Canadian Historic Sites: Occasional Papers in Archaeology and History No. 3

The Armstrong Mound on Rainy River, Ontario

by Walter A. Kenyon

The Excavation

The Armstrong Mound is situated on the edge of a relatively flat terrace that rises 26 ft. above Rainy River. (Elevations are relative to water level south of the mound on June 8, 1966.) About 215 ft. from the edge of the river, the terrace drops off sharply to an old flood plain which is also relatively flat, and which rises 10 ft. above the river. On a topographic map, the mound is located just below the 1,100-ft. contour line.

Although the mound stands on the edge of a clearing, brush and young trees had encroached upon its eastern edge. Here was a dense thicket of hazel and wild plum, with a few poplars up to 10 in. in diameter, and a scattering of young bur oak.

When the mound had been cleared of brush and trees, a stake was driven into what appeared to be the centre of the structure. Using this as a reference point, a 10-ft. grid was superimposed on the mound and a contour map was drawn with a one-foot contour interval. The basic grid was oriented so there would be a minimum of stumps and roots on these balks which we knew in advance would be profiled. (In this report, all directional references are grid references; grid N. bears 18-1/2° west of astronomic N., and 25° west of magnetic N.)

The mound was almost circular in outline, with a north-south diameter of 63 ft. and an east-west diameter of 67 ft. It was 6 ft. high, and contained approximately 756 cu. yd. of earth.

We knew from previous experience in the Rainy River District that the mound fill would be largely sterile; that is, we expected the fill to contain some habitation refuse, but we expected that the amount would be relatively small. We decided, therefore, to excavate the mound by shovel. This technique proved satisfactory and was used throughout the excavation, except for the detailed work at the base of the mound and the single intrusive burial that was encountered.

We had planned to leave a two-foot balk standing on the north-south and east-west centre lines, thus dividing the mound into four sectors, or quadrants, for excavation. Because of a technical error, however, the east-west balk was left one foot north of its planned position.

In order to define the archaeological problem as early in the excavation as possible, a line of 10-ft. squares was cut into the mound from its western edge, immediately south of the east-west balk. These squares were excavated to undisturbed subsoil. As we approached the centre of the structure, wood fibres were encountered on the upper surface of the old sod line at the base of the mound. Leaving approximately a foot of earth to cover the wood, we then excavated the upper portions of the trench to the north-south balk. Finally we removed the protective layer of earth and exposed the wood with trowels and brushes. The same procedure was followed in the other three quadrants, thus exposing a crude rectangle of logs which were found to enclose three separate burials (Fig. 4). (Actually, we did not proceed in such an orderly manner, but shifted constantly from one quadrant to another because of the fact that rain water tended to collect in our lower trenches.)

4 Floor plan at the base of the mound. (click on image for a PDF version)

In addition to these logs which were measured and plotted with reasonable accuracy, a thin layer of wood fibres covered the rectangle which was defined by the outer logs. This layer probably resulted from a layer of brush which had been scattered within the rectangle after the logs had been placed on the sod, but before the three burials had been deposited there. Evidently the mound-builders first outlined a rectangular plot near the edge of the upper river terrace, then scattered some brush or twigs within the rectangle. Next they placed three separate burials on top of the thin scattering of brush, and finally constructed the mound. This sequence of events was established by determining that the layer of brush — that is the short wood fibres which were randomly oriented — was superimposed on the logs¹ but was beneath the burials.

A vertical face through the mound (see Fig. 5) shows that the structure was built up as a unit with individual basket-loads of earth. The mottled effect results from the fact that some loads were largely a thick, black humus: some were the light tan clay which is the characteristic subsoil of which the terrace is formed, and others were a mixture of the two. We were able to identify the containers as "baskets" because we found, near the base of the mound, some four feet west of its centre, the clear imprint of birchbark separating a lens of clay from a lens of humus. Almost certainly, the bearer had decided that the basket was no longer serviceable and had simply tossed the container with its contents upside down on the growing heap of earth. The absence of lensing near the upper surface of the mound is probably due to leaching.

5 Vertical profiles through the mound. (click on image for a PDF version)