pLAXT-BKIiEDI.NC; FOR FARMERS. 159 



grain from each row and compare the yield per row with the yields from 

 the rows of the standard varieties and record the results. This will 

 give an indication of the yielding capacities of the different select strains. 



Third generation. — As in the preceding year plant 17-foot rows of all 

 of the second generation rows selected and of a few of the very best and 

 highest yielding strains plant larger plots to increase the seed for more 

 extensive planting. 



When the test rows of this year have been examined, threshed and 

 the production per row obtained and compared with the production per 

 row in the second generation tests, the experimenter should be in posi- 

 tion to form a fair idea as to which strains are the best yielders. Save a 

 few of the best for more extensive trials in the fifth year. Of some of 

 these best strains where small increase plots were grown, there should be 

 sufficient seed for comparatively large increase plots in the fourth genera- 

 tion. 



Fourth generation. — Plant 17-foot test rows of the strains retained in 

 the third generation in comparison with some standard varieties and where 

 the seed is abundant several of these test rows should be planted in different 

 parts of the test plot. Plant as large increase plots of the most promising 

 strains as the seed obtained will permit. 



Examine the test rows when ripe, harvest and weigh the product of each 

 separately, as heretofore, carefully preserving the yield records. Xow 

 compare the test row and increase-plot yields of each strain for the three 

 years with each other and with the yields of the standard varieties 

 which also should have been retained in the test for comparison. The 

 breeder should with the data now accumulated be able to determine 

 whether any of his strains have yielded better than ordinary varieties 

 and are thus valuable to retain for extensive trial. All but those which 

 he believes to be superior to any other grown and to the standard varie- 

 ties, should be discarded. It is safe at this point to discard all but two or 

 three strains, and if only one strain is markedly superior probably all but 

 this one should be discarded. All of the seed from test rows and increase 

 plots of the strains finally selected should be preserved and a large field 

 planted the next season. 



If the grower has reason to believe that he has secured a superior 

 variety of high yielding capacity he should as soon as possible have it 

 tested by his neighbors and distribute the seed as extensively as the results 

 obtained with the variety justify. If extensive tests prove it to be a 

 superior variety is should be given a distinctive name so that it will not be- 

 come confounded with other sorts. 



The above outline is based on selecting entire plants and getting the 

 total product of the different plants in beginning the selection. In 

 many cases, it is found handy to simply select large, fine heads which ap- 



