fsAUNDERs] CEREAL BREEDING ON EXPERIMENTAL FARMS 159 



milling and baking tests before being sent out. These tests are con- 

 ducted at Ottawa in laboratories provided with special apparatus. 



It is perhaps scarcely necessary to point out that in most of the 

 work of grain breeding and selection the chief objects have been of 

 what is called a practical rather than a scientific character. In other 

 words the production of valuable varieties with the least possible delay 

 and with the least possible expense has been the purpose, rather than 

 to study the laws of heredity and accumulate a store of facts which 

 might in the long run prove of immense value. 



I am not attempting to express any opinion as to the wisdom or 

 unwisdom of the course followed; but it seems necessary to mention 

 it in order to explain the curious circumstance that while contributing 

 so enormously to the immediate prosperity of the country, the work 

 that has been done has added comparatively little to our knowledge of 

 the fundamental principles which underlie the breeding of cereals. This 

 is much to be regretted, especially on account of the large amount of 

 valuable material which has been thrown away almost every year, 

 as soon as it appeared that it was not likely to be any practical use 

 to the Canadian farmer. 



Some deductions of value to science have, however, been drawn 

 from time to time, but I cannot refer to any of them in this paper. 

 The practical utility of these researches must remain their chief justi- 

 fication. And certainly that utility has already been amply demon- 

 strated. But we are much nearer the beginning than the end of this 

 work, and he would be a wise and daring prophet who could predict 

 within even a few hundreds of millions of dollars what will be the net 

 profits to Canada, during the first half of this century, from the modest 

 sums spent on researches in cereals. 



