As we have seen, testability is the heart and soul of any scientific theory. However, as crucial as it might be, it is never sufficient to test a theory against the observations that led to its creation. A scientific theory must predict unknown facts that can only be confirmed by fresh observations.
As a simple example let's assume that after you had tasted sugar, you concluded that white, small-grained powders are sweet. The unscientific way would be to use sugar as a proof for your theory. The scientific way, on the other hand, would be to seek other white, small-grained powders and taste them. If you stayed alive after your experiment, you would have to admit that your theory was wrong. You will have to find another explanation for the sweetness of sugar.
If anything the theory predicts is proven wrong, the theory is incorrect. This potential refutability is a powerful criterion that distinguisesh science from psedu-science. Nowhere it is as clear as in the case of astronomy with astrology.
For thousands of years astrologers had been using the apparent movement of the sun, moon, and the five known extraterrestrial planets (Mercury, Venus, Mars, Jupiter and Saturn) to foretell earthly events. Once a sixth planet, Uranus, was discovered in 1781, astrologers added the new planet to their charts and continued with their foretelling as before, without asking whether astrology is still valid with the new planet, or why they had not fotetold the existance of the new planet.
Astronomers, on the other hand, were fervidly forecasting the orbit of Uranus. Their calculations, however, based on Newton's theory of gravity, did not match the observed path of planet. There were two potential explanastions: either the theory of gravity was wrong, or that the gravity of an unknown heavenly object, farther from Uranus, was responsible for the deviation. Using the same law of gravity, astronomers forecasted where the body that distorts Uranus' orbit should be. And they were right. In 1846, as predicted, they discovered a new planet, Neptune.
Just as discrepancies in the orbit of Uranus had led to the search for Neptune, irregularities in the orbit of Neptune, and to a lesser extent in the orbits of Uranus and Saturn, had led scientists to suspect the existence of a ninth planet. Once more, it was theoretical calculations that led to the discovery of Pluto in 1930. Their calculations were confirmed, and so was the theory of Gravity -- at least until the next observation.