RED-CLOVER SEED PRODUCTION. 17 



The results given in Table IV are somewhat higher than the average 

 seed yield of the uncovered check of the experiments summarized in 

 Table II. It may be that the close proximity of the checks given in 

 Table II to heads covered with tarlatan kept bees from making as 

 many visits to those heads as they would otherwise have made. 



FLOWERS POLLINATED WITH POLLEN FROM ANOTHER HEAD ON THE SAME PRIMARY 



BRANCH. 



Since the amount of seed obtained in 1911 from self -pollinated 

 heads under cover was so small that it could be accounted for by bees 

 working through the tarlatan, it was not deemed necessary to 

 emasculate the flowers for cross-pollination work in 1912. With 

 this in view a series of heads was covered before any of the flowers 

 came into bloom and later pollinated with pollen from another head 

 on the same primary branch, 20 flowers on each of 11 heads being 

 pollinated in this manner. Not. a single seed was produced. 



FLOWERS POLLINATED WITH POLLEN FROM A HEAD ON A DIFFERENT PRIMARY 

 BRANCH OF THE SAME PLANT. 



Another experiment similar to the preceding one was conducted, 

 except that the pollen was taken from heads on different primary 

 branches of the same plant, 20 flowers on each of 10 heads being 

 pollinated in this manner. One seed was produced. 



CROSS-POLLINATION EXPERIMENTS. 



Alternately with the above two experiments 20 flowers on each of 

 13 heads were pollinated with pollen from a separate plant. An 

 average of 14.3 seeds per head was obtained. 



The results obtained in the last three experiments, as well as with 

 all preceding ones, show that clover is practically self-sterile and that 

 pollen must come from a separate plant in order to effect fertilization. 



BUMBLEBEES AS CROSS-POLLINATORS OF RED CLOVER. 



In view of the consensus of opinion that the bumblebee is responsi- 

 ble for the cross-pollination of red-clover flowers, and since no 

 investigator, so far recalled, has denied its ability to do this, it was 

 deemed desirable to study the relative efficiency of the bumblebee 

 as a cross-pollinator of this plant. 



For this work a cage 12 feet square and 6 feet high, made of wire 

 screen having 14 meshes to the linear inch, was erected shortly after 

 the first crop of clover had been cut. As soon as the second crop 

 started to bloom bumblebees were caught with an insect net and 

 placed in the cage. It was soon found that bees would live about 

 three days when confined in the cage and that six bees in confinement 

 would visit approximately as many flowers as one bee would have 

 visited had it worked nearly all the time. With this in mind, two 

 bumblebees were placed in the cage each forenoon until all the clover 



