1879 *] Electric Lighting by Incandescence . 171 
Of course, this is on the assumption that the heat con- 
ductivity does not vary with the electric conductivity, and 
this seems to a certain extent to be the case, since, while 
we find the electric conductivity of carbon varying between 
very wide limits, the heat conductivity, as far as results of 
experiments are available, seems to be much more constant. 
At first sight it might appear inconsistent with ordinary 
experience to say that for a carbon rod of given dimensions 
the smaller the specific resistance the greater the quantity 
of heat developed, but it must be remembered that here we 
are considering a constant electromotive force applied at the 
two ends of the rod, whereas in ordinary practice it is a 
constant current we have to deal with, since when even a 
Grove’s battery is employed to produce the current through 
a short continuous rod of carbon, the chief resistance is in the 
battery itself, so that the current is practically independent 
of changes of resistance in the carbon. 
Consequently, for this investigation, it is better to employ 
a form of carbon opposing a low resistance, than one offering 
a high resistance to the passage of the current. 
Substituting in our equation the known values of the heat 
conductivity, &c., of platinum and carbon, we arrive at the 
result that if our incandescent wire be about No. 20, Bir- 
mingham wire gauge, or have a diameter of about one 
millimetre, and if its length be five centimetres in the case of 
the platinum and two in the case of the carbon (this longer 
length in the first instance being necessary on account of 
the better conducting power of the platinum), then the 
electromotive force necessary to be maintained at the two 
ends of the wire is — 
0*2848 volts, or roughly one-third of that of a DanieiPs 
cell, in the case of the platinum, and 
0*46013 volts, or roughly one-half that of a DanieiPs cell, 
in the case of the carbon wire. 
It is, therefore, quite possible to produce a light with an 
electromotive force far less than that of a single DanieiPs 
cell, but not, of course, with a DanieiPs cell itself, since the 
resistance of this would be incomparably greater than that 
of our incandescent rod or wire. 
Although it requires 50 per cent more electromotive force 
to maintain the centre of our two-centimetre rod of carbon at 
iooo° C. than it does to maintain the centre of our five-centi- 
metre length of equally thick platinum wire at the same 
temperature, it is probable that the shorter carbon rod would 
give out more light — first, because, in consequence of the 
inferior heat conductivity of carbon, the temperature curve 
