specimens of Leda limatula, mixed with rema plants.

The following are the species of plants re these nodules :

1. Drosera rotundifolia, Linn. In a calcar from Green's Creek, the leaf only preserved. is common in bogs in Canada, Nova Scotia, foundland, and thence, according to Hooker, to circle. It is also European.

2. Acer spicatum, Lamx. (Acer montanu Leaf in a nodule from Green's Creek. Foun Scotia and Canada, also at Lake Winnipeg, ac Richardson.

3. Potentilla Canadensis, Linn. In nod Green's Creek; leaves only preserved. I have

difficulty in determini but believe they must b to the species above 1 to P. simplex, Michx., by Hooker and Gray to riety. It occurs in Ca New England, but I ha formation as to its ran ward.

4. Gaylussaccia resi rey and Gray. Leaf i at Green's Creek. Abu New England and in also on Lake Huron and the Saskatchewan, acc Richardson (Fig. 77).

5. Populus balsamifera, Linn. Leaves and in nodules at Green's Creek. This is by much common species, and its leaves are of small size, a trees growing in cold and exposed situations. Th is North American and Asiatic, and abounds in N land and Canada. It extends to the Arctic circl

emains of land

recognised

careous node

d. This pr tia, and N I, to the Art

abundant on the shores of the Great Slave Lake and on the McKenzie River, and according to Richardson constitutes much of the drift timber of the Arctic coast (Fig. 78).

6. Thuja occidentalis, Linn. Trunks and branches in the Leda clay at Montreal. This tree occurs in New England and Canada, and extends northward into the

Hudson Bay territories. It is a northern though not arctic species in its geographical range. According to Lyell it occurs associated with the bones of Mastodon in New Jersey. From the great durability of its wood, it is one of the trees most likely to be preserved in aqueous deposits.

7. Potamogeton perfoliatus, Linn. Leaves and seeds in nodules at Green's Creek. Inhabits streams of the Northern States and Canada, and according to Richardson extends to Great Slave Lake.

8. Potamogeton pusillus. Quantities of fragments which I refer to this species occur in nodules at Green's Creek. They may possibly belong to a variety of P. hybridus which, together with P. natans, now grows in

the river Ottawa, where it flows over the beds containing these fossils.

9. Caricea and Gramineæ. Fragments in nodules from Green's Creek appear to belong to plants of these groups, but I cannot venture to determine their species.

10. Equisetum scirpoides, Michx. Fragments in nodules, Green's Creek. This is a widely distributed species, occurring in the Northern States and Canada.

11. Fontinalis. In nodules at Green's Creek there occur, somewhat plentifully, branches of a moss appar

ently of the genus Fon

tinalis.

12. Algae. With the plants above mentioned, both at Green's Creek and at Montreal, there occur remains of seaweeds (Fig. 79). They seem to belong to the genera Fucus and Ulva, but I cannot determine

the species. A thick

stem in one of the nodules would seem to indicate a large Laminaria. With the above there are found at Green's Creek a

number of fragments of leaves, stems, and fruits, which I have not been able to refer to their species, principally on account of their defective state of preservation.

None of the plants above mentioned is properly arctic in its distribution, and the assemblage may be characterised as a selection from the present Canadian flora of some of the more hardy species having the most northern range. Green's Creek is in the central part of Canada, near to the parallel of 46°, and an accidental selection

from its present flora, though it might contain the same species found in the nodules, would certainly include with these, or instead of some of them, more southern forms. More especially the balsam poplar, though that tree occurs plentifully on the Ottawa, would not be so predominant. But such an assemblage of drift-plants might be furnished by any American stream flowing in the latitude of 50° to 55° north. If a stream flowing to the north, it might deposit these plants in still more northern latitudes, as the McKenzie River does now. If flowing to the south, it might deposit them to the south of 50°. In the case of the Ottawa, the plants could not have been derived from a more southern locality, nor probably from one very far to the north. We may therefore safely assume that the refrigeration indicated by these plants would place the region bordering the Ottawa in nearly the same position with that of the south coast of Labrador fronting on the Gulf of St. Lawrence at present. The absence of all the more arctic species occurring in Labrador should perhaps induce us to infer a somewhat milder climate than this.

The moderate amount of refrigeration thus required would in my opinion accord very well with the probable conditions of climate deducible from the circumstances in which the fossil plants in question occur. At the time when they were deposited the sea flowed up the Ottawa valley to a height of 200 to 400 feet above its present level, and the valley of the St. Lawrence was a wide arm of the sea, open to the arctic current. Under these conditions the immense quantities of drift-ice from the northward, and the removal of the great heating surface now presented by the low lands of Canada and New England, must have given for the Ottawa coast of that period a summer temperature very similar to that at present experienced on the Labrador coast, and with this conclusion the marine remains of the Leda clay, as well as the few

land molluscs whose shells have been found in the beds containing the plants, and which are species still occurring in Canada, perfectly coincide.

The climate of that portion of Canada above water at the time when these plants were imbedded may safely be assumed to have been colder in summer than at present, to an extent equal to about 5° of latitude, and this refrigeration may be assumed to correspond with the requirements of the actual geographical changes implied. In other words, if Canada was submerged until the Ottawa valley was converted into an estuary inhabited by species of Leda, and frequented by capelin, the diminution of the summer heat consequent on such depression would be precisely suitable to the plants occurring in these deposits, without assuming any other cause of change of climate.

I have arranged elsewhere the Post-pliocene deposits of the central part of Canada, as consisting of, in ascending order: (1) The boulder clay; (2) a deep-water deposit, the Leda clay; and (3) a shallow-water deposit, the Saxicava sand. But, although I have placed the boulder clay in the lowest position, it must be observed that I do not regard this as a continuous layer of equal age in all places. On the contrary, though locally, as at Montreal, under the Leda clay, it is in other places and at other levels contemporaneous with or newer than that deposit, which itself also locally contains boulders.

At Green's Creek the plant-bearing nodules occur in the lower part of the Leda clay, which contains a few boulders, and is apparently in places overlaid by large boulders, while no distinct boulder clay underlies it. The circumstances which accumulated the thick bed of boulder clay near Montreal were probably absent in the Ottawa valley. In any case we must regard the deposits of Green's Creek as coeval with the Leda clay of Montreal, and with the period of the greatest abundance of Leda