CHEMISTRY AND PHYSICS. 35 
Frequently the related facts of comparison are quite obscure and seldom con- 
sciously recognized. Even when they are sought out they may be missed alto- 
gether in some cases, and erroneous conclusions may be stated and find wide 
acceptance in explanation of certain phenomena. 
Consider the case of the apparently increased diameter of the sun or moon 
when near the horizon over what it is when nearer the zenith. An explanation 
frequently given for this phenomenon is the unconscious comparison which the 
observer makes of sun or moon with objects near the horizon. Why does the ob- 
server not remember these impressions and give the sun and moon the same value 
when overhead? Why should there be an ‘unconscious comparison”’ at all in 
the mind of the observer between terrestrial objects at the horizon and the diam- 
eter of the sun or moon or planet? This is one of those explanations which gain 
currency, but which do not explain. It is a false correlation of facts. The true 
explanation of the familiar phenomenon cited is as follows: To every observer 
the impression of the contour of the heavens is that of a flattened dome, and not 
a hemisphere. We all conceive of the heavenly bodies as traversing this dome, 
on the surface of which we naturally think of them as located. This is a child- 
hood conception of the heavens, and all subsequent education and knowledge of 
the varying immense distances of the heavenly bodies can do practically nothing 
to alter these natural impressions. The dome appears flattened because we con- 
ceive of the distance to the blue in the direction of the horizon as greater than 
that overhead. 
In the illustration herewith given, let HH: represent the horizon, E the posi- 
tion of the observer, HMHi a semicircle, and HdH; the apparently flattened 
dome of the heavens. The angle of vision at E, subtended by the moon’s diam- 
eter, is slightly greater when the moon is in the zenith than when it is near the 
horizon. So the moon should actually ap- 
pear larger when overhead than when rising ; 
for when it is directly overhead we are nearer: 
to it by the distance equal to the earth’s: 
radius. But the difference in visual angle due 
to this nearer approach is so slight as to pass: 
unnoticed, unless we take special pains to de- 
tect it, as may be done by looking at the moon 
through a roll of paper so adjusted as to 
just take in its disc when at the horizon. When at the zenith the whole moon 
will no longer be visible through the paper roll, proving it to be actually nearer, 
though to the eye apparently smaller. Neglecting this small difference, the vis- 
ual angle is practically the same wherever the moon may be, and therefore that 
body should always appear of the same dimensions; and it would so appear did 
we but refer its position to the surface of a sphere and therefore always at the 
same distance, instead of to the surface of a flattened dome, and consequently at 
varying distances. In the illustration, an object referred to S, Si, or Se will not 
change in apparent dimensions, but referred successively to D, D1, Ds, it will 
apparently grow smaller until directly overhead, and thereafter will seem to grow 
larger until again in the horizon at Hi. Thus we see that ‘‘ objects near the ho- 
rizon, such as trees, buildings, etc.,’’ having nothing to do with the apparent 
size of the moon, sun, or other heavenly bodies. The stretch of the earth’s sur- 
face far out toward the horizon, beyond which we must still think of the blue 
vault as located, gives us an impression of greater distance in that direction to 
the blue than directly overhead. Besides, the greater quantity of light that 
