70 
of this sort, two where it runs across a pair of high hard ridges 
of ga1)hro. and a very long one where it goes right through the 
heart of the high hilly country lying l)etween Marble Bar and 
Lalla Rookh, Again, close to Marl)le liar the Coongan River is 
found cutting its way through the liills quite similarly, 
emerging on the ])lain country through the picturesque Doolena 
Gorge as a wide tiat-bedded channel walled by precipitous rocky 
sides. It is difficult to find a consistent theory of the erosion of 
these gaps and gorges, occurring as they do on what would appear 
to be naturally the lines of greatest rather than of least resistance, 
and of the iirevalence of such gaps as a recurring characteristic in 
many different streams, by reference to sub-aerial erosion alone, 
but a good deal of the difficulty disappears if we regard the sur- 
rounding plains as liaving been occupied by the sea. and the hills 
and ridges as islands standing out of it. Doubtless the higher 
land would he cut deeply into by valleys formed by storm-erosion 
before the subsidence took place which converted them into islands, 
but once they were submerged to any considerable extent, so as 
to allow the sea to convert the higher ^leaks into groups of islets, 
the breaching action of the water, aided by currents round and 
between the islands, would be quite competent to account for much 
cutting down of channels lietween islands. As the succeeding 
elcvatory movement progressed and the islands arose more and 
more out of the sea, it seems i)rol)able that some of the main 
channels cut by the water between islands would tend to be cut 
dee])er as being the most direct outlets to large areas of water, 
partly imprisoned on the inland side of the ridges, and that these 
hows would determine the form of the basins left when the sea had 
entirelv retreated. 'I’he stormwater outfall would naturally follow 
the course shaped by the retiring sea waters. 
A very beautiful example of a gap in which there is no river, 
is seen on"th'‘ --oad from Peak Hill to Nullagine, about 2.') to oO 
miles out fioni Peak Hill. I’or this distance the track lies over 
verv level plains, cut into here and there by a "creek,’ which is 
one of the heads of the south l)ranch of the Ciascoyne River, and 
the shallow sections exposed in the watercourses show the soil to 
be underlaid by layers of ferruginous comi)acted sand and small 
pebbles. In one place numeiams thin Hakes of hard whitish lime- 
stone were picked up. showing the i)rcsence of beds of the Nulla- 
gine series. Several clay ])ans were seen, in which fresh water 
would accumulate during rains. While i)assing over those plains 
the road is heading towards a blue ratige of hills, visible for many 
miles, and as one gets closer ami closer to this range, it seems im- 
possible that it can be crossed excc])t by climbing over it. But 
suddenly a wide iiass opens right through it. and the plain con- 
tinues as a fiat wide strip of level ground, plentifully covered on 
surface by subangular stones, between stee]) slo])es on either side. 
Soon the road emerges again from the range on its north side, ano 
