512 REMARKS ON THE CLASSIFICATION OF MAMMALIA. 
such valleys as those of St. Davids and the Welland Canal. They 
at length emerged slowly, and portions of their edges were removed 
by the action of waves and currents by which cliffs were formed at 
successive heights, such as those I have already adverted to around 
Flamboro’ Heights. After this event another submergence under the 
sea occurred, and was followed by the glacial period, which it is 
to be remarked was of comparatively modern date. As soon as the 
table land between Lakes Hrie and Ontario emerged, subsequently 
to the deposition of the drift and successive formation of the sea 
beaches, the River Niagara came into existence, the basin of Lake 
Ontario still forming part of the sea. The cataract would then be- 
-at Queenston, falling directly imto the sea; and then would commence 
the retrograde course of the river, contimuing uninterruptedly as 
already described, till the present time. 
REMARKS ON THE CLASSIFICATION OF MAMMALIA. 
eras 
BY THE REV. W. HINCKS, F.L.S. 
PROFESSOR OF NATURAL HISTORY IN UNIVERSITY COLLEGE, TORONTO. 
I will introduce the few remarks I have to submit on this subject, 
by referring to an attempt which I made, at one of the meetings of 
the Canadian Institute, to give an improved expression of the affinities 
of the different groups in the class Birds, on the principle of a central 
body especially typical, with deviations of structure to suit special 
tendencies or special conditions. 
I then endeavoured to show that if the great body of Insessorial 
or perching birds, be considered as the truest representatives of bird 
structure, there would be found to be five principal deviations from it 
which have been deemed so important as to be regarded by most orni- 
thologists as forming distinct orders :—Raptores or birds of prey, 
Scansores or climbers, Rasores or poultry and game birds, Grallatores 
or waders, and Natatores or swimming birds. I pointed out also that 
there are families among the Insessorial birds, of greater or less extent, 
approximating to each of these deviative groups; and that there are 
