circular Polarization of Light in Liquids. © 243 
semi-transparent; so that the double refracting prism cannot 
be applied to the analysis of it. But this difficulty is at once 
provided for, and the compound “achromatized prism dis- 
pensed with, by the very simple eye-piece which I have 
adopted ; for the use of which the fluid need not be absolutely 
transparent if it only allows a sufficient quantity of light to 
pass. 
This eye-piece consists simply of a rhomb of calc spar in its 
natural state, the light being admitted through a small hole 
in the end or bottom of the short tube which contains it, of 
such a size, relatively to the thickness of the crystal which the 
light traverses, that the two emergent images of the hole shall 
not overlap each other; this is easily found by trial; as how- 
ever these images may be too small readily to follow the 
changes of tint in them, I magnify them by a lens of short 
focus fitted in a short tube sliding in the upper part of that 
containing the rhomb. It only remains to attach to the tube 
a graduated rim which can turn with it in an outer cell which 
is attached to the stand of the apparatus, and on which changes 
in azimuth, or the arcs of rotation of the rhomb about the ray, 
are measured. 
By the arc necessary to be revolved through, by the rhomb, 
in order to make the extraordinary image come to its mini- 
mum, as compared with the position for that effect, when no 
liquid is interposed, it is, that we estimate the rotatory power 
of the liquid. 
In the eye-piece thus constructed, it will be evident that the 
object at which we look is the small hole at the bottom of the 
rhomb. So long then as enough light enters that hole it is im- 
material how irregular the refractions of it may be in passing 
through the tube before it reaches the hole: we are independent 
of the distinctness of the image which it transmits, and the only 
material point is the intensity of light which the liquid allows to 
pass; and this is in fact one of the chief difficulties we have 
to contend with in these experiments; since many solutions 
which appear sufficiently transparent in small thicknesses are 
by no means so when we come to thicknesses of 12 inches or 
more. 
In my first construction the polarizing part of the apparatus 
consisted of a simple plate of glass inclined 35}° to the axis 
of the tube, and in order to have the polarized ray vertical, it 
was necessary to throw the light on this reflector by means of 
a small silvered mirror. 
In this part, however, I find it a most material improvement 
to substitute a Nicol prism for the plate of glass; this of course 
has its axis coincident with that of the tube, and the small 
R2 
