In an interview on Cosmology

Tell me about Darkness at Night, because, frankly, it doesn’t surprise me that nights are dark. What’s the point of the title?

Well, when people think about darkness at night they take it as obvious that it’s because the sun isn’t shining on that side of the earth. But actually it’s a lot more mysterious – in fact it’s a 400-year mystery first noticed by Johannes Kepler, the German astronomer. Exactly 400 years ago, Galileo looked through his telescope and wrote The Starry Messenger, and one of the things he saw was lots of faint stars that were not visible to the naked eye. Kepler read about that and wrote a reply, Conversation with the Starry Messenger, and pointed out something. He pointed out that if the universe goes on for ever, there would be stars behind the faint stars, and behind those stars other, still fainter stars. In fact, between every two bright stars there would be two fainter stars and between those another pair. So the whole sky should be papered with stars and as bright as the surface of the sun from horizon to horizon.

So why is it not?

Well, people have been wondering about it ever since, and oddly enough the first person who had an inkling of an answer was Edgar Allan Poe. Between 1840 and 1850 – some time around then – he conjectured that the light from the most distant objects might not have actually got here yet. And of course that was the explanation that everybody embraced in the late 20th century when we discovered the Big Bang. If the universe was born 14 billion years ago, it turns out that we can only see objects whose light takes less than 14 billion years to get here. So the answer to the mystery of darkness at night seems to be that the universe is not eternal: that it was born.

Amazing.

Except that’s not the answer. That’s what most professional astronomers think, but they’re wrong. And that’s why Harrison wrote Darkness at Night. There’s a flaw in Kepler’s logic, and the flaw is that he assumes stars burn for ever. In Kepler’s time, of course, they didn’t realise that the fuel stars burned would eventually run out. Conservation of energy was only discovered in around 1850. So the question is really, how long would it take the stars to flood the whole universe with light? Think of the universe as a bath and light as the water. How long would it take the stars to fill it? You can actually calculate that, and the fact is that stars don’t live that long. Kepler’s paradox is really no paradox at all, which is a bit of an anticlimax, but finding out why takes you on an amazing history of astronomy, as well as giving you the right answer to a question that 99 per cent of professional astronomers get wrong. It includes loads of historical and literary asides and really brings the subject to life.