38 PROFESSOR SWAN ON THE GRADUAL PRODUCTION 
taken the place of 4, the aperture m will have moved out of the line of vision, 
and thus no light can reach the eye through /. Of the four apertures in the 
wheel D we may employ one only, or two 180° apart, or all four, as may be found 
most convenient, according to the speed with which the train is driven, the 
apertures not in use being stopped by plugs of cork. 
The photometrical arrangement for measuring the brightness of the luminous 
impressions transmitted by the revolving disc, and its train of wheels, may either 
be that described in my former paper on the formation of luminous impressions, 
or the following arrangements may be adopted :— 
1st, Let C, D (fig. 3) represent screens having apertures covered with ground- 
glass, or glass rendered milky-white by arsenic or phosphate of lime, and A, B 
flames, by which the apertures C, D are illuminated. A piece of transparent 
parallel glass E is placed so that an observer at F views the aperture C by trans- 
mitted, and D by reflected light in apparent contact. The light A remains fixed, 
while B is moved, until the two apertures appear equally bright. The disc Gis now — 
made to revolve, when the aperture C, seen during successive short intervals of 
time, will appear less bright than before, and the light B must be withdrawn to 
a greater distance from the screen D, until the equality of the illumination is 
restored. If the lights emitted by the flames A, B remain constant, the ratio of 
the brightness of the impression of short duration, transmitted by the revolving 
disc to that of a complete impression on the eye, will be that of the squares of 
the distances of the light B from the screen D. 
2dly, An arrangement which, from an imperfect trial I have made of it, pro- 
mises to succeed well for comparing the brightness of the illuminated apertures, 
may be made by cementing together two equal and similar rectangular glass 
prisms A B C, B C D, so as to form a parallelopiped, by means of a small portion 
of Canada balsam, which, when the prisms are pressed together, expands into 
a circular thin film E. The illuminated apertures C’, D’ in the screens are placed 
opposite to the faces A C, C D, and the observer looks through the face B F. 
The light transmitted through A C, and falling on B ©, will be totally reflected, 
except the portion which falls on the film of Canada balsam at E, which will be 
nearly all transmitted to the eye of the observer. The light which is transmitted 
through the face C D will be totally reflected to the eye by the face B C, except 3 
what falls on the Canada balsam at E, which will be nearly all transmitted. The 
spot E will appear of a different brightness from the rest of the surface B C, 
except when the light totally reflected by B C is equal in intensity to the sum of 
the lights transmitted and reflected at E. The spot E will then disappear, owing 
to the whole surface.of B C, including the spot, becoming uniformly bright. 
Assuming that the light partially reflected at E has a constant ratio to that 
totally reflected by the rest of the surface B C, and to that transmitted by A C, 
it is obvious that the squares of the distances of the flame from the aperture D 
