on the subject is given in
the "Life and Letters," II., pages 102-5.)

Down, August 22nd [1857].

Your handwriting always rejoices the cockles of my heart; though you
have no reason to be "overwhelmed with shame," as I did not expect to
hear.

I write now chiefly to know whether you can tell me how to write to
Hermann Schlagenheit (is this spelt right?) (53/2. Schlagintweit.), for
I believe he is returned to England, and he has poultry skins for me
from W. Elliot of Madras.

I am very glad to hear that you have been tabulating some Floras about
varieties. Will you just tell me roughly the result? Do you not find it
takes much time? I am employing a laboriously careful schoolmaster, who
does the tabulating and dividing into two great cohorts, more carefully
than I can. This being so, I should be very glad some time to have Koch,
Webb's Canaries, and Ledebour, and Grisebach, but I do not know even
where Rumelia is. I shall work the British flora with three separate
Floras; and I intend dividing the varieties into two classes, as Asa
Gray and Henslow give the materials, and, further, A. Gray and H.C.
Watson have marked for me the forms, which they consider real species,
but yet are very close to others; and it will be curious to compare
results. If it will all hold good it is very important for me; for it
explains, as I think, all classification, i.e. the quasi-branching and
sub-branching of forms, as if from one root, big genera increasing and
splitting up, etc., as you will perceive. But then comes in, also, what
I call a principle of divergence, which I think I can explain, but which
is too long, and perhaps you would not care to hear. As you have been on
this subject, you might like to hear what very little is complete (for
my schoolmaster has had three weeks' holidays)--only three cases as yet,
I see.

BABINGTON--British Flora.

593 species in genera of 5 and 593 (odd chance equal) in
upwards have in a thousand genera of 3 and downwards have
species presenting vars. in a thousand presenting vars.
134/1000.* 37/1000.

(*53/3. This sentence may be interpreted as follows: The number of
species which present varieties are 134 per thousand in genera of 5 species
and upwards. The result is obtained from tabulation of 593 species.)

HOOKER--New Zealand.

Genera with 4 species and With 3 species and downwards
upwards, 150/1000.