12 BULLETIN 289, U. S. DEPARTMENT OF AGRICULTURE. 
ARTIFICIAL MANIPULATION OF CLOVER HEADS. 
Experiments were conducted to determine, if possible, the effect 
on seed production of various types of artificial manipulation of the 
clover heads while the flowers were in bloom (fig. 3). A sufficient 
number of heads were selected on each plant so that the work could 
be conducted on heads covered with tarlatan (fig. 4) and on heads 
exposed to the action of insects. The experiments on the heads 
exposed to the action of insects were to determine whether the 
artificial manipulation of the flowers would have any harmful effect 
on seed production. The different treatments given the heads 
covered with tarlatan were to determine whether fertilization could be 
Fic. 3.—A screen cage (in the background) in which bumblebees were confined. Hand-pollination work 
is in progress in the foreground. 
produced by any method of artificially manipulating clover flowers | 
from which insects were excluded. For this work, plants were 
selected bearing at least eight heads which would come into bloom 
at approximately the same time. These plants were taken at random 
and each marked with a stake, as shown in figure 5. The heads on 
each plant were labeled from A to H, inclusive, and treated as shown 
in Table I. 
These experiments were conducted in Iowa, in 1911 on 50 plants 
at Ames, and 25 at Altoona, and in 1912 on 70 plants at Ames. 
Table II gives the results obtained on 25 representative plants, 
selected from the entire number, and also the average seed yield per 
head of the entire 145 plants experimented with in 1911 and 1912. 
ee ey a ee Ce a 
a eee 
