Within our solar system the masses of the planets are practically negligible compared to the mass of the sun, and it is for this reason that they appear to revolve about the _center_ of the sun. As a matter of fact, no body in the universe revolves about the _exact geometrical center_ of another body, but two mutually attracting bodies revolve in orbits about their common center of gravity, which always lies between the two bodies on the line connecting them and at a distance from each of them that is in inverse proportion to the mass of the body. The moon does not revolve about the _center of the earth_, but about the _center of gravity_ of the earth and moon, which lies on the line connecting the two bodies and at a distance from the earth's center that is one eighty-first of the distance from the center of the earth to the center of the moon, since this represents the ratio of the masses of the two bodies. This center of gravity of the earth and moon, lies, then, about two thousand miles from the earth's center, and about this point both earth and moon trace out orbits of revolution that are identical in form and differ only in size. In the same way each of the planets of the solar system revolves about the center of gravity of itself and the sun, but the mass of the sun is so far in excess of the combined masses of all the planets that we may consider, for all practical purposes, that the planets revolve about the sun's center, the center of gravity of the system being within the sun, just as the center of gravity of the earth and moon is within the earth.
Prof. T. J. J. See found from the investigation of forty binary star orbits that the average eccentricity of double star orbits is twelve times as great as the average eccentricity of a planetary orbit, and that the masses of the component suns never differ very greatly. The center of gravity of a binary system, therefore, lies at a great distance from the centers of the stars, and about this point, as a focus, the stars move in orbits that are exactly similar in form but differ in size in inverse proportion to the ratio of the masses. Since the orbits of binaries are, moreover, very highly eccentric, the two suns are, as we have said, anywhere from two to nineteen times nearer to each other at periastron than they are at "apastron."