10 Jan., 1916.] Wheat Yield and Winter Rainfall. 39 



line representing the average rainfall. In other words, during this 

 period farmers never reaped less than 1 bushel per acre for each inch 

 of composite winter rainfall. 



Lest it be supposed that we are approaching the limit of our produc- 

 tion so far as yields per acre are concerned, it might be mentioned that 

 carefully-conducted tests on the water requirements of wheat at Ruther- 

 glen during the past two years demonstrate that 1 inch of rain is 

 capable of producing at least 2^ bushels of wheat, providing the whole 

 of the water is used by the crop and none dissipated by evaporation. 



At present, however, we in Victoria are securing less than half this 

 amount, and the more widespread adoption of better methods of cul- 

 tivation, systematic crop rotation, rational soil fertilization, and careful 

 seed selection will gradually raise the average yield per acre until it 

 approaches the above limit. 



Effect of Abnormal Seasons and Crop Conditions. 



The graph shows that the average yield per acre closely follows the 

 composite winter rainfall. This is particularly true for normal 

 seasons. Occasionally, however, we get an unusually wet season or a 

 disastrous drought. 



The two wettest winters on record during the period under review 

 are 1894 and 1906. In both these years the composite rainfall exceeded 

 14 inches at the ten selected centres during the six winter months. 

 Many of the crops were waterlogged during the winter months, and 

 ultimately the yields were much lower than in seasons of normal rainfall. 



During the two drought seasons, 1902 and 1914, there were whole 

 districts where the crop was an absolute failure. The rainfall last 

 year at Mildura during the six winter months was less than 1 inch. 

 Little wonder that no crop was reaped. 



Looking over the chart, it is apparent that the best average yield is 

 obtained when the composite winter rainfall lies between 10 inches and 

 13 inches. 



If a larger number of centres were chosen it is likely that the 

 approximation of the graphs of rainfall and crop yields would be even 

 closer. Absolute agreement could not be expected, however, for the 

 reason that crop yields are dependent on other factors than rainfall, 

 though rainfall is the dominating factor. For example, fungoid pests 

 such as rust, smut, takeall, affect the crops in some degree every season, 

 but in some years the seasonal conditions are highly favorable for fungus 

 diseases, and heavy toll is levied on the wheat crops. 



Again, the premature appearance of continuous liot winds just as 

 the croj) is filling depresses the yields. Untoward accidents, such as 

 widespread heavy hiiilstorms and violent winds at the time of ripening 

 of the crop, materially lower the yield. On the otluM- hand, a long 

 cool spring, witli mild spring temperatures and opportune showers, as in 

 the present season, following on a normal winter, would tend to un- 

 usually heavy crops. Finally, good seeding rains and e;irly Xoveinher 

 rains have a stimulating effect on the crop averages. 



Possibility of Forecasting the Crop. 



If the influence of these fjictors bo taken into considcnitinn, it is 

 apparent that the graph suggests a method of forecasting approximately 



