89 



fillings of the tubes of decayed roots of trees, burrows of insects 

 and other small animals. 



Another specimen of ieolian action, the only one in force in 

 tliis country is this : — It was observed that sand-hills generally 

 consisted of a white quartz sand of a very uniform grain ; whilst 

 the one in the valleys between the sand-dunes was found to be 

 sand mixed with gravel, this sand having ini'ariably a reddish 

 colour, and giving a striking contrast to the red sand-hills. I 

 should like to explain this fact by supposing the wind acts as a 

 natural air separator, as the white, and therefore pure, quartz of 

 a lower specific gravity has been removed easier by the wind than 

 the ferruginous red quartz, which is of course of higher specific 



gravity. 



The shape of 

 fig. 10, pi. iv. 

 rising ; whilst 

 course of wind, 

 height, leadina; 



tlie sand-hills is invariably the same as given in 



Tiie north-eastern slope is convex, gradually 



the south-western one avers to the prevalent 



and has a steep precipice of four or five feet in 



to a concave, gradually flattening, incline. 



and 



^•eiy 



These sand-dunes are really all parallel to each other, 

 seldom found confluent. 



The repeated proofs of their instability was observable on 

 account of gum-trees growing on top of the sand ridges. The 

 roots of these trees were often sticking far out on the lee side of 

 the dune, entirely exposed and uncovered, whilst the roots ou the 

 opposite side were buried deeply in the accumulated sand. 



Whei'e the sand-dunes are not grown over with vegetation, 

 splendid ripple marks may be noticed, covering especially tlie 

 highest part of the dunes, and representing in miniature a sys- 

 tem of ieolian formations, as the sand-dunes do en gros. 



When destitute of vegetation, these sand-dunes can be seen 

 extending sometimes in one long line almost as far as the eye can 

 reach. Their height is on an average 60 to 70 feet, except in the 

 southern portion of the great Victorian Desert, where, when 

 approaching the Queen Victoria Spring, sand-hills of 300 feet 

 were traversed. They were there, however, less regular, and 

 more tumbled and confluent. 



This, I have reason to believe, is jirobably due to disturbances 

 of the icrial influence, for, as explained in the meteor-ological 

 notes, a climatic boundary exists there, where the dry continental 

 N.E. winds and the coastal winds meet, causing thus a higher 

 rainfall, which was also able to create a strong soakage on a 

 favourable spot such as Queen Victoria Spi'ing has been at one 

 time. 



This meeting of two winds from opposite directions also 

 atteoted the accumulation of such immense sand deposits as 

 found there, by causing the continental winds, heavily laden witli 



