light a candle—"

"Yes, yes, I do see that. Don't begin about candles."

"Well, the result of that is that there is less pressure at A. In other words, there is more room for the air to move about. When that happens the air at the place B—"

"Where is that?"

"Oh anywhere. I told you to think of two places, A and B."

"No, you told me to think of a place A, and I am still thinking of it, because it is very hot there."

"Well, this is another place, where the pressure is simply frightful. When the air rises at A the air from B rushes over to A to fill up the gap, and that is what we call wind."

"I see."

"No, you don't. It isn't quite so simple as that. Now, the atoms of air rushing from B to A don't go straight there, but they travel in—in sort of circles."

"Why do they do that?"

"Well, the fact is that these atoms are so keen to get over to A, where there is plenty of room, that they jostle each other, and that makes them go round and round. If they go round and round against the clock, like that, they are called cyclones, or depressions, or low-pressure systems. If they go with the clock, like that, it is an anti-cyclone."

"Oh!"

"What do you mean—'Oh'?"

"What I said; but go on."

"Now suppose this air—"

"Which air?"

"The air from B. Suppose it is travelling in a cyclone—"

"But isn't a cyclone a low-pressure thingummy?"

"Yes."

"And didn't you say that B was a high-pressure place?"

"Yes."

"Then how does the air coming from B manage to be low-pressure stuff?"

"I see what you mean. There is an explanation, but it would take too long to hazard it now. Suppose the air is coming from B in an anti-cyclone, then ..."

"All right. I'll suppose that."

"... it rushes over to A and fills up the gap. There is more pressure at A and the barometer goes up—"

"Is it fine then?"

"No, it rains. You see, the air from B is colder than the air at A was before the air came from B."

"I don't see."

"Well, obviously it must be."

"How 'obviously'?"

"Well, the whole thing started with it being very hot at A, you remember, so that the air rose. If it had been hotter still at B just then the air would have risen at B instead, and it couldn't have rushed over to A. There'd have been a frightful muddle."

"There is."

"Well, it's your own fault for interrupting. This air, then—"

"Which air is this?"

"The air from B. The air from B cools the air at A—"

"But I thought the air at A had risen."

"Not all of it. And that makes it rain."

"Why?"

"Oh, well, I can't go into that. It's something to do with condensation. Air absorbs more moisture when it is hot than when it is cold—"

"So do I. I understand that."

"When the air cools the water condenses."

"Is it fine then?"

"No, it rains, you fool."

"When is it fine?"

"Wait a bit. The falling of the rain of course generates heat—"

"Why 'of course'?"

"I can't explain exactly, but you know perfectly well that it's always warmer on a cold day after the rain."

"Yes, but not on a hot day."

"Yes, it is."

"No, it isn't."

"It is, really. Anyhow, this is a cold day."

"No, it isn't. You said it was very hot at A."

"I'm not going to argue. You must take it from me that rain generates heat."

"All right. Is it fine then?"

"No. Heat being generated the air rises. The result of that is that there is less pressure at A—"

"Is it fine then?"

"I've explained already what happens then. The air from B—"

"Do we begin all over again now?"

"More or less, yes."

"So that at this place, A, it's always raining or just going to rain?"

"Yes, if it starts by being hot there, as it did just now, I suppose it is."

"What happens if it starts by being