358 
ME. W. CEOOKES ON EEPTTLSION EESULTING EEOM EADIATION. 
clearest. The best results seem to be obtained when there is a moderate amount of 
wind, and great white clouds are floating about. The method of observation was as 
follows : — Having adjusted the scale so that the index ray of light pointed to zero, and 
having brought the particular part of the spectrum I wished to examine opposite the 
aperture in the screen, I opened the shutter and watched the progress of the luminous 
index until it reached its furthest point, when the shutter was closed. As the scale 
was divided into millimetres, the number of these traversed by the index was taken as 
the value of the part of the spectrum which shone on the pith. The piece of pith 
was 13 millims. square; consequently each number so obtained must be considered 
the average of a portion of the spectrum 13 millims. long. 
189. I give below the actual results as they were obtained. 
Date. 
Portion of spectrum 
thrown on to pith*. 
Deflection of 
index ray 
of light. 
Observations. 
millims. 
July 26. 
c 
25 
,, 
D 
19 
E 
12 
b 
8 
F 
5 
,, 
between F and Gt 
2 
29. 
C 
61 
D 
34 
,, 
b 
12 
„ 
F 
9 
Aug. 4. 
A 
23 ^ 
| 
B 
24 
„ 
C 
35 
1 
y 
Sky hazy, and light varying very much. 
„ 
between C and D 
27 
f 
1 
D 
15 
1 
b 
10 
1 
5. 
between B and C 
40 
D 
29 
between D and E 
24 
b 
17 
F 
10 
between F and G 
6 
„ 
G 
3 
H, 
2 
13. 
ultra-red 
147 " 
121 
The sun was very bright and powerful. 
” 
B 
1 05 
L 
All the observations were taken between 
between C and D 
75 
f 
12 noon and 1 p.m. 
65 J 
i 
16. 
between A and B 
96 i 
D 
80 
„ 
between D and E 
57 
5? 
b 
33 
Sun very bright. No clouds. 
F 
25 ] 
r 
between F and G 
16 
» „ 
12 
” 
G 
5' J 
* The lines were caused to fall on the exact centre of the pith. Consequently the action is caused by the 
portion of the spectrum 6| millims. on each side of the portion here designated. 
t The exact position of the central line falling on the pith was in this and similar cases carefully taken and 
marked on a diagram (see Plate 35. fig. 14). 
