36 Wanted — Government Experimental Farms. 



the aid of the grass mixtures commonly used in rotation 

 husbandry, as with these from six to eight years would 

 be required to form a sod, and even then that would be 

 far inferior to the sod which can be produced in four, 

 or even three, years, with the aid of the mixtures I have 

 found to be most efficacious. 



This question now naturally arises — Why should I 

 have had to work out "my own salvation at my own risk 

 and cost ? Why should I not have known exactly what 

 to do when I first took over the farm alluded to ? And 

 why should farmers in Great Britain in general not 

 know exactly what to do in order to cope best with 

 the difficulties of these times, and how to do it in the 

 cheapest manner which combines with it the utmost 

 degree of efficiency? The answer to these questions 

 simply is because of the negligence of our Government, 

 for there cannot be the slightest doubt that had it 

 had experimental farms and agricultural schools the 

 principles I have laid down, and proved the success of, 

 would long ago have been brought to the notice of our 

 agriculturists and generally adopted. For, as we have 

 seen, the farmers in Normandy, aided as they were by 

 Government schools and farms, seem to have had no 

 hesitation in at once altering their systems in accord- 

 ance with the requirements of the times, and there can 

 be no doubt that the same results would have occurred 

 here had similar facilities existed for the diffusing of 

 agricultural knowledge, and full and timely information 

 as to all the world-wide causes which would necessitate 

 a complete change of farming system. 



But it may be urged that, as we have hitherto done 

 fairly well without Government schools and farms, 

 none are needed now. Such reasoning — and it is a too 

 common course of reasoning — shows how dangerous it 

 is to rely on the experience of the past for lessons for 

 the future, for it is seldom that the whole conditions of 

 the past are exactly repeated, and there is, therefore, 

 always a great risk run of applying to a diilerent set of 



