214 THE FERTILISATION OF FLOWERS. (PART 111. 
excludes all insects from the honey which are not strong enough 
to perform the actions necessary for fertilisation. 
Along with these obvious advantages, the firm closure of the 
flowers has the very important consequence that it makes the 
work difficult even for such bees as are able to reach the honey and 
to effect cross-fertilisation, and deters them from the plant when 
other more convenient flowers are at hand. In its original home 
the Pea no doubt adapted itself to some strong and at the same 
time diligent and skilful species of bee, which could easily depress 
the carina, and was plentiful enough in ordinary weather to act as 
the regular fertilising-agent. Under such conditions the advan- 
tages of firm closure would outweigh the disadvantages. In our 
climate, the Pea fails to find bees adapted for its flower, and it 
would be much better for it under these altered conditions to have 
its flowers less firmly shut. I have often watched beds of peas 
in bloom in my garden in sunny weather and have only occasionally 
seen a visitor, while beans, blooming at the same time in alternate 
beds, were abundantly visited by humble-bees. 
The only insects which I have seen on the flowers in the course of four 
summers are: (1) Eucera longicornis, L. ; (2) Megachile pyrina, Lep ;—the ¢ 
of both species, s.; the 9, s. and c.p.; both freq., but not abundant ; (3) Two 
specimens of Halictus sexnotatus, K. 9 ;—they collected pollen with difficulty, 
holding the edges of the carina apart anteriorly with their legs. 
Though most flowers remain unvisited by insects, they all pro- 
duce good fruit. The self-fertilisation which the structure of the 
flower necessitates must therefore be quite efficient; and this 
indeed has been shown experimentally by Dr. Ogle, who found the 
Pea as productive when insects were excluded as when left 
unprotected (633). 
Tribe Phaseolec. 
Amphicarpea, according to Torrey and Asa Gray (NV. Amer. 
Flora, t., p. 291), has fertile cleistogamic flowers, and also flowers 
which open, but are for the most barren. Both kinds of flowers 
are above ground. Darwin found that subterranean pods of 
Amphicarpea monoica which he received from Meehan, contained 
each a single seed, while the ordinary aérial pods, which he culti- 
vated himself, contained from one to three small seeds; these latter 
averaged only 7's of the weight of the subterranean seeds (167, 
2nd Ed.). 
The genera Neurocarpum, Desv., Martiusia, Schult., Glycine, L., 
