272 



ME. J. S. GARDNER ON THE LEAF-BEDS 



We propose to describe the more important outcrop first, then 

 those on the coasts of Carsaig and Burgh. 



The sedimentary rocks at Ardtun, of which the leaf-beds form a part, 

 appear between basalts along the seaward face of the headland for a 

 distance of somewhat over a mile, and their preservation is obviously 

 due to the circumstance that they have been entirely sealed up by 

 great overflows of trap (see fig. 1). They dip under the sea on the 

 west or Loch-na-Lathaich side, as well as to the eastward, but it is 

 probable that the width of the headland corresponds approximately 

 with their original limits ; for though the proper horizon reappears 

 again in the next headland, half a mile distant, no trace whatever 

 of them is visible there. Patches of pisolitic iron and bole are seen 

 here and there among the basalts along the shore up Loch Scridain, 

 but on higher horizons. They rest upon a mass of basalt about 

 80 feet thick, the upper half of which is amorphous and vesicular, 

 while the base exhibits the most beautifully formed and, for the most 

 part, slender columns. These are, in places, curved in every direction, 

 even lying almost parallel to the bedding *, and closely resemble 

 those of the "clam-shell" cave at StafFa, about seven miles distant. 

 This columnar trap is riddled with caves, which, though far from 

 rivalling Fingal's Cave, are still of great beauty and interest, the 

 resemblance between their masses being so complete as to render it 

 probable, as already inferred by the Duke of Argyll, that they 

 actually formed part of a single flow. Above the leaf-beds and 

 gravels is a second mass of trap, some 50 feet thick, and rudely 

 columnar in structure, forming a vertical cliff. This flow has appa- 

 rently been completely denuded off Staffa, but is represented at 

 Burgh Head as described further on. Above this, again, on the crest 

 of the headland, are fragments of a third flow of a similar kind, 

 neither of them being scoriaceous or amygdaloidal, or showing any 

 tendency to decompose. This so far simple stratification of the head 

 is complicated by the intrusion of a sheet of very dense trap (fig. 1, «), 

 which penetrates it at the sea-level on its east side, and after forming 

 extensive but deeply indented horizontal plateaux a little above the 

 sea-level, forces a devious course upward through the columnar and 

 other basalts and the sedimentary beds, and becomes, owing to the 

 extensive denudation it has been subjected to, exposed at the surface 

 near the front of the head. It dips again, however, almost imme- 

 diately, passing in a sinuous course downward into the lower basalt, 

 and, after forming a few small promontories, finally disappears under 

 the sea at the western side of the head. It is just possible, but not 

 probable, that the ravine in which the leaf-beds are exposed may be 

 the site of a feeder of this subterranean flow. The intrusive sheet 

 is of perfectly uniform thickness, and shows a starch-like weather- 

 ing on its exposed face. The lines separating it from the basalts 

 into which it is intruded are perfectly sharp, and its clean and resist- 

 less, but devious, passage alike through every quality of rock resem- 



* Macculloch, ' Western Isles of Scotland,' vol. i. p. 49G, mentions the 

 occurrence of columns parallel to the bedding, and the rule that columnar struc- 

 ture is developed at right angles to it does not apply in these cases. 



