I am now strongly inclined to believe that sterility is at first a selected quality to keep incipient species distinct. If you have looked at Lythrum you will see how pollen can be modified merely to favour crossing; with equal readiness it could be modified to prevent crossing.
It is this which makes me so much interested with dimorphism, etc. (152/2. This gives a narrow impression of Darwin's interest in dimorphism. The importance of his work was (briefly put) the proof that sterility has no necessary connection with specific difference, but depends on sexual differentiation independent of racial differences. See "Life and Letters," III., page 296. His point of view that sterility is a selected quality is again given in a letter to Huxley ("Life and Letters," II., page 384), but was not upheld in his later writings (see "Origin of Species," Edition VI., page 245). The idea of sterility being a selected quality is interesting in connection with Romanes' theory of physiological selection. (See Letters 209-214.))
One word more. When you pitched me head over heels by your new way of looking at the back side of variation, I received assurance and strength by considering monsters--due to law: horribly strange as they are, the monsters were alive till at least when born. They differ at least as much from the parent as any one mammal from another.
I have just finished a long, weary chapter on simple facts of variation of cultivated plants, and am now refreshing myself with a paper on Linum for the Linnean Society.
LETTER 153. TO W.B. TEGETMEIER.
(153/1. The following letter also bears on the question of the artificial production of sterility.)
Down, 27th [December, 1862].
The present plan is to try whether any existing breeds happen to have acquired accidentally any degree of sterility; but to this point hereafter. The enclosed MS. will show what I have done and know on the subject. Please at some future time carefully return the MS. to me. If I were going to try again, I would prefer Turbit with Carrier or Dragon.
I will suggest an analogous experiment, which I have had for two years in my experimental book with "be sure and try," but which, as my health gets yearly weaker and weaker and my other work increases, I suppose I shall never try. Permit me to add that if 5 pounds would cover the expenses of the experiment, I should be delighted to give it, and you could publish the result if there be any result. I crossed the Spanish cock (your bird) and white Silk hen and got plenty of eggs and chickens; but two of them seemed to be quite sterile. I was then sadly overdone with work, but have ever since much reproached myself that I did not preserve and carefully test the procreative power of these hens. Now, if you are inclined to get a Spanish cock and a couple of white Silk hens, I shall be most grateful to hear whether the offspring breed well: they will prove, I think, not hardy; if they should prove sterile, which I can hardly believe, they will anyhow do for the pot. If you do try this, how would it do to put a Silk cock to your curious silky Cochin hen, so as to get a big silk breed; it would be curious if you could get silky fowl with bright colours. I believe a Silk hen crossed by any other breed never gives silky feathers. A cross from Silk cock and Cochin Silk hen ought to give silky feathers and probably bright colours.
I have been led lately from experiments (not published) on dimorphism to reflect much on sterility from hybridism, and partially to change the opinion given in "Origin." I have now letters out enquiring on the following point, implied in the experiment, which seems to me well worth trying, but too laborious ever to be attempted. I would ask every pigeon and fowl fancier whether they have ever observed, in the same breed, a cock A paired to a hen B which did not produce young. Then I would get cock A and match it to a hen of its nearest blood; and hen B to its nearest blood. I would then match the offspring of A (viz., a, b, c, d, e) to the offspring of B (viz., f, g, h, i, j), and all those children which were fertile together should be destroyed until I found one--say a, which was not quite fertile with--say, i.