298 



INFRA-RED EMISSION SPECTRA. 



the zero reading. Generally, however, the "drift" was from i to 2 

 mm, per 100 seconds, and since it was always in one direction it was 

 easily eliminated form the total deflection. 



From this prolix description it will be seen that a radiometer of such 

 extraordinary sensitiveness as the present one is useful only for work 

 where the radiation is weak. In all other work a less sensitive, quicker- 

 period instrument is more serviceable. 



From Table I the reader will obtain some idea of the sensitiveness 

 of previous instruments used in radiation work. The paraffin candle 

 is not a very satisfactory comparison source, but it is the only one for 

 which data are available. The "sensitiveness" is expressed in deflections 

 in centimeters per square millimeter of vane exposed, for a scale and 

 a candle each placed at a meter's distance. 



Table I.— Sensitiveness of Different Instruments. 



LSensitveness is defined as the deflection in centimeters per square millimeter of exposed 

 surface, for a candle and a scale at a distance of i meter.] 



Obser^'er. 



Boys, radiomicrometer — 

 Nichols, first radiometer... 

 Lewis, radiomicrometer... 



Snow, bolometer 



Stewart, radiometer 



Another form, esti- 

 mated 



Nichols, radiometer for 



star work 



Drew, radiometer 



Coblentz : 



Radiometer for vacu- 

 um-tube work 



Radiometer for absorp- 

 tion-spectra work 



Deflec- 

 tion. 



cm. 

 6 

 6 

 2 



15 

 10.3 



I . I 

 30 



26 



Distance 

 of scale. 



cm. 



(?) 



1-3 



(?) 

 300 



63 



183 

 100 



140 



Distance 

 from can- 

 dle to in- 

 strument. 



cm,. 

 152 

 600 

 300 

 100 

 300 



811 

 200 



140 



Area of 

 surface. 



sq. m,m. 



4 

 2X1.5? 



1.4 



•35 

 30 



3.14 

 7 



i{=.095X9-8) 



Period. 



sees. 

 10 

 12 

 20 



14 

 40 



45 



50 



45 



Sensi- 

 tiveness. 



cm,. 

 0.9 

 7(?) 

 1-3 (?) 

 14.7 

 4.9 



12.5 

 17. 1 



52-3 



8(?) 



The test of sensitiveness of the radiometer used in the vacuum-tube 

 radiation is given in Table II. Since it was impossible to place the 

 candle at a sufficient distance to keep the radiometer deflection upon the 

 scale for the slit width, 0.7 mm, used in the regular work, the slit was 

 decreased to 0.095 X 9-8 mm. (as found upon subsequent measurement 

 under the microscope) and readings made for the candle at different 

 distances. Several commercial paraffin candles having a diameter of 2 

 cm. were tried. The height of the flame was 5 cm. By keeping the 

 wick w^ell trimmed there was no appreciable change in the deflections. 



