THE DESERT SANDSTONE. 327 



miles north-east of the telegraph station, the river flows through 

 a narrow gorge of sandstone conglomerate, which, in the few places 

 that I was able to examine, clipped away from the stream in the 

 manner described in the other localities. On the summit of all 

 these sandstone tablelands the strata seemed to be horizontal ; but 

 this was difficult to ascertain satisfactorily owing to the amount 

 of false bedding. I noticed the same curious dip in the McAclam 

 Range on the side of the Victoria River and the same has been 

 referred to by Gregory, Wickham and Stokes ; but I was not able 

 to examine the locality closely. We were but three Europeans 

 in a small steam launch, and the blacks are particularly numerous 

 and hostile, so that to land anywhere was to risk unnecessarily a 

 personal encounter. 



This then is the character of the fluviatile sandstones and it is 

 an unvarying one wherever I have met the formation. The 

 uniform dip, the hard flaggy nature of the deposit, and the included 

 waterworn conglomerate mark it unmistakably. I have never 

 met with it far from any river, but it is not always present. On 

 the Victoria River it is never absent ; but sometimes the range 

 or plateau recedes five or six miles away from the channel as in 

 the case of the McAdam Range. On the Katherine River there 

 are long stretches of country in which it flows through a uniform 

 sandy valley, and then succeeds another kind of channel which is 

 bounded by broken tablelands of this fluviatile sandstone, either 

 abutting on the stream or enclosed in valleys five or six miles in 

 width. The broken detached character of these plateaux gives 

 rise to a perfectly impassable country, with rugged rocky scenery 

 of grand and wild character. In all cases the stone is composed 

 in the same manner as already described with the usual waterworn 

 conglomerate. 



Now are we to attribute this singular formation to trap-rocks 

 and ash, or what is its nature 1 The difference from the Desert 

 Sandstone is the perfectly smooth and rounded conglomerate 

 which it contains, and more or less through the whole of it ; that 

 is waterworn quartz gravel or a conglomerate of boulders eight or 

 ten inches in diameter. The quartz is usually, but not always, 

 milk-white. It is certain that we have here evidence of the long 

 sustained action of running water rubbing down the very hardest 

 materials and perfectly rounding them. Moreover the quartz does 

 not belong to the sandstone formation, but rather to the palaeozoic 

 slates upon which it lies. 



On the whole the most probable interpretation of this formation 

 is that the conglomerate has been derived from a river channel 

 through the palaeozoic rocks which contain an abundance of quartz 

 reefs. The sand has been an ash deposit partly filling up the 

 channel and mingling with the conglomerate or covering it over. 



