III. Glaciation.

The greater portion of the county is occupied by strata belonging to the Old Red Sandstone formation, of which the most prominent subdivision is the well-known Caithness Flagstone series. A line drawn from Ben Rha, near Rcay, south-eastwards by Loch Scye, Loch More in Strathmore, to Morven and the Ord, marks the inland limit of this forma- page 8 tion. The geological structure of this tract has been fully described by Professor Geikie in his elaborate monograph on "The Old Red Sandstone of Western Europe,"^* to which reference may be made for details. Beyond the limit just indicated the strata consist of white quartzites forming the range of the Scarabens (2054 feet), orthoclase gneiss, mica schists, and occasional masses of granite; but to these we paid no special attention.

An important feature connected with the glaciation of Caithness is the complete divergence in the trend of the ice-markings, in the area occupied by the shelly boulder clay, and the district lying to the west towards the county boundary. In the former area the prevalent trend is northwest and south-east, which Mr Jamieson and Dr Croll clearly showed could not have been produced by any local radiation of the ice—a conclusion which is self-evident to any one who considers the physical features of the county. In the district lying to the west of this area towards the county boundary, the strife point E., E.N.E., N.N.E., N., and eventually they veer round to the N.W., along the inland margin of the shelly boulder clay. We shall first describe some of the examples met with in the area occupied by the shelly drift.

In the district between Reay and Thurso we observed numerous examples on the surfaces of the flagstones. By the roadside west of the granite ridge at the burn of Isauld, stria; are seen on the grey flagstones beneath a thin covering of red boulder clay, pointing W. 40° N". These agree with the instances noted by Mr Jamieson at Beay, running N.W. and W. 35° N. At Shebster we observed ice-markings pointing W. 30° N., and at Westfield, near the Forss Water, W. 20° to 25° N.

On the surface of the blue flags in the Achscrabster quarries the trend is W. 25° N. This locality is interesting on account of the evidence which it affords of an ice movement towards the north-west. Subangular blocks have been broken off the flagstones by the ice in its passage over them, which are tilted at a gentle angle towards the north-west, while the flags themselves have been bent over in the same

^* Trans. Roy. Soc., Edin., vol. xxviii., p. 406.

page 9 direction. In the Jamestown quarries, about two miles southwest of Thurso, the prevalent trend is W. 20° N., but on the same rock surface we noted a fainter set running N.E. These have been well-nigh effaced by the north-west movement, and for this reason they seem to be of older date. This was the only example we noted of a north-easterly trend away from the inland margin of the shelly drifts.

Between Thurso and Castletown, and eastwards to Brough, similar evidence is obtained of this north-west and southeast movement. On the surface of the grey flagstones in the Castletown quarries the direction is N. 20° to 23° W., and again by the roadside, south of St John's Loch, the trend is N. 10° to 15° W. Several examples were noted on the slopes between Brough and Dunnet Bays, on the surfaces of the coarse yellow sandstones of that peninsular tract. In the old quarries, on the slope to the north-west of Brough, the striæ, point W. 35° N., W. 30° N., and W. 25° N. A few yards to the west of the above locality another instance was observed pointing W. 25° N. Farther to the south, by the road leading to Dunnet Church, the direction is W. 25° to 30° N., and to the west of St John's Loch W. 35° to 40° N. A careful examination of the striated surfaces on this slope convinced us that they had been produced by ice moving towards the north-west. The gradual rise on the ground to the west of Brough towards Dunnet Head (346 feet) is due to the presence of coarse massive sandstones of Upper Old Red age. On the south-east slope the sandstones are finely moutonnéeM, and the smooth faces point to the south-east, indicating the direction from which the ice came. When we come to discuss the dispersal of the stones in the boulder clay, we shall see that additional evidence is obtained in support of this conclusion.

In the undulating plain between Halkirk and Westerdale the same north-west and south-east trend is observable. In a quarry south of Achies farm-house, about three miles south of Halkirk, the direction is W. 30° N., and alongside of these is a fainter set pointing nearly west.

