70 NORMAL AND EXPERIMENTAL POLLINATION. 



Racemes horizontal. — Bombus juxtus landed at the corner of the 

 mouth, turned sideways, and then sucked nectar in the normal position with 

 relation to the flower. Osmia pentstemonis tried to enter 3 flowers with the 

 corolla tip pointing up, but after landing and starting in, it backed out 

 each time without finding the anthers. The next individual went into these 

 flowers apparently without noticing the change in position. 0. melanotricha 

 was puzzled by those that had the corner of the mouth pointing up. 

 It landed by placing the right legs on the lower outside lobe of the upper 

 lip and the left legs on one of the lower, thus straddling the angle of the 

 mouth, which pointed down. The bee then moved back and forth, trying to 

 rub, and some time elapsed before it discovered that the anthers were 

 missing, when it flew away. 



MUTILATION. 



Cotton at the corolla mouth. — Absorbent cotton was placed in the 

 corolla mouth, thus obscuring the anthers and blocking the entrance to 

 the nectaries. The amount of cotton used was small and it was placed in 

 position lightly, so that insects could easily push it aside and gain entrance 

 to the flowers if they made any effort to do so. Half of the racemes were 

 left normal and those with obstructed corollas were arranged at various 

 places in the cluster. Two individuals of Osmia pentstemonis hovered at 

 the corolla mouth, but did not try to force an entrance. Osmia phaceliae 

 attempted repeatedly to get in past the corolla mouth but did not succeed, 

 while Vespa germanica noticed the change just as it landed and did not try 

 to enter. 



Styles, stamens, and staminode removed. — In flowers thus modified, 

 Bombus juxtus started to enter one, but before turning upside down noticed 

 the difference and went to the next flower. Osmia bruneri flew to the 

 corolla mouth, but noticed the change and did not enter, while 0. phaceliae 

 reached the point of getting its head in position to suck nectar before 

 flying away. Another individual went into the flower, worked as usual, 

 and left, apparently without having noted the change. 



Corolla split. — When the corolla was deeply split between the lips, 

 Andrena vicina visited 3 flowers without noting the difference. The first 

 Halictus pulzenus that arrived walked around the flowers, but without 

 entering. Another landed on the lowest lobe of the corolla and wandered 

 back and forth over the edge. It then walked over the lowest lobe 5 times, 

 trying to find the way into the corolla, and then flew away without entering 

 at all. Some individuals of Bombus juxtus stopped and sucked nectar as 

 if they noted no change, while others flew away at once as if frightened. 

 Clisodon terminalis worked in the normal way, and both Vespa germanica 

 and Pseudomasaris vespoides went directly to the nectary, took nectar, 

 and departed. One Osmia pentstemonis landed and went at once to the 

 nectary, but flew away without sipping. Another alighted on the lower 

 petal and walked around, but since the anthers were projecting in front of 

 the flower and the corolla lips were somewhat reflexed, it failed to find the 

 anthers and the tip of the staminode, which it usually uses as a guide to the 

 nectary. Osmia phaceliae went to 9 flowers, among which was one with 

 the corolla lips separate. This bee advanced to the nectar, but suddenly 



