knoll, it was found that those which were uppermost must have come from the N. W., otherwise they could not have got into the positions they occupy.

There were no boulders near the top of the knoll on the S.E. side; but at the base of the knoll on that side, several boulders were lying, which might have fallen from the top. They were not heaped on one another, as they were at the top of the knoll, but lying separate.

2. About 200 yards to the N.E. of the big boulder there is a boulder on smoothed rock which dips due north at an angle of 20°. The size of the boulder is 5 x 4 x 4 feet. The steepness of the rock surface on which it lies, is so great, that it would have a better chance of obtaining and retaining its position by coming from the north, than from any other quarter.

3. About 300 yards to the S.E. of the "big boulder" there is a boulder 8 × 6 × 5 feet, at a height of about 228 feet above the sea, shown on fig. 10. The boulder at its east end presses closely on a rock, which has prevented it moving further in an easterly direction.

4. On the N.W. side of Ben Erival, where its sides slope down steeply to the sea, there are numerous boulders, and many of them pressing in like manner against the rocks of the hill, in such a way as to show that they must have come from some point between west and north. They are at various heights, from 400 to 500 feet above the sea.

5. There is a low hill to the N.N.W. of Ben Erival, adjoining Traigh Vore," or Great Strand (a narrow neck of sand which here separates the east and west shores), through part of which an open fissure in the solid rocks runs for some distance. It has evidently been one of those rents alluded to by Macculloch in his Account of Barra, which had once been filled by trap, but "of which the exposed portions have been washed out." (Vol. i. p. 89.)

The height above the sea-level is about 120 feet.

For about 80 yards, this rent or fissure now presents two vertical walls of gneiss, from 11 to 12 yards apart, and from 8 to 14 feet high.

The direction of the rent is (by compass) N.W. and S.E. The rocks on the north wall are rounded, and in many places present

smoothed surfaces. The rocks on the south wall are rough and jagged. The appearances on the north walls can be naturally accounted for by the action of a strong sea current moving from W.N.W., which would, with any bodies floating in or swept along by it, grate against the north, but not against the south wall. (See fig. 11.)

6. Ben More is a hill on the farm of Eoligarry tenanted by Dr. MacGillivray. Its west end forms a steepish sea cliff, rising up to a height of 330 feet above the sea. Half way up this sea cliff, there is a boulder, 20 x 10 x 5 feet, resting on the rocky surface, which here dips towards the W.S.W. But the rock, judging by the marks on it, has been smoothed by something passing over it from the N. W., and the boulder is blocked at its S.E. end by a vertical portion of the hill, as shown on fig. 12.

7. At Castle Bay, which is at the south end of Barra, the hills are seen to be more covered with boulders on their N.W. sides than on any other. This observation, however, was made only from the steamboat.

Mr J. F. Campbell, in his paper on the "Glacial Phenomena of the Hebrides," states that, in Sept. 1871, he took rubbings of stria at Castle Bay, showing that the striating agent had come from N. by W. (magn.)

He mentions also that on the small island of Bernera, above 12 miles to the south of Barra, "the last of the Hebrides," he got striæ at a height of 720 feet above the sea, crossing the strike of the rock, from N.N.W. (Lond. Geol. Soc. Journal, vol. xxix.)

In coasting along the east shore of Barra it is perceivable, from the deck of the steamboat, that the rocks on the sea cliffs which face the N.W. have been smoothed, whilst the rocks facing the east are rough and jagged.

8. On the hill called Scurrival, whose west side rises abruptly up from the sea to a height of about 240 feet, the hard gneiss rocks present many proofs of grinding, and also of transporting agency from the N.W.

The rock-strata here are tolerably horizontal and form blocks lying about north and south. The vertical sides facing the sea present frequent smoothings, which could have been made by the action of a strong N.W. current, especially if loaded with

ice.

(See fig. 13.) The surfaces facing the east present no smoothings.

The examples are numerous on this hill of boulders blocked on their S.E. ends or sides. They are cases exactly similar to that shown on fig. 12. These boulders are within 200 yards of the open ocean, and less than 100 feet above its level. The situation and position of these boulders combine to show that they must have come from the westward-though in that direction there is only the wide Atlantic.

At the very top of the hill, which consists of well rounded and smoothed surfaces of gneiss, numerous boulders lie scattered-most of them on that part of the top facing W.N.W.

VI.-ISLAND OF SOUTH UIST.

1. Beginning near the south end, notice has to be taken of a well striated gneiss rock, recently exposed by the removal of materials for the high road. The spot is on the east bank of Loch Dunkellie and at the west side of a hill called Carshavaule, which is marked on the Admiralty map as 226 feet high. The striated rock is only about 20 feet above the sea-level.

