amounting to only 210 of a, second, is more than I can undertake to answer for. In consequence of that uncertainty, another method has been introduced, admitting of far greater accuracy; it is "by comparing two stars whose declinations are nearly the same. And here we fall upon another method, very similar to that which is used for measuring the distance of the moon. Suppose we have two stars S and S", and suppose that the star S" is at such an immeasurable distance that we cannot see in it any change of position. But suppose that I think it possible that the star S has a sensible parallax, these two stars being seen nearly in the same direction. I have already mentioned that we have obtained the parallax of the moon with the greatest accuracy, by comparing it with a fixed star, which is seen nearly in the same direction. We get rid of the uncertainty of refraction in this case, as the moon and the star are seen near to each other, and are therefore affected with almost exactly the same refraction. In like manner, if we compare a near star with a distant star, seen nearly in the same direction, we get rid of the uncertainty of refraction; and we also get rid of precession, nutation, and aberration; because they produce sensibly the same effect on both stars. Now, if I suppose S to be near, and S" to be at such an enormous distance that it will have no sensible parallax, then when the earth revolves round the sun, I have only at E' to observe the angle S"E'S between the two stars, and then in another position E’” to observe the angle S"E"'S between the two stars; and because E'S" is sensibly parallel to E'"S", the difference between these two measured angles, is the angle E'SE"'.
This is the method which the celebrated Bessel, of Königsberg, used for determining the distance of the
