RUBUS DELICIOSUS. 31 



Table 12. — Visits to natural flowers and flowers with fringed paper corollas. 



Species. 



Normal. Red. Yellow. Green. Blue 



Apis mellifica 



Bombus bifarius . . 



juxtus. . . 



proximus . 

 Andrena crataegi . 

 Osmia phaceliae . . 

 Musca sp 



Total. 



Bombus proximus. . . . 



bifarius 



Andrena crataegi .... 



madronitens. 

 Syrphus amerieanus. . 

 Halictus pulzenus. . . . 



Prosopis elliptica 



Musca sp 



Total . 



106 

 39 

 12 



ADDITION OF HONEY AND ODOR. 



Honey. — In this installation one-third of the flowers were supplied with 

 a drop of honey at the base of the filaments, another third with a drop at the 

 center of the stamen mass, and the remainder were left normal. Halictus 

 pulzenus visited 4 normal flowers and 10 with honey at the filament base. 

 Andrena crataegi visited but one flower of this sort; it actually landed at 

 others but flew away at once, as did Bombus proximus. The total number 

 of visits to normal flowers was 46, in contrast to 11 for those with honey 

 added, giving a ratio of 4:1. 



SUMMARY. 



A comparison of the tables obtained from the various calendars not only 

 reveals the differences arising out of time and place, but also throws light 

 upon discrepancies between the results of different observers and investi- 

 gators, making it clear that comparison is often misleading if not impossible 

 when different years and regions are concerned. There was much difference 

 in the number of visits on successive days in the same spot, and a marked 

 one when different portions of the flowering period were concerned, arising 

 chiefly out of the period of flight for the various genera. Neighboring 

 areas likewise showed a distinct divergence, which was strikingly evident in 

 the ratios of visits for the bees, namely, Apis 1:6, Bombus 17:1, Andrena 

 7:1, and Osmia 3: 1. 



The decisive effect of cutting the petals back to half their length is espe- 

 cially significant in connection with the question sometimes raised as to 

 the value of the corolla in attraction. Covering the stamens excluded 

 visits entirely and suggests that insects distinguish clearly between the 

 corolla and the stamen mass, or at least readily recognize that the latter is 

 changed. Although 49 visits were made to paper flowers of various colors, 



