214 B. N. A. BOUNDARY COMMISSION. 
485. The most instructive sections of this great plateau were, 
however, observed on the Roseau River, which rising a few miles west- 
ward of the Lake of the Woods and approximately following the forty- 
ninth parallel, though sometimes dipping to the south, finally flows into 
the Red River, north of the Line. From the eastern edge of the plateau 
where it fronts on the Red River Prairie, for about twenty-five miles 
eastward, high cliftlike banks with good sections, frequently overlook 
the stream. In their general character they all agree more or less 
closely, and the sections often present features almost identical for long 
distances. The lowest beds are of very fine sand, and arenaceous clay, 
often quite plastic. These generally show irregular and current bedding, 
and do not include large stones or boulders. Above these beds, and 
resting on their denuded edges, lie coarser horizontal deposits of rounded 
pebbles, and small boulders, chiefly of white limestone, and apparently 
indicating littoral conditions ; which as the country slopes gently west- 
ward, must have passed over it in that direction, as the former great lake 
of the Red River Valley grew smaller. Above these gravel beds, yellowish, 
or white, clean, fine sand, several feet in thickness, is very generally 
found ; over which is a second pebble-bed, the fragments generally of 
limestone, and smaller and better rounded than in the lower. On this 
rests the present soil, usually from one, to two, or three feet in 
thickness, and coloured with vegetable matter. 
486. Some sections showed near the water’s edge, clay beds filled 
with stones, which though not well exposed, probably represent the 
boulder-clay proper. From it must be derived a great part of the large 
boulders which so encumber the course of the stream. The false-bedded 
sands, and associated clays, thus seem to intervene between the boulder 
clay and the littoral lake deposits, and appear to indicate moderately deep 
water, with str®ag currents, but as a rule without ice. Careful search 
was made for organic remains, but with little success. Not a single 
Molluse of any kind was discovered, but one section showed a con- 
siderable quantity of carbonaceous and peaty matter along the irregular 
deposition planes of the sand, and about thirty feet from the top of the 
bank, a decayed tree trunk stood out. The wood is soft and crumbling, 
and stained yellow by oxide of iron, but from its microscopic characters 
appears to be a fragment of the ordinary cedar. (Thuja occidentalis.) 
487. Figures 1 and 2 Plate XI, illustrate the nature of the drift and 
superficial deposits of the plateau in this region. Section 1, is remarkable 
for the great thickness of the first gravel bed, and the large sized 

