764 Prof. H. Rubens and Mr. H. Hollnagel on 



Arrangement of Apparatus. 



Fig. 2 indicates schematically the arrangement of the 

 apparatus. A is a Welsbach burner, without the glass 



Fiff. 2. 



chimney, which served as a source of radiation ; B is a 

 concave mirror which projected an image of the mantle on 

 the central portion of the air-layer C. Directly to the left 

 of C is a shutter D, conveniently operated from the ob- 

 server's position ; on raising the shutter the rays enter the 

 enclosure K, which surrounds the four crystalline reflecting 

 surfaces F^ F 4 , another concave mirror, G, and a sensitive 

 radiomicrometer, H. By means of this concave mirror an 

 image of the Welsbach mantle could be focussed on the 

 thermo-element of the radiomicrometer. The instrument 

 was provided with a hollow, circular cone, highly polished 

 on its inner surface and extending to the immediate sur- 

 roundings of the thermo-element. This effected an increased 

 concentration of the energy in the direct proximity of one 

 thermo-junction. The inner aperture of the cone was sealed 

 with a lamina of mica, 1*3 /uu in thickness. The radio- 

 micrometer thus constructed gave a deflexion of 100 mm. 

 for a candle six metres distant with a scale distance of 

 3 metres and a deflexion period of 10 seconds. Under favour- 

 able conditions, especially in clear weather and no wind, and 

 a constant room temperature, the error in a single observation 

 rarely exceeded 0'2 mm., and by an accumulation of observa- 

 tions smaller deflexions of 2-3 mm. could be accurately 

 observed to at least a few per cent. The various series of 

 observations described later were recorded under such con- 

 ditions. When, however, the weather conditions were un- 

 favourable, and especially when the wind was high, the 

 zero-point of the instrument varied most irregularly, the 



