Alfred Harker — Suhaerial Erosion of Ski/e. 489 



vertical pi-ecipice of 140 feet into a steep-sided winding gorge, which 

 continues for some 300 yards. The fall is caused by an irregular 

 sheet of a fine-grained gabbro, more durable than the prevalent 

 coarse type in the neighbourhood, and the gorge, at least as regards 

 its upper part, results from the cutting back of the falL Here, if 

 anywhere in the district, we should find evidence of active erosion 

 at the present time, and one or two small stone-falls of evidently 

 recent date at the upper end of the gorge suggest at first glance that 

 tiie cutting back may be going on rapidly. A closer examination, 

 however, negatives this supposition, and proves that practically tbe 

 whole of the gorge must have been cut out in pre-Glacial times. 

 The boulder-clay by which it has been filled is for the most part 

 cleared out from the bottom of the gorge, though some relics remain 

 intact, the nearest being about thirty yards from the waterfall. The 

 post-Glacial recession cannot, however, have amounted to anything 

 like thirty yards, as appears from examining the rim of the gorge. 

 The head of it is expanded into a relatively wide cauldron (itself 

 a significant feature), and the drift is found to creep down the steep 

 slopes of this cauldron in such a way as to leave little doubt that it 

 has once been banked up against a face very nearly coincident with 

 the present precipice of the waterfall. 



The other locality alluded to is the Sligachan Eiver, in the neigh- 

 bourhood of the bridge, and its interest is with reference to the 

 transport of material downstream. This river is one of the principal 

 streams of the district, and is also liable to sudden floods in time of 

 rain, rising sometimes five or six feet in a very short time. Like all 

 the other streams, it is perfectly clear, even in the heaviest spates. 

 The idlest lounger on the bridge cannot fail to remark that while 

 the bouldei's in the bed of the river have dark blue and rusty brown 

 colours (basalt and gabbro), those in the tributary stream on the 

 right, which joins the river just below the bridge, are white or of 

 pale tints (granite). Observing further that below the confluence 

 the left bank of the river is dark and the right bank pale, he may 

 perhaps conclude that the main river drains an area of dark rocks 

 and the tributary burn an area in which the pale predominates, 

 A little more examination, however, will compel him to abandon 

 this explanation, for he will find that the pale boulders creep 

 obliquely across the river until they occupy.the whole of the bed, 

 and the broad shingle-flats between and near tide-marks in the 

 estuary below are composed essentially of white and yellow granite. 



The key to the puzzle lies in the fact that here, as throughout 

 the district, the boulders and gravel — coarse or fine — in the bed 

 of a stream have no direct relation with the solid rocks of the 

 drainage-basin, but depend solely on the constitution of the drift 

 deposits in the immediate neighbourhood. Where there has been 

 no drift, there are no pebbles in the burns. Where a burn has had 

 to re-excavate a channel through the drift, it has done so by mei'ely 

 removing the finer material, the enclosed stones, large and small 

 alike, remaining as boulders and gravel in the bed of the stream. 

 They may have travelled downstream for a certain distance, but 



