This situation began to change when, in the protogalaxies, the protostars started condensing, and I quickly realized where it would all end, with that temperature rising all the time, and so I said: ‘Now they’re going to catch fire.’

  ‘Nuts!’ the Dean said.

  ‘Want to bet?’ I said.

  ‘Anything you like,’ he said, and wham, the darkness was shattered by all these incandescent balls that began to swell out.

  ‘Oh, but that isn’t what catching fire means . . .’ (k)yK began, quibbling about words in his usual way.

  By that time I had developed a system of my own, to shut him up: ‘Oh, no? And what does it mean then, in your opinion?’

  He kept quiet: lacking imagination as he did, when a word began to have one meaning, he couldn’t conceive of its having any other.

  Dean (k)yK, if you had to spend much time with him, was a fairly boring sort, without any resources, he never had anything to tell. Not that I, on the other hand, could have told much, since events worth telling about had never happened, or at least so it appeared to us. The only thing was to frame hypotheses, or rather: hypothesize on the possibility of framing hypotheses. Now, when it came to framing hypotheses of hypotheses, I had much more imagination than the Dean, and this was both an advantage and a disadvantage, because it led me to make riskier bets, so that you might say our probabilities of winning were even.

  As a rule, I bet on the possibility of a certain event’s taking place, whereas the Dean almost always bet against it. He had a static sense of reality, old (k)yK, if I may express myself in these terms, since between static and dynamic at that time there wasn’t the difference there is nowadays, or in any case you had to be very careful in grasping it, that difference.

  For example, the stars began to swell, and I said: ‘How much?’ I tried to lead our predictions into the field of numbers, where he would have less to argue about.

  At that time there were only two numbers: the number e and the number pi. The Dean did some figuring, by and large, and answered: ‘They’ll grow to e raised to pi.’

  Trying to act smart! Any fool could have told that much. But matters weren’t so simple, as I had realized. ‘You want to bet they stop, at a certain point?’

  ‘All right. When are they going to stop?’

  And with my usual bravado, I came out with my pi. He swallowed it. The Dean was dumbfounded.

  From that moment on we began to bet on the basis of e and of pi.

  ‘Pi!’ the Dean shouted, in the midst of the darkness and the scattered flashes. But instead that was the time it was e.

  We did it all for fun, obviously; because there was nothing in it for us, as far as earning went. When the elements began to be formed, we started evaluating our bets in atoms of the rarer elements, and this is where I made a mistake. I had seen that the rarest of all was technetium, so I started betting technetium and winning, and hoarding: I built up a capital of technetium. I hadn’t foreseen it was an unstable element that dissolved in radiations: suddenly I had to start all over again, from zero.

  Naturally, I made some wrong bets, too, but then I got ahead again and I could allow myself a few risky prognostications.

  ‘Now a bismuth isotope is going to come out!’ I said hastily, watching the newborn elements crackle forth from the crucible of a ‘supernova’ star. ‘Let’s bet!’

  Nothing of the sort: it was a polonium atom, in mint condition.

  In these cases (k)yK would snigger and chuckle as if his victories were something to be proud of, whereas he simply benefited from overbold moves on my part. Conversely, the more I went ahead, the better I understood the mechanism, and in the face of every new phenomenon, after a few rather groping bets, I could calculate my previsions rationally. The order that made one galaxy move at precisely so many million light-years from another, no more and no less, became clear to me before he caught on. After a while it was all so easy I didn’t enjoy it any more.

  And so, from the data I had at my disposal, I tried mentally to deduce other data, and from them still others, until I succeeded in suggesting eventualities that had no apparent connection with what we were arguing about. And I just let them fall, casually, into our conversation.

  For example, we were making predictions about the curve of the galactic spirals, and all of a sudden I came out with: ‘Now listen a minute, (k)yK, what do you think? Will the Assyrians invade Mesopotamia?’

  He laughed, confused. ‘Meso- what? When?’

  I calculated quickly and blurted a date, not in years and centuries of course, because then the units of measuring time weren’t conceivable in lengths of