Dr. W. D. Matthew—Relationship of the Sparassodonta. 538 
Concerning the dental formula he observes: ‘‘ The dental formula of 
the cheek-teeth, in the opinion of Dr. J. L. Wortman and the present 
writer, is identical in the Sparassodonts, Carnivorous Marsupials, and 
Creodonts, and is therefore of no importance except to indicate the 
mutual relationship of all three groups.’”’ This assertion can only be 
taken to mean that the dental formula in Marsupials and Sparassodonts 
is derived from that of Creodonts by reduction of the successional 
premolars. his theory is very generally held. But for ‘ Creodonts’ 
might just as well have been substituted Carnivora or Eutheria in 
general. Accepting Dr. Lydekker’s view as to the derivation of one 
formula from the other, the fact remains that they are distinct and 
show no approximation among the oldest known Tertiary representa- 
tives of the two groups. Carnivorous Marsupials, perfectly typical as 
to dental formula, inflection of the jaw, etc., occur with the early 
Creodonts in the Eocene and Oligocene formations of North America 
and in the Oligocene of Europe. Indeed, if we may rely on the 
imperfectly known Jurassic mammals the distinction between the two 
formulas dates well back into the Mesozoic. 
In one or more genera of the Sparassodonta, Dr. Ameghino has found 
evidence that the canines-and the second post-canine tooth (premolar), 
as well as the third, is replaced by a successional tooth. In others, 
so far as known, the third premolar only is replaced, as in modern 
Marsupials. This does not affect the dental formula. The fourth post- 
canine tooth is always molariform, belongs to the lacteal series, and is 
never replaced by a successional tooth. In no Creodont is any tendency 
observed toward the retention of the fourth milk premolar in the 
permanent series; the number of permanent molars does not exceed 
three, and the replacement is of the normal Eutherian type. 
The absence of palatal vacuities among the Santa Cruz Sparassodonta 
and of epipubic bones in one genus cannot be considered as of equal 
importance with the dental formula and other characters listed above, 
since they are neither constantly nor exclusively present among 
Marsupials. Epipubic bones are vestigial in Zhylacinus, the most 
nearly related form among modern Marsupials; their presence in less 
specialized types of ancient and modern Marsupials indicates them as 
a primitive character of the order lost in certain specialized races. 
Even if this character were known to be constant throughout the 
Sparassodonta it would merely accord with the general trend of their 
specialization, and is surely no argument for removing them from the 
Marsupialia. As for the absence of palatine vacuities in Sparassodonts, 
this feature can hardly be given great weight since palatal vacuities 
are absent in several modern Marsupials and present in several 
Insectivores. Whether the vacuities are primary or secondary is not 
very clear; it does not follow, at all events, that because the Sparasso- 
donts are related to Thylacinus they are directly ancestral, and 
the presence of palatine vacuities in the modern genus may be an 
archaic character lost by the South American group but retained along 
with several other more primitive characters by the modern Australian 
genus. Palatal vacuities are absent in the primitive Erinaceid Proterix 
and in the Leptictide, from which family the hedgehogs are probably 
derivable, but are present in the modern hedgehog; this may be 
