398 M. MELLONI ON THE POLARIZATION OF HEAT. 
with a smaller aperture. A longitudinal opening made from end to 
end in the superior surface of the case enables us to see whether the 
axis of the pile is placed in the direction of the radiation. By gently. 
moving the case in one direction or another, the circular aperture is 
brought into the proper position relatively to the pile, which continues 
to occupy the same part of the table on which it has been placed. 
By means of this double case, the air surrounding the thermoscopic 
body is kept perfectly calm. At least the motions produced in it by 
its slow variations of temperature react upon the two sides of the pile 
with such uniformity that the index of the thermomultiplier stands 
exactly at the zero of the scale, and when a specific deviation has been 
produced by the influence of an exterior calorific radiation, returns to 
that point in some minutes after the communication of the rays has 
been intercepted by means of a metallic screen. 
Those who are in the habit of frequently using astatic galvanometers 
ust have remarked, no doubt, that these instruments, whatever may 
be the solidity of the table on which they stand, suffer, in consequence 
of the observer's changing his place or of the passing of vehicles in the 
neighbourhood, a slight tremulous motion, which, being transmitted to 
the suspension thread and to the needles, causes them to oscillate like 
a pendulum for a period of greater or less duration. For the purpose 
of preventing these eccentric oscillations, which prove so embarrassing 
in very delicate observations, we have but to fix the instrument on the 
marble slab of a chimney or any other horizontal plane firmly fastened 
to one of the solid walls of the building in which we are carrying on 
our operations. The wire will then always preserve its vertical direc- 
tion; the needles will be no longer susceptible of any other than hori- 
zontal movements ; and the index will deviate with so much regularity 
under the action of the electric currents, that, upon seeing it, one might 
fancy that, instead of being suspended by a thread, it moves on a 
pivot. 
In the particular instance under consideration, the deviation com- 
mences as soon as the rays of the calorific source, at a constant tempe- 
rature, have reached the anterior face of the pile through the aperture 
in the receptacle. The motion of the needles is at first very slow, but 
becomes gradually accelerated, and, having attained its maximum of ve- 
locity, is again retarded by imperceptible gradations, until at last it 
ceases altogether: the needles then return gently towards zero, de- 
scribe an arc of some degrees, resume the direction of the first motion, 
and, after having made three or four oscillations successively decreasing 
in extent, stand still and take a fixed position. Now this steady devi- 
ation is always a little inferior to the deviation indicated by the needles 
at their first departure from the point of repose. The difference varies 
csi a ’ 
