Another phenomenon supposed to prove rotundity, is found in the fact that Polaris, or the north polar star, gradually sinks to the horizon as the mariner approaches the equator, on passing which it becomes invisible. First, it is an ordinary effect of perspective for an object to appear lower and lower as the observer recedes. Let any one try the experiment of looking at a lighthouse, church spire, monument, gas-lamp, or other elevated object, from the distance of a few yards, and notice the angle at which it is observed: on going farther away, the angle will diminish and the object appear lower, until, if the distance be sufficiently great, the line-of-sight to the object, and the apparently ascending surface of the Earth upon which it stands will converge to the angle which constitutes the vanishing point; at a single yard beyond which it will be invisible. This, then, is the necessary result of the everywhere visible law of perspective operating between the eye-line and the plane surface upon which the object stands; and has no relation whatever to rotundity. It is not denied that a similar depression of a distant object would take place upon a globe; it is simply contended that it would not occur upon a globe exclusively. But if the Earth is a sphere and the pole star hangs over the northern axis, it would be impossible to see it for a single degree beyond the equator, or 90 degrees from the pole. The line-of-sight would become a tangent to the sphere, and consequently several thousand miles out of and divergent from the direction of the pole-star. Many cases, however, are on record of the north polar star being visible far beyond the equator, as far even as the tropic of Capricorn. In the _Times_ newspaper of May 13, 1862, under the head of “Naval and Military Intelligence,” it is stated that Captain Wilkins distinctly saw the Southern Cross and the polar star at midnight in 23·53 degrees of latitude, and longitude 35·46. [Illustration: FIG. 20.] This would be utterly impossible if the Earth were a globe, as shown in the diagram, Figure 20. Let N represent the north pole, E E the equator, C C the tropic of Capricorn, and P the polar star. It will be evident that the line-of-sight C D being a tangent to the Earth beyond the equator E must diverge from the axis N and could not by any known possibility cause the star P to be visible to an observer at C. No matter how distant the star P, the line C D being divergent from the direction N P could never come in contact with it. The fact, then, that the polar star has often been seen from many degrees beyond the equator, is really an important argument against the doctrine of the Earth’s rotundity.