The Science of Discworld Revised Edition

it work? Could you get ‘something for nothing’ in this way? Finding out was never going to be easy, because uranium-235 is mixed up with ordinary uranium (uranium-238), and getting it out is like looking for a needle in a haystack when the needle is made of straw.
    There were other worries
too
… in particular, might the experiment be too successful, setting off a chain reaction that not only spread through the experiment’s supply of uranium-235, but through everything else on Earth as well? Might the atmosphere catch fire? Calculations suggested: probably not. Besides, the big worry was that if the Allies didn’t get nuclear fission working soon then the Germans would beat them to it. Given the choice between our blowing up the world and
the enemy
blowing up the world, it was obvious what to do.
    That is, on reflection, not a happy sentence.
    Loko is remarkably similar to Oklo in southeastern Gabon, where there are deposits of uranium. In the 1970s, French scientists unearthed evidence that some of that uranium had either been undergoing unusually intense nuclear reactions or was much, much older than the rest of the planet.
    It
could
have been an archaeological relic of some ancient civilization whose technology had got as far as atomic power, but a duller if more plausible expanation is that Oklo was a ‘natural reactor’. For some accidental reason, that particular patch of uranium was richer than usual in uranium-235, and a spontaneous chain reaction ran for hundreds of thousands of years. Nature got there well ahead of Science, and without the squash court.
    Unless, of course, it
was
an archaeological relic of some ancient civilization.
    Until late in 1998, the natural reactor at Oklo was also the best evidence we could find to show that one of the biggest ‘what if?’ questions in science had an uninteresting answer. This question was ‘What if the natural constants
aren’t
?’
    Our scientific theories are underpinned by a variety of numbers, the ‘fundamental constants’. Examples include the speed of light, Planck’s constant (basic to quantum mechanics), the gravitational constant (basic to gravitational theory), the charge on an electron, and so on. All of the accepted theories assume that these numbers have always been the same, right from the very first moment when the universe burst into being. Our calculations about that early universe
rely
on those numbers having been the same; if they used to be different, we don’t know what numbers to put into the calculations. It’s like trying to do your income tax when nobody will tell you the tax rates. From time to time maverick scientists advance the odd ‘what if?’ theory, in which they try out the possibility that one or more of the fundamental constants isn’t. The physicist Lee Smolin has even come up with a theory of evolving universes, which bud off baby universes with different fundamental constants. According to this theory, our own universe is particularly good at producing such babies, and is also particularly suited to the development of life. The conjunction of these two features, he argues, is not accidental (the wizards at UU, incidentally, would be quite at home with ideas like this – in fact, sufficiently advanced physics is indistinguishable from magic).
    Oklo tells us that the fundamental constants have not changed during the last two billion years – about half the age of the Earth and ten per cent of that of the universe. The key to the argument is a particular combination of fundamental constants, known as the ‘fine structure constant’. 4 Its value is very close to 1/137 (and a lot of ink was devoted to explanations of that whole number 137, at least until more accurate measurements put its value at 137.036). The advantage of the fine structure constant is that its value does not depend on the chosen units of measurement – unlike say, the speed of light, which gives a different number if you express it in miles per second or kilometres per second. The Russian physicist Alexander Shlyakhter analysed the different chemicals in the Oklo reactor’s ‘nuclear waste’, and worked out what the value of the fine structure constant must have been two billion years ago when the reactor was running. The result was the same as today’s value to within a few parts in ten million.
    In late 1998, though, a team of astronomers led by John Webb made a very accurate study of the light emitted by extremely distant, but