126 THE AMERICAN NATURALIST. [Vor. XXXV. 
notably by Röse and Kükenthal,! of the former existence of a. 
complete double dentition in the Marsupialia, and also indi- 
cated their tooth series as directly comparable to those of the 
Placentalia. Prior to this the permanent antemolar series of 
marsupials had been regarded as homologous not with the per- 
manent series of placentals but with their milk series? The 
results of Wilson and Hill seemed accordingly to suggest a 
fundamental relationship between the two groups in respect 
to dentition. . 
Similarly, the brilliant discovery by Hill (97) of an allantoic 
placenta in Perameles, which was announced at about the same 
time, appeared to indicate a further relationship in respect to 
placentation, the Marsupialia having been previously regarded 
as entirely non-placental. 
These writers in discussing the bearing of their results, 
although maintaining a moderate position throughout, recog- 
nized these resemblances as primary, and indicated their pref- 
erence for a derivation of the Marsupialia, by retrogression 
both of milk dentition and placenta, from primitive Placentalia. 
Now, although it must be admitted that decisive proof of a 
former diphyodont condition in the Marsupialia would be of 
considerable confirmatory value in the presence of other evi- 
dence, it would not of itself suggest placental origin. The 
possibility of such a condition was clearly perceived by Huxley 
in advocating the reverse. On page 655 of his essay he says: 
* I think that there can be no reasonable doubt that the existing 
marsupials have undergone a like suppression of the deciduous 
teeth, in the course of their derivation from ancestors which 
possessed a full set." Huxley, in fact, astutely realized that, 
in order to successfully derive the Marsupialia and Placentalia, 
it was necessary to assume the existence of a diphyodont 
condition in their metatherian ancestors. 
We may accordingly pass over the evidence derivable from 
the tooth development of Perameles and consider in detail only 
that presented by Hill alone from a study of the placenta of 
! See Wilson and Hill (97, pp. 433-438). 
? Róse, Kükenthal, Leche ; see Wilson and Hill (97, p. 582). This view has 
been recently supported by Dependorf (98). 
