METALLIC WIRES BY ELECTRICITY. 465 
Mechanism of Incandescence. 
Incandescence also is not to be included amongst the mere 
calorific phenomena of electricity ; the mechanical effects which 
precede it, the formation of vapour, the vibrations, and more 
particularly the constant bends produced in the wire, prove this 
in a direct manner. Indirectly we arrive at the same conclusion 
from the facts, that the increase of temperature produced by 
altering the connecting circuit and the charge, the laws of which 
for low temperatures are known, is insufficient to enable us to 
calculate the temperature of incandescence (p. 434), and that, 
from the power of the current which produces incandescence in 
one metal, we cannot calculate the power required to produce 
incandescence in another metal. With wires of the same metal, 
differing only in their dimensions, the dependence of incandes- 
cence upon the power of the current was found exactly to corre- 
spond with what might be determined by the established laws 
regulating the excitation of heat in different wires. Like tempe- 
ratures in different measured wires presuppose the same relations 
of temperature in a constant wire, as were here really found for 
the incandescence. The variation from the course of the regular 
temperatures is here also perceptible, if the thermometrical 
changes in a constant connecting circuit, with different dis- 
charges, are compared with each other. For a constant circuit 
in which no part suffers mechanical action by the discharge, the 
thermometrical alteration 6, on the discharge of the quantity of 
electricity g contained in s jars, is expressed by the equation 
2 
- at, in which « remains so distinctly constant throughout 
a whole series of observations, that, as was the case in all my 
researches on heat, one value for « is sufficient to establish the 
whole series. This is no longer the case when a wire is included 
in the connecting circuit, which is mechanically affected, and is 
made incandescent by the discharge employed. The value of 
# decreases considerably from the first mechanical action upon 
the wire until it becomes incandescent, remains during the dif- 
ferent grades of incandescence nearly constant, and again in- 
creases on the wire being split in pieces and fused. This is 
obvious from the following experiments. 
Experiment 45.—A platina wire, 16 lines long, radius 0°0261, 
