291 
of the Rodentia to the Marsupialia. 
which fall out early in life, but the second pair is long 
retained. In Eypsiprymnus two incisors in the lower jaw 
work against six in the upper, and of the latter the first pair 
have grown considerably stronger, while the second and third 
pairs are of inferior size. In Phalangista the six incisors 
stand in an elegant horseshoe-like curve on the margin of the 
broad premaxillae ; but in Hypsipryrnnus the snout has 
become narrower, the premaxill^e being laterally compressed. 
Then the four smaller incisors curve more towards the middle, 
in order to function, as opposed to the upper teeth, in tearing 
off plants. A large series of skulls of Eypsiprymnus shows 
in what different ways this purpose can be attained. But the 
four teeth are too weak to be retained with advantage in 
adaptive groups ; hence they undergo the same fate as the 
corresponding teeth in the lower jaw of Phalangista . In this 
way it seems to me that the typical dentition of the Rodents 
with its two pairs of incisors has been produced. The trans- 
formation of the enamelled and root-bearing incisor into the 
persistently growing gnawing- tooth furnished with an 
enamel plate on one side only may also be easily traced in 
the stem of the Marsupialia. In the group of the Lago- 
morpha the dentition shows conditions which accord well with 
my speculations. In the upper jaw, behind the gnawing- 
teeth, the second pair of small incisors is quite pressed towards 
the median plane ; they are also changed and have acquired 
the power of persistent growth. The gnawing-teeth them- 
selves, in both the upper and the lower jaw, also have very 
short alveoli and a slight curvature. 
Notwithstanding the undoubtedly important part which 
the dentition plays in rapid systematic diagnosis I do not 
think that the notion of a direct blood-relationship can be 
founded with sufficient certainty upon the similarity of the 
dentition alone. Therefore I will adduce further proofs. 
The horizontally inward projection of the angle in the 
lower jaw of the Marsupials is well known as a very con- 
venient and striking character. If the Rodents be phylo- 
genetically related to the Marsupials this structure must also 
be still recognizable ; and in fact the comparison of many 
skulls has showm me that the often -described bending of the 
posterior angle of the mandible in Rodents, which occurs in 
variable degrees in different sections of the order, is derivable 
in a direct series from what is found in the Marsupials. I 
affirm most decidedly that Rodentia and Marsupialia manifest 
their relationship by the homologous behaviour of the angle 
of the mandible. In Muridse, Bciuridae, and Myoxidae this 
peculiarity is particularly clearly marked, although it has 
