82 THE FERTILISATION OF FLOWERS. [PART IIL. 
outer side with pollen; while in older flowers the same parts 
come in contact with the carpels which have elongated and spread — 
their stigmas more widely apart. Cross-fertilisation is the inevitable 
result. Thus the Columbine is admirably adapted for fertilisation 
by humble-bees: but to reach the honey they must have a 
proboscis at least 10 to 17 mm. long, even supposing that they 
stick their heads fully into the mouth of the spur and so shorten 
the passage by about 5 mm, This fully explains the actions of the 
insects that I have observed on this flower. Bombus hortorwm, 
L. 2 (with a proboscis 19 to 21 mm. long), is the only insect 
that I have found very abundant on it; B. agrorum, F. 2 (with a 
proboscis 12 to 15 mm. long) is found much more rarely, sucking 
the flowers in the regular way and accomplishing cross-fertilisation. 
I saw B. terrestris, L. 2 (with a proboscis 7 to 9 mm. long), fly on to 
the upper surface of a flower, and lick over the base of the sepals, 
and on finding nothing there, creep to the lower surface, and thrust its 
head into the spur. Then it again crawled on to the upper surface, 
and again licked fruitlessly the base of the sepals; and finally, 
biting a hole in the spur at the curved part, it introduced its 
proboscis and plundered the honey. Without further consideration 
it proceeded to secure the honey of the other petals and of other 
flowers in the same manner.. The numerous other individuals of 
B. terrestris, L. 9, which I before and afterwards observed thus 
perforating the spur, had probably first learned by trial how the 
honey might be won. 
I have often seen #. terrestris bite through the spurs of still 
unopened flowers, and so forestall all legitimate visitors. The 
hive-bee also, as Sprengel noticed, bites through the spur at the 
bend and steals the honey: it often takes advantage of the holes 
made by B. terrestris. | 
I have seen smaller bees, Halictus Smeathmanellus, K. 2, and 
HT. leucozonius, Schr. ? , collecting pollen on these flowers, which they 
would naturally fertilise by doing so. 
On a double garden-variety of the Columbine, in which seve- 
ral spurs are placed one within another, I noticed a hive-bee 
thrusting its head in the regular manner into the inner spur; and 
I could see, through the transparent flower, the tongue stretched 
— out to its fullest extent without reaching the honey. 
In default of insect-visits, self-fertilisation must easily take 
place, owing to the position of the parts of the flower. 
Aquilegia atrata, Koch, is also proterandrous, and visited by 
humble-bees (609, p. 137). 
