Page:Optics.djvu/135

 ocean, when nothing is seen "nisi pontus et aër," the horizon appears very much less distant than when there are lands or vessels in view.

We proceed to the fourth means, namely, the different appearances of known objects, or the known magnitudes of their least visible parts.

There is, perhaps, not a more common or a more convenient way of estimating the height of buildings at a distance, than comparing it with that of men or cattle standing near them; and another has been used with nearly equal ease and probably greater accuracy, which is counting the courses of masonry in a wall, when they are of a known height, or which is still more certain, the number of rows of bricks, when that is the material used. This, however, amounts to nothing more than measuring an unknown magnitude by a known one, for the eye is not employed to determine the height of the man or horse standing by a building, or of the depth of a course of bricks: these are taken for granted, and all that is required of the eye is to tell how many times the one magnitude is contained in the other.

Different degrees of brightness, and different colours in known objects cause a difference of apparent distances.

It is well known that the farther an object is removed from the eye, cæteris paribus, the less distinctly it is visible, and the more its colour approaches to the natural blue of the air. We say cæteris paribus, because it is plain that in point of distinctness much must depend on the state of the weather in the case of known objects, and of those with whose external appearance we are not well acquainted, the apparent distance will depend very much on the brightness and distinct markings of the surface. In foggy weather all objects seem farther, and consequently larger than ordinary; a westerly landscape, in a clear morning with the sun upon it, seems nearer than it does later in the day.

"From the several preceding observations," continues Harris, "it appears, that after joining together all the helps we can have, our estimations of distances, beyond a certain limit, are gross and uncertain; and this limit also varies in different circumstances. And the more certain estimates we always make of near distances, seem, as has been before observed, to prove that in these cases,