4 
264 BOTANICAL GAZETTE. | Oct., 
claimed for the view maintained in this paper, that it will account 
for pollination as readily in the one case as in the other and by 
the same mode. : 
H. Miller, who supposes that the whole foot enters the stig- 
matic chamber, says: “When the insect tries to draw its foot 
out in order to proceed further, the diverging claws are caught 
by the apposed edges of the anther-wings, and guided upwards 
in the slit, so that one or other of the two claws is brought with- 
out fail into the notch in the lower border of the corpusculum 
and there held fast®.” On the same subject Mr. Corry observes : 
“When the foot reaches the superior end of the alar chamber in 
which it has been guided, one at least of the two hooked claws 
upon it, or some part of the foot in the case of Diptera, must 
easily enter the hollow cavity of the corpusculum, which lies in 
such a position that this result is inevitable’.”. The importance 
which these authors attach to the view that the whole foot enters 
the chamber, in my opinion, rests on a misunderstanding of the 
mode of insertion of the pollinia, and has led them to overlook 
the precision with which a corpusculum comes to be fastened to 
a hair or claw. The corpusculum is placed so nicely at the top 
of the wings that its cleft is fairly continuous with the slit be- 
tween them (fig. 2 ¢.), and I can not conceive that the contriv- 
manner as a hair or spur. In a careful examination of the feet 
of 116 hive-bees which were killed by being caught on the flow- 
ers of A. Sullivantii, I have found that, with but two exceptions, 
when a foot was held by the wi ngs, only one claw was between 
them, the other being free, or less often the pulvillus was held 
between the wings and both claws were outside. 
When first withdrawn the pollinia lie in the same plane (fig. 
4.) Ina few minutes the twisting of the retinacula brings the 
pollinia into nearly parallel planes, but the upper ends are still 
®Tbid, p. 398, 
7Ibid, p. 188, 
