SIR W. DE AY. ABNEY AND PROF. AY. AYATSON ON 
112 
Table III. 
Angle subtended 
by disc. 
V 
8'. 
34'. 
17'. 
SSN. 
log 
reduction. 
log energy. 
log_ 
reduction. 
log energy. 
log 
reduction. 
log energy. 
60 
r 
L012 
1-429 
1-327 
1-744 
1-795 
2‘212 
55 
. 
3-873 
0-052 
2-308 
0-487 
2-876 
1 -055 
50 
c5 
4-848 
2-830 
3-333 
1-315 
2-031 
0-013 
45 
r* 
o 
4-293 
2-089 
4-676 
2-472 
3-521 
1-317 
40 
=-*— 1 
4-126 
3-724 
4-555 
2-153 
3-388 
2-986 
35 
§ \ 
4-198 
3-629 
4-660 
2-091 
3-531 
2-962 
30 
4-424 
3-677 
4-903 
2-156 
3-776 
1-029 
25 
O 
lO 
4-761 
3-840 
3-241 
2-320 
2-082 
1-161 
20 
3-155 
2-079 
3-677 
2-601 
2-440 
1-364 
15 
3 • 505 
2-261 
2-143 
2-899 
2-855 
1-611 
10 
2-165 
2-683 
2-606 
1-124 
1-306 
1-824 
60 
r 
2-976 
1-393 
1-129 
1 • 546 
P292 
1-709 
55 
d 
2-246 
0-425 
2-429 
0-608 
2-574 
0-753 
50 
<D 
!>• 
2-130 
0-112 ' 
2-378 
0-360 
2-565 
0-547 
45 
2-251 
0-047 
2-533 
0-329 
2-788 
0-584 
40 
<1 
2-470 
0-068 
2-800 
0-398 
1-064 
0-662 
35 
- 
2-920 
0-351 
1-219 
0-650 
1-505 
0-736 
A good deal of work on the influence of the size of the stimulus surface on the 
threshold has already been published (see Nagel in Helmholtz’s ‘ Physiol. Optik,’ 
3 edit., 2, p. 284). Thus for the dark adapted eye Loeser found that for the fovea the 
threshold varied inversely as the square of the diameter of the circular surface used 
as a stimulus. For peripheral regions of the retina Ftper found that the threshold 
varied inversely as the diameter. One of ns in a previous communication (‘ Phil. 
Trans.’ A, vol. 190, 168, 1897) found that for the fovea and for discs which did not 
subtend a greater angle than 4 degrees the threshold varied inversely as D/'G5 where D is 
the diameter of the stimulus disc or the angle subtended at the eye. If the threshold 
value, T, varies inversely as I)", then on plotting the logarithms of the threshold values 
against the logarithms of D we should obtain a straight line, the tangent of the inclina¬ 
tion of which to the I) axis will be equal to n. In figs. 16 a and 16 b dotted curves are 
drawn for values of n equal to 1, 1 '65 and 2, that is corresponding to the results obtained 
by Piper, Abney, and Loeser respectively. From figs. 16 a and 16 b it will be seen that 
the numbers obtained by W. by the method described in this paper give approximately 
straight lines at the fovea and at the red end of the spectrum at 5 degrees from the 
fovea, but depart very appreciably from a straight line throughout the rest of the 
spectrum at 5 degrees. At the fovea the inclination of the lines joining the observed 
points agrees fairly closely with that corresponding to the threshold value varying 
inversely as the diameter. At 5 degrees on the other hand the inclination of the 
lines joining the observed points is approximately that corresponding to the threshold 
