448 
CAPTAIN ABNEY AND MAJOR-GENERAL FESTING 
those portions of the spectrum on which he operated will, however, be seen by the 
figure to lie nearer to our curve than the former. But on consideration it would 
appear as if the curve should be drawn through the more luminous ends of the por¬ 
tions, as these would be the parts of the spectrum where the colour was actually 
last extinguished by the measured amount of white light. The curve drawn thus 
differs, as will be seen by reference to the figure, very considerably from ours. 
We believe that there are several disadvantages in Vierordt’s method, and much 
uncertainty attending its use, at least we have found more uncertainty than in the 
method of separating lines. We may call attention to the luminosity curve given by 
Sir W. Herschel in the Phil. Trans., 1801, p. 265, and even allowing for the impurity 
of the spectrum he worked with, it appears to us that Sir W. Herschel considerably 
over-estimated the intensity of the blue and violet. 
§ XX. Clerk Maxwell’s Diagrams of Colour Intensity. 
If reference be made to Clerk Maxwell’s paper in the Phil. Trans, of 1860, to 
which we have already alluded, it will be seen that his intensity curves of sunlight are 
very different in character to ours. There is a dip in the curve at about the position 
in the spectrum where our curve is most concave, but his curve rises again in the blue, 
indicating far more intensity than is the case in ours. It is not cur purpose now to 
discuss the cause of this difference, but it should be remarked that his intensity curve 
was arrived at by mixing light of different colours to compare in whiteness and 
intensity with the light, the source of which was a white screen illuminated by 
sunshine. * 
§ XXI. Intensity of Light passing through Turbid Media. 
We have thought that in connexion with this curve of luminosity of the light of 
the sun, as it reaches us after having passed through our atmosphere, it would be 
interesting to give a comparison of the curve derived from the light of the crater of 
the electric arc, after it had passed through a turbid medium, with that of the light 
passing through the same medium but unclouded. 
It also occurred to us that the theoretical results obtained by Lord Rayleigh 
regarding the colour of light passed through turbid media, as given in the Phil. Mag. 
(ser. 4, vol. xli. (1871), p. 447), required experimental verification. It is true that in 
the paper referred to he explains how he attempted to verify his results by comparing 
sunlight with skylight, and he remarks that “ considering the difficulties and uncer¬ 
tainties of the case the two curves ” (deduced from theory and experiment respectively) 
* Lord Rayleigh has kindly pointed out to us that the scale of ordinates chosen by Clerk Maxwell 
for the three colours is purely arbitrary. When his ordinates, at the three points where he supposes 
pure colour to be exhibited in the spectrum, are reduced to those of one normal curve, the agreement is 
better.—[July 23.] 
