On the edge of the shrubby woods the sand had the appearance of an inundation which was overwhelming them, terminating in an abrupt bank many feet higher than the surface on which they stood, and having partially buried the outside trees. The moving sand-hills of England, called Dunes or Downs, to which these have been likened, are either formed of sand cast up by the sea, or of sand taken from the land itself in the first place by the wind, and driven still farther inward. It is here a tide of sand impelled by waves and wind, slowly flowing from the sea toward the town. The northeast winds are said to be the strongest, but the northwest to move most sand, because they are the driest. On the shore of the Bay of Biscay many villages were formerly destroyed in this way. Some of the ridges of beach-grass which we saw were planted by government many years ago, to preserve the harbor of Provincetown and the extremity of the Cape. I talked with some who had been employed in the planting. In the “Description of the Eastern Coast,” which I have already referred to, it is said: “Beach-grass during the spring and summer grows about two feet and a half. If surrounded by naked beach, the storms of autumn and winter heap up the sand on all sides, and cause it to rise nearly to the top of the plant. In the ensuing spring the grass mounts anew; is again covered with sand in the winter; and thus a hill or ridge continues to ascend as long as there is a sufficient base to support it, or till the circumscribing sand, being also covered with beach-grass, will no longer yield to the force of the winds.” Sand-hills formed in this way are sometimes one hundred feet high and of every variety of form, like snow-drifts, or Arab tents, and are continually shifting. The grass roots itself very firmly. When I endeavored to pull it up, it usually broke off ten inches or a foot below the surface, at what had been the surface the year before, as appeared by the numerous offshoots there, it being a straight, hard, round shoot, showing by its length how much the sand had accumulated the last year; and sometimes the dead stubs of a previous season were pulled up with it from still deeper in the sand, with their own more decayed shoot attached,—so that the age of a sand-hill, and its rate of increase for several years, is pretty accurately recorded in this way. [Illustration: The sand dunes drifting in upon the trees]