in South-Central Western Australia. 311 



Lalie Cowan, Norseman. W. D. Campbell's maps and descriptions ^ 

 of this lake clearly show that rock-cliffs exist on the western shore 

 and blown sands on the eastern. 



Lake Barlee. — H. AV. B. Talbot's maps* show this lake to be star- 

 shaped; but where north and south trending portions of the lake 

 abut rock ridges, the latter are on the western shore of the lake. 



Lalce Gordon, near Kanowna. This lake has rocky cliffs on the 

 ■western shore and sands on the eastern, as personally observed by 

 the writer. 



Meaning of the Phenomena described. 



The facts stated above appear to decisively indicate that the 

 ocouirence of rock-cliffs and rock-floors on the western sides of 

 "dry" lakes, and their absence on the eastern sides (Avith replace- 

 ment on such eastern sides by sand and silt), are not merely 

 coincidences. There must be some agent now or formerly operating 

 Avith dominant power over wide areas in order to produce such 

 remarkable uniformity of conditions. Such power seems to be 

 restricted either to marine erosion, or to erosion by terrestrial waters, 

 or to wind erosion. These possibilities are now considered. 



Marine erosion does not seem to apply, for reasons that will be 

 briefly stated. (1) No evidence has been adduced that the country 

 has been recently submerged beneath the ocean farther north than 

 Norseman,^ which is about 120 miles north from the southern coast 

 of Australia. This distance, however, is short in comparison with 

 the north and south length of the belt occupied by the " dry" lakes. 

 (2) The marine theory claims that the forms of the cliffs are due to 

 marine erosion. It therefore assumes that these cliffs have practically 

 sustained no erosion since their supposed emergence from the sea; 

 an assumption the validity of which is at once questioned owing to. 

 the fact that normal subaerial erosion — apparently to a considerable 

 extent— has taken place near the coast since the last undoubted 

 emergence from the sea of the coastal lands, and consequently 

 subaerial erosion farther inland must have also occurred. (3) The 

 effect of present erosional processes is ignored. (4) As there are 

 long sub-parallel lines of cliffs forming the western shores of lakes, 

 such cliffs, on the theory of marine erosion, could only be produced 

 by many distinct uplifts, with pauses long enough to allow the lines 

 of cliffs to be successfully eroded. This means that the country 

 should rise by a series of marine-cut very wide benches one above 

 another, a state that, so far as all knowledge of the country goes, 

 does not exist. Moreover, as the cliffs face the east, the emergence, 

 on such theory, must have been from west to east, that is, away 

 from the present western coastline, the last land to emerge being 



1 Bull. 21, Geol. Surv. W. Austral., 1906, p. 21, and plate. 



- Bull. 45, Geol. Surv. W. Austral., 1912, pis. i, ii. 



" Well to the east of Norseman the sea in probably Miocene times was much 

 farther north, but this formerly sea-covered land is outside the area discussed 

 in this paper. 



