RUBUS DELICIOSUS — ROSA ACICULARIS. 101 



tive to the other bees as the raspberry and Scrophularia was about equal to 

 the latter. It is interesting that P. barbatus received neither visits nor 

 inspections from bees to most of which it was new, but this probably is to 

 be ascribed partly to color and partly to habit. In the second experiment 

 the honey-bee was again almost perfectly constant to Rubus, while the other 

 bees preferred Aquilegia in the ratio of 2:1. Bombus juxtus showed a 

 decisive preference for the latter, as Prosopis did for Mertensia; on the 

 contrary, the wasp Pseudomasaris went only to the raspberry. Taking all 

 bees into account, Rubus was much the most attractive, but, ignoring the 

 honey-bee, the attractive features of Pentstemon secundifiorus and Aquilegia 

 coerulea proved more potent than the control exerted by habit. 



RUBUS DELICIOSUS. 



Experiments. — In the studies with this species it was compared with 

 two plants of the same family, viz, Rosa acicularis and Prunus demissa. 

 The flower of the rose resembles that of the salmonberry closely in form, 

 size, and grouping, but contrasts sharply in color. On the other hand, 

 the flowers of the chokecherry, though white, are several times smaller 

 and arranged in close racemes. The color area is much the same in the 

 three, but in Prunus it is broken by the spaces between the flowers; the 

 raceme is much the most fragrant and produces more nectar, while the 

 amount of pollen appears to be much the same as in the single flowers of 

 the other two species. Since the plants grew in juxtaposition, it was only 

 necessary to bring the branches together to form an effective competition 

 group, though in the last case a bouquet of Rubus was used (table 69). 



Summary. — Of the total of 434 visitors, 302 went to Prunus, 120 to 

 Rubus, and 12 to Rosa. While this indicates the relative attractiveness, 

 it must be considered with reference to the fact that 387 of the total were 

 contributed by Andrena and Diptera. Had either Apis or Bombus been 

 abundant in the area at this time, the results would have been very different. 

 The marked preference of Andrena for Prunus is shown by the fact that 

 the latter obtained 225 visitors to 43 for Rubus and 4 for Rosa, all of the four 

 species exhibiting the same behavior. The preference was less marked in 

 the case of the flies, 75 going to Prunus and 40 to Rubus. The honey-bee 

 was all but perfectly constant to Rubus, 26 visiting this in contrast to one 

 to Rosa and none to Prunus. All told, the rose received less than a tenth of 

 the total number of visitors in the experiments concerned, affording clear-cut 

 evidence of the effect of habit in the face of greater total attraction. 



ROSA ACICULARIS 



Comparison. — Of the competitors employed, Rubus deliciosus most 

 nearly resembles the rose in size and form, but contrasts sharply with 

 its white color. Rubus strigosus and Opulaster opulifolius, though likewise 

 belonging to the same family and white in color, differ greatly in form, size, 

 and arrangement. Geranium caespitosum and Chamaenerium angustifolium 

 approach closely in color, but they are much smaller and yield much less 

 pollen, and this is largely true also for Cleome serrulata. Aquilegia coerulea 

 equals the rose in pollen production and excels it in the attractiveness of the 

 bright blue corolla, while the flowers of both species of Mertensia are rela- 



