pritchard's wedge photometer. 



321 



(c) may be considered in the case of a homogeneous ray as the constant of trans- 

 mission for a unit thickness of the absorbing material. As the writer has elsewhere 

 shown, it is not only not a constant for non-homogeneous rays, but for these latter 

 (though still reckoned for a unit thickness) it increases as the thickness from which 

 it is determined increases. These variations are shown in Fig. 6. 



% 



60 



50 



40 



30 



20 



10 







V- 



0.3 



0.4 



0.5 



0.6 



0.7 0.8 



Fig. 6. 



0.9 



1.0 



In these prismatic measures the light is nearly homogeneous, and here it will 

 be seen, by considering the ratios in the horizontal line corresponding to any 

 single wave-length, that the ratios are constant for each ray, except for the rela- 

 tively considerable errors of observation where the heat is feeble, and for the 

 variations due to fluctuations in the original solar radiation, which may be large 

 even with the clearest sky ; * but as we go from one ray to another the ratios differ, 

 and this difference grows very marked as we approach the red end at 0^.7. 



* That the irregularities due to variations in the solar beam itself are relatively very much greater than those 

 proceeding from purdy instrumental causes, may also be inferred from the probable errors attached to the mean 

 deflectlonr of Tables I. and II ; since, if these probable errors are expressed as percentages of the correspondmg 

 deflections, it will be seen that they are but slightly greater for the small deflections than they are for the large ones. 



