34 CHELTENHAM COLLEGE NATURAL HISTORY SOCIETY, 
that pirate Bombus terrestris thrusts his proboscis through, and steals 
the honey from the yet unopened flowers. In the true roses the parts 
mature together, but cross-fertilisation must often occur owing to their 
great distance apart. I have no time to speak of the little self-fertilised 
Agrimony, or the very interesting Burnets, but here as in the Ranun- 
culacee, we do not find that cross-fertilisation necessarily means abun- 
dance. This is partly accounted for by the prevalence of succulent 
fruits, which are dispersed by birds, and runners which ensure change 
of ground and surroundings. 
Another order that I cannot pass over is Geranacee. It possesses 
the peculiar interest of containing the first flower that Springel ob- 
served, but apart from this it illustrates exceedingly well the law that 
a cross-fertilised flower will be more conspicuous than a nearly allied 
self-fertilised one. In Geranium pratense cross-fertilisation is brought 
about in this way.—The ten stamens stand round the unopened pistil, 
and form a convenient alighting place for an insect seeking honey. 
First five of them dehisce, and when they have shed their pollen, 
wither and fall down; then the other five dehisce, and after they are 
gone the stigma comes to maturity. Thus self-fertilisation is impos- 
sible, but a bee going from flower to flower is certain to effect cross- 
fertilisation. In G. pyrenaicum the pistil comes to maturity before the 
second lot of stamens have withered, and so will be fertilised by their 
pollen, if insects have not previously removed it. In G. molle five 
stamens mature, and then the stigma before the other five, so self- 
fertilisation is much more probable than in the last case. In G. 
pusilium the stamens are reduced to five, and the stigma matures be- 
fore them. There are in all four stages: the flower is female at first, 
then hermaphrodite for a time, then as the stamens overtop and cover 
the stigma it is male, and lastly when they have withered, it is once 
more female. Few insects visit the plant, so self-fertilisation generally 
occurs in the second stage. Here we can clearly observe the effect of 
cross-fertilisation. G. pratense is more than one inch in diameter, and 
bright blue, and the other species show every degree of degradation, 
to the tiny G. pussidum with its faint pink petals a quarter of an inch 
across. But for all this, the Dovesfoot cranesbill (G. AZod/e) is much 
more abundant than its brighter brethren. The Storksbill (Z7odzum) 
is fertilised much like G. AZol/e, but it has honey guides. H. Muller 
noticed a curious case of want of adaptation in an insect visiting it. 
A ladybird bent on securing honey, perched on a petal and made a 
grab, but in the effort it broke off the petal, and was left hanging to 
the next ; here it repeated the process, and so continued until it fell 
