382 
DR. T. R. MERTON AND MR. S. BARRATT 
width of the Helium line /15015 A as a check on the accuracy of the method. The 
choice of lines in the secondary spectrum is limited in the first place by their brightness, 
and in the second by the possibility of their being isolated completely from neighbouring 
lines in the spectrum by means of the constant deviation spectroscope which was used 
for preliminary analysis of the spectrum. It was necessary that the exposures should 
be comparatively short to avoid disturbances due to the effect of changes of temperature 
on the echelon grating. An inferior limit to the temperature of the gas in the discharge 
tube Avas found in the following manner. The discharge Avas passed through the vacuum 
tube for a considerable time, in order to reach a state of equilibrium between the heating 
of the gas by the discharge and the cooling of the walls of the tube by radiation and 
convection. The temperature of the Avails of the tube was then measured by putting 
specks of organic substances of known melting point on to the Avail of the tube, and 
observing Avhicli of them melted. 
The temperature of the outer Avails of the tube being thus measured Ave deduce the 
temperature of the gas as folloAvs :—It is clear that Avhen thermal equilibrium has been 
established the temperature of the outer Avail of the capillary must lie between that of 
the radiating gas within the tube and the temperature of the room. An inferior limit 
to the temperature of the radiating gas is obtained by assuming that the interchange of 
energy takes place by radiation only, since an undue allowance for the effects of con¬ 
vection currents might lead to too high a value for the temperature of the gas. It is 
further assumed that the conduction of heat by the glass tube is infinite compared Avith 
that of the gas and of the surrounding air. The temperature of the radiating gas in the 
discharge tube is thus given by T 4 gas — T 4 glass = T‘ glass — T 4 room , and it will be noted that 
the assumptions are such as to lead to an inferior limit for the temperature of the 
gas. This is an important consideration, for an inferior limit to the theoretical limit to 
the widths of the lines under the conditions of experiment is required. It is also 
important to note that any ill-adjustment of the apparatus, resulting in a loss of defini¬ 
tion, will give rise to too great a value for the determined half-widths of the lines, and 
therefore to too small a value for the mass of the radiating particles. It folloAA^s 
that when values for the half-width and the temperature have been determined in the 
manner described, the masses of the radiating particles must exceed a certain specified 
amount. 
With a discharge of suitable intensity it was found that small specks of cinnamic acid 
were just melted. The melting point of this substance is 133° C., and applying the 
correction in the manner described above, the temperature of the radiating gas is found 
to be 456° Absolute. In the following Table are gwen the results obtained for the three 
secondary Hydrogen lines and also for the green Parhelium line. In the latter case the 
temperature of the radiating gas was probably someAvhat higher, as the gas was con¬ 
tained in a vacuum tube Avith a narroAv capillary, the walls of which were much thicker 
than in the case of the Hydrogen tube, but the theoretical half-width has been calculated 
on the assumption that the temperature was the same in both cases. 
