another Observatory at another part of the earth (as the Cape of Good Hope); and then, by adding these two angles together, if the sum exceeded 180 degrees, we should consider that excess as due to the parallax, and should calculate the sun's distance in the same manner as in the case of the moon. But then comes in refraction, which is here a serious matter. The sum of the parallaxes of the sun at the two stations that is to say, the angle GMC, Figure 40, (supposing M now to represent the sun), does not exceed eight or nine seconds; it is an exceedingly small angle. The uncertainty of refraction in the observation of the sun is always two or three seconds; the air being hot at the time the observations are made, and the sun almost always appearing, in a telescope, tremulous and ill-defined. It is plain, therefore, that where there is such an uncertainty, it is useless to attempt to determine an angle which is not greater than eight or nine seconds. You may say, perhaps, that we could do the same thing with the sun as with the moon, by referring it to the stars. But there is this difference: we can see the moon, and we can see small stars close to the moon, and at the time that we observe the moon and the stars the air is not disturbed by the heat of the sun; everything is steady and is seen well; but in the heat of the day, objects are unsteady and are never seen well: we cannot see a faint star at all; and we can only see bright stars at a distance from the sun; therefore we are cut off entirely from observing the parallax of the sun in that way. We cannot observe its angular distance from the North and South Poles from the uncertainty of refraction, and we cannot compare the sun with the stars, because we cannot see stars near to the sun when the sun is shining.
