TRANSACTIONS OF SECTION A. 411 
chamber to prevent the deposition of frost. Cooling was effected by liquid 
air applied from below to a metal tube at the upper end of which was the 
substance, while the lower end was submerged to various depths in a Dewar 
flask. Temperatures were determined by means of the resistance of a coil 
of pure copper wire surrounding the capsule containing the substance. 
The measurements consisted in comparing the brightness of the 
fluorescence spectra, wave-length by wave-length, with that of an acetylene 
flame which had previously been similarly compared with the spectrum of 
the light from an ideal black body of known temperature. 
The results are exhibited by means of three sets of curves. 
(1) Curves giving the variations of brightness of the fluorescence 
spectrum with wave-length at a given temperature. These show the form 
of the fluorescence bands, their changes in intensity and breadth and their 
shift in wave-length as the temperature diminishes. 
(2) Curves giving the variations of intensity of a given wave-length 
with change of temperature. 
(3) Curves giving the shift in wave-length of regions of equal brightness 
in the fluorescence spectra with change of temperature. 
6. Absorption and Fluorescence of Canary Glass at Low 
Temperatures. By R. C. Gress. 
The specimen of glass studied was obtained from the Geophysical 
Laboratory at Washington, and was prepared according to the following 
formula: SiO, 70 per cent., K,O 24 per cent., CaO 6 per cent., to which 
2‘5 to 3:0 per cent. sodium uranate is added. 
In order to study the absorption and fluorescence at low temperatures 
the glass, a rectangular plinth, was mounted in an unsilvered cylindrical 
Dewar bulb with proper screening, to avoid stray light. The observations 
were made with a Lummer and Brodhun spectro-photometer. For measuring 
the absorption an acetylene flame was placed so that by means of reflectors 
light could be made to illuminate the comparison slit, and also by another 
path to pass through the glass to the other slit of the spectro-photometer. 
A Cooper-Hewitt mercury lamp, placed so as to illuminate the surface of 
the glass parallel to the direction in which the light passed through the 
glass when studying the absorption, was used to excite fluorescence, the 
intensity of which was compared with that of an acetylene flame. 
The intensities of transmission and fluorescence were measured at five 
or six different temperatures, chosen at fairly regular intervals, from room 
temperature down to —175° C. Cooling was secured with liquid air, and 
the temperature measured with a thermo-junction by the potentiometer 
method. During any particular run through the spectrum the temperature 
was kept constant within two degrees. Extreme care was taken to eliminate 
frost or moisture from the surface of any of the glass through which the 
light passed. 
The glass shows considerable absorption throughout the spectrum, but 
the maximum occurs in the violet end and extends up to about ‘51. The 
decrease in temperature produces a decrease in absorption in all parts of 
the spectrum measured, the change being very slight in the yellow and red, 
while from ‘45 » to ‘54 » the transmission intensity increased from 15 to 
25 per cent. 
The main fluorescence band extends from ‘48 to ‘59 p. At room 
temperature the fluorescence corrected for absorption shows a rather broad 
band with a slight indication of two maxima between ‘51 p and ‘535 p. 
The curve is steeper on the violet side than on the red. With decreasing 
temperature the fluorescence for the most part increases, the maximum 
~ 