On the eastern seaboard, between Duncansbay Head and Dunbeath, numerous instances occur which harmonise with page 10 the general trend now described. On the top of the cliff, near Skirsa Head, Mr Jamieson noted striæ pointing W. 25° N.; in the bed of the Freswick burn, N 35° to 40° W.; and at Keiss, N. 35° to 40° W. Numerous examples are met with in the neighbourhood of Wick, varying from N. 15° W. to N.W. One of the best instances occurs on the cliff top opposite the Old Man of Wick, which is specially noteworthy, inasmuch as it presents certain appearances which could only have been produced by ice coming from the southeast. At this locality the flagstones dip inland (N. 30° W.) at a gentle angle, thereby forming tiny escarpments along the outcrops of the successive beds. On the dip slopes the strice point N. 10° to 20° W., but opposite each small escarpment they are deflected, the trend being still more northerly, while in the succeeding dip slope the normal direction is resumed. The slight deflections observable in these small escarpments point to an agent moving inland from the southeast.

A more striking example, in proof of the ice having moved inland from the North Sea, is to be met with in a goe between Dunbeath and Latheron, close by Latheronwheel. On the north side of a narrow inlet, which is about 100 feet in depth, the face of the cliff is finely polished and striated; the striae beginning near the water-level, and ascending the cliff obliquely. It ought to be borne in mind that there has been a large amount of denudation along that rocky coastline since glacial times, which is greatly accelerated by the system of jointing so characteristic of the Caithness flagstones. These cliffs usually present clean-cut faces, owing to the removal of huge slices of rock along the joints by the combined action of the sea and atmospheric influences. Hence it is difficult to find instances of striated surfaces rising from underneath the water-level. Moreover, in the bays at Freswick, Wick, Lybster, Latheronwheel, and Dunbeath, the boulder clay descends to the shore, thus indicating that the streams had cut down to the present sea-level, and probably below it in pre-glacial times.

From the foregoing examples, it is evident that all over the broad flats of Caithness the general trend of the ice- page 11 markings is north-west and south-east; but to the west of the inland limit of the shelly boulder clay the trend is widely different.

Following the shore northwards from the Ord of Caithness towards Berriedale, the stria; run in an easterly direction, as if produced by ice moving off the high ground in the southeastern part of the county. North of Berriedale, however, towards Dunbeath, they gradually swing round and creep inland from the sea, the trend varying from N. 10° to 15° E.

Again, in the valley of Strathmore, between Loch More and Dirlot, the same curious deflection of the ice-markings is observable. By the roadside between Loch More and Strathmore Lodge, several pits have recently been opened, which have brought to light finely striated surfaces. Not far from the north-west corner of the loch, in a quarry on the north side of the road, there is a beautifully-polished surface which has been produced by ice descending the strath. The striæ occur on red sandy flags and grits, pointing E. 20° N. The smooth face of the roche moutonnée confronts the high grounds to the west, leaving no room for doubt as to the direction from which the ice came. On the opposite side of the road well-marked striae are also exposed, pointing E.N.E., and the same trend is observable a short distance to the east of these localities. Again, about a quarter of a mile from the loch, in the direction of Strathmore Lodge, we noted several instances trending N. 5° E. Further down the valley, and a little way east of Strathmore Lodge, ice-markings were found, pointing N.N.E. and N.W., on the same glaciated surface, underneath a thin covering of boulder clay. We shall have occasion to point out in a subsequent paragraph that the shelly boulder clay does not extend further up the valley than Strathmore Lodge. It is apparent, therefore, that the deflection increases as we approach the margin of the shelly drift. It cannot be ascribed to any elevated mass of ground in the neighbourhood of Strathmore Lodge, because no eminences intervene which could possibly give rise to such a phenomenon. On the contrary, it points to the existence of a powerful opposing agent, which must have operated over the whole of the broad Caithness plain.

Again, on the moor between Dalnawillan Lodge and Altnabreac Station well-marked striæ were observed on granite, trending E. 10° to 15° N. This example is in harmony with the striations near Loch More. This locality is situated about four miles from the county boundary, and the markings were clearly produced by ice moving off the adjoining high grounds.

