PENTSTEMON GLABER. 69 



Since there were about 7 times as many flowers in the normal position 

 as inverted, the latter were relatively about a third more attractive, the 

 ratio being 118:169. 



In another series of observations the behavior of certain individuals 

 was noted in detail, especially for Osmia, Vespa, and Apis, The first 

 individual of Osmia phaceliae landed on the new lower lip, then climbed to 

 the upper, went in upside down, and scratched its thorax back and forth 

 on the stamens. At the next inverted flower it landed in the same way, 

 looked around, and backed out. The second went to 4 inverted flowers, 

 where it scraped its thorax against the inverted staminode and then pushed 

 in to where the nectar should be. The third went straight in and tried 

 to find the opening to the nectary, but did not succeed. The fourth indi- 

 vidual entered the inverted flowers just as the normal ones and stayed the 

 same length of time, while the fifth attempted to enter 3 inverted flowers 

 without success. The sixth went into 3 inverted flowers and obtained 

 nectar with but little more difficulty than in normal ones, but the seventh 

 entered upside down, hence in its usual position with respect to the flower 

 parts, and consequently worked as usual. Four others then did the same 

 in succession. Osmia phaceliae often went into inverted flowers and 

 scratched its thorax against the staminode, evidently without realizing 

 that it was not touching the anthers. Some of this species at once recog- 

 nized that the flowers were inverted and landed accordingly. Osmia 

 melanotricha also went into such flowers upside down and secured nectar. 

 It entered and scraped pollen as in normal flowers. 



Vespa germanica twice tried unsuccessfully to find nectar in these flowers. 

 It landed, but before entering noticed something different and went to the 

 next flower, which was also inverted, where it behaved in the same way. 

 A second individual entered hesitatingly in such a manner that it was upside 

 down and then took nectar. A third one flew above and examined 6 inverted 

 blossoms but did not land, while a fourth wasp entered the flower, and 

 reached into the nectary, but did not stay long enough to get nectar. It then 

 flew past the mouths of other inverted flowers, but did not land. Apis melli- 

 fica went into 2 inverted flowers, but departed at once. It stopped several 

 times at such flowers later, but noticed the difference and flew away. 

 Finally, it came back to a stem bearing inverted flowers and obtained pollen 

 without standing upside down. Another individual stood upright and 

 gathered pollen from the anthers underneath these flowers instead of 

 standing upside down. Clisodon terminalis visited 10 normal flowers 

 and 4 inverted ones, but obtained nectar in but 2 of the latter, while Andrena 

 vicina secured it as usual. 



When the two lobes of the lower lip in an inverted flower were cut into 

 separate petals, Osmia melanotricha landed easily, but the stamens and style 

 which projected in front of the flower confused it so that it could not find 

 the entrance to the nectar. Osmia pentstemonis likewise landed on the in- 

 verted upper lip, but did not find the anthers and nectar. In the case of 

 inverted flowers with the corolla mouth closed by wilting, it succeeded in 

 opening the corolla with difficulty, but failed to find the anthers and so 

 flew out at once. 



