The first recorded artificial insemination was by the eighteenth century Scottish anatomist John Hunrer who used a syringe to impregnate a woman whose husband had a deformed penis. Since then, to assist the work of nature has become commonplace. Artificial insemination outside the body had to wait until 1978, when sperm met egg in a test-tube to produce Louise Brown. The technology is less simple than it sounds, as eggs in the right stage of development must be retrieved from a potential mother; but, even so, about one in four attempts succeed. After hormone treatment, eggs are sucked from the ovary with a fine needle and fertilised with the relevant sperm. This need not happen at once as eggs can be frozen for later use. After a few divisions, the fertilised egg is returned to the uterus; either of the natural mother or, if she has reproductive problems, into a volunteer. Often, more than one is used (which sometimes leads to several children being born); and, somerimes, the ball of developing cells is screened to check whether it carries a genetic abnormality before deciding to continue. About one British birth in a hundred is a test-tube baby and there are about half a million such children in the world today.

Often, the problem lies with the male. Perhaps his sperm is of such low quality that it cannot penetrate the egg. Sometimes, indeed, it is quite unable to move and cannot escape from the testes. In such cases, sperm can be extracted with a needle and sperm heads injected into the egg. Fertilisation over, the egg is implanted or frozen for later use. Some suggest, indeed, that given the increase in genetic damage in the children of older parents it might be wise for a woman to freeze a sample of her eggs during her teens to ensure the health of future children. For a man the task would be even easier.

All this has led to controversy (including questions as to who might own a dead man's sperm) but is becoming part of medical practice, with hundreds of clinics available across the world. Genetics is often involved, with a check for defects in the fertilised egg. Surrogate motherhood, too, has become common since the first fertilised egg was implanted into an unrelated female in Iceland in 1989. It contains some unpleasant reminders of social reality. In almost every case the surrogate is poorer and less educated than the egg-donor (which, at up to $50 000 a pregnancy in the United States, is not surprising). For all these procedures the yuk factor has been forgotten, but for cloning it remains.

I write as a clone and the son of a clone; and one of the few British citizens entitled to commit incest. A clone, of course: we are all one of those, for the billions of cells that we contain are — each of them — copies of the fertilised egg that made us, reproduced without benefit of sex. My mother, as it happens, is an identical twin (one of two hundred thousand or so in Britain), so that another individual shares all her genes. Her twin — her clone — in turn, has a daughter who is legally my cousin, but in genetic terms a half-sister. The question has never arisen, bur there is no legal impediment to marriage, close relative though she is.

My mother and aunt are different individuals, so why the fear of clones? Cloning is, after all, common. It is sex that is rare — to persuade two cells to fuse to make one is, in some ways, the antithesis of reproduction (which, in its clonal version, involves one cell splitting into two). Plenty of creatures, from fungi to lizards, manage without it (and even turkeys can be persuaded to lay eggs without benefit of males). Potatoes are clones and animals can, in principle, be multiplied in the same way. Take an early embryo, split it into pieces and, sometimes, each will grow into an identical twin. This has been successful in rhesus monkeys and, no doubt, could also be done on ourselves. Sheep and cattle have been split at the eight-cell stage, to give (so far) a maximum of five identical offspring. Cows from the same herd may differ greatly in the amount of milk they yield, and as it takes several years per generation it is much more efficient to clone the champion rather than mating her with some favoured bull. In the 1990s, the method was used by breeders in the United States. It failed because the calves tend, for some reason, to develop much larger than usual and either die or demand an expensive caesarian birth.

Cloning of the Dolly kind is more sophisticated, with the movement of nuclei between cells, but is, after all, just another form of reproductive technology. It follows in the tradition of the great Italian biologist Spallanzani who artificially inseminated a bitch in 1782. Cloning itself — the growth of an organism from an egg containing a foreign complement of genes-began in the 1950s with frogs. They have lazy embryos, because frog eggs are so stuffed with food that they divide to make several thousand cells before they use any of their own genes. Their eggs hence scarcely notice the insertion of another piece of DNA as they are well on in development before genetic information is needed. Sheep wait for just four cell divisions — the 16-cell stage of the embryo — before switching on their genes, while humans DNA becomes active after three division, pigs after two, and mice even before cell divisions begins. This slight delay might explain why sheep proved easier to clone than work on mice (which were recalcitrant about accepting alien genes) had suggested.

Apart from simple vanity or dislike of the opposite sex, cloning might be useful as an aid to the infertile. Perhaps a male is unable to make sperm: and one of his cell nuclei might be inserted into his wife's egg to make a clone. Perhaps one partner has a genetic illness and prefers to use the other's genes to avoid the risk to a child. There are also various eccentric ideas about armies made of cloned copies of some dictator. Whether they would obey orders is another question, and any cloned child, identical as it is to its parent, is likely to prove a particular disappointment should it fail to live up to expectations

None of these possibilities is legal — or feasible — at present. All the fuss about what might be done should be tempered with realism. Cloning, even of sheep, is a complicated business. First an egg must be harvested and the cells that are to provide the nucleus made ready. This is fused to an egg that has lost its own nucleus — with a Frankenstein touch — a burst of electricity, and stored in the reproductive tract of a second sheep. Those that pass the test are then moved to the surrogate mother herself. Dolly was the only one of three hundred experiments that worked and most cloned cattle, sheep and mice have been born dead or deformed. For humans, for the time being, moving cell nuclei around is just too risky. Even so it seems almost certain that cloning will, at least in one form, become part of medical practice.

In the early 1980s it was found that cells from embryonic mice could be kept alive in the laboratory and that some, instead of moving down the path to adulthood, stayed forever young: ready to develop into any tissue when prompted to do so. They have been kept as perpetual adolescents for up to ten years. The technique involves a certain trickery, with various growth factors added to the culture. These embryonic stem cells, as they are called, when injected into another developing embryo, are happy to develop into blood cells, nerves and so on; or, if they find themselves in the right place, into the precursors of sperm or egg. The recipient grows up as a chimaera; a mixture of cells with different genes — in effect, a mouse with four parents.