•244 Mil ALFRED KAIIKEK ON 



in the drift on the right side of the valley and gabbro boulders on the left. A line 

 drawn down the valley through places where the two rocks are equally represented 

 is what we have styled the mid- stream line of that branch of the ice which took this 

 direction. This line can be drawn with considerable precision, since the relative pro- 

 portions of the two rocks vary rather rapidly as we cross the floor of the valley. Thus, 

 at or near Loch Dubh, the percentages of granite and gabbro are 76 and 15 at 70 

 yards to the right (i.e., E.) of the line, 86 and 4 at 250 yards, and 98 and 2 at 450 

 yards; while to the left (or W.) of the line there is an equally rapid change in the 

 opposite sense. The 100 per cent, is made up by boulders of basalt (including dolerite) 

 and, at this place, volcanic agglomerate. We have not always found it possible to 

 discriminate with certainty between the basalt lavas and the basalt and dolerite dykes 

 and sheets, but at this place the latter must supply the chief contribution. 



Proceeding down the glen, we find that the line can be traced throughout with 

 sufficient accuracy. Near the outfall of Allt Coire Riabhach, for instance, the per- 

 centages of granite and gabbro are 55 and 36 to the right and 7 and 67 to the left, 

 the two spots being only 40 yards apart. Here, as in other places where it is most 

 sharply defined, the line coincides exactly with the junction of granite and gabbro in 

 place. Indeed, there are many circumstances which seem to indicate that the 

 immediately subjacent rocks have contributed an important part of the boulders. 

 This becomes very clear when, about 1^ mile above Sligachan, we come on to the 

 basaltic lavas. At once the proportion of those miscellaneous basaltic boulders which 

 we have grouped together as the third element begins to increase rapidly, and 700 or 

 800 yards lower down they already make up 65 to 80 per cent, of the whole, instead 

 of 10 per cent, or less. It is clear, too, that the bulk of these basalt boulders are of 

 the lava type, and they must indisputably have been torn from the floor of this broad 

 open strath. 



From Sligachan bridge, being now clear of the mountains, the line curves away 

 in accordance with the general westward deflection already considered, sweeping round 

 b}^ Loch M5r na Caiplaich into the Drynoch valley, where for some distance it runs 

 very near the D un vegan highroad. The percentage of basaltic boulders has now risen 

 to 90 or 95 or even 98 or 99, mostly amygdaloidal lavas, so that it becomes necessary 

 to examine a larger number of boulders in order to determine the proportions of granite 

 and gabbro with sufficient exactness.* Before reaching Drynoch our line takes a curve 

 to the south-west, rising to about 500 feet and coming down to the valley of the 

 Vikisgill Burn. Crossing this, it passes southward up a tributary valley, Allt na 

 Qreadha, over a third watershed (about 550 feet) into the valley of Allt nam Fitheach, 

 and so south-west to the sea at Loch Eynort, having traversed a semicircle from 

 Sligachan. 



To the left of the line as thus traced, gabbro boulders are in force, while those of 

 granite become rare ; though a few of the latter are still found as much as 1\ miles 



* Farther west the basalt has often been discarded and only the granite and gabbro boulders counted. 



