Alfalfa in Kansas. 197 



Having selected a considerable number of mother plants of desirable 

 types to begin with, and having increased them by the methods described, 

 it now remains to sow the seed of the close-fertilized offspring of these 

 mother plants side by side in nursery rows, to be tested for yield. The 

 field should be level, of uniform texture, and should have been previously 

 subjected to uniform cropping treatment. A field previously divided into 

 small strips sown to different kinds of plants should be avoided, on ac- 

 count of the consequent differences resulting in soil fertility and soil mois- 

 ture content, which will materially affect the yield. It is well to have the 

 series of rows in duplicate, with a check row of a single variety (which 

 should be as pure a strain as can be gotten) placed at the beginning and 

 at the end of each series. The rows should be three feet apart, and the 

 plants should be thinned until they stand two feet apart in the rows. The 

 rows should all contain the same number of plants say one hundred. 

 Each row should be cut separately, and the forage yield determined. 



It is advisable to continue these comparisons of forage yield for three 

 years in succession, if possible, since a single year's test is not conclusive. 

 A further test is necessary which should be carried on at the same time 

 as the general yield tests, viz., a determination of the relative proportion, 

 by weight, of leaves to stems. The most valuable part of the alfalfa 

 plant is in the leaves, and in breeding a high-yielding plant the object 

 should be to secure the greatest possible weight of leaves as compared 

 with the weight of the stems. 



Having thus determined which rows are the highest yielders of gross 

 forage, and which give the highest percentage weight of leaves, the time 



has come to eliminate the inferior 

 rows, and these should be plowed up. 

 Every effort should now be devoted 

 to increasing the superior strains. 

 Cuttings should be made in quanti- 

 ties, and, if possible, rooted in a 

 greenhouse in the fall. They can 

 be grown indoors over winter and 

 transplanted into the field in the 

 spring. The second year a consider- 

 able quantity of close-pollinated seed 

 may be obtained from these plants. 

 A field plot should now be sown 

 for increase. The danger of contami- 



f W cf JrtH f'wi nation from foreign pollen brought 

 ing petals, w, w ; and standard, s. by insects now begins, but this can 



not well be avoided. The best that 



can be done is to have the breeding plot in a locality as remote from 

 other alfalfa as possible. Yield tests on the fractional-acre plan should 

 follow, and finally the seed should be distributed among reliable, pro- 

 gressive farmers for cooperative seed-growing purposes. 



Breeding for Yield of Seed. 



The method outlined above will hold in this case, except that the yield 

 of seed by weight takes the place of the yield of forage. It may be said, 

 in general, that the yield of seed is in an inverse ratio to the yield of 



