144, 145.]
§21. On the Intensity of the Light Sensation
173
luminosity. Let the luminosity of a white surface illuminated by a dim
light be denoted by h. Now interpose an opaque body that casts a
shadow on the surface so that the parts of it in the shadow do not get
any light from the source. Then add a second light which by itself
would produce the luminosity H; which may be varied by varying the
distance between the surface and the second light. Then the objective
luminosity in the shadow is H, and outside the shadow it is (H+h).
Now if the luminosity H is very low, the eye will detect the shadow,
that is, it will distinguish between the luminosity H and the luminosity
(.H+h). But no matter how great h may be, it appears that there is
always a higher luminosity H for which the shadow is invisible; for
which, therefore, the difference h in the objective luminosity no longer
produces any perceptible increase of sensation.
A light equivalent to moonlight makes a perceptible shadow on
white paper; but if a good lamp is brought near the paper, the shadow
disappears. Again, the shadow made by the lamplight disappears in
sunlight. Indeed, the luminosity of the surface of the flame of a bright
oil-burner with a circular wick can hardly be distinguished by the
eye from double the luminosity. A flame of this sort is transparent
enough to show this, as may be easily seen by looking at its faint image
reflected in a plate of glass, and then shoving a second flame behind the
first. The outlines of the latter can be quite distinctly recognized.
But if both flames are viewed by the naked eye, the farther one will not
be seen through the other, at least not through its brightest part ;
unless, perhaps, the intensity of the sensation gets blunted by looking
at the lights too long. It is just as hard to tell by the naked eye that
the edge of the flame as seen lengthwise through the glowing film of
gas is very much bright er than the middle where the film is least deep ;
but it is easy to see this by looking at the image of the flame reflected
in a plate of glass. The same explanation applies also to the disappear¬
ance of the stars in the daytime, and to the disappearance of images in
a glass plate when it reflects other light, etc.
So far the difference of luminosity has been kept constant while the
absolute value of the total luminosity has been varied; but we may
also let the difference increase in the same proportion as the luminosity.
Suppose a drawing is made on a transparent plate of glass in very dilute
black india ink; or suppose the plate is covered with a thin film of
lampblack and lines drawn on it; or, best of all, suppose we have a
photograph made on transparent glass, with soft shadows in some
places and deep ones in others. If a diagram of this sort is held against
a bright background of steadily increasing luminosity, the soft shadows
will be found to be invisible when the luminosity is low; and then as
the luminosity gradually increases, they will begin to be seen and con-