'Complete reflection now. What shall we do? Go back to one centimetre?'

'No, Harry,' said Kingsley. 'My revolutionary suggestion is that we go upstairs to the sitting-room, where we drink coffee and where we listen to music played by Ann's fair hand. I'd like to switch off for an hour or two and come back later.'

'What on earth is the idea, Chris?'

'Oh, just a hunch, a crazy idea, I suppose. But perhaps you'll indulge me for once in a way.'

'For once in a way!' chuckled Marlowe. 'You've been indulged, Chris, from the day you were born.'

'That may be so, but it's scarcely polite to remark on it, Geoff. Come on, Ann. You've been waiting to try out the Beethoven Opus 106 on us. Now's your chance.'

It was an hour and a half or so later, with the opening chords of the great sonata still ringing in their heads, that the company made its way back to the transmitting lab.

'Try the one-metre first, just for luck,' said Kingsley.

'Bet you that one-metre is completely trapped,' Barnett said as he clicked on various switches.

'No, it's not, by John Brown's body,' he exclaimed a few minutes later, when the equipment had warmed up. 'It's going through. It just isn't believable, and yet it's as plain as a pikestaff on the tube.'

'What's your betting, Harry, on what's going to happen next?'

'I'm not betting, Chris. This is worse than "spot the lady".'

'I'm betting it's going to saturate.'

'Any reasons?'

'If it saturates I'll have reasons, of course. If it doesn't there won't be any reasons.'

'Playing safe, eh?'

'Signal going up,' sang out Barnett. 'Looks as though Chris is going to be right. Up it goes!'

Five minutes later the one-metre signal saturated. It was completely trapped by the ionosphere, no power getting away from the Earth.

'Now try ten centimetres,' Kingsley commanded.

For the next twenty or thirty minutes the equipment was watched keenly, all comment silenced. The earlier pattern repeated. Very little reflection was obtained at first. The reflected signal then increased rapidly in intensity.

'Well, there it is. At first the signal penetrates the ionosphere. Then after a few minutes the ionization rises and we get complete trapping. What's it mean, Chris?' asked Leicester.

'Let's go back upstairs and think about it. If Ann and Yvette will be kind-hearted and make another brew of coffee, perhaps we can do something towards licking this business into shape.'

McNeil came in while coffee was being prepared. He had been attending a sick child while the experiments had been going on.

'Why the air of great solemnity? What's been happening?'

'You're just in time, John. We're going to run over the facts. But we've promised not to start until the coffee arrives.'

The coffee came, and Kingsley began his summing up.

'For John's benefit I'll have to start a long way back. What happens to radio waves when they're transmitted depends on two things, the wave-length and the ionization in the atmosphere. Suppose we choose a particular wave-length for transmission and consider what happens as the degree of ionization increases. To begin with, for low ionization the radio energy streams out of the atmosphere, with very little of it getting reflected. Then as the ionization increases there is more and more reflection until quite suddenly the reflection goes up very steeply until eventually all the radio energy is reflected, none of it getting away from the Earth. We say that the signal saturates. Is that all clear, John?'

'Up to a point. What I don't see is how the wave-length comes into it.'

'Well, the lower the wave-length, the more ionization is needed to produce saturation.'

'So while one wave-length might be completely reflected by the atmosphere, some shorter wave-length might penetrate almost completely into outer s.p.a.ce.'

'That's exactly the situation. But let me go back to my particular wave-length for a moment, and to the effect of rising ionization. For convenience in talking, I'd like to call it "pattern of events A".'

'You'd like to call it what?' asked Parkinson.

'This is what I mean: 1. A low ionization allowing almost complete penetration.

2. A rising ionization giving a reflected signal of increased strength.

3. An ionization so high that reflection becomes complete.

'This is what I call "pattern A".'

'And what is pattern B?' asked Ann Halsey.

'There won't be any pattern B.'

'Then why bother with the A?'

'Preserve me from the obtuseness of women! I can call it pattern A because I want to, can't I?'

'Of course, dear. But why do you want to?'

'Go on, Chris. She's only pulling your leg.'

'Well, here's a list of what happened this afternoon and evening. Let me read it out to you as a table.'

Transmission wave-length Approximate time of switching on Event 1 metre 2.45 p.m.

Pattern A taking approximately half an hour.

10 centimetres 3.15 p.m.

Pattern A taking approximately half an hour.

1 centimetre 3.45 p.m.

Complete penetration of ionosphere over a period of three hours roughly.

10 centimetres 7.0 p.m.

Pattern A taking approximately half an hour.

No transmissions from 7.30 p.m. to 9.0 p.m.

1 metre 9.0 p.m.

Pattern A taking half an hour.

10 centimetres 9.30 p.m.

Pattern A taking half an hour.

'It certainly looks horribly systematic when it's all put together like that,' said Leicester.

'It does, doesn't it?'

'I'm afraid I'm not getting this' Parkinson.

'Nor am I,' admitted McNeil.

Kingsley spoke slowly.

'As far as I'm aware, these events can be explained very simply on one hypothesis, but I warn you it's an entirely preposterous hypothesis.'

