90 DR. ROBERT BELL 
of moving them little by little from the central point towards the periphery. After the 
lapse of a sufficient length of time, all the boulders would be deposited around the 
circumference of the pond, as we see them. It is a curious circumstance that when a 
large boulder happens to occupy the central point of a pond, it remains undisturbed, while 
every other boulder is removed. 
DYKES OF BOULDERS AND SHINGLE.—These are found at high-water mark around 
the shores of islands and points in many of our northern lakes of large size which freeze 
over in winter. They are particularly observable where the water is shallow and the 
shores low and shelving, as in Lakes Winnipegosis and Manitoba, and St. Martin Lake. 
Some of the islands in these lakes are completely surrounded by such dykes, which are 
quite steep on both sides, and have an almost uniform height of about five or six feet. 
They are clearly due to the shoving of the rafts of ice after it breaks up in the spring, and 
as these impinge on different parts of the shore of an island in different years, a dyke is 
at length formed around it. The evidence of recent ice-shoves may be plainly seen here 
and there every year, the freshly upturned earth and boulders marking the sites of those 
of the preceding spring. Similar dykes may sometimes be observed around the upper 
portions of alluvial islands in rivers, and they have been ascribed to human agency by 
those unfamiliar with their mode of formation. 
PERPETUALLY FROZEN SoiL.—The limit of perpetual frost in Canada appears to be 
placed too far south. In the banks of our northern rivers, even as far south as the Nelson, 
it is not uncommon in the middle of summer for a shell of clay to peel off and expose a 
frozen face of thirty or forty feet in height. This, however, does not imply that the 
ground in the neighborhood is everywhere frozen to this depth. The bank-face being 
destitute of snow and exposed to the full force of the winter winds, the frost penetrates 
into it horizontally to a greater depth than it does vertically from the level surface of the 
ground—a depth sufficient to preserve the frost all summer. The writer having called 
Sir Henry Lefroy’s attention to this circumstance, he admitted that, owing to the radiation 
which takes place both vertically and horizontally from steep banks, we may reasonably 
conclude that the frost penetrates to much greater depths in such situations. In more 
southern latitudes, the writer has observed two instances in similar positions, in which the 
ground remained frozen long after it had thawed out everywhere else in the surrounding 
country. One was near the brink of St. Francis River, at Trenholmville, in the province 
of Quebec, and the other in a similar situation on the bank of Nipigon River at Red 
Rock, Lake Superior. In both cases, the discovery was made in digging holes to plant 
posts for mooring boats, and in each, the circumstance was a matter of surprise to the 
inhabitants. In the country around York Factory on Hudson Bay, where the soil is, by 
some, supposed to be perpetually frozen, the swamps are full of water all winter, and the 
snow even prevents thick ice from forming at the surface. Water escaping from these 
swamps trickles down the banks of the rivers, even in the middle of winter. The fact 
that in these regions there is plenty of water under the ice in the small streams, and that 
the beaver inhabits them as far north as Fort Churchill, would appear to show that the 
frost does not penetrate to such a depth as to remain throughout the year. <A test was 
made by the writer in the swamps just behind York Factory in August, 1879. A heavy, 
