Juty 18, 1912] 
was much less affected. The blackening occurred to 
a much less extent if the tube was shielded. It seemed 
as though particles, possibly electrified, were shot off 
{ 
Loe. Amperes 
02+ 
Amperes —— 
20 
095 
P ——s_] 
1500° 1700° 1900° 
Temp. 
5-—The full line curve shows a relation between ionisation current | 
and temperature, for an applied potential of 2 volts on a1 cm. gap | 
between the electrodes. The dotted straight line is plotted from the | 
log. of the current and the temperature. | 
2100° 
2300°C. 
Fic. 
from the carbon walls of the furnace across a space 
of some 5 or more mm. into the material of the re- 
fractory tubes. 
Dr. G. W. C. Kaye and I were led to investigate 
the cause of these phenomena, and 
yesterday we gave an account of some 
of the results to the Royal Society. 
I propose to devote the remainder of 
my lecture to a description of the 
methods employed and the results 
obtained in what proved to be a very 
interesting investigation. 
Many experiments have been con- 
ducted, notably by Prof. O. W. 
Richardson, on the corpuscular emis- 
sion of electricity from carbon at very 
low pressures, but scarcely anything 
is recorded for pressures approaching 
atmospheric. Positive ions and 
material particles are also discharged 
by carbon, as well as by hot metals, 
at suitable temperatures and pres- 
sures. 
It is to be understood that in all 
the experiments now to be described 
the pressure remained atmospheric, 
and alternating current was employed 
for heating. Access of air to the in- 
terior of the furnace was prevented 
by windows at each end, perforated 
as required. 
In the early experiments we in- 
serted within the carbon-furnace tube 
two insulated carbon electrodes, one 
of them being hollow, so that a 
Siemens optical pyrometer could 
be sighted through it. The two electrodes were 
joined externally to an ammeter and a_ battery 
of cells (see Fig. 2), and we proceeded to 
determine current-voltage curves at various furnace 
NO. 2229, VOL. 89] 
200 
00 
IONIZATION CURRENT 
MICROAMPERES 
applied. 
NATURE 
S17 
temperatures. Some of these curves are shown in 
lig. 4 for an electrode gap of 1 cm. No appreciable 
current could be detected below 1400° C. with applied 
potentials up to 8 volts, but as the temperature rose 
the current rapidly increased until at 2500° or more 
currents up to 10 amperes were recorded. At the 
lower temperatures the currents soon attain what 
appear to be saturation values. At higher tempera- 
tures there is a linear relation between potential and 
current. As the length of the gap increased the 
current diminished at a regular rate, but the decrease 
was small. 
Fig. 5 exemplifies the exponential relation between 
temperature and current for a 1 cm. gap and an 
applied potential of 2 volte The dotted straight line 
was plotted to axes of temperature and logarithm of 
current. 
The magnitude of the currents made it evident 
that in spite of the high pressure the atmosphere of 
the furnace was ionised to an unusual degree at high 
temperatures, and we were led to investigate the 
effect of temperature alone. The battery was accord- 
ingly cut out, and one of the two carbon electrodes 
was mounted on a sliding carriage so that it could, 
at will, be moved in or out of the hot part of the 
furnace, i.e. away from the fixed electrode or back 
towards it. The movable electrode would thus be 
temporarily cooler than the fixed electrode which re- 
mained steadily in the furnace. The ammeter in the 
circuit indicated a current which amounted to 2 milli- 
amperes at 1400°, and nearly 2 amperes at 2500°; the 
cooler electrode was the positive one. The currents 
died away when the two electrodes attained the same 
temperature. 
The production of an alternating current of very low 
frequency is thus rendered possible by the use of some 
periodic device. In one form of the experiment 
(shown) the movable electrode is attached to a crank 
| 
ECTRODES Sax 
TIME 
;. 6.—Relation between ionisation current and time with a steadily rising temperature. The 
“enld” electrode was water-cooled ; the hot electrode was of new carbon. 
No potential was 
which, rotated slowly by clockwork, performs. the 
necessary displacement of the electrode within the 
furnace. The ionisation currents produced are 
sufficient to make a nest of small glow-lamps light up 
