Modifying Climate. 



117 



August to November, though the last might possibly have been 

 'better with November out of it, as in some years the grass is 

 -dry, and the crops nearly ripe long before the month is ended. 

 The rainfalls for these periods were expressed as percentages 

 of the annual amount, and these are shown in the following 

 tables : — 



(a) Stations too Far North or North-West to Benefit by 

 Spread of Cultivation. 





Swan Hill. 



Kkrang. 



Tyrbll Downs. 



Period. 



Percentage of Annual 



Percentage of Annual 



Percentage of Annual 





Rainfall. 



Rainfall. 



Rainfall. 





Sept.-Oct. Aug. -Nov. 



Sept.-Oct. Aug. -Nov. 



Sept.-Oct. Aug. -Nov. 



1885-1894 - 



18-1 - 38'4 



184 - 357 



18-6 - 393 



1895-1904 - 



177 - 327 



17-9 - 360 



158 - 350 



1905-1914 - 



161 - 31-5 



157 - 304 



17-4 - 323 



(b) Stations Benefitting, if at all, by Spread of Cultivation. 





Lakk Marmal. 



WVCHITKLLA S. 



Charlton. 



Period. 



Percentage of Annual 



Percentage of Annual 



Percentage of Annual 





Rainfall. 



Rainfall. 



Rainfall. 





Sept.-Oct. Aug.-Nov. 



Sept.-Oct. Aug.-Nov. 



Sept.-Oct. Aug.-Nov. 



1885-1894 - 



18-4 - 38-5 



1676 - 3625 



17-7 - 36-5 



1895-1904 - 



195 - 36 5 



178 - 34-5 



1916 - 345 



1905-1914 - 



1905 - 35-1 



18-5 - 339 



1796 - 33-5 



In the first group the September-October rains declined by 

 2.0, 2.7, and 1.2 per cent., giving a mean decrease of 2 per cent. 

 In the second group there were increases of 0.65, 1.84 and 0.26 

 per cent., giving a mean rise of 0.9 per cent., or a relative gain 

 for that period of about 3 per cent, of the annual total, or about 

 half an inch of rain. For the four months' period for (a) we 

 get decreases of 6.9, 5.3 and 7.0 per cent., giving a mean of 

 6.4 per cent. For (b) the declines were 3.4, 2.3 and 3.0 per 

 •cent., giving a mean of 2.9 per cent., or a relative gain of 3.5 

 per cent., thus suggesting that the benefit is not quite limited to 

 .the two months, September and October. 



The assumption with regard to this selection is, of course, that 

 in (a) we get simply the effects of periodic seasonal variation, 

 and with (b) we get the periodic variation plus that due to the 

 •change of environment. If the deduction is correct, we have 

 quite a marked improvement in spring conditions at Lake Marmal, 

 Wychitella South, and Charlton, owing, we will say, to the eva- 

 poration from wheat and grass lands lying to north and west of 

 them. 



It might be said that any gain in the spring rainfall would be 

 at the expense of the rains of other parts of the year. It is not 



