ROBERTSON FLOWERS & INSECTS, ASCL. TO SCROPH. 595 



Smaller bumble-bees, and some other bees which never or 

 rarely try to suck, hang under the anthers and work out the pollen 

 by striking the trigger-like awns. They reverse of their own ac- 

 cord, since they are so small they are not compelled to do so on 

 account of the form of the flower. The tube is large and the nar- 

 row part only about lO mm. long, so that most bumble-bee work- 

 ers could easily reach the nectar if the tube were not curved in 

 the opposite direction from that of most flowers, and if the an- 

 thers did not obstruct the entrance. On one occasion, I saw 

 Bombus separatus S alight in the mouth of the corolla, and, not 

 knowing how to enter, fly away without obtaining any nectar. 

 Prof. W. W. Bailey mentions that small bumble-bees, which he 

 saw sucking at holes in the flower, were baflied when they at- 

 tempted to go in at the mouth. If these bees made the holes at 

 which they were sucking, it was no doubt because they did not 

 know how to reach the nectar in the proper way. 



Although the flower secretes nectar and is enabled to increase 

 its set of visitors in this way, the chief attraction is the pollen. 

 Early in the morning the bees are busy collecting the pollen, but 

 as soon as it is gone they are less attentive, aud the flower soon 

 withers. 



The following visitors were observed on eight days, between 

 Aug. 21 and Sept. 14: 



Hymenoptera. — Aj)idce., (i) Bombus virginicus Oliv., § , s. & c.p. ; 

 (2) B. separatus Cr. ^ , c.p. ; (3) B. vagans Sm. §,s., once; (4) B. 

 americanorum F. (^ 9 § 1 s. & c.p., ab ; (5) Megachile mendica Cr. $ , c.p. 



Birds. — Trochilidm, (6) Trochilus colubris L..once; front and base 

 of bill white with pollen. It is interesting to note that, while the bees 

 must alight and reverse to get the nectar, the humming-bird can obtain 

 it while on the wing. Halictus coiifusus Sm. 9 visits the flower for stray 

 pollen. 



Gerardia purpurea L. — The form of this flower resembles 

 that of the preceding, but the flower is much smaller and of a 

 handsome purple color. The upper wall of the corolla is straight, 

 while the lower is curved and longer, so that, when the flower 

 stands horizontally, the mouth looks upward and forward. As 

 in G. pedictilaria^ this form of the corolla requires the bees to 

 reverse to reach the nectar. On one occasion I saw a worker of 



