462 ALFALFA FARMING IN AMERICA. 
under way at the Arlington Experimental Farm, near Washing- 
ton, D. C., at Pullman, in the eastern part of the state of Wash- 
ington, and elsewhere tend to prove that heavy-seeding propensity 
is heritable to a marked degree. In consequence of this, a race 
of unusual excellence could readily be secured by propagation of 
the progeny of individuals selected on this basis. 
When alfalfa is grown in rows to permit of intertillage, it is 
much easier to make selections than in broad-casted stands, chief- 
ly because individuals in rows have better opportunity for expres- 
sion of their normal character. In addition, the comparative isola- 
tion of the plants gives readier access to them. 
At first thought it might appear that in thinning out stands of 
row-cultivated alfalfa, only individuals of the greatest seed-produc- 
ing capacity should be left. A second thought quickly reveals the 
fallacy. of this idea, as the ultimate purpose of all alfalfa growing 
is hay production. Selection based on seeding habits alone will 
develop this side of the plant unduly at the expense of its forage- 
producing capacity. The highest type of alfalfa for use in areas 
where seed production is the primary purpose in growing the crop 
is one that combines satisfactory hay and seed producing quality 
in symmetrical proportions. 
It is recommended that the selection of desirable plants com- 
mence as soon as the preliminary seeding has developed plants 
large enough to show their value. The field should be inspected 
row by row, and seed of the selected plants should be gathered in 
advance of the regular harvest. The relatively small quantity of 
seed secured in this way should be sown with great care to make 
it cover the greatest possible area of ground. The plat of alfalfa 
thus secured will produce seed of much greater value than that 
obtained from unselected plants. If this method is carried out, 
materially increased crops of seed may be secured without de- 
tracting from the hay value of the strain. Indeed, both the hay 
and the seed producing capacity may be increased by the process. 
If it is impracticable to secure sufficient seed from selected 
plants for all of the new seedings that one desires to make, the 
selected seed should be planted separately, and that harvested 
from this plat should be used for. subsequent seeding. This 
method will also afford an opportunity for demonstrating the rela- 
tive value of selected as compared with unselected seed. 
Conclusion—The results obtained by farmers on a field scale, as 
well as of the experiments thus far conducted, indicate that the 
growing of alfalfa in cultivated rows for seed in the semi-arid 
regions offers every promise of success. The method is recom- 
mended particularly for those sections where on account of the 
