206 THE ROYAL SOCIETY OF CANADA 



spectrum could be recorded photographically at least as far up in 

 the infra-red as X = 11137A. It was also shown that intensity measure- 

 ments on line spectra radiation could be made with considerable ease 

 by the use of one of the thalofide cells devised by T. W. Case^, of the 

 Case Research Laboratory, Auburn, N.Y., U.S. A The active part 

 of such cells is a preparation of thallium-oxy-sulphide, fused on the 

 surface of a quartz plate; the latter being securely mounted within 

 an evacuated cylindrical flask about 2.5 cm. in diameter. Evacua- 

 tion was found to increase the sensitivity of the cell and to prevent 

 deterioration through oxidation. The thalofide cell has been found 

 to be photoelectrically sensitive in the near infra-red region from 

 X = 6000A to X = 12000A. The sensitivity curve is given by Coblentz 

 shows a sharp rise from X = 6000A up to X = 9000A and then a further 

 rise to X = 10000A. From this wave-length upwards the sensitivity 

 falls off rapidly. The photoelectric sensitivity of these cells is brought 

 into evidence by a lowering of the electric resistance of the active 

 preparation when the latter is exposed to radiations comprised within 

 the limits cited above. Though the cell has a maximum or specific 

 sensitivity at or near X = 10000A, it will be seen from the curve in 

 Fig. 22 which was prepared for Professor McLennan by Mr. V. P. 

 Lubovich from the spectro-photoelectric currents obtained with 

 various wave-lengths in the radiation emitted by a 400 watt nitrogen 

 filled incandescent filament lamp that the cell can be used with 

 advantage for certain types of work within the spectral region .5 f^ 

 to 10 fx. For example, it was thought that it might be used to spot 

 the wave-lengths in the infra-red grating spectrum of micrcury though 

 it was clear that the readings taken with it could not be taken directly 

 as a measure of the energy associated with these wave-lengths. In 

 this regard the thalofide cell is at a disadvantage compared with a 

 linear thermopile. 



To investigate this point the procedure adopted by McLennan 

 and Shaver was followed. The grating used (with Eagle mounting) 

 had a radius of one metre and a ruling of about 7.5 cm. with a total 

 of 46,167 lines. Several photographs were taken of the first order 

 infra-red spectrum of the light from a quartz mercury arc lamp 

 bearing a current of about 4 amperes, the overlapping radiations from 

 higher orders being cut off by the use of a Wratten filter known as 

 No. 22. A reproduction of one of these photographs is shown in 

 Fig. 24. From measurements on the lines of these plates the wave- 

 lengths of the radiation producing the spectrum were deduced. 



2T. W. Case, Phys. Rev. (2), Vol. 15, p. 289, 1920. 



