150 E. T. Quayle: 



The improvement in the run-off may be made more obvious per- 

 haps if we limit the comparison periods to the Winter months and 

 begin each flood season only with those months prior to which suf- 

 ficient rain has fallen to make the soil fairly moist. This will re- 

 quire, say, two inches of rain during the preceding month. This 

 means that the period will rarely begin before May and will end not 

 later than October. Adopting this procedure and using the rainfall 

 data from the five stations selected by myself, which are slightly 

 wetter than the average for the basin, the following results were 

 obtained. 



The years 1890 to 1909, omitting the great drought year, 1902, 

 gave a total of 102 months with an average rainfall of 2:53 inches,, 

 and average mean monthly river flow, of 152 eft. per second. The 

 years 1910 to 1921, omitting the drought years 1914 and 1919, gave 

 50 months, with, an average rainfall of 2:75 inches, and average river 

 flow of 270 eft. per second. That is, unless the run-off has improved, 

 an increase of 22 points in the rainfall gives an increased run-off of 

 118 eft. per second, or a 10 per cent, rainfall increase means a 60 

 per cent, run-off increase, which seems hardly possible. 



In order to get a series of earlier years with approximately the 

 same rainfall as that of the later series, it will be necessary to re- 

 ject, in addition to 1902, the rather dry years, 1895, 1896, 1897, 1899, 

 1904, and 1907. The remaining 13 years give an average monthly 

 rainfall of 273 points, and an average mean river flow of 187 eft.. 

 per second. Therefore, for practically the same rainfall we get an 

 increase in the run-off of nearly 45 per cent. 



The very wet months in these periods, or of over 4 inches, were 

 as follow: — 



1890, June, 431 pts. 1909, August, 695 pts. 



1893, May, 660; June, 485 pts. 1911, September, 406 pts. 



1894. October, 605 pts. 1912, September, 536 pts. 



1898, June, 520 pts. 1915, Sept., 547; June, 419 pts. 



1899, June, 456 pts. 1916, June, 410; Sept., 457 pts. 



1900, August, 403 pts. 1920, August, 505 pts. 

 1906, May, 448; June, 438 pts. 1921, September, 44 pts. 

 1903, July, 481, and Sept., 409 pts. 



This gives 12 for the 20 earlier years, with an average of 

 503 pts., and 8 for the later 12-year period, averaging 466 pts.; which 

 rather favours the earlier period as regards flood possibilities. 



This 45 or 50 per cent, gain may be apportioned to the two main 

 sources of increased flow, (a) the amount gained by springs from the 

 destruction of the trees, and given to maintain a constant flow; (b) 

 the increased run-off due to the better drainage resulting from erosion 

 of channels, silting up of water holes, formation of paths by stock, 

 moister condition of subsoil, quicker saturation of the subsoil due to- 

 killing of the trees, the raising of the water table, etc. It has already 

 been shown that the former results in an average increased minimum 

 flow of 8:7 eft. per second. This gives for one year, 6300 acre feet. 

 That leaves, therefore, as the gain due to the second group of causes- 

 about 20,000 acre feet. 



