part ut] THE MECHANISMS OF FLOWERS. 505 
lanceolata which I have gathered at Lippstadt, and on which as 
well as on the intermediate forms I have observed pollen-collecting 
insects, are: (1) a short form with short spike from the sunny 
caleareous slopes of the Rixbecker Hill; (2) a tall, long-spiked 
form from the rich alluvial meadows of the Lippe. 
(1) In the most stunted specimens from the hill the scape 
is scarcely four inches long, the spike is spherical, about 5 mm. in 
diameter, and the individual flowers are convex anteriorly, and 
flat or slightly concave on the side turned towards the axis; of the 
four sepals the two inferior are coherent almost to the tips (Fig. 
170, 5). While the lobes of the corolla are still erect, contiguous 
(Fig. 170, 2), and covered by the sepals (1), and while the stamens are 
still inclosed in the bud, with their anthers not nearly ripe and their 
filaments scarcely beginning to develop (3), already the stigma is 
tully mature, and protrudes for more than 1 mm. beyond the bud 
(Fig. 170, 1, 2). As the stigma becomes withered and brown, 
the filaments begin to grow rapidly, and the other parts of the 
flower more slowly. At length the lobes of the corolla fold back, 
and the stamens with the now ripe anthers protrude 5 to 6 mm. 
from the flower, which is itself now 3 (in the female period only 2) 
mm. long (Fig. 170, 4). The filaments are quite straight, but thin 
and flexible enough to be swayed backwards and forwards by the 
slightest breeze, scattering a cloud of dry pollen. Nevertheless, 
this pollen is collected by the honey-bee. The pollen-grains are 
smooth, ‘016 to ‘(020 mm. in diameter when dry, and swelling to 
a diameter of ‘021 to ‘026 mm. when moistened. 
(2) The tall form from the manured alluvial meadows has 
a scape a foot to eighteen inches long, a spike 15 to 30 mm. long, 
and stamens projecting 6 to 7 mm. from the flower. In the com- 
plete proterogynous dichogamy, in the cohesion of the two lower 
sepals, in short in all other characters, it agrees perfectly with the 
dwarf form from the Rixbecker Hill. I have observed pollen- 
seeking insects indifferently on both extreme and intermediate 
forms. 
A. Hymenoptera—Apide : (1) Apis mellifica, L. $, very ab.,c.p. I wrote 
the following account of the honey-bee’s actions in the summer of 1868, and 
assured myself of its accuracy in the three following summers: ‘‘ The honey- 
bee flies buzzing to a spike, and while it hovers in the air it spits a little honey 
on the exserted anthers. Then, still hovering and buzzing, it brushes pollen with 
the tarsal brushes of its forefeet off the anther, the tone of its hum becoming 
suddenly higher ; in the same instant one sees a)cloud of pollen rise from the 
shaken anthers. After placing the pollen on ith hindlegs the bee repeats the 
operation on the same or other spikes, or if it is tired it alights on the spike 
