43 



by means of a photometer, whose peculiar construction can only be 

 understood by a long description, the relation between the visible 

 illuminating power of the same portion of zenith sky and that of 

 the total heavens was determined. As, in the case of lights from 

 the same source but of different degrees of intensity, the chemical 

 actions are proportional to the visible illuminating effects, it was 

 only necessary, in order to obtain the chemical action produced by 

 the total diffuse light, to multiply the chemical action of the zenith 

 portion of sky by the number representing the relation between the 

 visible illumination of the total sky and that of the same zenith 

 portion. 



The laws according to which the chemical rays are dispersed by 

 the atmosphere can only be ascertained from experiments made 

 when the sky is perfectly cloudless. In the determinations made 

 with this specially-arranged photometer, care was therefore taken 

 that the slightest trace of cloud or mist was absent, and the relation 

 between the visible illuminating effect of a portion of sky at the 

 zenith and that of the whole visible heavens, was determined for 

 every half-hour from sunrise to sunset ; the observations being made 

 at the summit of a hill near Heidelberg, from which the horizon was 

 perfectly free. 



The amount of chemical illumination which a point on the earth's 

 surface receives from the whole heavens, depends on the height of 

 the sun above the horizon and on the transparency of the atmo- 

 sphere. If the atmospheric transparency undergoes much change 

 when the sky is cloudless, a long series of experiments must be made 

 before the true relations of atmospheric extinction of the chemical 

 rays can be arrived at. The authors believe, however, founding 

 their opinion on the statement of Seidel in his classical research on 

 the luminosity of the fixed stars, that the alterations in the air's 

 transparency with a cloudless sky are very slight ; and they there- 

 fore think themselves justified in considering the chemical illumi- 

 nation of the earth's surface, on cloudless days, to be represented 

 simply as a function of the sun's zenith distance. Although, from the 

 comparatively small number of experiments which have been made, 

 owing to the difficulty of securing perfectly cloudless weather, the 

 constants contained in the formulae cannot lay claim to any very 

 great degree of accuracy, the authors believe that the numbers ob- 



