ough I cannot give
any of my reasons.

The Leguminosae are my greatest opposers: yet if I were to trust to
observations on insects made during many years, I should fully expect
crosses to take place in them; but I cannot find that our garden
varieties ever cross each other. I do NOT ask you to take any trouble
about it, but if you should by chance come across any intelligent
nurseryman, I wish you would enquire whether they take any pains
in raising the varieties of papilionaceous plants apart to prevent
crossing. (I have seen a statement of naturally formed crossed Phaseoli
near N. York.) The worst is that nurserymen are apt to attribute all
varieties to crossing.

Finally I incline to believe that every living being requires an
occasional cross with a distinct individual; and as trees from the mere
multitude of flowers offer an obstacle to this, I suspect this obstacle
is counteracted by tendency to have sexes separated. But I have
forgotten to say that my maximum difficulty is trees having
papilionaceous flowers: some of them, I know, have their keel-petals
expanded when ready for fertilisation; but Bentham does not believe
that this is general: nevertheless, on principle of nature not lying, I
suspect that this will turn out so, or that they are eminently sought by
bees dusted with pollen. Again I do NOT ask you to take trouble, but if
strolling under your Robinias when in full flower, just look at stamens
and pistils whether protruded and whether bees visit them. I must just
mention a fact mentioned to me the other day by Sir W. Macarthur, a
clever Australian gardener: viz., how odd it was that his Erythrinas in
N.S. Wales would not set a seed, without he imitated the movements of
the petals which bees cause. Well, as long as you live, you will never,
after this fearfully long note, ask me why I believe this or that.

LETTER 586. TO ASA GRAY. June 18th [1857].

It has been extremely kind of you telling me about the trees: now with
your facts, and those from Britain, N. Zealand, and Tasmania I shall
have fair materials for judging. I am writing this away from home, but
I think your fraction of 95/132 is as large as in other cases, and is at
least a striking coincidence.

I thank you much for your remarks about my crossing notions, to which,
I may add, I was led by exactly the same idea as yours, viz., that
crossing must be one means of eliminating variation, and then I wished
to make out how far in animals a