42 



ADDRESSES, 



been of much value to the farmers conducting them, but from perhaps the 

 majority of them came fiothing but tales of woe. How could such a condi- 

 tion exist? 



At first we were disposed to blame the trouble on the method of conducting 

 the experiment and accordingly tried a number of methods, as some of you may 

 know. We had small hand experiments worked out with the greatest of care. 

 We had small experiments in connection with which machines were used. We 



ftvcfocf/on of Woo/ 



Jby Townships /'ft 



FIG. 3. 



had large experiments made under field conditions. In fact, we tried almost every 

 plan we could devise, not only one year but for several years and yet the results 

 were unsatisfactory. Varieties or methods of the excellence of which we had not 

 a doubt would fail comparatively in a large number of cases no matter how we 

 would conduct the test. We could not blame the result on the carelessness of co- 

 operators, because we knew from visits at some of their farms that such was not 

 the case or rather was not the case in enough instances to explain all our trouble. 



