FERTILISATION OF VARIOUS' FLOWERS BY INSECTS. 16/5 
the carina not thus united, the style and stamens would not 
remain in the cavity ; for their elasticity tends to make them 
spring upwards, so that they are always pressing from within 
against the upper side of the carina. Consequently, if a flower 
be gathered and the carina be gently opened above with a 
needle, the pistil and stamens will be seen to spring from their 
prison. The same result will ensue if simply downward pres- 
sure be made on the carina, for this is so united to the axis as 
to admit easily of depression. Should, then, a bee settle on 
the upper surface of the carina, the weight will tend to depress 
this. This will assist the upward pressure of the organs within, 
and if the weight be sufficient they will break through the 
cohesion of the petals which opposes them and spring upwards, 
while the carina will sink downwards. In the mature flower 
the weight of a bee is often insufficient to effect this ; either the 
cohesion of the petals being stronger or the upward pressure of 
the style stamens being weaker than at a later period. So that 
one may frequently see a bee visit a just opened flower without 
liberating the reproductive organs. Probably these are not 
then in that state of maturity which would render their libera- 
tion advantageous. But if a flower be mature, the first time a 
bee visits it the carina is depressed, the stamens and pistil fly 
upwards and strike the under surface of the insect with some 
force. A little cloud of pollen is seen to be thrown off by the 
shock, some of which will adhere to the insect, and cannot fail 
to be frequently carried to another flower, while the rest is 
scattered over the neighbouring flowers, in many of which the 
stigmas are already exposed by previous visits of bees. 
I have spoken as though the bee settled directly on the 
carina. In reality it settles not on this, but upon the side 
petals — alae — the legs of tire one side of its body upon the 
one, those of the other side upon the other. These alae afford 
a better landing-place than would the narrow edge of the carina. 
The weight, however, still acts upon the carina as much as if 
the bee had settled directly upon it. For there are, on the 
outer surfaces of the carina and on the inner surface of each alae, 
certain projections and recesses which fit into each other, so 
that alae and carina are locked together, and depression of the 
former implies necessarily depression of the latter. 
In many species of Leguminosce this jointing together of alae 
and carina is replaced by actual coherence. Such, for instance, 
is the case with the Sweet-pea, Scarlet Runner, &c. In some 
species the upward spring of the stamens and pistil when set 
free from the carina is even more marked than in Ulex — as in 
Broom, or still more conspicuously in Mcdicago. In most, how- 
ever, no such motion can be observed. In these cases the 
contact of pollen and stigma with the bee’s body is brought 
