[Q] George Williams ends his 1975 book about sex in pessimistic vein, suggesting that he can understand sex for those animals which he calls high fecundity, but for low fecundity animals like us, he more or less throws up his hands in despair and says, 'Well, it's just a frozen accident. Once it's... once you're stuck with sex, you can't get rid of it.' Do you think that's what happened with, say, mammals, I mean are we stuck with sex because we just can't get rid of it?

Well, I think it's very possible. I do remember attending a meeting on parthenogenesis in Scandinavia, this must have been 20 years ago or so now. And I wrote in 'News and Views' for Nature, when I came back, just describing the meeting, as one sometimes does. I commented that we had recognised at this meeting that there were two major taxa which, as far as we knew, were never parthenogenetic. One are mammals, and I'm ignoring the one observed, or one claimed observed case, because I'm not entirely convinced.

[Q] You mean Jesus Christ or Dolly the sheep?

Oh, Dolly the sheep is fine, no, I was meaning Jesus Christ. I think the evidence in the case of Dolly the sheep is fairly good. But... the other actually, one are the mammals, the other actually is the conifers. There are no parthenogenetic conifers, so far as I know. Almost the next week, there appeared in Nature a letter saying, 'How can Maynard Smith be so ignorant as to not to know why mammals were never parthenogens.' And quoting this curious phenomenon of so-called imprinting of genes, the evidence is that in mammals, a few genes, it's not most of them but just a small number of genes, are labelled, in a sense, as to whether they came from father or mother, in the foetus, and it's called imprinting, you know, a little stamp saying, 'I came from mother.' And in some tissues, only mother's gene is active and the gene from father is not, and in other tissues, only father's gene is active and mother's is not. And I think, actually, David Haig has now given us a very nice evolutionary explanation as to why this imprinting works the way it does, but that certainly wasn't known in those days. But given that that's true, for the foetus to develop properly it must have one father and one mother, at least. If it hasn't got a father then the father's gene tissues fail and vice versa. I felt a little cross about this complaint because the data hadn't at that time, about imprinting, had not even been published, so the fact that I didn't know it seemed to be excusable. It was some years later, actually, that I discovered why conifers are not parthenogens, it's much more obvious. I was reading a review about the inheritance of chloroplasts and mitochondria in plants, and it turns out that in the coniferous plants, the chloroplast is inherited through the pollen grain. And, clearly, if you don't have a pollen grain you're not going to have a chloroplast, you're not going to grow. Now, I've called these things sexual hang-ups, but... you call them frozen accidents, I think it's basically the same idea; that if a group has been sexual for a long period of time, then other, quite secondary, and in some ways quite trivial, secondary adaptations may be attached on to the male-female differentiation, and once that's happened, the reverse mutation may be exceedingly difficult. And... I'm inclined to think that mammals... we don't get parthenogenetic mammals because of sexual imprinting, we don't get parthenogenetic conifers because of the enhancement of chloroplasts.