MR. LOUIS VESSOT KING ON THE SCATTERING AND 
43-J 
Diagram XIII. —Quality of Sky Iladiation Calculated at Mount Wilson and Compared luith Ohservations. 
The curves (i), (ii), (iii) of Diagram XIII. are plotted from the results of Table X., and show graphically 
how observed values of sky radiation agree with values calculated from mean attenuation coefficients at 
Mount Wilson. The curves corresponding to the extreme and mean values of the solution of the integral 
equation are retained throughout. Observed points are denoted by small circles, and lie fairly close to the 
calculated curves. 
Diagram XIV. — Polarization of Zenith. Sky Calculated from irashington and Mount Wilson Ohservatwns. 
The curves given in Diagram XIV. are drawn from data calculated in Table XIII., and represent the 
ratio of the component intensities of zenith sky polarized in the principal plane to the component polarized 
at right angles to that plane. This ratio is represented roughly by the formula 
Ty(o^) ^ i + Q(o, 0/P(Q,C) 
T,{0,0 cos^C + Q (0,0/P (0,0 
Curves are given for various zenith distances of the sun, making use of the constants for Mount Wilson 
and Washington. In the latter case the effect of “ dust ” gives rise to a discontinuity in the neighbourhood 
of 'GlO/r, represented in an actual case by a peak in the curves (represented by dotted lines). The 
characteristic just mentioned is actually observed in curves of polarization of zenith sky obtained by 
Nicols.* In comparatively “ dust-free ” air as at Mount Wilson this peak is wanting, and the curve 
represents a type obtained by NicoLsf which rises with increasing wavedength. Since the forms of the 
curves just described depend essentially on the effects of self-illumination, these will be extremely 
sensitive to the presence of “ dust,” and will therefore vary greatly with time and place. Reflection of 
solar radiation from the earth’s surface {e.g., a snow-covered landscape) will affect in a marked manner 
the polarization of sky radiation independently of effects of self-illumination, and may therefore account 
for the great variety of results obtained by observation on this point. 
* Nicols, loo. cit., p. 508, fig. 10, Curves (J) and (It'). 
t Nicols, Ioc. cif., fig. 10, Curve (c). 
