Discovery of Cen X-3 as Binary Pulsar

The following diagram is something I created in 1971, when first proving that Cen X-3 was a binary system. I wrote a program called “sinefit” that did a best fit of a sine curve to the data and hit the jackpot.

First the sine curve was an exact fit to the data.

Second there was a big eclipse in exactly the right place in the curve.

Third, the long base line enabled us to increase the accuracy from about 1 part in a thousand to 1 part in a million, which enabled the determination that the 4.8 rotation period was getting shorter with time, which means the Cen X-3 pulsar was spinning faster as time went on. Completely the opposite of all other pulsars, known at the time, which were all slowing down.

Note: this diagram originally appeared in the discovery of Cen X-3 as a binary system paper: Cen X-3 Binary paper.