana. This is 

 well shown in the inside view. In the placental Mammalia, in which the first 

 small false molar is similarly separated by a diastema from the second, the 

 first false molar has only a single fang. In the present fossil, the empty 

 socket of the first false molar proves that the tooth had two fangs, as in the 

 marsupial Perce and Insectivora. There is nothing in the size or form of the 

 socket, posterior to the implanted tooth of the fossil, to forbid the supposition 

 that it contained a false molar, resembling the one in place ; had it been the 

 socket of a true molar, then the fossil could not have belonged to Didelphys, 

 or to any other known marsupial genus, because no known marsupial animal, 

 which presents the posterior false molar of a similar form and in like juxta- 

 position with the true molars as the tooth in the present fossil (on the sup- 

 position that it immediately preceded the true molars), has the next false 

 molar so small as it must have been in the fossil on that supposition. 



Upon the whole, the conclusion that the present eocene tertiary fossil is 

 marsupial is the most probable one, but the evidence is insufficient to demon- 

 strate its truth. Cuvier, however, might have failed to convince contem- 

 porary naturalists that the corresponding formation in France contained the 

 remains of a Didelphys, unless he had had the good fortune to demonstrate 

 the marsupial bones in their natural connections with the pelvis. 



Genus Phascolotherium. 

 With regard to the fossil on which the genus Phascolotherium is founded, 

 the maxillary evidence is more complete. Two rami of lower jaws, one con- 

 taining the whole dental series, have been discovered in the oolitic slate at 



