248 THE AMERICAN NATURALIST: [VoL. XXXV. 
are two reasons for regarding these as degenerate in dentition. 
First, there is a prototypal tooth pattern in marsupials, which 
is shared alike by the Didelphyida and the insectivorous mem- 
bers of the Dasyuridze, and of which the tooth patterns of the 
carnivorous Dasyuride and the Peramelide are undoubtedly 
progressive modifications, while those of Myrmecobius and 
Notoryctes are aberrant. The tooth patterns of both of these 
forms are more primitive in aspect than those of any other of 
the Australian forms, but the animals are specialized in other 
respects. Thus Myrmecobius is typically dasyurid in foot 
structure and in its incisor formula of 4, while Notoryctes, 
according to Dollo (99), is specialized in a phalangerine direc- 
tion in foot structure. Second, both Myrmecobius and Noto- 
ryctes are ant-eating forms,! such animals being usually 
degenerate in dentition, judging from Echidna and the placen- 
tal Myrmecophagide and Manidz, in which the teeth have 
been entirely lost. 
Excluding Myrmecobius and Notoryctes, we find the primi- 
tive dental characters of marsupials distributed as follows : 
The least reduced dentition is found in the Peramelide, 
where the formula is: 7.8, ¢.4, 5.8, m.42 
The least elaborated molar teeth are found in the insec- 
tivorous members of the Dasyuride. The upper teeth (Pl. I, 
Fig. d)? are triangular in shape, with three main cusps (proto- 
cone, paracone, metacone) * and an outer row of styles. There 
are no conules at the base of the protocone, such as are com- 
monly met with in placental types. The metacone exceeds the 
paracone in size, and its posterior border is produced into a 
trenchant spur. The lower teeth (Pl. II, Fig. d) are of the 
! For habits of Notoryctes see Stirling ('91, p. 158) and Spencer (96, p. 50, 
Mammalia). 
2 Or (m)i.3, mce4, mp. and 72.$, m.3, we Lydekker (99). The upper 
incisor bey may contain teeth of both seri 
is type might be more appropriately donaei by the teeth of Phascolo- 
cis khi pic. or Antechinomys, Dasyurus viverrinus being more carnivorous 
abit. Thomas, however, remarks that the teeth of the three former genera 
worn one another very closely, and differ only from those of Dasyurus in being 
more wp d cuspidate. See Thomas ('88, pp. 27 3» 298, 309). 
* Following the nomenclature of the tritubercular theory. See Osborn 
(88, '91) 
