DELPHINIUM SCOPULORUM. 25 



a paper sack to protect the crepe-paper from the rain, this also preventing 

 'bees from securing the nectar until the time of observation. B. juxtus was 

 the only visitor and the several individuals went to 9, 3, 2, 3, 5 normal 

 flowers respectively, some flying low over the colored ones as if inspecting 

 them, but not alighting. 



Painted flowers. — A further study of the reaction of pollinators to 

 color was made by means of water-colors (plate 1). Three colors, green, 

 red, and yellow, were represented by 1 1 flowers each, and these were arranged 

 so that each one adjoined a normal flower. The first visitor was Bombus 

 juxtus, which went to all the green flowers once and to some twice, but 

 visited no others. B. proximus visited 3 normal flowers in one group, 4 

 flowers in a nearby cluster, and then 2 green flowers. The last visitor 

 was a very persistent individual of juxtus, which came to the flowers in 

 the following succession: 10 normal, 3 yellow, 2 green, 1 green, 3 green, 

 25 normal, 21 normal, 3 normal, 3 normal, 2 green, 4 normal, 4 green, 4 

 yellow. In the next experiment, 12 normal flowers and 6 each of red, 

 yellow, green, and brown, were used, but without being arranged in any 

 definite order. Three visitors came during the hour of observation with 

 the following results: 2 individuals of juxtus visited 17 and 19 normal flowers 

 respectively and edwardsi 3 normal ones, none paying attention to the colored 

 flowers. Five days later, when the same number of colored and normal 

 flowers were used in the same type of arrangement, the results were quite 

 different. The first visitor stopped at a normal flower only, but the next 

 one visited the flowers as follows: normal, green, blue, and normal, while 

 edwardsi visited 2 normal flowers alone. The last juxtus went to normal, 

 blue, yellow, normal, normal, red, blue, red, and blue flowers in order. 

 In the next experiment 5 flowers each were painted red, green, blue, yellow, 

 or purple, on both sides of the corolla, and 25 normal flowers were left on 

 the same stalk for comparison. B. edwardsi visited 127 normal and 3 blue 

 flowers, but no red, yellow, purple, or green ones, flying low enough over all 

 those painted to distinguish the colors. B. juxtus visited 116 normal, 4 

 blue, 2 green, and 1 yellow, but no red flowers. 



The total number of visits to painted flowers was 51, in contrast to 323 

 to normal ones, the ratio being 1 : 7. The individual range in behavior was 

 of the greatest, one juxtus visiting green flowers alone and another none but 

 normal ones, while a third did not discriminate between them and a fourth 

 went to twice as many painted as natural flowers. When the colors em- 

 ployed were red, yellow, and green, the latter was given a marked preference, 

 but when blue was added, this was visited more than all others combined. 



ADDITION OF ODOR. 



Perfume. — Rose perfume on absorbent cotton was wrapped around the 

 pedicels under the flowers on one-half of the stalk. The flowers perfumed 

 were on the side away from the wind, in order to make the odor more notice- 

 able on the one half. Bombus juxtus visited 9 normal and 2 perfumed 

 flowers, these being the only visits during 2 hours of observation. On 

 another day, carnation perfume was placed on cotton wads wrapped around 

 the pedicels of one-half of the stalk. Nearby, powder made from camphor 

 balls was sprinkled at the center of the stamens on flowers in a similar 



