PART 111. | THE MECHANISMS OF FLOWERS. 207 
Owing to the small elasticity of the ale and carina in the bean- 
| flower, they fail to return to their former place if they are 
thrust far down, but leave the end of the style with its stigma and 
i brushes and even the empty anthers exposed. 
Visitors: A. Hymenoptera—Apide: (1) Bombus hortorum, L. Q (21) ; 
(2) B. senilis, Sm. 9 (14—15) ; (3) B. confusus, Schenck, ? (14) ; (4) B. lapi- 
darius, L. 2 (12—14) ; (5) B. silvarum, L. 2? (14),—all sucking in the normal 
way, ab.; (6) B. terrestris, L. 9 (7—9), sucking honey through holes bitten 
the tube ; (7) Apis mellifica, L. 9 (6), sometimes c.p., sometimes sucking 
through the holes bitten by B. terrestris ; in the former case it effects cross- 
fertilisation as well as those bees which suck in the normal manner ; (8) An- 
drena convexiuscula, K. 9, c.p. ; (9) A. labialis, K. ¢, seeking in vain for honey ; 
(10) Osmia rufa, L. 9, sucking normally,—it creeps so far into the flower that 
its whole head is hidden under the basal part of the vexillum. B. Coleoptera 
_ --Malacodermata: (11) Malachius bipustulatus, F., feeds upon the stamens 
after they have been laid bare by repeated visits of humble-bees. 
Darwin found the fertility of the bean reduced to a third when 
insects were excluded by a net; but if the flowers were shaken 
| they produced good and well-filled pods though still protected from 
insects (152). 
Vicia amphicarpa, Dorthes., has, besides its ordinary flowers, cleis- 
_ togamic flowers without petals on subterranean shoots (399, 531). 
120. LATHYRUS PRATENSIS, L.—This plant affords us another 
example of a papilionaceous flower, in which when the carina is 
depressed the tip of the style only emerges, and sweeps part of the 
pollen out of the apex of the carina by means of a brush, applying 
it to the under side of the bee, afterwards returning within the 
carina when the pressure is removed. 
The style, which ascends vertically from the end of the 
horizontal ovary, curving slightly inwards, expands immediately 
below the oval stigma into an elliptic lamina. This lamina is 
x Bickod, not only on the edges, as Delpino says (178), but on the 
whole surface turned towards the base of the flower, with hairs 
| pointing obliquely upwards; it lies in the conical apex of the carina, 
which points directly upwards (5, Fig. 70), and its hairy surface, 
_ turned towards the base of the flower, faces also the free edges of 
if me tip of the carina. Between it and the margins of the carina is 
a deep pouch on each side (p, 5, 6), which is separated by a deep 
fold (a’) from the margin, and has its only entrance at the apex of 
the cone (m, 4, 5, 6). In the bud, this pouch contains all the 
_ anthers, which only dehisce immediately before or during the 
. coeemancas es ere 
