WCP2975

[1]

Netherby

Victoria Road, S.

Southsea.

Jan 5th 1910.

Dear Dr Wallace

I think you have found my letter difficult only because you have thought that there was more in ^it my letter than there really was. Dont[sic] read this if you find it at all difficult, and above all dont[sic] answer it. Isn't all the following true in normal breeding — e.g. human breeding.

(1) Offspring tend to blend parental non-sexual characters — e.g. colour of skin, shape of hands.

(2) But they do not blend sexual characters — e.g. [illeg.] sexual traits — except in rare abnormal cases — e.g. human hermaphrodites (so called).

[2]

(3) The sexual characters of the male are latent in the female & vice versa — e.g. the daughters of heavily bearded men tend to have heavily bearded sons, bulls transmit the good milking qualities of their mothers, aphides in summer transmit the male traits, sometimes females develop male characters in old age.

(4) It is reasonable [to] suppose that if blending is the rule in the non-sexual characters, it is also the rule in the more sexual characters; but here the latent male characters of the female blend with their homologues in the male — i.e. his patent traits, & vice versa. That is to say the patent uterus of the female blends with the (latent) uterus of the male, & so on. This, of course, is quite possible for ^all blending occurs in the germ plasm. Sometimes blending occurs abnormally between the female & the male characters as and this is especially the case when [3] varieties are crossed. Sometimes also, especially when varieties are crossed, male & female characters are mixed, as has happened in the case of bees, which when crossed have ^{produced individuals} been partly male & partly female.

(5) Natural Selection has established the sexual characters in sets — a male set & a female set — and keeps them ^sets separate (one patent & the other latent). When, abnormally, some male & some female characters are patent together, the individual tends to perish. Man, however by artificial selection has succeeded in several cases in transferring male ^characters to females characters & vice versa, as in the case of the chest[?] of the Polish fowl & the plumage of the Sebright bantam. They are then called Mendelian characters when varieties are crossed.

(6) Now when a mutation appears it tends to be eliminated in nature for it does not fit in with the other characters, but under artificial selection it is often carefully preserved. There is ample evidence that the [4] inheritance of mutations follows the mode of sexual characters mode. That is to say ^{they are} it is patent or latent in descendents. Thus suppose, in a variety of wild-grey rabbits, an albino appears & is crossed with one of the ordinary type, then the descendents tend to be grey or white, not blends. ^{The white or grey may be latent for long periods & then} reappear, as the male traits in aphides¹.

(7) There is, however, this distinction between the inheritance of mutations & sexual characters. Nature has established the latter through ages of selection. Therefore they rarely blend (the male ^traits with the female) & they do not intermix (that is, they remain in their seperate[sic] male & female sets). But mutations, not having been established by selection tend more to blend. Therefore, in Mendelian traits there is almost always some blending. Moreover there is no inheritance in distinct sets. Thus the albino colone[sic] appears indifferently (in patent) in male & female. This is the so-called independent inheritance of which the Mendelians make so much. Lastly, the alternation of sexual characters, having been established by natural selection is more perfect & regular.²

[5]

(8) As an albino may appear in a ^pure race[?] of grey rabbits by m mutation, so a black individual may appear. But if these are crossed we with the grey — as they always are for mutations are rare — then grey is latent in the albino descendents of the one & the black descendents of the other. If then a white descendent of the one is crossed with a black descendent of the other, grey reappears. The conditions under which latency occurs are disturbed as on the advent of cold weather disturbs the conditions under which the male characters of aphids are latent.

(9) Grey is what is termed dominant. That is, if a grey rabbit is crossed with a white or a black, or if a white is crossed with a black, the grey tends more [6] than any other colour to appear. The dominant is usually the ancestral trait which has been established by selection. Almost all human abnormalities are transmitted in this way — hair-lip, club-foot, imbecility, deaf-mutism, haemophilia, colour-blindness. The last two however, as a rule, only become patent in males, showing the close connection between Mendelian & sexual traits. Normality is dominant.

(10) Artificial selection has been in many cases been founded on the selection of mutations. But not in all cases. Thus race-horses have been selected by the test of speed ^{in which thousands of structures co-operated}. A mutation would throw the animal out of gear. ^{It would ruin}[?] ^{co-adaptation} Therefore the selection here has been of fluctuations — just as nature selects. Therefore, if you cross a race-horse with an ordinary horse, the result is a blend. You dont[sic] get some of the descendents race horses & some ordinary horses.

