206 Jie ooo (DeGs i. cBunek: | [ Mar. 7, 
obtained at the end of the period of darkness. As a check, the two positions 
of extinction are marked with both F and G. 
After coming out of the dark room the pins are carefully adjusted to each 
of the marks in succession, using a magnifying lens, and the corresponding 
readings taken. As there is no possibility of mistaking the order in 
which the marks were made, owing to the cards having been moved on 
between each observation, the Nicol may be turned back some 20° or 30° so 
that there may be nothing to prejudice the observer in his next attempt. 
Practically I found this made no difference and is not advisable in dealing 
with very feeble illumination, because it exposes the eye to an image which 
may be several times stronger than the minimum visible. 
Experiment 5.—Made at night, after walking a mile and then spending hal f- 
an-hour in the gas-lit laboratory. 
Having arranged the apparatus with the room lighted, I reduced the 
illumination of the spectrum to the minimum visible and adjusted the Nicol 
till the red R, and the blue-violet V, were equally bright, and pricked a 
record. I then moved the lamp A (Fig. 2) so much further from the paper 
reflector that I could see nothing whatever in the spectroscope, shut myself 
in the dark room, turned out the lights, and waited. After 15 minutes I 
could just detect the blue-violet band, and after 30 minutes the red also was 
visible, the blue appearing most luminous. I then adjusted the Nicol until 
both were equally bright, and pricked the record. 
The first record was 13°77 and the second 37°2 from the position of 
extinction of the red. The intensities of the two images were, therefore :-— 
s 
V,D,~? cos? 13°7 : R,D,~* sin? 13°7 in the first case, 
and V,D,~* cos? 37:2 : R,D,~* sin? 37-2 in the second, 
where D, and D, are the distances of the lamp A from the paper in the 
first and second cases respectively. That is to say, after half-an-hour’s rest 
in darkness the sensitiveness of the eye had increased to red, and had increased 
also to blue-violet, but in so much greater proportion that it now required 
9-7 times as much red to match the feeblest visible blue as it did before the 
eye was rested. 
Experiment 6.—Made on a dull day after a walk of three-quarters of a 
mile, followed by half-an-hour in the laboratory. 
The Nicol was adjusted till the red and the blue-violet appeared equally 
bright, all the blinds being up and the gases lit. The intensity was just too 
low to enable me to see the cross-wires. Taking this ratio as unity, at the 
end of 5 minutes the red had to be increased to 6 times, and after 20 minutes 
