394 ' A. Wade—Formation of ‘Dreikante’. 
a fossil earthquake because the jar the explosion must have given to 
the earth’s crust must have produced a quite appreciable earthquake 
at the time of its occurrence, but in the present instance it is 
interesting as an example where a perfectly circular vent may be 
drilled. through the rocks by means of the eruption of volcanic 
material without a fracture, but this only occurs where the molten 
rocks come sufficiently near the surface of the earth’s crust to allow 
the explosion of gases. The Cradock fossil earthquake is in direct 
connexion with a fissure filled with dolerite which was once molten, so it 
seems to me probable that the Herdubreid volcano likewise is con- 
nected with a volcanic fissure rather than that it should have pierced 
the whole thickness of the solid earth’s crust down to where some 
reservoir of molten material might have existed. 
I1V.—On tHe Formation oF DREIKANTE IN Desert REGIONS. 
By A. Wave, B.S&c., A.R.C.S., F.G.S. 
(PLATES XXXI AND XXXII.) 
HEN one considers the part Britons have played in exploring 
and investigating the various desert regions of the world, one 
is surprised at the relatively small amount of literature that there is 
in the English language dealing with the phenomena which are 
peculiar to desert conditions. Many hazy, and even erroneous, ideas 
are still held with regard to the processes at work and the effects 
produced by the various agents of denudation. 
The three-edged stones known as ‘dreikante’, which are so 
characteristic and abundant in most desert regions, have been 
subjected to little careful investigation whilst in actual process of 
formation, and many ideas which appear to be incorrect are current 
with regard to their mode of origin. 
Originally it was thought that they had been cut by river action, 
or by the movements of sand in the beds of rivers. Steenstrup 
proved that this was impossible,! and it is now known that they are 
essentially the products of wind-blown sand. In shape they usually 
resemble very closely a Brazil nut, and it is commonly thought that 
this characteristic form is produced as follows. 
Wine, il, 
The travelling sand-grains strike a stone or a pebble lying on the 
desert. They are in consequence divided into two streams which pass 
along the sides of the object, wearing away the sides and producing 
a pointed snout to the pebble and a sharp ridge along the crest. 
(See Text-fig. 1.) It is difficult to see, however, why the ridge 
should remain equally sharp from front to back, and why the end 
1 Geol. Foren. Stockholm, x, p. 485 ; xiv, p. 498. 
