or three numbers together. I am really ashamed to think of your having
given me such a valuable work; all I can say is that I appreciate your
present in two ways--as your gift, and for its great use to my species-
work. I am very glad to hear that you mean to attack this subject some
day. I wonder whether we shall ever be public combatants; anyhow, I
congratulate myself in a most unfair advantage of you, viz., in having
extracted more facts and views from you than from any one other person. I
daresay your explanation of polymorphism on volcanic islands may be the
right one; the reason I am curious about it is, the fact of the birds on
the Galapagos being in several instances very fine-run species--that is, in
comparing them, not so much one with another, as with their analogues from
the continent. I have somehow felt, like you, that an alpine form of a
plant is not a true variety; and yet I cannot admit that the simple fact of
the cause being assignable ought to prevent its being called a variety;
every variation must have some cause, so that the difference would rest on
our knowledge in being able or not to assign the cause. Do you consider
that a true variety should be produced by causes acting through the parent?
But even taking this definition, are you sure that alpine forms are not
inherited from one, two, or three generations? Now, would not this be a
curious and valuable experiment (16/1. For an account of work of this
character, see papers by G. Bonnier in the "Revue Generale," Volume II.,
1890; "Ann. Sc. Nat." Volume XX.; "Revue Generale," Volume VII.), viz., to
get seeds of some alpine plant, a little more hairy, etc., etc., than its
lowland fellow, and raise seedlings at Kew: if this has not been done,
could you not get it done? Have you anybody in Scotland from whom you
could get the seeds?

I have been interested by your remarks on Senecia and Gnaphalium: would it
not be worth while (I should be very curious to hear the result) to make a
short list of the generally considered variable or polymorphous genera, as