i905-i9°6.] Formation and Flora of a Shingle Island. 295 



to account for the presence of the sand-drifts. First, that the 

 river, by the time it reaches this part, has already dropped all 

 the heavier shingle on the eastern bank, and has nothing left 

 but small shingle and sand to deposit here. Some colour is 

 given to this theory by the fact that at the junction of the 

 two parts a good deal of small shingle is deposited ; and also 

 by the fact, already mentioned, that the eastern bank is 

 receiving fresh accumulations of large shingle. On the other 

 hand, why should not the sow^A-western bank be sandy too ? 

 His other suggestion is that, as the north-western part is 

 some feet lower than the eastern, the water in flood-time lies 

 comparatively deep and still over it, and can therefore deposit 

 the sand it is carrying ; whereas in the shallower parts there 

 is a stronger current, and the sand is swept along. Perhaps 

 the north-western part gets the benefit of an eddy, in which 

 the south-western part does not share. But the river is 

 constantly shifting its course and its currents, and it is not 

 easy to form a definite opinion on the subject. 



This overgrown north-western part is much less regular 

 than the rest of the island, being full of large clumps of 

 vegetation, deep holes, scattered sand-drifts, and depressions. 

 On rushing over a clump of vegetation in flood-time, the river 

 pours down beyond it with added force, and so scours out a 

 hole immediately below. It is perhaps twenty years ago that 

 plants hegan to grow here, but only within the last five or 

 six years that the vegetation has become noticeable. 



All these islands — Eilean a Phortaire proper, its western 

 continuation, and the Shingle Island — reach a maximum 

 height, westwards, of 6 feet, which is the same level as that 

 of the adjoining fields. 



Floods and fiood-drift. — In summer the Shingle Island is 

 left high and dry, and one can walk to it dryshod from any 

 part of the adjoining river-bank. In fact, owing to fresh 

 shingle having lately been piled up near the breakwater, it is 

 now (September 1906) less of an "island" than when we first 

 visited it in the summer of 1905. In rainy weather, however, 

 it is isolated by a good-sized stream ; and it is subject to 

 frequent floods, which occur chiefly in autumn and spring. 

 Sometimes it is completely submerged, together with the 



