in South-Central Western Australia. 313 



This conclusion thus bears out, to some extent at least, Woodward's 

 original idea that such lakes are due to wind action. 



In their general trend, the lakes lie along drainage lines — more or 

 less dismembered and probably in part deformed — which have been 

 formed under existing conditions or have belonged to an old river 

 system when the climate was moister and perhaps the country iower 

 than at present, as suggested by Gibson and J. W. Gregory. That 

 they are the remains of an ancient river system formed under 

 different climatic conditions than those of to-day has not yet been 

 demonstrated; but if it were, such demonstration would not, in the 

 Avriter's opinion, invalidate the conclusion that portions at least 

 of the lakes and the striking characteristics, the subject of this paper, 

 are dominantly due to wind erosion. 



Few observations have as yet been recorded as to the dominant 

 direction of the wind, but from the writer's personal, although 

 limited, observations in the Comet Vale- Goongarrie district, made 

 since writing his physiography of "Western Australia, and the paper 

 referred to above in the Geographical Journal, it may be said that 

 although there is perhaps no prevailing wind, yet the dominant wind 

 appears to be from the west (north-west, west, or south-west). As 

 shown above, Honman and Blatchford also hold that the dominant 

 wind is from the same quarter. 



Woodward' has noted the retarding effect of the ground-water 

 table on wind erosion on lake-floors, and this feature may help to 

 explain the very even surfaces of the billiard-table rock-floors, the 

 wind receiving at least a temporary check on reaching the more or 

 less saturated zone. Honman^ also believes that during lake 

 migration the rock-floor is kept level by moisture. The silts of the 

 lakes are always moist, except perhaps, at times, the actual surface 

 film. A moist surface, in addition to preventing the removal of 

 materia], may also, as pointed out by A. W. Grabau, with regard to 

 the American playas,^ catch dust particles carried across the surface 

 by the winds, and by this means the thickness of the deposits may be 

 increased. Hence the silts of the eastern portions of the "dry" 

 lakes may be partly wind-blown, and not entirely due to deposition 

 under water, this wind-blown material being probably derived from 

 the western areas.* If such wind transportation does take place it 

 shows that, despite its retarding effect, the hygroscopic character of 

 the silt is not an absolute bar to wind transportation, probably 

 because the actual surface may (at various times at least) be 

 sufficiently drj' to allow such transportation. In anj'' event, however, 

 from the facts set out in this paper, it seems clear that the wind can 

 and does remove material from the western sides of the " dry " lakes. 



' Op. cit., p. 365. 



2 Bull. 66, Geol. Surv. W. Austral., 1916, p. 36. 

 ^ Principles of Stratigraphy , New York, 1913, p. 603. 



■* Dust is no doubt also caught by the lake waters when they are in 

 existence, and this is then deposited as an aqueous sediment. 



