LECTURES. 29 
ined was Geranium sylvaticum (Wood Cranesbill). He noticed that 
little hairs covered the lower parts of the petals, and confident “ that 
a wise Author of Nature had not created a single hair without a de- 
finite purpose,” he soon assured himself that their purpose was to 
protect the honey of these flowers from rain for the sake of the in- 
sects which visited the ower. By a laborious course of experiment, 
he further assured himself that the majority of conspicuous flowers 
showed the following features (1) a honey gland or nectary which se- 
cretes honey ; (2) a receptacle for the honey; (3) a contrivance to 
shelter the honey from rain; (4) lines or spots to guide the insects in 
their search for honey; (5) means to prevent the flower from fertilising 
itself without the help of insects. He even recognised that insects 
carry pollen from one flower to another of the same species, and ob- 
served that the two sets of sexual organs in the same flower are often 
not developed simultaneously. So near was he to the recognition of 
the fact that self-fertilisation leads to worse results than cross-fertilisa- 
tion, but he failed to show why this uncertain fertilisation by means of 
insects was more advantageous than the obvious and certain self- 
fertilisation, and so the weakness of his theory threw all his patient 
observations and interpretations into obscurity. 
The subject was revived after more than half a century by Knight 
and Darwin, who laid down the law that in no plant does self-fertilisa- 
tion occur for an unlimited number of generations. This was found 
to be too wide, and later Darwin only concluded that in no organism 
does the structure or situation of the reproductive organs prevent an 
occasional cross. The discovery of the extreme ingenuity of devices 
to ensure cross-fertilisation tended for a time to throw too much stress 
upon it, and Henslow’s work on self-fertilisation, which pointed out 
that many of our most persistent species (e.g. chickweed and groundsel) 
depended on it, was a useful antidote. Axell’s statements that though 
cross-fertilisation is better than self-fertilisation, the latter is infinitely 
better than nothing at all, seem to sum up the case. Finally, H. 
Miller, by collecting all the researches of others and adding an im- 
mense number of his own (in his book he has a record of nearly 14000 
insects belonging to 2000 species which had been taken on flowers) 
set the seal to the conclusions of former observers, and once more re- 
stated Darwin’s law in a more restricted form, “ Wherever a cross- 
fertilised species comes into conflict with a self-fertilised one, it will 
always be victorious.” ‘The fact seems to be that a constant change of 
conditions of life is necessary for a species, so there is a struggle be- 
tween the new vigour which springs from cross-fertilisation, and the 
certainty of reproduction which self-tertilisation alone can give. 
