328 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1959 
air as affected by temperature and pressure; the other governs 
changes in the speed of light, and, incidentally, its direction, as it 
passes from an air layer of one density into that of another. There 
are two other factors involved—junior partners that have a definite 
hand in the business, to be discussed later. Watching a mirage, one 
often gets the impression that it is a directed performance, and one 
well done. 
The widespread belief that a mirage is something unreal, a sort of 
trick played on the eyes, is wrong. The picture a mirage presents is 
real but never quite accurate. The effect of the mirage is to change the 
seeming location or appearance of an object—not infrequently, both. 
The seeming shift in position may be hardly noticeable or may amount 
to many miles; change in appearance may be slight or extensive. Both 
of these changes are brought about by refraction; that is, a bending in 
the course of the light path from object to observer. The bending 
occurs when light passes from dense to less dense air, or vice versa. 
Most of us remember the example of refraction we were shown in 
school when a spoon was placed in a glass full of water. Looked at 
from an angle, the handle seemed to bend sharply where it entered the 
water. Whereas the difference in the densities of air and water is com- 
paratively great, that between two layers of air is very slight, so that 
the bending would be almost infinitesimal. A succession of such bends, 
however, could bring about an appreciable change in direction of the 
light path. To do this—and it is necessary for mirage making—a 
mass of air must have a progressive and abnormally great density 
change within it. The function of such an air mass as a bending or 
refracting agency, similar to that of the lenses in our optical instru- 
ments, seems to justify calling it an air lens, at least until a better name 
offers. Physically it bears little resemblance to our optical lenses. Its 
depth may vary from an inch or two to a hundred feet or more and its 
area from a few square feet to many square miles. Because it is com- 
posed of air and is surrounded by air, except where it rests on the 
ground, its boundaries are often hard to ascertain. The evidence at 
hand indicates that the lens is likely to be approximately flat and to 
operate chiefly in a horizontal position, since bending of the light 
path is mostly up and down, not sideways. 
It is not easy to accept the fact that a certain mass of air, surrounded 
by other air, can be maintained as a separate entity and do things the 
surrounding air cannot do; that it can continue as a thing apart for 
some time, strongly resisting integration. A demonstration of this 
type of phenomenon came to my notice in the Norris Basin of the 
Yellowstone Park a number of years ago. Norris Basin is an area of 
many small geysers which intermittently spew up steam and water a 
few feet into the air. Runoff from the many small geyser basins is 
