404 PROCEEDINGS OP SECTION E. 



obtained. But there are also cases both in India and New South 

 Wales where water can be obtained at a moderate depth in river 

 channels which have become filled up within a comparatively 

 recent period. 



I have already described the manner in which a river flowing 

 through alluvial soil I'aises its bed and banks by degrees till at 

 last, after breaking through one of its banks, it seeks the lowest 

 land, and flows there. When that takes place in a district where 

 there is drifting sand, the old river channel becomes choked up, and 

 not only so, but on account of the obstruction to the wind caused by 

 the trees and other vegetation on the river banks, the sand accu- 

 mulates to some height above the surrounding plains ; so that 

 the old course of the river is indicated by a sandy ridge. In 

 .such cases the old river channel carries a small proportion of the 

 original discharge and this is sensibly augmented by tlie rainfall 

 on the ridge ; for it is found that in heavy rain over eighty per 

 cent, of the rainfall percolates into clean sand. Instances of 

 this description are more common in India than in New South 

 Wales ; but there are good instances to be met with here, parti- 

 cularly in the district of the Castlereagh River. 



It might at first sight be supposed that a fair appi'oximate 

 estimate could be formed of the proportion of the rainfall which 

 passes to underground streams, but in reality tliis is a very com- 

 plicated question. I have ascertained that in 1884, which was an 

 exceptionally dry year, the quantity of water which flowed past 

 Wao;ga Wagga was less than one-tenth of the rainfall on the 

 catchment area of the Murrumbidgee above that place. But on 

 examination of Mr. Russell's "Rainfall Return," I find that during 

 the year the rain must have fallen chiefly in light showers ; so 

 that evaporation and the requirements of vegetation would 

 account for nearly the whole quantity. 



By giving due consideration to the principles described we dispel 

 many apparent anomalies in connection with our rivers. For in- 

 stance, at first sight it seems incredible that the Clarence, with a 

 catchment area of only eight thousand four hundred square miles 

 should in a high flood discharge more water than the Murray 

 below Wentworth, when both that river and the Darling are in 

 flood ; yet such is a fact. The reason is that in the case of the 

 Clarence, several causes combine to make the catchment area 

 highly eflfective, while the conditions of the western rivers are 

 exactly the reverse of this. The Clarence has a rapid fall and a 

 short course, a compact catchment area, and a rainfall about fifty 

 per cent, in excess of that on the more favoured parts of the 

 Darling basin. The result of these conditions is that a large 

 proportion of the rainfall on the Clarence basin runs off with a 

 high velocity ; or that while the duration of a flood on the 

 Darling is reckoned by months, that of a flood on the Clarence is 

 measured by hours, or at most by days. 



