COBLENTZ: BISMUTH THERMOPILES 357 



be published elsewhere. Further interpolations are being made 

 for the purpose of obtaining a complete table of this important 

 function. 



Credit is due Messrs. R. Weinstein and A. G. Seller for the 

 thoro and careful manner in which the computations have been 

 made. 



PHYSICS. — Summary of tests made on bismuth thermopiles. Wm. 

 W. CoBLENTZ, Bureau of Standards. 



During the past year various tests have been made in con- 

 nection with the improvement of the radiation sensitivity of 

 thermopiles made of bismuth and some other metal, e.g., silver, 

 iron, etc. The results of these tests are now published with the 

 hope that they may be useful to others. 



The term radiation sensitivity is to be emphasized for the 

 thermo junction giving the highest thermal sensitivity (emf.) ; it 

 is not necessarily the one having the highest radiation sensitivity. 



With reference to the radiation sensitivity, Johansen^ has 

 shown theoretically: (1) that the radii of the two wires of the 

 element should be so chosen that the ratio between the heat 

 conductivity and the electrical resistance is the same in both; 

 (2) that the heat loss by conduction thru the wires should equal 

 the heat loss by radiation from the junction; and (3) that the 

 radiation sensitivity is proportional to the square root of the 

 exposed surface. Johansen's practice, however, was poor, in that 

 he joined the ''cold," unexposed, junctions directly to the metal 

 posts supporting the instrument and unfortunately some manufac- 

 turers of thermopiles are now following his example. 



In the construction of thermopiles the most important desid- 

 eratum is symmetry of the "hot" and the "cold" junctions, just 

 as is true of bolometers. The unexposed junctions must be cov- 

 ered with receivers having the same size and emissivity as the 

 exposed junctions. Furthermore all the junctions must be free 

 from the supports so that they may quickly take the temperature 

 of the surrounding air which becomes heated by the incident 



ijohansen Ann. der Phys. (4) 33: 517. 1910. 



