sent to Darwin by F. Muller; it
seems possible, therefore, that Hedyotis was written by mistake for some
other Rubiaceous plant, perhaps Oldenlandia, which John Scott sent him
from India.); the two other genera in the same sub-family with Villarsia
are dimorphic in Europe and Ceylon; a sub-genus of Erythroxylon (675/3.
No doubt Sethia.) is dimorphic in Ceylon, and Oxalis with you and at the
Cape of Good Hope. If you can find a dimorphic Oxalis it will be a new
point, for all known species are trimorphic or monomorphic. The case of
Convolvulus will be new, if proved. I am doubtful about Gesneria (675/4.
Neither Convolvulus nor Gesneria have been shown to be dimorphic.),
and have been often myself deceived by varying length of pistil.
A difference in the size of the pollen-grains would be conclusive
evidence; but in some cases experiments by fertilisation can alone
decide the point. As yet I know of no case of dimorphism in flowers
which are very irregular; such flowers being apparently always
sufficiently visited and crossed by insects.

LETTER 676. TO F. MULLER. Down, April 22nd [1867].

I am very sorry your papers on climbing plants never reached you. They
must be lost, but I put the stamps on myself and I am sure they were
right. I despatched on the 20th all the remaining copies, except one for
myself. Your letter of March 4th contained much interesting matter, but
I have to say this of all your letters. I am particularly glad to hear
that Oncidium flexuosum (676/1. See "Animals and Plants," Edition II.,
Volume II., page 114. Observations on Oncidium were made by John Scott,
and in Brazil by F. Muller, who "fertilised above one hundred flowers of
the above-mentioned Oncidium flexuosum, which is there endemic, with
its own pollen, and with that taken from distinct plants: all the former
were sterile, whilst those fertilised by pollen from any OTHER PLANT of
the same species were fertile.') is endemic, for I always thought that
the cases of self-sterility with orchids in hot-houses might have been
caused by their unnatural conditions. I am glad, also, to hear of the
other analogous cases, all of which I will give briefly in my book
that is now printing. The lessened number of good seeds in the
self-fertilising Epidendrums is to a certain extent a new case.
You suggest the comparison of the growth of plants produced from
self-fertilised and crossed seeds. I began this work last autumn, and
the result, in some cases,