158 Intelligence and Miscellaneous Articles. 
ization is at right angles to the plane in which the reflexion ought 
to occur. For on examining the cylinder of light round its axis in 
the direction of the maxima of polarization, it was found that the 
light proceeding from it was polarized tangential to that point of the 
surface of the cylinder towards which the polariscope was directed. 
Whether the plane of polarization becomes removed by its repeated 
reflexions from the gas molecules, or whether the action of gases 
under certain circumstances is analogous to that of refracting bodies, 
are questions which hitherto I have not been able to decide by ex- 
periment. 
I endeavoured to depolarize the light completely on its entry into 
the dark room, by allowing it to pass through a thin sheet of white 
paper; but the phenomena, with the exception of the intensity of 
the light, were quite the same. 
Light polarized by reflexion from a black glass experienced no 
perceptible change by the action of the smoke, and its plane of 
polarization always retained its original direction. 
It is possible, by suitably regulating the incident quantity of po- 
larized light, to succeed in finding a limit to the action of the gas 
molecules, beyond which the original polarization of the pencil pre- 
ponderates over the molecular forces of the medium which the light 
has to penetrate. 
The relations which these facts may possibly bear to the pheno- 
mena of atmospheric polarization, and perhaps also to fluorescence 
and the peculiar colour of bodies, have induced me to publish these 
observations, spite of their incompleteness. 
Some days after the preceding experiments had been laid before 
the Academy, I repeated them with more sensitive polariscopes, and I 
found exactly the same facts; I can further state that the plane of 
polarization of diffused light suddenly rotated 90°, on passing the 
direction in which I had seen all trace of polarization disappear in my 
previous experiments. 
Thus on receiving in the polariscope the rays emanating from the 
luminous track produced by the passage of the sun’s light, or of the 
electric light, through the smoke of incense, it is found that under a 
small inclination (the angles being measured from the luminous 
source) the polarization of diffused light is already very perceptible ; 
that it increases up to a certain angle, which is the maximum; it then 
decreases, and at the normal it is almost nil. Up to this point the 
plane of polarization is perpendicular to the plane which passes 
through the source of light, the place observed, and the eye or the 
polariscope. Above 90°, the polarization, although very feeble, reap- 
pears, but its plane is then perpendicular to the first plane. Still 
further it diminishes very rapidly, and the diffused light soon shows 
no sensible traces of polarized rays. 
I have investigated the smoke of tobacco in the same way; and 
the results were the same, though the angle at which I found the 
neutral point and the reversal of the plane of polarization was perhaps 
a little less than with the smoke of incense. 
It is possible that the nature of the diffused particles has an ap- 
preciable influence on these phenomena, and that different gases (if 
gases do diffuse light), vapours, and powders may in this manner be 
