BUMBLE BEES AS FERTILIZERS. 327 



with netting, that no one could say the difference in yield of 

 seed was due to the fact that one lot was covered and the other not 

 covered. It will be seen, that where bees were observed to work 

 on the flowers the yield of seeds was nearly four times that where 

 the bees were kept away. But perhaps the two plants would 

 not have yielded the same number of seeds had they been treated 

 in every way precisely alike. 



Iir reply to this suggestion I can offer the following, which 

 shows that in six examples, selected at random, only one was 

 found in which the yield of seeds was nearly twice the number 

 in the heads containing the fewest seeds. On September 13th, 

 1882, 1 selected of the second crop of red clover five plants within 

 ten feet of each other, which seemed to be much alike. They 

 had not been covered in any way. The seeds from 50 good heads 

 of each plant were shelled out with the following results : 1, 260 ; 

 1, 275; 1, 4GO; 1, 485; 1, 1,820. It will be seen that 50 heads 

 from plant number five contained only about one-third more 

 seeds than 50 heads from plant number one. 



In another place, 50 heads selected from one plant yielded 

 2,290 seeds, nearly twice as many as plant number one in the 

 first lot. 



Mr. C. Darwin covered one hundred flower-heads of red clover 

 by a net and not a single seed was produced, while 100 heads 

 growing outside yielded by careful estimate 2,720 seeds. He 

 says: "It is at least certain that bumble bees are the chief fer- 

 tilizers of the common red clover." 



It may not be out of place to say here that experiments with 

 white clover show that visits of honey bees increase the yield of 

 seeds enormously. In one case 



8 protected heads yielded 5 seeds 



8 visited by bees yielded 236 seeds 



This is an increase of over 47 fold in favor of the bees. . 



A large number of carefully conducted experiments made by 



