It is when we come to consider the impressions and sensations described by sky voyagers of bygone times that we find them curiously at variance with our own. As an instance, we may state that the earth, as seen from a highflying balloon, used to be almost always described as appearing concave, or like a huge basin, and ingenious attempts were made to prove mathematically that this must be so. The laws of refraction are brought in to prove the fact; or, again, the case is stated thus: Supposing the extreme horizon to be seen when the balloon is little more than a mile high, the range of view on all sides will then be, roughly, some eighty miles. If, then, a line were drawn from the aerial observer to this remote distance, that line would be almost horizontal; so nearly so that he cannot persuade himself that his horizon is otherwise than still on a level with his eye; yet the earth below him lies, as it seems, at the bottom of a huge gulf. Thus the whole visible earth appears as a vast bowl or basin. This is extremely ingenious reasoning, and not to be disregarded; but the fact remains that in the experience of the writer and of many others whom he has consulted, there is no such optical illusion as I have just discussed, and to their vision it is impossible to regard the earth as anything but uniformly flat. Another impression invariably insisted on by early balloonists is that the earth, on quitting it, appears to drop away into an abyss, leaving the voyagers motionless, and this illusion must, indeed, be probably universal. It is the same illusion as the apparent gliding backwards of objects to a traveller in a railway carriage; only in this latter case the rattling and shaking of the carriage helps the mind to grasp the real fact that the motion belongs to the train itself; whereas it is otherwise with a balloon, whose motion is so perfectly smooth as to be quite imperceptible. Old ideas, formed upon insufficient observations, even if erroneous, were slow to die. Thus it used to be stated that an upper cloud floor adapted itself to the contour of the land over which it rested, giving what Mr. Monck Mason has called a "phrenological estimate" of the character of the earth below; the clouds, "even when under the influence of rapid motion, seeming to accommodate themselves to all variations of form in the surface of the subjacent soil, rising with its prominences and sinking with its depressions." Probably few aeronauts of the present time will accept the statement.