[7]

(11) When we select mutations we render the ancestral characters latent. When we cross artificial varieties we disturb the conditions of latency & the ancestral trait reappears. Therefore, the reappearance of ancestral traits is common when artificial varieties are crossed. But nature eliminates mutations. Therefore they ^{ancestral traits} are not found when natural varieties are crossed. Mendelians & mutationists have experimented almost exclusively with artificial varieties. Hence these results. They have done nothing more than study those abnormalities of sexual inheritance — or rather reproduction, for the traits that remain latent are reproduc inherited as well as those that become patent — which occur under conditions of artificial selection³.

(12) The proof that there is alternative latency & patency of mutations (not alternative inheritance) is afforded by the fact that pure-bred domestic varieties (e.g. pigeons) sometimes reproduce the ancestral trait. We do not [8] know what has disturbed the conditions of latency here; but at any rate there ^{has been} is latency.

(13) When we cross natural varieties latent ancestral traits never reappear. But if the varieties are sharply contrasted in any particulars (i.e. if the differences are big like sexual differences) the reproduction of them may be Mendelian. The most interesting case I know is that of human eye-colour. Black is dominant. If we cross white with negro the offspring are all black-eyed. And thus black colour differs continues indefinitely in mulattos so long as they breed together. But after repeated infusions of white blood (the black being very strongly dominant) the blue eye suddenly appears & descendents have black or blue eyes, not a blend., Eye-colour, as shown by its attractive force on us, is largely a sexual trait. Except North Europeans & their descendents all ^{varieties of} animals have eyes of a uniform colour. This shows that North Europeans are a [9] mixed race — mixed by those old invasions of prehistoric & early historic times, Celts, Goths, Huns, &c.

(14) Now surely there is vast significance in the fact that latent ancestral traits are so common in va artificial varieties & so rare or unknown in natural varieties, and also in the fact that Mendelian reproduction is common in artificial varieties, when crossed & relatively very rare when natural varieties are crossed.

(15) We must admit that sexual traits are not reproduced as non-sexual traits are. There is alternative latency & patency here. There is clear evidence that the same is the case with Mendelian characters. You will agree with me [10] that there is clear evidence that natural varieties & species have originated through the selection of fluctuations. You will agree also that much there has been much selection of mutations in the creation of artificial varieties. ^{e.g. Ancon & Manchaump}[?] ^sheep. There is clear evidence that the inheritance of mutations tends to be Mendelian (i.e. in the sexual mode). Is it not clear, then, that in Mendelian "inheritance" we have rarely a case of non-sexual traits transmitted in the sexual mode. Herein lies the distinctive peculiarity of artificial selection. And because Mendelians and mutationists have studied practically only artificial varieties they have been deluded into the belief that all inheritance is Mendelian, and all evolution founded on mutations. That is what I meant when I declared that they are victims of a vast practical joke played by the human breeder.

[11]

(16) The points I venture to ask you to consider are these

(a) The resemblance between the reproduction of sexual and Mendelian characters — the lack of blending & the alternative patency & latency, both being more perfect in the case of the sexual traits since they have been established by Nat. Select. Moreover sexual characters are caused by Nat. Select. to have together in male & female sets; not so mutations.

(b)The fact that man does tend to select mutations the inheritance of which is Mendelian, but nature does not.

(c) The fact that blending is common when natural varieties are crossed, but alternative reproduction when artificial varieties are crossed.

(d) The fact that natural varieties, more display latent ancestral traits when crossed.

Surely in view of all these facts we cannot but conclude that the peculiarity of artificial selection is that it has [12] established a lot of abnormalities the inheritance of which is in the sexual mode.

The "Central doctrine of Mendelism is segregation" (Bateson⁴). If I can prove that there is no segregation (no alternative inheritance), but only patency & latency (alternative reproduction), then Mendelism & the Mutation theory are smashed. It will have been shown that these people have been merely studying the abnormalities of sexual reproduction that have arisen under artificial selection.

Please forgive me for writing at this length. And do not answer on any account. I am already sufficiently ashamed. I will tell my publisher to send you an early copy of my book. You will see there how conclusive is the evidence

Yours very truly | G. Archdall Reid⁵. [signature]

We believe that living beings are adaptational [illeg.]. We can explain mutations as exaggerated variations. But how are we to explain Mendelian reproduction except by assimilating[?] it to sexual reproduction. It undoubtedly exists.