1010.] Homology of the Molar Cusps in Zalambdodonts. 239 



is suggestive of the Jurassic Dryolestes. In the lower molars (\n^, \x\^ 

 the talonid is very small, })lace(l on a very low level and lacks a hypoconid. 

 In nig the talonid is better developed. The lower jaw and lower molars 

 show significant resemblances to those of Apternodus mediaevus Matthew 

 (c/. p. 259). 



In Solenodon, which in most other characters also is more specialized 

 than Microgale, the internal basal ledge in the molars is more pronounced 

 (Fig. 17, no. 3). In Pofamogale the internal basal ledge assumes the form of 

 a low V-shaped cusp, while, in correlation with the antero-posterior broaden- 

 ing of the crown (see p. 189 above), the main or outer V is more open and 

 tends to bud off an incipient posterior V (Fig. 17, no. 4). These features are 

 more strongly expressed in the milk dentition (Fig. 17, no. 5), a fact of 

 doubtful significance. 



The low internal cusp in the Pofamogale molar (Fig. 12, B) corresponds 

 in function and position to the protocone of normal tritubercular teeth 

 because as shown by examination of a specimen in the National Museum 

 (No. 124327) it clearly fits into the basin of the talonid, which in the 

 Pofamogale molars is much more developed than in any Centetid. 



And yet in spite of this correspondence in function with a protocone it 

 seems probable that this internal basal cusp in Pofamogale is not homologous 

 with the protocone of Icfops and of normal tritubercular teeth, but with 

 the basal internal cingulum, which is seen in various stages of development in 

 Tupaia, Pfilocercus, all the Zalambdodonts, Leptictids, Erinaceids, Mijo- 

 gale, Proscalops (Fig. 17, no. 9) and the modernized shrews and moles. And 

 although the high internal V in Microgale did not fit directly into the basin 

 of the talonid like a normal protocone, this appears to be because there is 

 properly speaking no basin of the talonid for it to fit into; this is owino- to 

 the lack of a hypoconid, which in fully developed tritubercular teeth fits 

 outside the protocone, which, in turn, is. thus received into the vallev between 

 the entoconid and the hypoconid. 



If then the basal internal cusp of the Pofamogale molar be homologous 

 with an upgrowth of the basal internal cingulum, while the main V-shaped 

 cusp be homologous with the protocone of Icfops we have a general analogy 

 with the fourth premolar of such a Creodont as Chriacus. According to 

 this view the para- and metacones of Pofamogale are represented not by 

 the main V and its incipient posterior offshoot but by the reduced external 

 cusps which may have been derived by reduction from the paracone and 

 metacone of such a form as Icfops thomsoni (Fig. 17, no. 6). 



Pofamogale a specialized Cenfefid. The validity of the conclusion that 

 the basal internal cusp of the Pofamogale molar is a cingulum upgrowth, 

 not homologous with the true protocone of other mammals, depends largely 



