No. 410.] THE MAMMALIAN SUBCLASSES. 123 
recognized a genetic succession of Prototheria, Metatheria, 
and Eutheria, the Metatheria being intermediate with re- 
' spect to the possession of certain prototherian and certain 
eutherian characters, among the former notably a non-placental 
allantois. To the Marsupialia, in passing from the purely 
metatherian condition, he assigned the adoption of the following 
special characters: a prehensile foot, associated with arboreal 
habit; a marsupium ; a reduced milk dentition; a foreshortened 
period of uterogestation. 
Huxley therefore implied the following plan of succession : 
the Prototheria giving rise to the Metatheria, and becoming 
specialized as Monotremata; the Metatheria, in turn, giving 
rise to the Eutheria, and becoming specialized as Marsupialia ; 
the Eutheria becoming specialized as Placentalia. 
In opposition to this Dollo and others regard the Metatheria 
as giving rise to the Eutheria or primitive Placentalia, and the 
latter as giving rise, on the one hand, by retention of the pla- 
centa, to the modern Placentalia, and, on the other, by loss of 
the placenta, to the Marsupialia. Other investigators, notably 
Osborn, Hubrecht, and Seeley, recognize a separate origin of 
the main mammalian stems. 
With this introduction we may examine certain features of 
the evidence on which these views are based. 
I. The Question of a Placental Origin of the Marsupialia. 
In the first place, what is the bearing of arboreal ancestry ? 
If we assign to arboreal habit any value as a differential 
character of the Mammalia, we may recognize two possible 
arrangements, based on the views of Huxley and Dollo, in 
which this factor might appear. For the sake of TEM 
these may be indicated in tabular form as on following page. 
According to the former plan arboreal habit would be asso- 
ciated with the differentiation of the Marsupialia and Placen- 
talia. According to the latter it would also be an index — 
following Dollo of a placental (eutherian) origin of ‘the 
Marsupialia. This will make it clear that we have two distinct 
Problems to decide: one as to the value of arboreal ancestry 
