32 
of both types. The metaconid has nearly the same relative size in both. 
In fact the agreement is so striking and the molar type so peculiar that 
it seems probable that Diaphorodon and Eodelphis were closely related, 
the latter perhaps ancestral to the former. 
Eodelphis is clearly a marsupial, as shown by Matthew, and, in common 
with most Lance marsupials, its resemblance to the later didelphids 
is so striking that the writer prefers to place it in the Didelphiidae, at least 
until it is much better known. Within the family, however, it occupies 
an isolated position, certainly not ancestral to any known Tertiary or 
Recent phylum. Aberrant specialization is seen in the stout jaw, the 
three incisors with the median one enlarged, the crowded nature of the 
premolars and enlargement of Pa,^ the relative enlargement of the paraconid, 
and reduction of the metaconid. It is one of the aberrant members of 
the varied Cretaceous didelphid group. More generalized forms are 
common in the Lance, but have not yet been found in the Belly River. 
iPi larger than Pt is a primitive character seen in most fossil didelphids. but in Eodelphis Pi is onsually large 
and heavy, although less so than in Thlaeodon, for instance. 
