PLANT IMPROVEMENT 159 



examined for quality, size, and trueness to type, and only 

 those units that come up to a good standard should be re- 

 tained. Ten of the best tubers of the selected units are to 

 be used for the next year's test. 



The second year each ten tubers saved from the 

 desirable units of the first year are quartered and planted 

 to a unit of forty hills. The same procedure is followed as 

 in the first season. The best tubers of the best forty-hill 

 units should be saved for planting the third season. By this 

 time, high-yielding strains of desirable tubers will be es- 

 tablished, with enough tubers to plant a large plot. But 

 forty or more exceptionally good tubers from the best forty- 

 hill units should be saved to continue the best pure lines or 

 units. At the end of the third year enough seed of high- 

 yielding strains will be secured for most of the planting needed 

 for the farm. Besides, the extra tubers may be sold for seed 

 at high prices. 



The investment of time and labor in developing a superior 

 strain of potatoes by this method is small compared with 

 the results that may be attained. For example, one farmer 

 was able in a few years to establish a strain which gave him 

 a yield averaging 282 bushels per acre for nine years, while 

 his neighbors secured an average yield of approximately only 

 150 bushels. The difference of 132 bushels per acre would 

 seem to be a good return for the time and trouble needed to 

 establish a high-yielding strain of potatoes. 



Improving other plants by selection. This same meth- 

 od of getting better plants by selection may be applied to 

 other crops such as wheat, oats, barley, etc. In this case 

 smaller plots should be used, otherwise the procedure is 

 essentially the same as for corn. 



In several states plant-improvement associations have been 



