ALFALFA SEED PRODUCTION. 31 
GENERAL CONCLUSIONS. 
The numerous researches of previous investigators on the pollina- 
tion of the alfalfa flower have resulted in somewhat divergent con- 
clusions. In but few cases has any attempt been made to determine 
the relation of pollination to the resultant crop of seed. 
The opinion has prevailed that insect pollinizers are of vital im- 
portance and that in the absence of these in adequate numbers the 
resultant seed crop is necessarily small. 
It has, however, been generally recognized that climatic conditions 
are important, as practically all the commercial seed is raised in 
regions having a semiarid climate, at least during the time the seed 
crop is made. 
Alfalfa flowers remain fully susceptible to pollination from the time 
of opening until the petals wither. Pollination is ordinarily effected 
when the elastic stamineal column has become “‘tripped.”’ No 
evidence was found to favor Burkill’s theory that tripping effects 
the rupture of the stigmatic cells and that this is an important factor 
in fertilization. Flowers tripped in various ways to prevent any 
stimulation or rupturing of the stigma by contact set pods equally as 
well as those tripped naturally. 
Flowers tripped artificially, and therefore self-pollinated, set pods 
freely. In one series of experiments on 77 plants at 7 different places, 
9,074 flowers set 2,784 pods when artificially tripped (a percentage 
of 30.68), while 8,939 flowers on the same plants exposed to natural 
conditions set 1,499 pods (16.76 per cent). The pods from artifi- 
cially tripped flowers contained an average of 1.72 seeds each, while 
those from naturally tripped flowers averaged 2.22 seeds each. 
Pollination from a different flower on the same plant is no more 
effective than self-pollination, but pollen from another plant increases 
both the proportion of pods set and the number of seeds per pod. It 
makes but little difference whether the pollen parent be the same or a 
different variety. 
Tripping in alfalfa flowers may be automatic or may be effected by 
insects and other external agents. Untripped flowers form pods and 
seeds only in rare instances. Automatic tripping is a normal phe- 
nomenon. On two plants at Chinook, Mont., in 1909, 33 out of 57 
marked flowers became automatically tripped on one plant and set 
21 pods, and under the same conditions 64 flowers on the second plant 
produced 16 pods from 36 automatically tripped flowers. The per- 
centage of pods to flowers on the first plant is 36.84 and on the second 
25. These are quite as high as normally occur under natural field 
conditions. In a similar experiment at Pullman, Wash., in 1910, 60 
flowers out of 775 on 9 plants became self-tripped, and 21 of these 
set pods. In this case only 7.74 per cent of the flowers automati- 
