COBLENTZ: NEW DESIGNS OF RADIOMETEES 473 



PHYSICS. — Some new designs of radiometers. W. W. Coblentz, 

 Bureau of Standards. 



In continuing the improvement of stellar radiometers several 

 new designs of instruments were considered and some of the pre- 

 liminary tests of their efficiency appear to be of sufficient impor- 

 tance to warrant publication. 



When a very thin strip of blackened metal, e. g. a bolometer 

 strip, is exposed to radiation it becomes warmed and it in turn 

 emits radiation. In previous investigations of the diffuse reflect- 

 ing power of various substances 1 and of the behavior of an abso- 

 lute thermopile 2 it was found that this warming of the receiver 

 is quite appreciable, and that this receiver can be a very efficient 

 secondary source of radiation which, in turn, can be used to 

 operate a radiometer. The utilization of this secondary source 

 of radiation can be accomplished by placing the receiver at the 

 center of an accurately ground hollow sphere having an opening 

 to admit radiation. In this case one would utilize the "re- 

 radiation" which has to be very carefully excluded in diffuse 

 reflection measurements. 3 



Another logical method for utilizing this radiation is the em- 

 ployment of multiple receivers, one being placed back of another ; 

 for example, a thermopile receiver back of a bolometer strip, 

 or one bolometer strip (or thermopile receiver) back of another. 

 It is with this method that the present paper is chiefly concerned. 



1 Bull. Bur. Stds., 9: 283. 1913. 



2 Bull. Bur. Stds., 11: 157. 1914. 



3 Paschen (Ber. Berliner Akad., p. 409. 1899) appears to have been the first 

 one to use a hemispherical mirror in front of a bolometer in order to "blacken" 

 it. The device has been used extensively by the writer (Bull. Bur. Stds., 4: 

 392. 1908). In spectral radiation work care must be exercised to avoid reflec- 

 tion of radiation from the adjacent parts of the spectrum upon the bolometer 

 strip. In investigations where it is unimportant whether some of the incident 

 beam of radiation falls upon a reflecting surface at the rear of the receiver before 

 it falls upon the receiver (i.e. in cases where it is unimportant whether the com- 

 plete beam of incident radiation is completely intercepted by the receiver) it 

 is possible to place the receiver at the center of an accurately made hollow sphere 

 as just mentioned. Pfttnd (Phys. Rev. 34: 288. 1912) claims a very large 

 increase in sensitivity as the result of using a thermoj unction at the focus of a 

 spherical mirror. 



