between Radiation, Energy, and Temperature. 225 
ments which we had some time ago initiated on the radiation 
from incandescence lamps might lead to results less liable to 
vitiation from external causes than those obtained from plati- 
num wire. 
Having in our possession incandescence lamps of many 
different patterns, we made careful simultaneous measurements 
of the radiation and of the energy in each when currents of 
varying strength were passed through the filaments. 
A Grove's battery was employed in preference to a dynamo, 
because of the greater steadiness of the current obtained. The 
current was measured in some cases by a tangent galvanome- 
ter, and in others by one of Sir W. Thomson's current-meters, 
or by both in circuit, the difference of potential between the 
terminals of the lamp being measured by a Thomson's potential 
galvanometer, and the radiation from the lamp by a thermo- 
pile the receiving surface of which was coated with lampblack, 
and which was connected with a Thomson's reflecting galva- 
nometer of '5-ohm resistance. The readings were taken of 
the first deflection caused by the thermoelectric current and 
checked by the total deflection. The former is the more rapid 
mode of proceeding, and, we believe, the more accurate of the 
two, as there is no change of zero-point during the time of 
observation. A pair of cardboard screens with narrow slits in 
them were placed between the lamp and the thermopile in 
such a way as to cut off as much as possible of the radiation 
from the glass of the lamp, and yet to allow to pass the radia- 
tion from a certain length of the filament. Another cardboard 
screen was interposed in front of the lamp after each observa- 
tion to cut off all radiation. 
The lamps were also heated in an oven to different tempe- 
ratures up to 350° C, and the change of resistance of the 
carbon filament measured directly by means of resistance- 
coils. We hope to be able in a subsequent paper to refer to 
the measurements of higher temperatures, our investigations 
in this respect being not quite complete. 
The results so far obtained may be summed up as follows: — 
The current can be expressed as a function of the potential ; 
the radiation, after a certain temperature of the filament has 
been reached, bears a simple proportion to the energy expended 
in the lamp ; the resistance can be formulated as a function 
of the energy and therefore of the radiation ; and the tempe- 
rature appears to be nearly a simple function of the resistance. 
The curves described by using as ordinates and abscissas the 
amounts of current and potential in the different experiments 
have the same general form, which appears from inspection 
to be that represented by the equation c = ap + bp^ ; where c is 
Phil. Mag. S. 5. Vol. 16. No. 99. Sept. 1883, S 
