308 IMPROVING BY CROSS-FERTILIZATION. 



in one neighborhood, and the other in another neighborhood 

 fifty or more miles distant. In alternate rows plant 'the kernels 

 taken from one or two ears of each lot. Before flowering thin 

 out all poor stalks. As soon as the tassels begin to show them- 

 selves in all the rows of one lot, pull them out, that all kernels 

 on the ears of those rows may certainly be crossed by pollen from 

 the other rows. Save and sow the seeds thus crossed and an in- 

 creased yield may be expected the next year. The benefits of 

 such crossing will gradually diminish and probably disappear in 

 a few years. All species which freely intercross by the aid of 

 insects or the wind can be crossed as follows: Procure a quan- 

 tity of seed grown for some years at some distance away and mix 

 with seed kept and raised for some time at the place where the 

 experiment is to be tested. " The two stocks will intercross, 

 with a thorough blending of their whole organizations, and with 

 no loss of purity to the variety ; and this will yield far more fa- 

 vorable results than a mere exchange of seeds." (Darwin). 



In brief, mix seeds of the same variety grown in different lo- 

 calities to grow your seed. 



The late Charles Darwin in his book on The Effects of Cross 

 and Self -Fertilization of Plants records the results of experi- 

 ments made on fifty-seven species of fifty-two different genera of 

 thirty families. These experiments were continued and re- 

 peated for ten years. He generally found the plants raised from 

 seed crossed with foreign stock were the most vigorous, the 

 largest, the hardiest, matured the earliest, yielded the most 

 seed, and such seeds were the most certain to germinate and ger- 

 minate soonest. 



In 1877 the writer began some experiments of this kind with 

 Indian corn and with beans, and has since made others. The 

 advantage shown by crossing corn with foreign stock was as 151 

 exceeds 100, and in the case of black wax-beans it was as 236- 

 exceeds 100. Other experiments have always shown a large gain 



