(167 ) 
the same experiment was carried out in liquid hydrogen, we were 
unable even with the most intense ultra-red to get the green to 
appear; in this case only the blue appeared, and hence the local 
temperature of -f- 300° C. cannot be reached when the temperature 
of the substance is in the neighbourhood of —250° C. 
§ 6. Spectral distribution of excitation and temperature. In a 
former communication (‘‘1904” p. 471) the supposition was advanced 
that the excitation-distributions corresponding with the various phos¬ 
phorescence-bands were independent of the temperature. At that 
time, however, only the region extending from ordinary to higher 
temperatures had been investigated. Now, however, in a mixture of 
carbon dioxide and ether (about —50° C.), in liquid air (—180° C.) 
and in some cases also in liquid hydrogen (—250° C.) an intense 
spectrum has been thrown upon the “phosphor” spread out flat, 
and by this method of illumination the result given above was 
confirmed. Sr Cu Na, Sr Zn FI, Ca Mn Na, Ba Cu Li were investigated, 
the last “phosphor” in liquid hydrogen. Here it appeared that the 
characteristic excitation-distribution which is shown in PJ. Ill N°. 39 
loc. cit ., and which reaches far into the visible region, holds good at 
that just as at ordinary temperature. In this case the spectrum (from 
a carbon arc) exciting the phosphorescence had to pass through two 
vacuum glasses (cf. fig. 1), and by this some of the finer details 
were necessarily destroyed. In the mixture of solid carbon dioxide 
and ether and in liquid air, the “phosphors” were enclosed in flat 
quartz tubes and placed about 1 mm. below the free surface of the 
bath; the spectrum of a quartz amalgam lamp was then projected 
upon it from above through a quartz prism; in this way there was 
less opportunity for disturbance. Here, then, there remains only the 
great difficulty to which attention has already been drawn (“1904” 
p. 464) that the exciting spectrum is discontinuous, so that in a 
certain measure only the rougher features of the charge distribution 
always appear, while further finer characteristics must necessarily 
be obscured by the line-grouping in the spectrum which is used to 
excite the phosphorescence. It is these rougher features of the excitation- 
distribution — which were already reproduced PI. Ill “1904” l ) — 
that remain unaltered also at low temperatures. 
h One of these excitation-distributions, viz.: that of Ba Gu (PI. Ill N°. 40) must 
— as has since been discovered — undergo a correction. We intend later to deal 
in a separate paper with a new investigation of the excitation-distribution by means 
of a continuous spectrum. 
