OF HIGH REFRAN GIBILITT UPON GASEOUS MATTER. 
349 
in the latter brilliant. When the line of vision was transverse, the colours seemed 
mainly limited to red and green. 
The tube was swept with dry air and exhausted. Half an inch of air and benzol 
vapour was admitted, and after it half an atmosphere of air and hydrochloric acid. A 
line blue colour soon appeared, and as long as it continued the direction of maximum 
polarization was along the normal. But a luminous white cloud was rapidly generated, 
the normal polarization becoming feeble and the oblique strong. The distant end of the 
cloud, however, continued blue, and in passing from it to the white cloud the plane of 
polarization changed 90°. 
The tube was again exhausted, and a quarter of an inch of air and benzol vapour was 
permitted to enter it, followed by a quarter of an atmosphere of air and hydrochloric acid. 
The incipient cloud showed an exceedingly line blue, the polarization along the normal 
being a maximum. The cloud gradually thickened at the centre, and finally the polari- 
zation there disappeared. As before, when the normal polarization became feeble the 
oblique became strong. 
The tube was once more cleansed and one-tenth of an inch of air and vapour was ad- 
mitted, followed by one-tenth of an atmosphere of hydrochloric acid and air. The blue 
of the incipient cloud was here superb, and it lasted longer than in the last case. The 
selenite tints produced by the normally polarized light were exceedingly brilliant ; but 
they faded gradually as the cloud passed from blue to whitish blue. At the centre of 
the cloud the normal polarization first fell to nil a«cl then reappeared, having changed, 
however, from positive to negative, the two ends remaining as before. The influence of 
attenuation on the production of the blue colour is here strikingly exemplified. 
The tube containing the benzol vapour was again cleansed and exhausted, and the 
last experiment was repeated. That is to say, one-tenth of an atmosphere of the air and 
vapour was mixed with one-tenth of an atmosphere of hydrochloric acid. After ten 
minutes’ action the actinic cloud was found divided into five segments, alternately 
blue and white. Every two adjacent segments of the cloud were oppositely polarized, 
being divided from each other by a section of no polarization. The rectangle (fig. 2) 
represents the several divisions of the cloud ; the letters B and W denoting the blue and 
white segments respectively. The transverse lines represent the neutral sections. 
Fig. 2. 
B 
W 
B 
W 
B 
On the 9th of December, 1868, some experiments were made with the nitrite of butyl 
which merit a passing notice. 
Atmospheric air was permitted to bubble through the nitrite until the experimental 
tube was quite filled with the mixture. Fifteen minutes’ exposure produced a very slight 
action, an exceedingly scanty and coarse precipitate being formed. When due care is 
taken the action entirely disappears. 
3 A 
MDCCCLXX. 
