"There's something satisfying, I think," Evans said, "about the idea of light traveling for millions of years through space and just at the right moment as it reaches Earth someone looks at the right bit of sky and sees it. It just seems right that an event of that magnitude should be witnessed."

Supernovae do much more than simply impart a sense of wonder. They come in several types (one of them discovered by Evans) and of these one in particular, known as a Ia supernova, is important to astronomy because it always explodes in the same way, with the same critical mass. For this reason it can be used as a standard candle to measure the expansion rate of the universe.

In 1987 Saul Perlmutter at the Lawrence Berkeley lab in California, needing more Ia supernovae than visual sightings were providing, set out to find a more systematic method of searching for them. Perlmutter devised a nifty system using sophisticated computers and charge-coupled devices—in essence, really good digital cameras. It automated supernova hunting. Telescopes could now take thousands of pictures and let a computer detect the telltale bright spots that marked a supernova explosion.

In five years, with the new technique, Perlmutter and his colleagues at Berkeley found forty-two supernovae. Now even amateurs are finding supernovae with charge-coupled devices. "With CCDs you can aim a telescope at the sky and go watch television," Evans said with a touch of dismay. "It took all the romance out of it."

I asked him if he was tempted to adopt the new technology. "Oh, no," he said, "I enjoy my way too much. Besides"—he gave a nod at the photo of his latest supernova and smiled—"I can still beat them sometimes."