M. MELLONI ON THE POLARIZATION OF HEAT. 327 
by means of a magnet or a piece of soft iron held at a proper distance, 
and is then left to itself, it resumes its natural position of equilibrium, 
and stands perfectly steady at the zero of the scale after the lapse of 
three or four minutes. This quick return to a state of rest is the effect 
of the neutralizing action of the copper disc placed for that purpose 
beneath the divided circle; for, since the beautiful discovery of 
M. Arago, it is well known that the action of this disc diminishes the 
amplitude without changing the time of the oscillations ; so that the 
oscillatory motion of the needles about the point of equilibrium is con- 
siderably reduced in duration, though the forces by which it is pro- 
duced have undergone no change of intensity. The time required for 
the return is still less when the galvanometer is left in communication 
with the pile, and the deviation of 35° or 40° is produced by the calo- 
rific radiation ; for then about two minutes are sufficient to cause the 
index to resume its stationary position at zero after the interception of 
the radiation. In this case, the heat still remaining for some time on 
that face of the pile which has been exposed to the radiation may be 
said to sustain the needles in their fall, and prevent them from oscil- 
lating to the opposite side of zero, at which they arrive almost exactly 
within the time required for the re-establishment of the equilibrium of 
temperature between the two faces of the pile. 
But in order to observe the latter periods with the utmost possible 
exactness, it was necessary to secure the pile from the differently heated 
currents of air which come successively into contact with each of the 
two appendages. These differences of temperature, though exceedingly 
small and imperceptible with the best common thermoscopes, are sufhi- 
cient to produce, in the instrument used by me, such deviations to the 
right and left as occasionally amount to some degrees, and interfere 
most inconveniently with the action of the calorific rays. To prevent 
this inconvenience, I placed the pile and its stand in a large metallic 
receptacle or case resembling a trough with the bottom turned up- 
wards, and measuring eighteen inches in length by eight in breadth. 
The sides of this trough (which are double) are of such a height that the 
bottom may not touch the most elevated parts of the mounting of the 
pile; and the interval between the interior and exterior side is filled 
with cotton at the lower edge, in order as much as possible to prevent 
the entrance of the external air. The wires which establish the com- 
munication between the pile and the galvanometer are passed under 
_ the edges, which for this purpose are grooved at one of the extremities 
_ of the case. 
In one of the shorter sides there is, at the same elevation as the pile, 
a circular aperture about an inch in diameter; but this may be reduced 
_ to any degree of minuteness by sliding into an exterior frame a plate 
; z2 