Another traverse from the hills round Loch Scye, eastwards by Loch Shurrery, Ben Dorrery, to Scotscalder, furnishes remarkable proof of the north-easterly trend of the local ice and its gradual deflection near the limit of the shelly drift. Between Achsteenalate and Loch Scye some finely glaciated surfaces have been recently exposed by the roadside. About a mile and a half to the west of Loch Shurrery well-preserved stria; are visible pointing N. 15° E. on a granitic breccia, which here forms the base of the Old Red Sandstone. To the east of this locality and about half a mile west of the same Loch, near Achsteenalate, the trend is N. 15° E., and a similar direction was noted in the bed of the stream flowing into the Loch on the west side. In these instances the roches moutonnéecs indicate a movement towards the N.N.E. Crossing the south shoulder of Ben Dorrery by the road leading to the Dorrery farm-house, several examples were noted by us pointing due N., and in one case N. 20° E. From the manner in which the south slope of this hill has been glaciated, it is evident that the ice-markings were caused by ice moving towards the north. Again, in the long railway cutting west of Scotscalder Station we found several examples pointing N. 5° W., N. 20° W., and N.W. Towards the middle of the cutting on the south side "cross hatches" were observed pointing N. and N.W., while a few yards to the west two instances point towards the N. Now, these "cross hatches," like the instance already quoted at Strathmore Lodge, occur near the inward margin of the shelly drift, and as they are situated in the midst of an undulating plain, no one can for a moment contend that such remarkable proofs of the deflection of the local ice are due to the contour of the ground.

Again, in the extreme north-west of the county similar page 13 evidence is obtained. About two miles south-west of Reay, on the south slope of Ben Rah (795 feet), less than a mile from the county boundary, there are fine examples of glacial abrasion to be seen. The two peaks on this hill are composed of outliers of coarse granitic breccia resting on fine white sandstones, which are beautifully polished on the intervening col and on the south slope. Indeed, the striations are as fresh as if the ice had but recently passed away, and they leave no room for doubt as to the direction from which the ice came. The instances noted by us point N. 10° E., N., and N. 10° W.; and they are situated to the west of the limits of the shelly boulder clay. It is evident, therefore, that the local ice must have been powerful enough to override hills of considerable elevation near the sea-level. Descending the slope to the burn of Isauld the striae swing round to the north-west as already indicated.

From the evidence now adduced it is clear that these two diverging systems point to the existence of two opposing streams of ice. From the billy ground along the county boundary and the heights in the east of Sutherlandshire the local ice flowed E.N.E. and N.N.E. towards the Caithness plain, while near the Ord and at Reay it flowed into the sea. But along the line indicated by the inland margin of the shelly drift, the local ice was compelled to veer round to the N., and eventually to the N.W. in harmony with the general movement all over the great plain. The "cross-hatching" met with along this line indicates a sustained conflict between the opposing streams, which resulted in favour of the north-westerly ice-flow. On both sides of this line the boulder clay is of two distinct types, as will be presently described. Occasionally there is a commingling of the material belonging to the rival ice-streams, but in general the features are totally different. Indeed, the phenomena now referred to are quite analogous to those met with in the central valley of Scotland, so well described by our friend and colleague Dr James Geikie. In this latter case, the great ice-streams from the Highlands and Southern Uplands coalesced in the midland valley, producing similar "cross-batching" and the same commingling of the moraine profonde.

Dr James Geikie reminds us that similar phenomena have been recorded by geologists in Scandinavia, "cross-hatching" having been observed many years ago by Forchammer in Denmark, and in more recent years again and again in Norway and Southern Sweden by various geologists. Quite recently they have been detected also in North Germany by Dr Fenck. "The intercrossing of boulders derived from different quarters,"^* he continues, "has likewise been frequently noted by continental geologists in the drift deposits of various parts of Europe. Perhaps among the most remarkable examples are those described by MM. Falsan and Chantre in their magnificent work on the ancient glaciers of the basin of the Rhone. In the region lying between Bourg and Grenoble (Ain, Savoy, and Isere) the glacial deposits display again and again the most remarkable examples of erratics which have crossed each other in their journey at all angles. For example, in the neighbourhood of St Paul, a little to the west of the beautiful Lac du Bourget, the glacial deposits are charged with blocks of dolomitic limestone, which have descended from Mont du Chat—i.e., in a N.N.W. direction. But in the same district occur many blocks of various rocks which have come from Upper Savoy—i.e., from N.E. Here the one set of boulders has crossed the other nearly at right angles. And numerous other similar examples are pointed out by the French glacialists. Still more striking is the fact that now and again erratics have travelled in precisely opposite directions—one set having been carried up, while another has been brought down, one and the same valley. Thus, in the Val Romey (S.E. of Nantua) erratics of local origin may be traced south as far as the Lyons and Geneva Railway, while boulders derived from the Alps have travelled up the valley for a number of miles! And the same peculiar phenomena are repeated in the case of many other valleys in Dauphiny and the adjoining regions. The origin of this 'intercrossing' of erratics is very simply explained by MM. Falsan and Chantre. They point out that before the great glacier of the Rhone and the Arve had

