no relics of the nature of grave-goods, though fragments of human skulls, teeth, and bones were seen.

The ground plan of the island (fig. 4) gives a complete survey, and a section across its centre from north to south on the line A B C. For this survey and section I am under obligation to Mr J. H. Cunningham.

The graves marked Ia, I, VI, VII, X, and XI were either laid bare during this storm, or have since been made apparent by the wearing away of the soil of the island under a similar process.

The depth of soil on the island varies from 18 to 24 inches, after piercing which the graves seem to have been scooped out of the soft natural rock to the required size. They are usually floored and lined

VOL. XL.

5

with three or more slabs on each side, and with a single slab at each end. They were also apparently covered with like slabs, and the junctions of the slabs seem to have been filled in with smaller stones on the outside. With two exceptions after mentioned, the general size of the graves points to burial in each of one adult person in an extended position, without a coffin. The graves were entirely filled with soil and gravel, in which respect they resemble the Gladhouse graves, and differ from those of Belhaven. Whether this condition has resulted from natural processes, or was brought about by intention at the time of burial, it is difficult to say. I note that Mr Abercromby, in describing the Gladhouse graves, inclines to the latter view.

From the position of the side stones in one of the graves, No. III (fig. 10), it would almost appear as if the body had first been laid on the prepared floor of the grave, and the sides and ends thereafter put into position, so closely is the general outline of a human figure followed.

The plan (fig. 4) shows (first) all that is left of the natural surface of the island; (second) the denuded gravel and rock of the foreshore, with a few detached pieces of turf-covered soil washed down from the main soil surface, also the high-water level of the reservoir last winter, it never having been full; and (third) the water-mark of the reservoir when the survey was made, the depth of the water in the reservoir being then only some 32 feet-an abnormal condition, brought about by the drought of the past summer, which was of some importance in the work of exploration and excavation.

The section is self-explanatory, and is useful in showing that the graves at high-water are beneath the water-level, a fact which, in the porous condition of the gravel and rock through which the water must freely percolate, may perhaps account for the silted-up condition of the graves, and certainly abundantly accounts for the almost entire absence now of human remains therein.

The plan further shows that the graves are all approximately, though not exactly, in an oriented position; and the dotted lines which I have

put on the plan show that the graves lie in three rows, with possible indications of a fourth to the west.

Grave VII (fig. 11) is a very perfect structure, and in itself proves the existence of the third row, without putting too much reliance upon the washed-out spaces marked VIII and X as indicative of the sites of graves. By excavation at the point marked IX (shown in figs. 2 and 3 by a spade handle), I satisfied myself that there is there another grave in correct alignment with VII and VIII, but the trunks and roots of the surrounding trees prevented more than a small opening, which disclosed covering stones, two or three side stones, and two stones close together on edge in a perpendicular position at right angles to the side. stones, which were suggestive of the head stone of one grave and the foot stone of another. The spot could only be very partially excavated, however, and the growth of the trees had no doubt disarranged the structure. I saw no vestige of human remains, nor anything of the nature of grave-goods; and a photograph could not be obtained.

As to the probability of a fourth row I cannot speak with certainty. The washed-out space marked XI looks like another grave-site. The experimental cuttings made at X and Y were perhaps not carried deep enough to reach any graves which may be there.

An enlargement of part of the survey is given in fig. 5 to show in more detail the structure and measurements of five of the graves, which were carefully excavated and examined, and which are numbered on the plan (fig. 4) Ia, I, II, III, and VII.

The first row in the plans (fig. 4 and fig. 5) is represented by one grave, Ia, which is the lowest of those shown in the view, fig. 6. It is of very small size, and is obviously the place of interment of an infant. The external measurements are length 24 inches, breadth at west end 14 inches, and at east end 12 inches. The internal measurements are - length 21 inches, breadth at west end

9 inches, and at east end 8 inches.

This was apparently the only grave which had not been floored with slabs, the bottom consisting solely of the natural rock. The covering

stone or stones were not in position; and, as will be seen from the photograph, the construction is rude and irregular.

The second row contains five very complete graves and the remains of a sixth, with a space in which a seventh may yet be found between the graves marked IV and V on the plan, fig. 4.

Grave I is shown in the middle distance in fig. 6, the third grave

in the top right-hand corner (marked by an axe) being the fine grave VII of the third row, to be afterwards illustrated.

Grave I, as shown in fig. 7, is of rude construction. Its external measurements are 4' 4" in length, with breadths of 10" at the west end and 14" at the centre and east end. Internally it measures 3' 10" in length by about a foot in breadth. The ends are each composed of a single stone, and there are three slabs on each side and four on the bottom, the two in the centre being of red Cairn-hill sand

stone.

The large slabs lying on each side of the grave may have