6 ; 
tinuously. It is slowly rising all the time, but at a grade imper- 
ceptible in travelling. In 81.) miles from Port Hedland to Coongan, 
the rise is ■!<.)() feet, or an average grade of only 1 in IJoT. or 4^ 
feet per mile. As we go inland the isolated ridges and small peaks 
of granite become larger and more common, but the plain runs 
continuously round them as before, until it strikes against the sides 
of the main island-like mass of hills lying between Marble Bar and 
Lalla Rookh. Many of the more prominent granite hills rise quite 
steeply from the plain, often with a talus of large rocks at their 
base, exactly resembling graiute islands off the coast. From a high 
hill the general appearance of the plain being a sea and the hills 
islands rising from it is most striking. The plain sweeps away to 
the north of the main island, well to the east of the Coongan 
River, and through the Dooleiia Gorge an easy way opens into the 
hills and enables us to reach Marble Bar (railway level COd feet). 
To the south of the island the jdain sweet)S round continuously, 
nroken by projecting ridges running out from the main mass of 
hills, by separate islands like the Wodgina Range, and by high 
black doleritic ridges, to Corunna Downs, and v\ell on towards 
Xullagine. Here it must be fully “<I0 feet above sea level, probably 
more, but it is practically continuous right from the sea beach up 
to this level. I'or the most part the bed rock appears to be quite 
close to the surface, though doubtless there are many places where 
the stqterficial material is of considerable depth — -usually, how- 
ever, much less than BIO feet, as is shown by the numerous wells 
— and we therefore see that the whole ol this huge peneplain may 
be taken as cut fairly evenly from the solid granite rock. It is 
very nodceable also that where the rock is bare at surface, and in 
the outstanding 'slands and ridges, the granite generally is very 
little weathered, hut on the contrary is fresh and hard. In this 
it resembles the wave-swept granites on the coast, and numerous 
granite bosses on the eastern goldfields, hut this condition is diffi- 
cult to reconcile with any theory of purely snbaerial erosion, 
whether by rain and river action, or by wind erosion — both of 
which require softening of the rock by weathering before there is 
much perceptible removal of material, d'hc features are exactly 
those which would result if the country were to subside gradually 
until the sea reached the vicinity of Xlullagine. and then rose again 
to its present level The ]jlain is practically continuous from the 
font of the high land inland to well out to sea from the coast, the 
sea bottom continuing to shelve seawards iii continuation of the 
plain. At Kicoll Bay near Roebourne, there is an inland extension 
of the beach which is said to he only very rarely covered by the 
sea. and which shows as a bare mud flat of great extent sloping 
very gently seaward. At its inland margin there are some small 
sandhills and then the mud flat continues iulatid on much the same 
grade. The sandhills at the head of the bay, and the small 
marginal cliffs where the beach sets in against the higher 
