When I became a Geneticist in 1947 three main problems were the centres of interest; the nature of the gene, how genes controlled development and morphogenesis and the function of sex with which went the problem of the deleterious effect of inbreeding.

The first is solved. But it is interesting in retrospect to recall that its solution depended essentially on the pedantry of matters of definition. It was the realisation, and the experimental demonstration, that the gene (the atom of biology) as defined as a unit of action is not the same as that defined as a unit of recombination or the unit of mutation which led to the knowledge that the gene had a structure, and hence to our present understanding. This should be a warning to anyone who uses pedantry as a pejorative word.

The second problem is by no means solved but is well on its way. but the third still stands out. Indeed mathematical population genetics tells us that recombination does not facilitate evolution. This has to be nonsense because sexual reproduction is ubiquitous as also are special means to minimise crossing between close relatives.

I believe that the answer lies with a weakness of theoretical population genetics that it does not allow for environmental change. Whereas, of course, environmental change has to be a consequence of evolution. An evolutionary change in one species is an environmental change in every other species that interacts with it. My definition of Biological Progress and long term fitness (see references 11 & 31) are relevant to these issues.

The studies of sex and inbreeding in the 1950's concentrated on the problem of the advantage of heterozygosity. Did heterozygotes have a physiological advantage as such, or was it that only the heterozygotes have been improved by selection whereas homozygotes being rare have not. Much of the relevant experimentation centred round the question of developmental homeostasis ( the ability of an individual to control its development despite accidental or environmentally caused disturbance).

The results showed that any departure from normal conditions, whether environmental change, close inbreeding, strong selection (See Ref 32 ), led to deterioration in developmental homeostasis. The conclusion was reached that a natural population in its natural environment consisted of a wide variety of coadapted chromosomes, adjusted by selection to collaborate well in the control of development ,and to produce mostly harmless recombinants.

This still does not explain why the sexual reproduction that results in the heterozygosity should be there in the first place.

I believe that with present knowledge we can offer a double explanation. First, I think sex was of the essence the origin of life as we know it, which is another topic. Second, I think that the ubiquity of sex is explained by environmental change. Now we know that genetic stability (as I called it) depends on most complicated sets of machinery to ensure that DNA replication is perfect, all the conclusions we reached about Developmental Homeostasis, are applicable to Genetic stability, (or Genetic homeostasis as Lerner called it ).

Any serious disturbances of the status quo, especially such as are involved in periods of mass extinction, must lead to a break down of the mechanisms that repair replication errors, control transposable elements, restrict crossing over etc. Hence at such times the spectrum of mutations would be totally different from that observed in stable populations. Provided sexual reproduction was available to re-assort such variants the population that survived such crises would be able to regenerate a wide variety that could fill the newly freed ecological niches.

So we envisage populations in a stable world gradually losing heterozygosity and often sex as theoretical population genetic predicts, but periodically those that have done so will be extinguished by some crises, and those that have not will repopulate the world. Every time this happens the relative merits of heterozygosity will be reinforced.

Artificial selection experiments generally showed one other unexpected thing that agrees with this. Selection did not seem to use up the genetic variance. Instead the lines stopped responding when they became very difficult to maintain, and had to be nursed carefully while viability recovered (see ref 41). It seemed that this was a period when coadaptation was renewed, after which a new fairly stable situation was established. And the resulting stable population still maintained genetic variance.Indeed as Mather and Harrison showed, back selection could return phenotypes to the starting point

There isomething odd about all this. We treat selection as a force and think of it as such. Yet looking back one realises that in fact 'selection' does nothing to the selected flies, except free them from competitition with their norrmal relatives, and prevent them from mating with their normal relatives. Yet the intrinsic variance produces such profound results.

This adds importance to what was always clear to us. There is no single thing that is the Drosophila genome or the Human genome . There are as many genomes as there are individuals. The within population variation may only involve a small part of the total DNA, but it is of the utmost importance. It is the source of evolution, and of the uniqueness of the individual. Without it societies such as ours could never have developed, for they depend on individual recognisability.

© J M Thoday 2006

Addition 2007

At the beginning above I said the first problem, the nature of genes, is solved. Now I think that this was naive. So impressive was the work that sorted out the DNA - RNA - Protein process of heredity that it was too easy to assume this was all there was to genetic variation. Hence the labelling of most DNA as 'Junk'.

But it is now becoming obvious that this could not be all the story. There is far too much DNA for that, and simple changes in enzyme sequences cannot readily account for the diversity of species and the individual variance that we showed was so ubiquitous .

Work which is beginning to suggest control roles for so called 'Junk DNA', is consonant with old work on heterochromatin and is surely the opening of a new phase in genetics that is long overdue.

The literature of experimental population genetics of the 1950's to 80's is full of phenomena that may involve roles for 'Junk DNA' which might well now be called the 'Dark Matter' of Biology.

It is time to repeat some of the work, such as simple selection experiments, that revealed this underlying cryptic genetic variance, with the addition of sequencing to show what is involved in individual variation and in evolution.