a

with the aid of a camera lucida, in 1827, has a claim for accuracy by no means to be disregarded. In the use of the camera lucida the draftsman sees the landscape as though faintly pictured on a sheet of paper, and at the same time sees the pencil with which he traces its outlines. Before photography this method was the most accurate known for recording the outlines of a landscape, and in skillful hands it gives results of notable precision, There is much internal evidence that Captain Hall's sketches at Niagara were made with care and fidelity, and in view of these facts I have thought it worth while to endeavor to combine his record with the records by mapping. He tells us that his principal sketch of the Horseshoe Fall (Pl. III) was made from the upper veranda of Forsyth's Inn, on the Canadian shore, and the relation of the veranda to the inn is shown by a contemporary drawing by Mrs. Trollope. The inn itself long since disappeared, but its position is still marked by the ruins of its foundations. Through the courtesy of Mr. James Wilson, Superintendpent of Victoria Park, who caused the necessary scaffolding to be constructed, I was enabled, in 1895, to place a photographic camera within a few feet of the position once occupied by the camera lucida, and this position has also been located on the map of 1905 (Pl. II, p. 47). A comparison of the two pictures made from that position yielded the identification of a common point on one of the Three Sister Islands (S, Pl. III), and with the aid of the orientation thus secured it became possible to draw upon the modern map the line XY in fig. 4, representing a direction from Captain Hall's point of view tangent to the head of the Horseshoe Fall. It will be observed that this line passes very near to the head of the curve as drawn in the map of 1842, the implication being that at the head of the gorge very little recession had occurred in the fifteen years intervening between 1827 and 1842. I am not sure that this single line, obtained by so circuitous a method, should be allowed to influence the result based on two topographic surveys, but to whatever extent it is given consideration its tendency is to reduce the estimate of the annual rate.

a Trollope, Mrs., Domestic manners of the Americans, vol. 2, London, 1832, frontispiece.