Health and Disease 
What statistical constraints are there between the parameters ? 
How big is the selective job, and could it have been done in the 
time? This ought in principle to be an answerable question 
in terms of ‘‘amount of information’’. In other words, the 
discipline of information theory applies to the selective power 
of natural selection. Unless we argue in these terms, the issue 
is left to the man with the loudest voice. 
Huxley: Muller made a big contribution to that in the 
Scientific Monthly in about 1930. He went a long way towards 
quantifying what changes could be achieved with a reasonable 
rate of mutation, and reasonable intensity of selection over the 
geological time-span available. The improbability of obtaining 
the observed results of evolution by chance mutation alone, 
without natural selection, is incredibly high. 
Haldane: In a couple of papers which I called “The Cost 
of Natural Selection” I once tried to work out how many 
organisms would have either to die or to have their fertility 
reduced in order to get a single gene substitution. I came to the 
conclusion that if two mammalian species differ by something 
of the order of a thousand gene pairs, that agrees reasonably well 
with the time that the geologists think man has lived on earth. 
MacKay: ‘The difficulty is that each estimate depends on 
our present understanding of the complexity of the matching 
process involved. Each new discovery on the details of Szent- 
Gyorgyi’s lock and key, if I may use that metaphor, may affect 
our calculations by an order of magnitude. The auditing of 
the evolutionary account seems to me to be an on-going process, 
not one that we can expect to close now. All we can say is that 
if things were as simple as in such-and-such a model, then there 
would be such-and-such a selective power in natural selection 
for a given time. But even this would be much better than base- 
less dogmatism. 
Huxley: Surely it is a question of studying processes, and 
this is where I think that work like Waddington’s is so illuminat- 
ing. He points out that just as biological evolution by natural 
selection is essentially self-directed, so also is the epigenetic 
process of individual development. It is Jargely canalized into 
243 
