as an explanation of the superficial layers of stones a certain 
amount of concentration of the heavier material by water removal 
of fine silt as well as by the subsequent accentuation of the same 
action by rain and wind after the lake waters have gone and the 
beaches have become dry land. 
The “made ground"— to use the miners’ term — quite often 
shows alternations of more sandy and more clayey material as 
would be expected from de])osits in water, and in no case has the 
writer come across any considerable mass of such ground which he 
could regard as even probably derived from accumulations of wind- 
borne dust and sand. The physical character of the superficial soils 
overlying the bed rock seems to him to be quite against the theory 
of wind deposition. 
SAND HILLS AND SAND PLAINS. 
Ill this statement, however, sand-hills and siqierficial sand- 
drifts must be excei)ted. but these also when examined present 
notable features which are of much consequence in arriving at a 
theory of their formation. It may seem the “most imkindest cut 
of all" to the suptiorters of the wind-erosion theory to claim their 
sand hills in supiiort of that of water planation, but it is in the 
position and behaviour of the sandy areas that the writer has found 
what appear to him to be powerful arguments in favour of the 
theory of lacustrine or marine planation as against that of wind- 
erosion. Sand-hills and sand-patches are found very commonly 
all over the country at all levels, but it is very noticeable that the 
biggest and best formed sand-hills, exce])t those close to the coast, 
are usually, so far as the writer's own experience goes, on high land 
surrounding the lake plains. One of the most extensive sandy 
areas encountered on the goldfields, is crossed on the road from 
Gabanintha to Wiluna, a short distance east of the rabbit-proof 
fence. Here the ridge between the Lake Annean to Lake Austin 
lake i)lain and that surrounding Lake Way and Lake Violet rises to 
a considerable height above these plains, and the sand is very 
heavy to traverse, ft is formed often into well-marked sand dunes 
of typical shapes, now covered with spinifex and small scrub, and 
rarely shows signs of having been in motion as drifting dunes for 
a very long time. 'Lhe dunes are practically fixed by the vegeta- 
tion, and wind appears to have Init little effect on them. The 
sand-ridges are found southward to Birrigrin and Barramhie, being 
very difficult to traverse between these two centres, and the rabbit- 
proof fence from Barramhie to Gum Creek passes through them 
for many miles. Rverywhere they are on the high ridge on the 
edges of the lake plains, and are practically fixed in position by a 
coating of vegetation. 
The Barr Smith Range, passed over on the old road between 
Kathleen Valley and Cork-tree well, on the way to Wiluna. is 
another very characteristic sand-plain, with only very small dunes, 
now well fixed by vegetation. It is high ground on a ridge between 
