Nevertheless this description-good-dominant, bad-recessive-was in essence correct; it synopsized the mechanisms by which a race conserved its favorable mutations and destroyed (eventually) its unfavorable mutations. "Bad-dominant" was almost a contradiction in terms, as a thoroughly bad mutation which was dominant killed itself off (along with the unfortunate zygote inheriting it) in one generation, either lethal in womb or so damaging to the zygote that it failed to reproduce.

But the usual weeding process involved bad-recessives. These could remain in the gene pool until one of two events happened, each controlled by the blind laws of chance: Such a gene could pair with a gene like it when sperm fertilized ovum and thereby eliminate itself by eliminating the zygote-hopefully before birth, or-tragically-after birth. Or this bad-recessive might be eliminated by chromosome reduction at meiosis and the result would be a healthy baby who did not carry this bad gene in its gonads-a happy outcome.

Both these statistical processes slowly weeded out bad genes from the race's gene pool.

Unfortunately the first of these processes often produced babies viable but so handicapped they needed help to stay alive-sometimes needing economic help, born losers, who never managed to support themselves; sometimes needing plastic surgery or endocrine therapy or other interventions or supports. When Captain Aaron Sheffield had been practicing medicine (on Ormuzd and under another name), he had gone through stages of increasing frustration over these poor unfortunates.

At first he had tried to practice therapy by the Hippocratic Oath-or close to it; he was by temperament unable to follow any man-made rule blindly.

Then he had had a period of temporary mental aberrance during which he had sought a political solution to what he saw as a great danger: reproduction by defectives. He tried to persuade his colleagues to refuse therapy to hereditary defectives unless they were sterile or sterilized or willing to accept being sterilized as a precondition for receiving therapy. Worse yet, he had attempted to include in the definition of "hereditary defective" those who displayed no stigmata save that they had never managed to be self-supporting--on a planet not overcrowded and which he himself had selected centuries earlier as nearly ideal for human beings.

He got nowhere, he encountered nothing but fury and contempt-save for a few colleagues who agreed with him privately and denounced him publicly. As for laymen, tar-and-feathers was the mildest medicine they prescribed for Dr. "Genocide."

When his license to practice was lifted, Lazarus regained his normal emotional detachment. He shut up, realizing that. grim old Mother Nature, red of tooth and claw, invariably punished damfools who tried to ignore Her or to repeal Her ordinances; he need not interfere.

So he moved and changed his name again and started to get ready to go off-planet-when a plague hit Ormuzd. He had shrugged and gone back to work, an unfrocked physician whose services were temporarily welcome. Two years and a quarter of a billion deaths later he was offered his license back-subject to good behavior.

He told them what to do with that license and left Ormuzd as quickly as possible, eleven year later. He was a professional gambler during that wait, that being the handiest way he could see at the time for saving up the necessary.

Sorry, Minerva, I was talking about those mirror twins. So the silly little wench was knocked up, which caused me to slip back into my baby-cotching, country-doctor persona, and I stayed up all night worrying about her and her brother and the baby they were going to have-unless I did something about it. To find out what I should do, I had to reconstruct what had happened and from that what could happen. Having no certain data, I had to follow that old rule for finding a lost mule.

First I had to think like that slave factor- A man who auctions slaves is a scoundrel but too smart to risk a caper in which he might wind up a slave himself, or dead if he was lucky-which is what would happen to one who played fast and loose on Blessed with the authority of a bishop. Ergo, the scoundrel had believed what he had said.

That being so, I could table the question of why this factor was commissioned to sell these two, while I tried to think like a priest-scientist engaged in human biological experimentation. Forget the chance that these two were ordinary siblings-no point in picking such a pair even for a swindle. Forget the chance that they were unrelated in any fashion, as in such a case it would simply be a normal case of breeding. Sure, sure, any woman, can give birth to a monster, as even with the most genetically hygienic of breeding a bad mutation can show up-and an alert midwife may neglect to give that first lifegiving spank~-and many have.

So I considered only the third hypothesis: complementary diploids from the same parents. What would this experimenter do? What-would I do?

I would use as near perfect stock as I could find and not start the experiment until I had both a male and a female parent who tested "clean" genetically in the most subtle ways for which I could test-which on Blessed meant quite sophisticated ways, for that century.

For a selected gene site and an assumption of 50-50 in the Mendelian distribution of 25-50-25, this pre-experiment testing would chop off the 25 percent chance of reinforcement of a bad recessive and leave a distribution of one-third bad, two-thirds good, at the parent generation-possible parents of possible Joes and Llitas, that is.

Now I start putting together mirror twists in my persona as a priest-experimenter. What happens? If we consider the minimum number of gametes needed to represent this one-third and two-thirds distribution, we get eighteen possible "Joes," eighteen possible "Llitas"-but in both male and female two of them show up as "bad"-the bad recessive has reinforced and the zygote is defective; the experimenter eliminates them...or he may not need to; the reinforcement may be lethal.

We wind up at this point with an 8 & 1/3 percent improvement, or a total improvement of 25 percent in favorable chances for Llita's baby. I felt better. If you add the fact that I am the sort of a midwife who is too busy helping the mother to stop to spank a monster, the favorable chances went way up.

But all that this shows is that bad genes tend to be eliminated at each generation-with the tendency greatest with the worst genes and reaching 100 percent whenever reinforcement produces a lethal-in-womb-while favorable genes are conserved. But we knew that-and it applies also to normal outbreeding and even- more strongly to inbreeding, although the latter is not well thought of for humans as it hikes up the chances of a defective by precisely the same amount that it weeds-that being the hazard, that I was afraid of for Llita. Everybody wants the human gene pool cleaned up, but nobody wants its tragic aspects to take place in his own family. Minerva, I was beginning to think of these kids as "my family."

I still did not know anything about "mirror twins."

I decided to investigate a more probable incidence of bad recessives at a given site. Fifty-fifty is far too high for a really bad gene; the weeding is drastic, and the incidence drops to a lower percentage each generation, until the incidence of a particular bad gene is so low that reinforcement at fertilization is a rare event, as reinforcement is the square of the incidence; e.g., if one-in-a-hundred haploids carry this bad gene, then it will be reinforced one-in-ten-thousand fertilizations. I speak of the total gene pool, or in this case a minimum of two hundred adult zygotes, female and male; random breeding in such a pool will, bring together that bad reinforcement only by that long chance-a chance happy or unhappy depending on whether you look at it impersonally in terms of -cleaning the gene pool or personally in terms of individual human tragedy.