64 NORMAL AND EXPERIMENTAL POLLINATION. 



side to the position in which it was accustomed to work, looked around, and 

 found no slit. It had difficulty in holding on to the under side in the absence 

 of the slit. It then moved to the dorsal side and found a few grains of 

 pollen, but did not seem satisfied. Again it tried the ventral side of the 

 corolla and looked all around for the slit and the lower lip. It went to the 

 dorsal side again, but failed to locate the slit. For a third time it moved 

 to the under side of the corolla, but not finding what it was looking for, 

 flew away to a flower in a normal position and located the slit at once. 

 A fourth one landed with its head toward the style, walked to the end of the 

 latter and over it, the style touching the pollen loads on its legs, and then 

 walked back again. It worked for several seconds with its head and the 

 first pair of legs in the attempt to open the slit, and finally succeeded. 

 A fifth individual alighted on the lower lip and immediately went down into 

 the corolla and sucked nectar. Then it turned around, walked out to the 

 end of the style, which swayed under its weight, and around it. It returned 

 to the end of the corolla and then backed out onto the stamens, where it 

 collected pollen, placing it on the hind legs before flying away. 



Selasphorus platycercus landed upon a low tree branch and sucked 

 nectar from the nearby Castilleia flowers, while resting its wings. Normally, 

 it keeps its feet close to the body and sustains itself in the air by rapid wing 

 motion while sucking nectar. In one flower the corolla tube was too small 

 even for its slender bill. The bird found no trouble in thrusting the bill 

 into the corolla, but was unable to pull it out again. It then lowered 

 itself so that the corolla tube became pendant and sipped more nectar, 

 finally freeing itself after further effort. 



The normal visitors to Castilleia are not numerous, observations for a 

 period of 2 hours and 20 minutes yielding only the following: Halictus 

 pulzenus 6, Sphecodes sp. 2, Selasphorus platycercus 3. 



Merritt (1897:21) states that the flowers of Castilleia affinis have abun- 

 dant honey, and that pollination is usually effected before the flowers attain 

 their full length. The exserted capitate stigma is struck by the visiting 

 humming-bird, while the bill receives a fresh supply of pollen from the 

 anthers just below in the galea. Longyear (1909:79) has described the 

 structure and pollination of the flower of C. linarifolia. 



EXPERIMENTS. 



MUTILATION. 

 Upper lip removed, spike inverted. — The upper lip of the corolla 

 was cut off even with the lower, the bract cut back to the same length, 

 and the cluster then inverted so that the lower lip of the flower was turned 

 up. Halictus pulzenus landed and remained quiet at the lower lip for 

 some time, with its head inside the corolla. It was probably sucking 

 nectar, but there was no movement of the abdomen to indicate this. It 

 then walked out on the filaments to the anthers and ate the pollen that 

 was still there. It next went to the end of the style, arching its body over 

 in such a way that the stigma rubbed against the abdomen. 



PAINTED FLOWERS. 

 Painted bracts. — The bracts of various clusters were painted purple, 

 green, or yellow with water-colors. A humming-bird sucked nectar from 



