RELATIVES OF THE HERB ROBERT 1389 
although not entirely, free from the attacks of grubs, 
molluses, fungi, and other pests, but the Meadow and 
the Cut-leaved* Cranesbill are sometimes spotted 
with rust and have their leaves eaten by caterpillars, 
snails, and slugs. I have also seen the Dovesfoott 
and the Round-leaved Cranesbill distorted by what 
was, I think, fungal disease. 
At one time or another I have had nearly all 
the British species and some of their varieties in my 
garden, but it was quite the exception for any of them 
to be eaten, at least until the autumn. The dying 
and the dead leaves are devoured by molluscs, and the 
almost omnivorous larve, such as the caterpillar of 
the Great Yellow-underwing, resort to the foliage 
when other food is scarce. 
There is no vegetative propagation, reproduction 
being effected only by seed, although at any rate 
some of the perennial species can be propagated 
artificially by division. 
The structure of the flower is of course the same 
throughout the genus, but it goes without saying that 
there are many differences in the details. There 
is, too, great uniformity in the pollination modes, 
cross-pollination being the rule, but the Small Doves- 
foot { has only five perfect stamens instead of ten, and 
it is much more dependent upon self-pollination than 
its brethren. 
The genus is of great interest in this respect. 
It was in the Wood Cranesbill that insect-pollination 
was observed, although its full significance was 
not understood, by Conrad Sprengel in 1787, one 
hundred and thirty-one years ago; and moreover 
* Geranium dissectum, Linn. ~ Geranium molle, Linn. 
t Geranium pusillum, Linn. 
