The Saxicava sand sometimes rests on the Leda clay or Boulder. clay and sometimes directly on the rock, and the latter is often striated below this deposit; but in this case there is generally reason to believe that Boulder-clay has been removed by denuda. tion.

4. Terraces and Inland Sea Cliffs.

These are closely connected with the deposits last mentioned, inasmuch as they have been formed by the same recession of the sea which produced the Saxicava sand. At Montreal, where the isolated mass of trap flanked with Lower Silurian beds, constituting Mount Royal, forms a great tide-guage for the recession of the Post-pliocene sea, there are four principal sea margins with several others less distinctly marked. The lowest of these, at a level of about 120 feet above the level of the sea at Lake St. Peter, may be considered to correspond with the general level of the great plain of Leda clay in this part of Canada. On this Terrace in many places the Saxicava sand forms the surface, and the Leda and Boulder-clay may be seen beneath it. This may be called at Montreal the Sherbrooke Street Terrace. Another, the Water-work Terrace, is about 220 feet high, and is marked by an indentation on the Lower Silurian limestone. At this level some Boulder-clay appears, and in places the calcareous shales are decomposed to a great depth, evidencing long sub-aerial action. Three other Terraces occur at heights of 386, 440, and 470 feet, and the latter has, at one place above the village of Cote des Neiges, a beach of sand and gravel with Saxicava and other shells. Even on the top of the Mountain, at a height of about 700 feet, large travelled Laurentian boulders occur. On the Lower St. Lawrence, below Quebec, the series of Terraces is generally very distinctly marked, and for the most part the lower ones are cut into the Boulder and Leda clays, which are here of great thickness. I give below rough measurements of the series as they occur at Les Eboulements, Little Mal Bay and Murray Bay, where they are very well displayed. I may remark in general with respect to these Terraces, that the physical conditions at the time when they were cut must have been much the same with those which exist at present, the appearances presented being very similar to those which would occur were the present beach to be elevated.

TERRACES LOWER ST. LAWRENCE.

Heights in English feet, roughly taken with Locke's Level and Aneroid.

With reference to the differences in the above heights, it is to be observed that the Terraces themselves slope somewhat, and are uneven, and that the principal Terraces are sometimes complicated by minor ones dividing them into little steps. It is thus somewhat difficult to obtain accurate measurements. There seems, however, to be a general agreement of these Terraces, and this I have no doubt will be found to prevail very extensively throughout the Lower St. Lawrence. It will be seen that three of the principal Terraces at Montreal correspond with three of those at Murray Bay; and the following facts as to other parts of Canada, gleaned from the Reports of the Survey and from my own observations, will serve farther to illustrate this:

Kemptville, sand and littoral shells, 250 feet.

The evidence of sea action on many of these beaches, and the accumulation of shells on others, point to a somewhat long residence of the sea at several of the levels, and to the intermittent elevation of the land. On the wider Terraces, at several levels it is usual to see a deposit of sand and gravel corresponding to the Saxicava sand.

In the following table I have endeavoured to represent to the eye the facts observed in the internal plain of the great Lakes, and in the marginal area of the Atlantic slope, with the mode of accounting for them on the rival theories of glacier ice and floating ice.

It will be observed that the theoretical views diverge with respect mainly to the Boulder-clay and the striation under the Erie clay, and to the cause of the erosion of valleys in the Pliocene land. I would merely remark, in addition to the considerations already advanced, that the occurrence of drift-wood in the Erie clay, and of sea shells in the Boulder-clay, are both most serious objections to the glacier hypothesis, reserving for the sequel a more full discussion of the rival theories.

While the marginal marine area strictly corresponds to the marginal areas of Europe, I have no distinct evidence that the internal plains and table lands of the old continent correspond in their formations to the internal lake area of America.

An interesting fact with reference to the Erie clay, stated in the Report of the Survey of Canada is, that these clays burn into

white brick, while the marine Leda clay burns into red brick. The chemical cause of this I have already referred to, but whether it implies that the inland clays are fresh-water, or only that they have been derived from a different material, is uncertain. The gray clays of the Hudson River series in Western Canada, might, according to Mr. Bell, have afforded such clays.

Under the theory of a glacial sea immediately succeeding the elevated Pliocene land, the great amount of decomposed rocks which must have accumulated upon the latter constitutes an important element in the estimation of the rate of deposit of the Erie and Boulder and Leda clays. It is also to be observed that this glacial sea might have had to scour out of the lake basins of Canada only the soft mud of its own deposition, the rock-excavation having apparently been in great part effected in the previous Pliocene period. On this subject I find that Dr. Sterry Hunt had, before the publication of Dr. Newberry already alluded to,* shown that not only channels but considerable areas about Lakes Erie and St. Clair had been deeply excavated in the paleozoic rocks and filled with Post-pliocene deposits. The Devonian strata, he remarks, "are found in the region under consideration. at depths not only far beneath the water level of the adjacent Lakes Erie and St. Clair, but actually below the horizon of the bottom of these shallow lakes." He shows that around these in various localities the solid rocks are only met with at depths of from one to two hundred feet below the level of the lakes, while "the greatest depth of Lake St. Clair is scarcely thirty feet and that of the South-western half of Lake Erie does not exceed sixty or seventy feet, so that it would seem that these present lake basins have been excavated from the Post-pliocene clays, which, in this region, fill a great ancient basin previously hollowed out of the paleozoic rocks, and including in its area the Southwestern part of the peninsula of Ontario."

It would thus appear that in the Pliocene period the basin of the lakes may have been a great plain with free drainage to the sea. Whether or not it was afterwards occupied by a glacier, this plain and its channels leading to the ocean were filled with clay at the beginning of the Post-pliocene subsidence; and at a later date the mud was again swept out from those places where the Arctic current could most powerfully act on it.

(To be continued.)

• On the Geology of South-western Ontario. Am. Jour. Sci. 1868.