W. Hallock — Very on Atmospheric Radiation. 231 



placed at its center of curvature. The coucave mirror was of 

 slightly larger angular aperture than the bolometer case and was 

 kept at a constant temperature. The difficulties of manipulation 

 and the comparatively small effect due to the differences of tem- 

 perature of the air rendered this method unsatisfactory. 



I a I 



1_* 



i — r 



j i i 



14* 



•B' 



Fig. 1, diagram of "Method A." M is a concave spherical mirror, concentric 

 with the bolometer strips at B. S S is the last screen in the opening of the 

 bolometer case. C C G is one of the two air tubes interposed between the 

 mirror and the bolometer. 



Considerable valuable data was obtained by " Method B," 

 where a vertical current of air of varying rectangular cross sec- 

 tion and temperature was made to rise in front of the bolometer. 

 The air current always subtended a greater angle than the aper- 

 ture of the bolometer case. This method is similar to that used 

 by Hutchins,* whose conclusion that " radiation only takes place 

 when there is a fall of temperature within the limits of molecular 

 action " appears to Prof. Very not altogether warranted. In this 

 method the investigator is confronted with the question as to 

 what extent a part of the radiating mass of air absorbs the 

 radiations from another part. There can be no doubt that this 

 takes place and that nothing is gained by increasing the thick- 

 ness of the mass beyond a certain limit. It must not be forgot- 

 ten, moreover, that this absorption is selective and does not affect 

 all wave lengths equally (cf. fig. 4). 



3 



i 



n n 



/ft. 



Fig. 2, diagram of " Method C." C C is the air-tight cylinder. S is the disk 

 which can be moved near to, or far from, the rock-salt window W, by means of 

 the rod R, which runs through a stuffing box. The bolometer is located at B. 



Perhaps the most satisfactory contributions were made by 

 "Method C," fig. 2. In this case an air-tight cylinder, five feet 

 long and one foot in diameter, was provided with a rock-salt 

 window at one end, and a stuffing box at the other. Through 

 the latter ran a rod carrying a disk on its inner end which nearly 



* This Journal, vol. xliii, p. 35*, May, 1892. 



