34 
ALFALFA 
it was almost a complete failure. Contra¬ 
dictory assertions have been repeatedly 
made by various observers that the flowers 
are self-fertile or self-sterile in the ab¬ 
sence of insects. For the purpose of set¬ 
tling this question definitely, numerous ex¬ 
periments were conducted by Piper and his 
assistants, the results of which were pub¬ 
lished by the Bureau of Plant Industry in 
1914. 
More than 24 species of wild bees, be¬ 
sides many butterflies, flies, and beetles, 
have been observed on the flowers; but 
many of these are useless as pollinators. 
In localities where alfalfa is nectarless it is 
almost entirely ignored by honeybees, but 
where it secretes nectar freely they are 
attracted in great numbers. Usually they 
obtain tbe nectar thru a hole in the side of 
the flower without tripping it. Out of 500 
visits observed by Westgate a flower was 
tripped in only one instance. In California, 
according to McKee, few flowers are trip¬ 
ped by honeybees; but in England, Burkill 
saw them tripping the flowers in great 
numbers. Even if an individual honeybee 
tripped a flow-er only oceasionallv, the ag¬ 
gregate exploded in a day over a large area 
of alfalfa would be large. In Colorado, 
western Kansas, and Nevada, where bee 
culture has been greatly developed in recent 
years, it is claimed that the alfalfa seed 
crop in fields near apiaries is much heavier 
and of better quality than that of fields a, 
few miles away. In the former fields the 
amount of seed was at least 50 per cent 
greater than in those which were remote 
from colonies of bees. In one instance in 
Nevada it is claimed that the presence of 
bees increased the crop of seed by over 300 
pei* cent. Bumblebees are more important 
than honeybees, and trip the flowers fre¬ 
quently— in Washington and Montana 
about 30 per cent of the flowers visited. But 
the leaf-cutting bees (Megachile) are the 
most efficient pollinators. M. latimanus 
trips 9 out of every 10 flowers visited, and 
three of these bees were observed to trip 
flowers at the rate of 552 per hour. Butter¬ 
flies are common visitors, especially in Cali¬ 
fornia, where the alfalfa butterfly (Evrg- 
mus eury theme) is abundant; but they ob¬ 
tain the nectar thru the orifice in the side of 
the flower without depressing the caring, Tn 
South America small birds called honey- 
suckers visit the flower for nectar. Pollina¬ 
tion by the wind does not occur. 
Untripped flowers seldom produce seed, 
but in the absence of insects a large per¬ 
centage may explode automatically. At 
Chinook, Montana, 33 out. of 57 marked 
flowers on one plant became self-tripped, 
and set 21 pods; and on a second plant 36 
flowers out of 64 tripped automatically, 
and produced 16 pods. In the first case 
63 per cent of the self-pollinated flowers 
produced pods, and in the second 44 per 
cent. In rare instances flowers develop 
pods without tripping. Variability in self¬ 
tripping is strongly influenced bv climatic 
factors, as temperature, humidity, and 
bright sunshine. A single alfalfa plant 
was screened from insects for 10 days or 
longer until it was in full bloom. The 
screen was then removed for 15 minutes on 
a very warm clear day. The (lowers quick¬ 
ly began exploding with a snapping sound, 
at times three or four being heard simul¬ 
taneously. It was estimated that more than 
one-half of the flowers were self-tripped 
before the screen was replaced. In the 
West, automatic self-tripping probably re¬ 
sults in the production of as many pods as 
insect pollination. This observation is im¬ 
portant since it explains the production of 
a large crop of seed in the absence of in¬ 
sects. 
A series of experiments was conducted 
to test the comparative effects of self-pol¬ 
lination and cross-pollination of alfalfa 
flowers. Nine thousand and seventy-four 
flowers were artificially tripped, and con¬ 
sequently self-pollinated, and set 2,784 
pods. The diffei-ent plants on which the 
flowers were thus self-pollinated varied 
greatly in the production of pods, 68 per¬ 
cent of the flower's in one instance setting 
pods, while others yielded none. Pollina¬ 
tion from a different flower on the same 
plant is of little advantage over self-polli¬ 
nation. Five hundred and thirteen self- 
pollinated flower's set 165 pods, or 32 per¬ 
cent; while 437 flowers pollinated from 
another flower on the same plant set 134 
pods, or 30 per cent. When 446 flowers 
were each cross-pollinated with pollen from 
another plant. 206 pods were produced, or 
46 per cent. The average number of seeds 
per pod of the self-pollinated flowers was 
1.4; of the flowers pollinated from another 
