468 The Ohio Naturalist. [Vol. IX, No. 5, 



Test I. May 1. Old apple tree in prime of bearing age, 

 just beginning to open blossoms. Number of blossoms on test 

 limb, 86. Number of blossoms on check limb, 88. Very cool 

 day and no insects working. May 5, warmer and insects work- 

 ing. May 16, found on test limb, 31 young fruits, some of them 

 dried and just ready to fall, while on the check limb adjoining, 

 55 fruits, most of them plump and growing. This makes the per- 

 centage of fruit setting 62.5% for the check limb, while on the 

 test limb with corollas removed only 36.04%. 



Test II. May 8. Wild Crab Apple Tree. Blossoms pink. 

 Corollas removed with fine pointed scissors. Test limb of 304 

 blossoms. Check limb of 200 blossoms. These blossoms were 

 not yet fully opened but were still partly enveloping the stamens 

 and pistils. On May 29, found on the check limb 140 fruits out 

 of the original 200, while an the test with corollas removed there 

 were onlv 145 out of the original 304, making the percentages 

 70% for the check and 47.6% for the test. 



On June 4, this was again visited and found on check 35 

 plump fruits and on test only 7, 17.5% for check and 2.3% for 

 test with corollas removed. 



Test III. May 9. Crab apple tree, near preceding test tree. 

 Test limb of 71 blossoms. Check limb of 72 blossoms. Flowers 

 not opened, but parts enveloped by corollas. Observations were 

 made on May 21, May 29, May 30 and June 4. Frost morning of 

 28th of May. On June 4 on check limb there were seven fruits 

 while on test only one, or 9.8% of fruit has set on the check 

 limb with corollas untouched while on the test limb with corollas 

 removed 1.4%. 



Test IV. Nasturtiums in the green-house. A large number 

 of flowers were secured for this test, there being three groups, 

 a middle group used for a check with corollas untouched and on 

 either side a second and third group from which the corollas 

 were kept removed. In these green house flowers no difference 

 was observed in test and check groups, fruit setting in as great 

 abundance on the group with corollas removed as on those flow- 

 ers with corollas left intact. 



The results of more tests could be given to show the same 

 lessening of the number of fruits setting on the groups of flowers 

 with corollas or other showy parts removed. From this we see 

 that a failure to set fruit on such flowers thus deprived cannot 

 be attributed to lack of insect visitations. 



WHY A FAILURE TO SET FRUIT? 



As may be surmised from what has previously been said of 

 the weather conditions as playing a part in the final results of the 

 experiment, it may be said that it is this factor that has played 

 so prominent a part, in that the flower has been deprived of a 



