bevelled slope is on the average 115 inch long, and is so arranged that the porous part of the under-side is taken away, so as to guard against the implement being broken when used. It is singular that in this group of implements the sloping cut surface has not a single line or scratch upon it, but, on the contrary, it is in some cases slightly hollowed. The cross cut here is more of an ellipse. They vary greatly in size and thickness, like those of the first group. The largest implement of this kind is broken off at the point, but it is more than 712 inches long, and very nearly an inch wide; very probably its length originally was twice as much. The smallest lance-heads are about 312 inches long and 14 of an inch thick. There are all sorts of varieties between the larger and the smaller kinds. One singular specimen is 814 inches long and about 910 of an inch wide, and on the surface it is regularly burnt, so that one might almost venture to suppose that it had been used as a spit. An exceptional specimen of this group is the lance-head, drawn Plate VI. fig. 26, which has a narrow furrow running lengthwise ⋅0586 inch dec. wide, and ⋅0868 inch dec. deep, and on the usual bevelled part there is a round perforation rather more than 110 of an inch in diameter, apparently as a means of hanging up the implement; but whether as a trophy for the owner, who possibly may have made some splendid shot with the weapon, or for some other purpose, it is of course difficult to decide. At all events it is a curious and unique specimen; the length of it is about 412 inches. The chisel-shaped bevelled part of all these lance-heads must have been intended to be stuck into a split wooden shaft, which was then bound round by some material, and thus the lance-head was made fast. The whole number of lance-heads found in the excavation was about ninety-three, but of these only about a fifth are in good preservation. The larger proportion are only in fragments, but still they all show quite clearly the chisel-shaped bevelling. It may easily be imagined that the points have suffered most; in fact they are wanting in all the broken specimens. Very singularly two fragments of implements of this group have been found made out of a mammoth's tusk. On some specimens of this class, which unfortunately are not quite perfect, the workman has spent more time than usual, and has tried his skill in drawing. Parallel lines grouped in various modes, sometimes on the sides and sometimes on the top, and also zig-zag lines
