748 ENERGY AND VISION 



sKt (20 + 3 mm.), in front of the glower. The reflection from the 



heater and porcelain support was suppressed by fixing the glower 



by means of its platinum wires at the end of two leads. The inside 



of the chamber was blackened with soot. A voltmeter was placed 



across the terminals, and an ammeter in series, so as to know exactly 



the input in watts. Under normal conditions, it was found to be 



87.5 volts X 1.05 amperes = 91.875 watts. These 91.875 watts 



are not all transmitted by radiation, but part of them are taken away 



by conduction and convection by the air. Lux and others give 



, . input ^ , r 1 T«- 



the ratio — : — : — . But as these figures may correspond to diner- 

 radiation. 



ent types of lamps, it was found safer to measure it directly. Besides, 

 this would allow us, by a simple calculation, to check our radiation 

 data against those published previously on the Nernst lamp. 



It was first attempted to use a specially made mercury thermome- 

 ter, with a known weight of mercury in a known weight of glass, 

 blackened on the bulb of which the rays emerging from a 0.1 sq. cm. 

 slit were concentrated by means of a fluorite lens of short focus. 

 This process showed a lack of sensitivity and it was necessary to 

 check it by means of an electric method. Although less difficult 

 to handle than a bolometric device, the following apparatus required 

 a great deal of care and time. A thermopile was made of copper 

 and constantan wires, with ten elements, disposed linearly; the cold 

 ends were simply bent out. On the top of the welded ends, carefully 

 planed and ground, a thin piece of tin-foil exactly 1 mm. wide and 1 

 cm. long was applied and fixed with a very thin layer of shellac. 

 Then the tin-foil was cut carefully between the welded ends, leaving 

 a little square table of very nearly 1 sq. mm. on each thermocouple. 

 These were blackened with soot, and the whole thermopile fixed 

 in a thermostat. The rays were allowed to fall on the pile through 

 an adjustable sHt, and the distance between the source and the 

 couples made equal to 1 meter. The method consisted in compen- 

 sating the heat generated by the incoming radiation, by the current 

 sent in a strip of constantan placed near the cold ends of the ther- 

 mocouples, in a tiny calorimeter, 2 cc. in capacity, filled with oil, 

 and well isolated. The following formula was used: 



0.1 sq. cm. sec. 



