The founder effect can be seen again and again among the descendants of those who colonised the world from Europe. Sometimes, the settlements are isolated by miles of ocean. Tristan da Cunha, a tiny island in the South Atlantic, was settled by a garrison sent to guard Napoleon, then in exile on St Helena. A few soldiers stayed on after the guard was withdrawn. They obtained wives by advertising, and a few shipwrecked sailors and others joined the community over the years. It went through a second bottleneck when several men drowned in a fishing accident and some families moved away, with the advice of a gloomy pastor. Now, the island is still a week's journey by ship from the mainland, but a few hundred people can stand the isolation. Again, they share names, seven altogether, and those — Bentley, Glass and Swain — of three of the first founders, are still common. Just five lineages of mitochondria! genes exist, and the island has its own genetic abnormality, a hereditary blindness brought by one of tlu- original wives.

Some migrant communities are isolated by social rather ih.in physical barriers. The United States has many religious groups whose founders emigrated to avoid persecution. They have grown into large populations which exclude outsiders. The Pennsylvania Amish have a unique inheritance. Almost a hundred babies have been born with six fingers and restricted growth, a condition almost unknown elsewhere. Every one of the affected children descends from Samuel King, a founder of the community.

To trace the movement of a gene around the world also shows the importance of chance. Huntington's Disease is relatively common among Afrikaners. Most cases descend from a Dutch man or his wife who emigrated in the 1650s. All copies on Mauritius are the legacy of a French nobleman's grandson, Pierre Dagnet d'Assigne de Bourbon, and more than four hundred patients in Australia have inherited their gene from a British immigrant, Mrs Cund-ick. Wales has a patch of the disease in the Sirhowy Valley, around the house of a mason who settled there in the nineteenth century and who must have carried the Huntington's gene. The largest kindred in the world (which was used to map the gene) is in Venezuela around an arm of the sea called Lake Maracaibo. Ten thousand descendants of one Maria Concepcion, who died in about 1800, have been traced. Four thousand either have the illness or are at a high risk.

Such accidents of colonisation must have happened again and again as humans spread across the world. Even without a written history, the surnames of the Afrikaners make it possible to estimate how many people were in at the beginning, three hundred and more years ago. Genes can do the same job. Patterns of variation show how many people founded a population, or whether it went through a bottleneck in the distant past.

Inherited diversity shows clear global patterns. Africans are more variable than are the rest of the world's peoples. Their cell-surface antigens (the cues recognised by the immune system) show about twice as much variation as do the equivalent genes in Europe, and many nf its variants are unique to Africa. Africans are more van.ibli- for blood groups, proteins and DNA sequences as well. 1'or mito-chondrial DNA, the average difference between two Africans is twice that found elsewhere. Venezuelan Indians, in contrast, whose ancestors were near the end of the long history of movement across the world from Africa, have almost no variation in their mitochondria! DNA.

The decrease in diversity outside Africa, humankind's native continent, may be because genes were lost as small bands of people moved, split and founded new colonies in the trek across the globe. Just as for Afrikaner surnames the number of variants dropped each time a new colony was founded. The high levels of diversity among Africans is evidence that Homo sapiens has been in that continent for longer than anywhere else. Its decrease at the tips of the evolutionary branches in South America and Polynesia shows how human evolution was driven by chance as the migrants passed through a succession of bottlenecks.

A comparison of the genes of Africans with those of their descendants elsewhere in the world makes it possible to guess at the numbers involved in those early colonisations. The order of bases along a short length of DNA is in some ways a 'genetic surname', a set of inherited letters which pass together as a group down the generations. The name written in nucleic acids around one of the haemoglobin genes has been looked at in detail worldwide. The results are quite unexpected.

All populations outside Africa, from Britain to Tahiti,  share a few common sequences. Africa itself has a different pattern of distribution. Just like the names in the Johannesburg telephone book compared to that of Amsterdam, the shift in pattern from the ancestral continent to its descendants may be a relic of a population bottleneck at the time of migration — this time from, rather than to, Africa. We can do some statistics (and make quite a lot of guesses) to work out the size of this hundred-thousand-year-old group of emigrants. They show that the whole of the world's population outside Africa may descend from fewer than a hundred people. If this is true, non-Africans were once an endangered species.

Science has two cultures: one (to which most scientists belong) uses mathematics and the other understands it. Such guesses about ancient population bottlenecks demand statistical acrobatics. They also depend on one crucial, and perhaps quite mistaken, assumption; that the genes involved do not alter the chances or survival or of sex. Molecular biologists tend to assume that small changes in the structure of DNA are unimportant. It is just as possible that they do have an effect on fitness. If, for example, Africans have more variation on the surfaces of their cells because it helps to combat disease, then to claim that a reduction elsewhere is due to an ancient bottleneck is simply wrong.

Any attempt to reconstruct the distant past is bound to suffer from ambiguities such as these. Genetics has not yet revealed just how many Adams and Eves there may have been, but shows that much of the human condition has been shaped by accident: an observation that might at least instil a certain humility into those whose genes have defeated the i iws of chance by surviving to the present day.

Renaissance painters on religious themes hat! a problem: when they showed Adam and Eve, should they have- navels? If they did, then surely it was blasphemous as it implied that they must have had a mother. If they did not, then it looked silly. Although some compromised with a strategic piece of shrubbery, that did not resolve matters. And where was the Garden of Eden? Various theories had it in Israel, Africa and even the United States. When it existed seemed obvious because to add up the ages of the descendants of the primal couple as given in the Bible set the start of history as 4 October 4004 BC.

The reason for leaving Eden was also clear. Its inhabitants had, with the help of an apple, learned forbidden truths, and as a punishment were forced out into the world. No longer could they depend on a god-given supply of food falling into their hands. Instead, they had to make a living. The first economy was born.

The escape from Eden — the colonisation of the Earth — showed how genetic change is linked to economic development. Economics is often seen as a kind of enlightened self-interest. The desire to increase one's own wealth may, as Adam Smith has it, be the invisible hand which is at the foundation of all social progress. The same argument is used by some evolutionists. Genes are seen as anxious to promote their own interests, even at the expense of their carriers. In its most naive form, this view of life is used to explain (or at least to excuse) spite, sexism, nationalism, racism and the economic and political systems that grow from them.