The rock had been covered by a bed of coarse sand intermixed with clay, so that its surface had been protected from the weather. The protecting cover contained numerous pebbles, hard and angular, the pressure of which on the rock, if they passed over it, would probably cause striæ.

The rock consists of strata which dip W.S.W. at an angle of about 10°. They were thus conveniently situated for being struck and pressed on by any striating agent from the west.

The lengths of the blocks rounded and striated were respectively, 4, 7, and 5 feet.

The striæ run in a direction N.W. by N. and slope up towards S.E. by S.

If these striæ were caused by rough stones carried in a strong current flowing from the N.W., or pushed by floating ice, the striæ would slope upward in the above direction, because the current would in this low lying spot have to rise, to pass through a valley situated close at hand, immediately to the south of Carshavaule hill. Mr J. F. Campbell in his paper (before referred to) states that in

a quarry by the roadside of Boisdale in South Uist, he observed "striæ running from N. 40° W. (magn.) pointing at a gap in the hills." This is probably the same spot as that noticed by the Convener. It was shown to him by Mr Drever, factor to Mrs Gordon of Cluny.

2. Loch Boisdale, a sea loch, is situated on the east coast. On the north side of the loch, there is a hill called Kennet, reaching to a height of about 890 feet.

The rocks on its N.W. side, from bottom to top, present numerous examples of flattened and rounded surfaces. The surfaces facing the S.E. on all sides of the hill are rough and angular. On the west side of the hill, at various levels between the bottom and the top, there are numerous boulders, some of them, by the way in which they lie, affording unmistakable evidence of the direction from which they came.

(Fig. No. 14 shows boulders there is a

For example, there are two boulders on a narrow shelf of rock which slopes down S. W. at an angle of 40°. The shelf is 96 feet above the sea, and quite close to the sea. The shelf is on the sea cliff, which is so steep, that the wonder is, how the boulders could have found a cleft in it to hold them. these boulders.) On the east side of the projecting ledge, against which the eastmost boulder (A) presses, and which had stopped its farther progress eastward. Another boulder (B) lies upon (A), and which, to get on the top of (A), must have come from some westerly point,-probably the N.W. A line through the chief points of contact and the centres of bulk runs in a direction N.N.W. A study of the boulders on the spot showed that, if they had been brought to this site from any other direction, they would inevitably have slid down the steep rocky bank into the sea. These blocks are nearly equal in size, viz., about 5 × 3 × 2 feet.

Fig. 15 shows a large boulder of coarse granite resting on a wedge of gneiss rock. The wedge or knob is under the boulder at its east end, and tilts up the boul ler slightly so as to show daylight under the boulder at that end. It rests on the ground chiefly at its west end. By this wedge (a in the figure) the boulder has evidently been stopped in its progress from the N.W. From its rounded shape, one might infer that the boulder had been rolled or

VOL X.

R

pushed for some distance before it was stopped. The west side is much rounder and smoother than any other side; so, probably after it had stuck, the current which brought it, beat and chafed on its west side, and smoothed it. This boulder lies on a level plateau of rock about 202 feet above the sea. It is all open country towards the N.W. and N.E., whilst the Kennet hill, reaching to a height of 890 feet, is within half-a-mile of the boulder to the S.E. and E.S.E.

On the west slope of this hill, at a height of 300 feet above the sea, the gneiss presents a rocky surface sloping down towards the west at an angle of about 10°. A boulder of coarse granite, 7 × 6 × 4 feet, rests partly on it and on another smaller boulder underneath. This boulder, at its S.E. end, abuts against the rock. It has come, therefore, almost certainly, from some north-westerly point and stuck there. (Fig. 16 represents this case.)

Not far from the top of the hill, viz., at 712 feet above the sea-level, there is a very large angular boulder on a flat ledge of rock, on the N.W. side, with open country in that direction. This boulder is 19 x 13 x 8 feet. Its further progress eastward has evidently been stopped by a projecting cliff of the hill on its south-east side, as shown in fig. 17.

3. Several large boulders may be seen at a small village, where the Free Church and Roman Catholic Church are situated at a junction of the roads from Barra and Loch Boisdale, about two miles to the south of Askernish. There is here a whole cluster of boulders. One, 16 x 6 x 5 feet, leans slanting upon the others, and must have come from the N. W. to attain its position.

4. On the hill to the east of Askernish, and on its side facing the west, there is a surface of rock, sloping down W.S.W. at an angle of 30°, well smoothed. A boulder rests on this slope, partly on the surface of the rock and partly on some smaller boulders which lie between the rock and it, near its S. E. end. The boulder has evidently obtained its position by coming from the N.W.

This is more clearly proved by a number of ruts or striæ, visible on the rock a few feet below the boulder, which run, as shown on fig. 18, by the arrows, in a direction from N.W. to S.E. That the striating agent first struck the rock from the N. W., is made evident. by the circumstance that most of the striæ are deeper and wider at