'Chris, will you please stop trying to be dramatic, and tell us in simple words what this preposterous hypothesis is?'

'Very well. In one breath that on any wave-length from a few centimetres upwards our own transmissions automatically produce a rise of ionization which continues to the saturation point.'

'It simply isn't possible.' Leicester shook his head.

'I didn't say it was possible,' answered Kingsley. 'I said it explained the facts. And it does. It explains the whole of my table.'

'I can half see what you're driving at,' remarked McNeil.

'Am I to suppose that the ionization falls as soon as you cease transmission?'

'Yes. When we stop transmission the ionizing agent is cut off, whatever it may be perhaps Bill's electrical discharges. Then the ionization falls very rapidly. You see the ionization we're dealing with is abnormally low in the atmosphere, where the gas density is large enough to give an extremely rapid rate of formation of negative oxygen ions. So the ionization dries up very quickly as soon as it isn't being renewed.'

'Let's go into this in a bit more detail,' Marlowe began, speaking out of a haze of aniseed smoke. 'It seems to me that this hypothetical ionizing agency must have pretty good judgement. Suppose we switch on a ten-centimetre transmission. Then according to your idea, Chris, the agency, whatever it is, drives the ionization up until the ten-centimetre waves remain trapped inside the Earth's atmosphere. And here's my point the ionization goes no higher than that. It's all got to be very nicely adjusted. The agency has to know just how far to go and no further.'

'Which doesn't make it seem very plausible,' said Weichart.

'And there are other difficulties. Why were we able to go on so long with the twenty-five centimetre communication? That lasted for quite a number of days, not for only half an hour. And why doesn't the same thing happen your pattern A as you call it when we use a one-centimetre wave-length?'

'b.l.o.o.d.y bad philosophy,' grunted Alexandrov. 'Waste of breath. Hypothesis judged by prediction. Only sound method.'

Leicester glanced at his watch.

'It's well over an hour since our last transmission. If Chris is right we ought to get his pattern A, if we switch on again at ten centimetres, that is to say, and possibly at one metre also. Let's try.'

Leicester and about half a dozen others went off to the lab. Half an hour later they were back.

'Still complete reflection at one metre. Pattern A on ten centimetres,' Leicester announced.

'Which looks as if it supports Chris.'

'I'm not sure that it does,' remarked Weichart. 'Why didn't the one metre give pattern A?'

'I might make some suggestions, but in a way they're even more fantastic, so I won't bother with 'em just for the moment. The fact is, and I insist it is a fact, that whenever we have switched on our ten-centimetre transmitter there has always been a sharp rise of atmospheric ionization, and whenever we switched off there has been a decline of ionization. Does anyone deny that?'

'I don't deny that what has happened so far agrees with what you say,' Weichart argued. 'I agree that no denial is possible there. It's when it comes to inferring a causal connexion between our transmissions and the fluctuations of ionization that I dig my toes in.'

'You mean, Dave, that what we found this afternoon and this evening was coincidence?' asked Marlowe.

'That's what I mean. I grant you that the odds against such a series of coincidences are pretty big, but Kingsley's causal connexion seems to me an out-and-out impossibility. What I feel is that the improbable can happen but the impossible cannot.'

'Impossible is too strong,' insisted Kingsley. 'And I'm sure that Weichart couldn't really defend his use of the word. What we're faced with is a choice between two improbabilities I said that my hypothesis seemed improbable when I first trotted it out. Moreover I agree with what Alexis said earlier on, that the only way to test a hypothesis is by its predictions. It's about three-quarters of an hour since Harry Leicester did his last transmission. I'm going to suggest that he goes right now and does another ten-centimetre transmission.'

Leicester groaned. 'Not again!'

'I predict,' went on Kingsley, 'that my pattern A will be repeated. What I'd like to know is what Weichart predicts.'

Weichart didn't quite like the turn of the argument, and he attempted to hedge. Marlowe laughed.

'He's pinching you, Dave! You've got to stand up and take it. If you're right about it being coincidence before, you've got to agree that Kingsley's present prediction is very unlikely to be right.'

'Of course it's unlikely, but it might happen that way all the same.'

'Come off it, Dave! What do you predict? Where d'you put your money?'

And Weichart was forced to admit that he put his money on Kingsley's prediction being wrong.

'All right. Let's go and see,' said Leicester.

While the company were filing out, Ann Halsey said to Parkinson: 'Will you help me to make more coffee, Mr Parkinson? They'll be wanting some when they get back.'

As they busied themselves, she went on: 'Did you ever hear such a lot of talk? I used to think that scientists were of the strong silent type, but never did I hear such a gibble-gabble. What is it that Omar Khayyam says about the doctors and saints?'

'I believe it goes something like this,' answered Parkinson: 'Myself when young did eagerly frequent Doctor and Saint, and heard great argument About it and about, but evermore Came out by the same door as in I went.'

'It isn't so much the volume of talk that surprises me,' he laughed. 'We get plenty of that in politics. It's the number of mistakes they've made, how often things have turned out differently to what they've expected.'