^* Extract from a letter by Dr J. Geikic, F.R.S., to one of the authors of this paper.

page 15 attained its maximum development, all the mountains of Savoy, Dauphiny, etc., had their local and independent glacier systems, some of which were very considerable. These local glaciers flowed down the valleys, as a matter of course. By-and-by, however, when the united glaciers of the Rhone, the Arve, the Isere, and the Drac, with their aflluents reached their greatest extension, so as to cover all the region between Bourg, Lyons, Vienna, and Grenoble with a vast mer de glace, the formerly independent glacier systems of Dauphiny, etc., were overwhelmed, and their flow arrested, and in many cases actually reversed. In other words, the united mers de glace of the Rhone, the Arve, etc., sometimes overflowed the summit-levels from which the local glaciers had descended, while in other cases they simply dammed back the local ice and protruded long tongues of ice into the lateral valleys formerly occupied by independent glaciers. And thus alpine rocks were often carried in very different directions to the course followed by the débris of the local moraines. But when the great mer de glace declined in importance, the local glacier systems came again into existence, and rocks of local origin travelled clown the valleys as before.

"German geologists have long been familiar with the fact that' intercrossings' of erratics are not uncommon in the so-called Northern Drift; and I may refer you to 'Prehistoric Europe,' pp. 203, 564, and Plate D, where you will find some account of the general results arrived at. The 'intercrossings' of boulders in the drift deposits of Lancashire, Cheshire, etc., so ably described by Mr Macintosh, are, I do not doubt, to be explained in the same way. If those who still cling to the iceberg origin of our boulder drifts can be induced to study MM. Falsan and Chantre's work, they will pause before appealing to the distribution of boulders in the northwest of England in support of the marine theory of the drifts. To me that distribution is eloquent of the successive changes of ice-flow which took place during the gradual increase and decrease of the mer de glace which enveloped that part of England. Long before that mer de glace attained its full development, the glaciers of North Wales and the page 16 Cumbrian Lake Country must have flowed outwards freely in many directions, which they could not afterwards follow when the united mer de glace came to fill up the basin of the Irish Sea and advance inland upon Cheshire, etc. At the period of maximum glaciation the path of the ice would often be at right angles to what it was before that maximum was reached, and to what it again became after the mer de glace was on its final decline."

Mr Jamieson states in his paper that where he observed "any indication of a stoss-seite it was on the north-west side."^* But no instance is quoted save one about two miles south of Berriedale where some masses of conglomerate "crag" to the east, as if produced by ice moving Seawards. This example is situated at the southern margin of the shelly drift, and was evidently caused by the local ice already described.

The absence of any well-marked roches moutonénes in the area occupied by the shelly drift may be satisfactorily accounted for by the peculiar mode of weathering of the Flagstone series. The flagstones were not capable of assuming the dome-shaped contours so characteristic of highly glaciated regions. In many instances they broke up into subangular blocks underneath the ice, a striking example of which has already been described in the Achscrabster quarries. The very same features we found to obtain in Orkney in the tracts occupied by this series. But notwithstanding this mode of weathering, we have adduced several examples which indicate a movement towards the north-west, and when these are viewed in connection with the remarkable deflection of the local ice-stream, it must be admitted that the evidence derived from the striated surfaces and the roches moutonnées is clearly in favour of this conclusion. This view receives additional support from the evidence supplied by the boulder clay.