1912-13.] 
273 
The Scattering of Light. 
In this connection it is instructive to note that the turbid solutions 
nos. 9 and 10 scatter the light so much that the maximum angle has not 
been reached by the available data (only a few measurements were made 
for the sake of contrast with the ground-glass results, and with some 
difficulty on account of the faintness of the light). 
The actual determination of the angle at which maximum total emission 
occurs in any case is not easy. I tried determining the constants of 
a function of 6 so that it should fit the relative intensity curve as well 
as possible in the neighbourhood of the region where, after multiplication 
by sin 0, the curve attains its maximum value. Multiplication of the 
formula by sin 6, and subsequent differentiation by 0, then led in the 
usual way to the value of 0 giving the maximum value. This procedure, 
however, proved to be tedious, and lacking in sufficient accuracy. Finally, 
the maximum was determined in each case by reading off from the relative 
intensity curve the value of the intensity for each successive degree, multi- 
plying it by the corresponding value of sin 0, and simply picking out the 
maximum from the values so obtained. The result for each experiment 
will be found under the heading “Angle of Maximum Total Emission.” 
In § 2, Table II., it may seem needless to set down the results of 
multiplying the relative intensities by sin 0, and the value of 0 at which 
the maximum occurs, for of course these figures have here no direct 
physical significance ; but it has been done simply for the sake of com- 
parison, and to bring out more fully the nature of the functions obtained 
from the measurements. 
In considering the figures tabulated above, the first point that suggests 
itself for investigation is the question of how the effectiveness of the scatter- 
ing action depends on the kind of surface. It will be seen that, if we take 
the suggested course of using the angle of maximum total emission as our 
criterion, the order of the surfaces, beginning with that giving the smallest 
amount of scatter, is as follows : (1) coarse ground glass, (2) acid-etched 
ground glass, (3) ordinary ground glass, (4) specially finely ground glass, 
and (5) the turbid solutions of resin. (See Plates I. and II., figs. 1 (4, 5), 
2, and 3 respectively. It is to be remembered, of course, that two plates 
of the “ordinary” ground glass were used.) The reason for this result 
is perhaps as follows. It is evident from the photomicrographs that the 
coarser, and less scattering, varieties of glass have pits of larger area, and 
it seems likely that the larger the area of the pit the less steeply do its 
sides slope, and therefore the less the deflection of the light from the 
normal direction. As the etched glass has pits with much larger areas 
than have any of the other sorts of glass, it would naturally be supposed, 
VOL. XXXIII. 18 
