extensive valleys. On the north end of the east side of the sound is Loch nan Uamh, and at the south end Loch Ailort.

At the mouth of Loch nan Uamh, on the north side, stands the vitrified fort of Ard Ghaunsgail, on a small headland from which the valley at the head of the loch is open.

At the mouth of Loch Ailort, on a small island, stands the double fort of Eilean nan Gobhar. This island is close to the shore on the south. These two forts are in view of each other; but, if they were to some extent to be used for sentinels, it is evident that they could give little notice of a fleet from the north, as the north side of the sound is covered by the projecting headland which ends in Rhu Arisaig. But, on making close inquiry, I found that there were vitrified remains on an island near the north-west angle of the sound called Eilean na Ghoil. There is on this island a fine example of the raised beaches so common on the west coast. On the top are remains of a fort with sporadic vitrification, and on the low beach other traces of vitrification. The places in both cases are more or less triangular. From this point the forts at Ard Ghaunsgail and Eilean nan Gobhar are visible, but there is a very limited seaview to the north. Further inquiry was made, and an islet called Eilean Port na Muirach was ultimately found with vitrification on the top, whence a fine open view was got of the sea from Ardnamurchan to the Sound of Sleat. Thus, then, there could have been telegraphic communication by smoke or fire to the valleys at the mouths of the Lochs nan Uamh and Ailort of a ship rounding Ardnamurchan Head from the south, or coming down the Sound of Sleat from the north. Now, while on no system of defence would it have been reasonable to isolate a small garrison on Eilean na Ghoil, and while in the case of Eilean Port na Muirach there was no space for a fort or garrison, these islands provided a perfect system for warning the inhabitants of the valleys round the Sound of Arisaig and its eastern lochs of the approach of an enemy, and this, I suggest, was their purpose.

The group of forts just described is a very small one, and I think

that progress in the investigation of the questions connected with vitrified forts would be much facilitated if the interdependence and correlation of the forts were more fully studied. For instance, a good deal has been somewhat loosely written about the extent to which the known vitrified forts along the great line of lochs between Inverness and Fort William are visible one from the other.

Fig. 3. Map showing Vitrified Forts in Arisaig and along the Caledonian Valley.

In trying to find out what the line of communication was-if, as is probable, there was one-I was driven to examine the promontory on which Castle Urquhart now stands. That point seemed to me the necessary site of a vitrified fort, if the assumptions on which I was proceeding were correct. It was gratifying, therefore, on a pretty close search being made, to find among the débris on the shore several water. worn pieces of vitrification which had formed part of a fort which must have been removed when the now ruined castle was built.

This discovery permits of the following suggestion being made as to the line of military signalling and communication from the north-east along the line of the lochs.

Starting from Craig Phadric, which commands a far view of the Moray Firth, communication could be carried to Dunreachie Fort on Ashie Moor. The stones of which Dunreachie is formed are of a very refractory nature, as the district belongs to the Old Red Sandstone, and no vitrification appears there.

From Dunreachie the fort at Castle Urquhart was visible, and would be now, were it not for the tall trees in Erchite Wood.

From Castle Urquhart, Dunjardel is well in view, and there is a clear though long line to Tor Duin, near Fort Augustus. It may be that there was another fort somewhere near Glen Moriston. But the discovery of that and of the communication south-west from Tor Duin remain for further investigation.

II.

NOTES ON A HOARD OF ELEVEN STONE KNIVES FOUND IN
SHETLAND. BY ROBERT MUNRO, M.D., LL.D.

At the request of their owner, R. C. Haldane, Esq., seven of these knives are now presented to the National Museum.

The following letter from Mr Haldane sufficiently explains the reasons for which these interesting objects were put into my hands, as well as the conditions on which they henceforth become national property.

LOCHEND, OLLABERRY, SHETLAND, 24th March 1905.

MY DEAR DR MUNRO,-1 send to-day, by parcel post, seven scrapers. They were found at Esheness, Northmavine, in making a road, 9 inches deep in a gravelly subsoil. They were packed closely together with the edges uppermost. There were eleven in all, but several were broken, and I bought the best. The other four I did not see, and do not know what became of them. Nothing else was with them, and they appear to have been a store which had not been disturbed. There was no trace of any building near, but the Brough of Priesthoulland was about half a mile distant. Before these were found the superincumbent peat had been removed. The finder thought a total depth of 4 feet of peat and soil had covered the scrapers. There are said to have been two or three Picts' houses not far away, one at a place near the churchyard called Saebrig and one at Hogaland. They were found in the year 1900.

When you are done with them, if they are of sufficient interest, please present them to the Museum from me. If they do not care to have them, I will keep them. Should they keep them, I would like them all kept together. I am afraid I can give you no more information about them, and must leave it to you to spin out their history.--- Yours sincerely, R. C. HALDANE.

A mere glance at these objects shows that they possess certain characteristics which place them in a special category among ancient stone implements. They are large thin blades made of volcanic rock. known as rock-porphyry, irregularly oval or subquadrangular in form, and highly polished on both surfaces, with the margin all round ground to what may be called a cutting edge.

With the assistance of Mr B. N. Peach, LL.D., F.R.S., whose knowledge of the geology of Shetland is unrivalled, I have drawn up the following descriptive details of each specimen in this hoard, so as to

make them available for comparison with similar discoveries elsewhere recorded :

No. 1. The rock of which this implement is made is quartz-porphyry, and shows double pyramids of quartz and porphyritic crystals of orthoclase in a crypto-crystalline ground-mass. It is subquadrangular in form, with a portion broken off at one end, and measures 5 by 5 inches.

No. 2. This specimen (fig. 1) has the same composition as No. 1, but in addition shows platy flow-structure oblique to the flat surfaces of the

(4.)

implement. Its shape is oval, with a good cutting edge all round, and it measures 6 by 4 inches.

No. 3. All the remarks made on the mineral structure of the two former are applicable to this specimen, with the addition that some layers are spherulitic. Oval in form, with one end nearly straight. diameters are 53 and 4 inches.

Its

No. 4. Made of very fine-grained quartz-porphyry with few porphyritic elements, suggesting that it came from the chilled outer edge of the volcanic mass. This implement (fig. 2) is irregularly quadrangular, three sides being nearly straight, and the fourth curved out