OBSEEVATIONS ON THE DENTITION OE CHRYSOCHLORIS. 135 



responds to two ordinary molar cusps conjoined or to a single 

 one ? Mivart and Gidley have both held that the trigon of 

 the upper molar of Chry sochloris represents a fusion of the 

 two external triangles of typical mammals^ such as Talpa 

 or Dasyurus, and they unquestionably bring forward some 

 sti'ong evidence. In Po tamo gale and My gale we seem 

 to have an intermediate stage, Avhere the two triangles are 

 partly fused. I, however^ rather side with Woodward in 

 believing that the single triangle of the Chry sochloris 

 molar corresponds to only the anterior triangle of typical 

 molars. In almost every instance in osteology where two 

 structures in an ancestor are replaced by one in a descendant, 

 it is by the loss of one, not by the fusion of the two, and we 

 have reason to believe that this holds good for molar cusps. 

 The last molar in primitive mammals is at least not simpler 

 than the anterior ones, but in large numbers of the descen- 

 dants we find the last molar reduced. In Dasyurus the last 

 molar consists of a single triangle with internal protocone — a 

 tooth, in fact, almost of the Chry sochloris type. If we 

 look at the last molar of Perameles we find a well-developed 

 triangle, a small protocone, and a small posterior cusp, which 

 represents all that remains of the posterior triangle. Even in 

 Talpa the last molar does not show a fusion of triangles, but 

 a dwindling of the posterior one. In Pros cal ops the pos- 

 terior triangle is almost gone. In Sorex it is quite gone. 

 Even Po tamo gale seems to me to show, not a fusion, but a 

 dwindling of the posterior triangle. The lower molars of 

 Dasyurus may be looked upon as two triangles in much the 

 same way as the upper, yet the evidence is quite conclusive 

 that the single triangle in Chry sochloris asiatica is not 

 formed by the fusion of the two, but by the gradual reduction 

 and ultimately complete loss of the posterior taloii. 



Though the lower molars of Chrysochlosis are extremely 

 like those of Amphitherium, and the upper molars some- 

 what like some of the early types, it seems probable that both 

 types ai*e really degenerate derivatives of more normal types, 

 such as seen in Dasyurus, and they are of particular value 



