tongues of the Antarctic Sea, preserving in the main its characteristic temperature. The explanation of this seems simple. For some cause or other as yet not fully understood, evaporation is greatly in excess of precipitation in the northern portion of the land-hemisphere, while in the water-hemisphere, and particularly in its southern portion, the reverse is the case: thus one part of the general circulation of the ocean is carried on through the atmosphere, the water being raised in vapor in the northern hemisphere, hurried by upper-wind currents to the zone of low barometric pressure in the south, where it is precipitated in the form of snow or rain, and welling thence northward in the deepest channels, on account of the high specific gravity dependent on its low temperature, it supplies the place of the water which has been removed.
A New Form of Stereoscope.—It is known to many people that, by placing the axes of the eyes parallel, it is possible so to see stereoscopic pictures without any instrument as if we were looking at them through stereoscopic lenses. To do this, we may make a small hole in the centre of each picture, and hold the paper in such a position that each eye looks through the hole at a distant object. But with ordinary stereoscopic pictures this object needs to be very far off, so that this contrivance is
not a very convenient one; nor is it desirable to make a hole in the centre of every picture we wish to see stereoscopically with our naked eyes. Another method is to make one familiar with the muscular sensibilities of the eyes according as their axes converge or diverge, and to make him acquire such control of his eyes that, without looking at any distant object, he will be able to place the axes of the eyes parallel. Though either of the above methods would answer for experimental purposes, neither would serve as a substitute for the ordinary stereoscope; for if one requires such an inconvenient arrangement, the other requires too much trouble in explaining. But I thought that, perhaps by some happy, simple contrivance, we could see stereoscopic pictures unaided by any lenses, yet without any conscious straining of our eyes; and after repeated experiments I found out that, if corresponding parts of two pictures were apart from one another only one inch and a half or so, and if by means of a partition the two pictures were so separated from one another that the right eye will see only the right picture and the left eye the left picture, the two pictures will combine just as easily as with an ordinary stereoscope.
The following is my explanation why two pictures combine so easily when corresponding parts of the two are apart only one inch and a half or so, while to make ordinary stereoscopic pictures combine requires so great effort:
To combine ordinary stereoscopic pictures unaided by any lenses, the axes of the eyes must be placed parallel; but, since it is not an habitual position of our eyes to have their axes parallel, this can only be accomplished by a great effort. But, if to place the axes parallel requires such an effort, to make them too convergent requires equally great effort. Thus to see a single image of a finger put on the tip of our nose is very difficult, and to see a single image of the tip of the nose itself is almost an im-
