27 
Of the fifteen spaces, ten, it will be observed, are vacant, because in 
the features those spaces represent there is practically no difference between 
the two tribes. In three of the five occupied spaces it is stated that a 
difference exceeds its probable error more than six times, which is tanta- 
mount to saying that the chances are over 19,300 to 1 that the two groups 
of Indians under consideration differ in the items those spaces represent. 
These are the cephalic index, the head width (biparietal), and the cephalo- 
facial index. Of the two indices, the cephalic, of course, denotes the 
relation the biparietal diameter bears to the glabella ad maximum; the 
cephalo-facial, the relation the bizygomatic bears to the biparietal diameter. 
Since the spaces that represent the glabella ad maximum and the 
bizygomatic diameters are blank, and the biparietal alone is occupied, it 
is evident that the biparietal diameter (or head width) is solely responsible 
for the high ratio of the two indices. And though the Oxford House 
Indians and the half-blood Sioux differ from each other in these three 
respects, one factor only, namely the width of head, and not three, is 
concerned. 
At the bottom of the column it is seen that though the figures repre- 
senting the width of the nose and the nasal index are slightly under 3 • 0, 
that which represents the height of the nose is less than 2-0, from which 
it may fairly certainly be assumed that the nose of the half-blood Sioux 
is slightly narrower than that of the Oxford House Indians. From the 
above, then, we surmise that the Oxford House Indians have likely as 
great an admixture of white blood as have the half-blood Sioux, and that 
of all the physical features with which we have to deal, only in breadth 
of head and in breadth of nose do they materially differ the one from the 
other; and especially in breadth of head. 
Comparing the Oxford House Crees with the pure Sioux, we gather 
from the third column of table IV that these two tribes could be distin- 
guished from each other by all the features represented in the table except 
the proportions of their trunks and limbs, lengths of heads, and their 
cephalo-facial indices. As we have already pointed out, we must beware 
of this index and regard it only in association with its component parts. 
When these are taken into account, it will be seen that the Oxford House 
head and face are both approximately 4*5 mm. narrower than the Siouan 
head and face. The absolute diameters differ very appreciably, but they 
differ to an equal extent, with the result that the index is unaffected. 
Accepting the Oxford House Crees as part-breeds, similar in almost 
all their general features to the part-breed Sioux, we may note that the 
features of these part-breed Crees depart from those of the Island and 
Gods Lake Crees and Saulteaux in the same general direction as the part- 
breed Sioux do from the pure Sioux. These directions, especially perhaps 
in the case of the Cree, are towards the Nordic type. Notably is this true 
of the stature, the width of head, and the width of face, the shortness of 
the face, the breadth of the hand, and to some extent the breadth of nose 
and mouth. The high arm stretch of the pure Indian seems to be a dom- 
inant trait which persists in the breeds. 
The male Crees and Saulteaux of Island lake may be differentiated 
from the pure Sioux in most of the tabulated features, the length and 
781S5-3 
