6 
Colorado Agricultural Experiment Station 
Tests and Experiments 
Seed Selection .—One of the first points to attract attention on 
investigating alfalfa seed production was the fact that there were in¬ 
dividual plants that bore seed heavily in fields where most of the plants 
were failing to form seed. It seemed that if selections of seed from 
these well filled plants were made, that immediate results would fol¬ 
low by establishing a new strain of seed-producing alfalfa. 
Selections of this kind were made in 1904. The results, however, 
were disappointing, owing to the fact that the selections were made 
from some ordinary alfalfa which afterwards proved to be much in¬ 
ferior in point of forage production as compared to other strains which 
were tested with these selections. Plate No. 1, is a view of this first 
test. The four rows to the left in the picture are the progenies of 
these first selections; the six rows in the center were Choice Turkestan 
plants. Owing to this unfavorable comparison the selections were 
abandoned for more promising ones that were subsequently made from 
a larger comparative test of sixty-four different strains of alfalfa 
from the different sections of the world secured from the U. S. Depart¬ 
ment of Agriculture. These varieties were sown in adjacent plats and 
all received the same care and were under the same conditions as 
nearly as possible. Each plat was thinned to single plants in order 
to observe the characteristic traits of the different individual plants 
and types. 
The results of this comparative test have been interesting and 
valuable. It was shown that there are other contrasts, of qualities 
in alfalfa that are more valuable than mere seed production. For in¬ 
stance, the important questions of hardiness, disease-resistance, and 
the quality and quantity of forage produced, are points that are funda¬ 
mentally important in seed selection. Hence our efforts have since been 
directed towards developing and establishing a type of alfalfa that will 
combine all the desirable traits as far as possible. The results have been 
encouraging. Plants have been found among the best hay types that 
produced as high as two ounces of seed per plant. 
Systematic seed selection has proved to be efficient in establish¬ 
ing greater uniformity in the types and qualities desired in alfalfa. The 
leaf characters, the stooling habits, and the flower colors have been 
reproduced true to type in the progenies of many of our selections. 
The seed producing traits of different plants in almost every strain 
tested, have shown inherent tendencies, which clearly indicate that seed 
selection will be one of the important factors in improving the produc¬ 
tion of alfalfa seed. It has been found possible to produce seed 
from good hay types if the proper cultural conditions are supplied. 
The results of a great number of selections that have been made during 
the past eight years have demonstrated that under favorable conditions 
of growth, the best types of alfalfa for forage are not inclined to 
produce much seed, and that the best seed yielding plants are not 
as a rule the best types for hay production. In other words, a strong 
