364 
IOWA ACADEMY OF SCIENCE Vol. XXIV, 1917 
cage. It was previously determined that a mesh of this size 
would permit a honey bee, or any insect smaller than a honey 
bee, to pass through, but would not permit bumble bees to do 
so. Two weeks before the clover came into bloom a small colony 
of honey bees was placed in one corner of this cage. The 
bees soon learned to pass through the screen. By the time the 
clover began to bloom the bees had become accustomed to the 
cage and while most of them worked on- flowers outside, yet some 
could always be seen at work on the clover within the cage. Bees 
working on the clover within the cage were observed to collect 
pollen from the flowers and carry it to the hive. 
As soon as all the flowers in the cage were mature, an area 4 
feet square was measured off and all heads within this area 
were collected, kept separate and threshed by hand. Of the 
623 heads collected from this area an average of 37.2 seeds per 
head 'was obtained. 
The higher yields of seed obtained in the honey bee cage than 
in the bumble bee cage may be attributed at least in part to the 
larger number of bees which had access to this clover. How- 
ever, the ratio of honey bee to bumble bee was no greater in 
the cages than in the clover fields in the vicinity of Ames, in 
1911. 
In 1911 the precipitation at Ames, Iowa, was 2.48, 3.83, and 
0.39 inches below normal for June, July and August respect- 
ively. When the clover was in bloom very few nectar producing 
plants were to be found. Whether the honey bee would work 
on red clover to this extent in a year of normal rainfall when 
the number of other nectar-producing plants is larger, is prob- 
lematical, but our observations and results show that the honey 
bee is able to spring the keels of red clover flowers and thereby 
cross pollinate them. 
A repetition of the experiment was made at Ames in 1913 
on first crop clover. It was found that 940 heads from the 
cage containing honey bees and excluding bumble bees yielded 
an average of 0.8 seeds per head, while 79 heads of the uncov- 
ered clover yielded an average of 12.9 seeds and 970 heads from 
a cage containing bumble bees yielded an average of 13.3 seeds. 
The summer was rather dry, but rains were more uniformly dis- 
tributed than in 1911. Perhaps first crop clover is less visited 
by bees than second crop. In 1914 practically the only time 
honey bees were seen on red clover was during August. 
In 1915 the experiment was repeated. This summer proved 
to be exceptionally cool and moist, quite the opposite of that of 
1911. The cage was erected June 26 and the crop harvested 
three months later. These three months had an aggregate pre- 
