104 NATURAL SCIENCE [August 
journey to varieties of the same species. In numerous cases I have 
seen bees visit two, three, and even four species in the course of a 
minute or two. The general results of my observations on this 
point are as follows :-— 
Hive bees are much more constant than wild hees, yet they pass 
freely from variety to variety, and not by any means rarely from 
species to species. As to the latter, take any wild bee, and if you 
can follow its movements for twenty visits or more, the chances are 
something like ten to one that it will be seen to change its species. 
of flower. If we suppose that the bee of the past acted as the bee 
of to-day, then it seems to me that in this habit alone we have a 
complete refutation of the theory. 
Another of the foundations of the theory is the benefit supposed 
to result from the cross-fertilisation effected by the bee in flying 
from flower to flower. Darwin’s well-known experiments on 
cross-fertilisation point to the conclusion that the seedlings of 
cross-fertilised plants are more numerous and vigorous than those 
of the self-fertilised. Without wishing to throw doubt on the 
general deductions from these experiments, I may be permitted to 
point out that certain facts regarding fertilisation in nature render 
them of doubtful support to the theory. First, there is the fact 
that certain species of flowers which are habitually self-fertilised are 
among the most numerous and vigorous of our native plants. Such, 
for example, are Polygonum aviculare, the least visited by insects, 
and yet the most abundant of its genus: Veronica hederaefolia, 
one of the commonest of the veronicas, yet very seldom visited by 
insects, as H. Miiller points out: while among the geraniums, G. 
molle and G. pusillum, which Miiller states to be the most fre- 
quently self-fertilised, and perhaps the most common of their genus 
with the exception of G. robertianwm. Professor Henslow, indeed, 
has gone so far as to state that “in nature whenever self-fertilisation 
can be effected more seed is borne than by the forms requiring 
crossing.” Among the orchids again, some species exhibit the most 
complicated arrangements for avoiding self- and securing cross- 
fertilisation ; others exhibit equal complications for securing the 
former and avoiding the latter. And if the inference is that the 
contrivances in the former case were evolved because  cross- 
fertilisation was an advantage, then it follows equally that in the latter 
case they were evolved because self-fertilisation was an advantage. 
Darwin, in accordance with his general views on cross-fertilisation, 
believed that such self-fertilised orchids were dying out, but the 
increased number of such now known seems to forbid this view, 
and it is difficult to understand how such self-fertilised orchids can 
have been evolved from a race specially fitted for cross-fertilisation 
on the supposition that this latter method is always beneficial. 
