THE DESERT SANDSTONE. 329 



layers of volcanic and sedimentary material that gradually fill up 

 the old valleys. The older waterworn gravels have been met with 

 in the several localities from sea-level to an elevation of 4,000 

 feet ; but some of these may not have owed their origin to fluviatile 

 action. Hardly any other mode of accumulation will satisfy the 

 requirements of the drifts of North Australia. 



It is very singular that we should find such constant and 

 unvarying evidence of the greater height at which the rivers 

 flowed, when we should be led to expect immense subsidences in 

 the volcanic tract. After such outpourings of lava and such an 

 enormous transfer of material to the surface in the form of ash, 

 one would expect to, meet the same evidences of subsidence which 

 we find in all volcanic regions. There is certainly no evidence 

 whatever of any upheaval. Possibly it may be explained by 

 supposing that the most of the river valleys were filled up in part 

 by the volcanic ash deposit, until gradually the rivers cut down 

 the loose material to its former level. 



It seems to me probable that the continent, before the volcanic 

 period, was higher above the sea than it is now. The period 

 itself may have been connected with the destruction of the 

 ancient land fauna of Australia, which does not apparently overlap 

 the existing fauna, at least as far as the terrestrial mammalia are 

 concerned. In Queensland the evidence about the Darling Downs 

 and main range is that violent and sustained volcanic disturbance 

 was followed by floods which swept into heaps fragments of the 

 remains which volcanic action had destroyed. The deposit at 

 King's Creek is an extremely abundant collection of broken bones, 

 mingled together with indescribable confusion, on a few square 

 yards of ground. In this there are gigantic marsupials — kangaroos, 

 wombats, and opossums, with water-birds, large crocodiles, gigantic 

 lizards and turtles. 



With regard to the age of some of the later volcanic phenomena, 

 we have very clear evidence in Moreton Bay that it belongs to 

 the most recent period. There is satisfactory evidence afforded 

 in a lava or ash stream which runs into the sea at Cleveland. It 

 has flowed over shells and corals which differ in no way from the 

 existing marine fauna. There is a basaltic flow at Lytton a few 

 miles further inland which appears as old as most of the volcanic 

 formations of the higher table-lands. The Glasshouse Mountains 

 on the north side of the bay, of which figures are here given (see 

 plates xix. and xx.) are apparently no older. 



The volcanic emanations of Mount Gambier, Mount Shanck, 

 and Tower Hill belong also to the recent period, and are connected 

 with the existing fauna. 



Altogether the evidence afforded by the volcanic rocks throughout 

 Australia is that there were at least two distinct periods of 



