T. O. Bosworth— Wind Erosion in Mull. 359 
windward end of each little ridge. A few wind-cut pebbles also lay 
upon the sand. . i 
_ Such surfaces as these, which thus at every point bear record of the 
exact direction of the eroding sand drift, can only be formed in places 
where the direction of the wind is constant. Probably this condition is 
here almost perfectly realized; the direction of the air currents being 
determined by the shape of the bay, so that changes in the wind 
outside affect only their intensity and not their direction. The 
bearings quoted here are from measurements on one particular 
hummock ; they vary somewhat from place to place in the bay, and 
in the arm, which branches off to the east, the dunes and corrugations 
are from west to east. 
Higher up among the rocks at the south end of the bay is a slightly 
inclined 9 foot sheet of mica trap which projects out from the rock 
side forming a bench whose foot is reached by the banked-up sand. 
The front of this sill is nearly vertical and is divided up by the joints 
into some four or five layers of rectangular blocks. The blown sand 
strikes obliquely against this wall of rock, and at one place every block 
has had its front top windward corner eroded away (Pl. XXIX, 
Fig. 4). In some cases several small facets have been cut, meeting in 
fairly sharp edges. A few of the facets are furrowed, each minute 
furrow beginning in a ferromagnesian mineral, but generally the corner 
surfaces are instead almost covered with small pittings, and are con- 
spicuous on account of their peculiar colour. Normal weathering gives 
the rock a red-brown rough surface, but these corner surfaces appear 
from a slight distance of a dark-purple plum colour, to which a little 
polish has added the appearance of bloom. The purple effect is due 
to the steel black colour in the pittings. It is removed to some 
extent by acid yielding a solution containing iron. Possibly it is due 
to a thin film of fine magnetite dust imparted by the magnetite grains 
in the blown sand. 
_ Thus wind-blown sand is here ever at work on the rocks in advance 
of the waves, lengthening the bay at its landward end, and so playing 
some small part in the coast erosion. 
EXPLANATION OF PLATES XXVIII AND XXIX. 
Prats XXVIII. 
Fie. 1. Two of the small hummocks, conical in front, with points towards left-hand 
lower corner of picture. Both are about 18 inches high. 
_,, 2. Front of a rock tacing the wind. Lower middle part of picture shows 
projecting knobs of quartz; middle of picture is a more horizontal 
surtace with corrugations ; upper middle portion is a vertical surface 
facing wind, and studded with knobs of quartz. Scale, 6 inches. 
Pratt XXIX. 
_,,.8. Two specimens of corrugated rock surface. Left-hand specimen was 
a a surtace which met wind obliquely ; right-hand specimen an horizontal 
surface. Scale, 4 centimetres. ; , 
~,, 4. The sill of mica trap which is met obliquely by the wind. Top front 
' - windward corner of each block worn away. Scale, 1 foot. 
