Effects of UiiifoniLlt/f and Contract o)i Light. 
135 
In the dark room, on looking at the print the aperture of the pupil 
becomes set to a certain width as long as one does not move the eye away 
from the book. At the same time one has the feeling of the surrounding 
blackness, and the great contrasts prevailing make the eye roam instinc- 
tively. As soon, however, as it encounters the blackness the pupil 
extends. In the next instant the eye strikes again the brilliantly illu- 
minated paper, and consequently the pupil contracts, and apparently 
more than is necessary, because it experienced a glare. But too much 
contraction makes the illumination insufficient, and so the eye has to 
expand again. In other words, the aperture of the pupil alternately 
expands and contracts, making a higher average illumination necessary. 
In a room with white walls and ceiling, and especially with uniform 
illumination, the eye does not experience any great contrasts, and it may 
consequently set itself definitely to a certain aperture, which does not 
alter even if the eye roams about the room. A smaller illumination is 
thus required, because the pupil can extend to a maximum without having 
to be afraid of glare. 
It might be added here that even the ordinary black print on white 
paper causes some glare, and I have no doubt that for the preservation of 
the eyes coloured print would be preferable. 
In conclusion, I add a few curves (Fig. 2) which deal with the illumina- 
tions derived from various types of lamp shades. The superiority as 
regards uniformity of the prismatic glass or holophane shade is strikingly 
illustrated. It might be further improved by lengthening the shade or 
surrounding it with a deep, slightly obscuring frosted globe. The long 
frosted shade, although it diffuses the light badly, hides the whole lamp 
bulb, and the filament is therefore invisible. The intrinsic brilliancy is 
thereby greatly reduced." 
All the tests were carried out on a horizontal plane 1 metre above the 
floor, as this is now recognised as the standard test plane. Testing the 
illumination of inclined planes would make it impossible to take into 
account the illumination from all sources taking part in the lighting of 
the regions to be tested. 
* The holophane shades are somewhat expensive. Fig. 2 shows, however, that with 
the ordinary conical opal shade a uniformity superior to that of holophane globes may be 
obtained. The lamps must, however, be hung close to the ceiling, as the shades do not 
obscure the lamps. 
