196 On tlie Thermoelectric Properties of Quartz. 
The following experiments seem to me peculiarly adapted to 
support my theory. 
A ring of tinfoil, of internal diameter 2 centim. and exterior 
diameter 4 centim., was cemented to a homogeneous plate of 
quartz cut at right angles to the principal axis ; the ring was 
then cut through radially at six points in the direction of the 
axis of no piezoelectricity, so that six pieces of the ring, 
insulated from each other, were obtained. The first, third, 
and fifth pieces were connected by wires with one half-ring of 
a Kirchhoff-Thomson electrometer, the second, fourth, and 
sixth pieces and the other half-ring of the electrometer being 
connected with the earth. 
If then, the plate possessing, to begin with, the temperature 
of the room, the central uncovered portion was heated by 
placing a warm metallic cylinder upon it, or by radiation 
from a flame or from a heated piece of metal, or by a current 
of hot air, or in any other way, then the portions of the ring 
became electrified, so that each portion of the ring acquired 
the same electricity as the end of the secondary axis lying 
next it would have acquired if there had been increase in 
a pressure acting in the direction of the corresponding 
secondary axis. Cooling of the central portion produced in 
any way, on the other hand, always produced the opposite 
electricities. 
If now, in a subsequent experiment, it was not the central 
portion of the plate, but that surrounding the tinfoil ring 
which was heated or cooled as the case might be, then, in case 
of heating, the electrometer showed the presence of the same 
electricity previously found by cooling the centre, and con- 
versely, in case of cooling, the same electricity as was produced 
by heating the centre. 
These results are not surprising if we start from the view 
that changes in tension in the crystal are the cause to which 
the evolution of electricity is due. In the first and fourth 
cases, for example, heating at the centre, or cooling at the 
periphery, produces tensions in the plate which are of the 
same kind as those produced by a uniformly distributed pres- 
sure exerted upon the edge. In the second and third cases 
central cooling or peripheral heating produces a condition 
of tension analogous to the condition brought about by a 
uniformly distributed tension exerted upon the edge. In all 
cases, during the time immediately following the heating or 
cooling as the case may be, a rapid increase of the tensions pro- 
duced takes place. Consequently in the two cases first men- 
tioned there must be the same distribution of electricity as 
would correspond to an increase of the pressure exerted upon 
