26 Dr. E. N. da C. Andrade an the Electrical 



pyrometer (Lummer and Kurlbaum type) was used, the 

 observations being made through a red glass. By observation 

 of the current required to make the filament in the instrument 

 glow at the brightness of the luminous streak, so as to be 



invisible when against it, we can obtain the temperature of 

 the " black body which would radiate the given radiation 

 with the observed intensity. Curves (by the Phys. Tech. 

 Reichsanstalt) are supplied with the instrument which 

 enables this to be read off directly. Knowing this we can 

 calculate the relative amounts of energy radiated for two 

 different readings. Using Planck's radiation formula 



"2 



e XT -l 



wehavo 5l _ „X \X~tJ, 



%~ e 



where r 2 is a constant having the value 14200 if X be expressed 

 in fi, and Tj, T 2 are the u black-body " temperatures in the 

 absolute scale. 



The amounts of salt vaporized arc very small, fractions 

 of a nig. per hour. A microbalance to weigh these amounts 

 was constructed by fixing a light horizontal arm of aluminium 

 at right angles on a stretched horizontal quartz thread. To 

 one end of the arm a hook was hung by means o\' a very 

 fine silk fibre, while at the other was a counterpoise, which 

 caused the arm to rest agains] a stop when the balance was 

 unloaded, while when the platinum wire was hung on the 

 hook, the arm swung free. Readings were made with a 

 horizontal microscope provided with a scale in the eyepiece: 

 a whole sealo-division (tenths could bo read) corresponded 

 to a change in weight of 0*2 mg. The whole was enclosed 



in an air-tight ease and kept Avy with PgO fi . 



The strontium sail (SrClj) was always glowed before the 

 experiment, SO as to be in the state o\' oxide, tli- 1 wire o\ tbe 

 loop being covered with a thin layer of -alt : a bead of sail 

 in the loop could not be used, as with so mueh sail tbe 

 luminosity varies too quickly, and there is also danger of 

 the nms< breaking away partly from tbe wire. I be wire 

 was generally in tbe flame for an hour, observations with the 

 pyrometer being made at regular interval-. The luminosity 



was altered from experiment to experiment by varying the 



small amount o\' sail on the wire (by prolonged glowing); 



