I 



RIVERS OF NEW SOUTH WALES. 399 



•densely covered with vegetation, a heavier rainfall is required 

 before any surface flow takes place. Hence as a general rule the 

 nominal drainage areas of rivers require to be divided into 

 effective and non-effective areas ; and it must be borne in mind 

 that the ratio which the tirst of these divisions bears to the second 

 is constantly changing. Thus in the case of the Darling, while 

 the total catchment area is about 232,000 square miles, the portion 

 of this which can be classed as effective in ordinary seasons does 

 not exceed 54,600 square miles; and it is doubtful whether in a 

 dry season the effective catcliment can be estimated at more than 

 a third of the latter area. It is clear from these considerations 

 that only a small portion of the rainfall ever reaches the western 

 rivers ; and in addition to this fact it has to be remembered that 

 the water ffowing in these rivers is subject to constant diminution. 

 This is due tirstly to evaporation, secondly to absorption by vege- 

 tation in the channel or on its banks, and thirdly to percolation. 

 In connection with the tirst of these sources of loss it is necessary 

 to remark that evaporation takes place not only from the water 

 surface, but from so much of the margin on each side as is kept 

 moist by capillary attraction. The latter point is of much greater 

 importance than might at tirst sight be expected ; for Mr. H. C 

 Russell, Government Astronomer of New South Wales, has ascer- 

 tained by experiment that the rate of evaporation fi'om soil 

 destitute of vegetation is from two to three and a-lialf times 

 greater than from water, and the rate from a grassy surface from 

 one and a-half to two and a-half times greater than from water. 

 In a Report on the Macquarie River, which I presented to 

 the Royal Commission on the Conservation of Water, in May, 

 1885, I pointed out as the result of actual experiments that 

 ■when the discharge of that river at Dubbo was one hundred 

 and fourteen and a-half cubic feet per second, that at Warren 

 was only tifty-two cubic feet per second, and that when 

 the discharge at the former place fell to twenty cubic feet 

 per second, the current ceased at about eighteen miles above 

 Warren. I also showed that of the total loss of water be- 

 tween these places, sixteen and a-half per cent, in the former 

 case and thirty-eight per cent, in the latter might be due to 

 evaporation, while by absorption the trees on the river bank could 

 account for twelve and a-half cubic feet per second. The opinion 

 which I formed regarding percolation in that part of the Macquarie 

 was that, at least when that river is low, very little loss is due to 

 that cause. It is, in fact, unlikely that any considerable propor- 

 tion of the waters of our western rivers is lost by percolation, 

 excepting in the higher parts of their courses, for the natural 

 tendency of a river ffowing tlirough alluvium is to tamp up all 

 interstices in its channel. There are, doubtless, cases in which 

 loss of water from river channels is sustained by percolation 

 through alluvium. For instance, in that portion of the Murray 



