238 CHARLES S. EIDGWAY 



trum. Dr. Bacon cites several published articles tending to 

 show that solutions of both uranium salts and oxalic acid pro- 

 duce absorption bands in that region of the spectrum charac- 

 terized by short wave lengths, and makes the statements: 

 . . . .it may be considered as being fairly well estab- 

 lished that the active rays from the sun measured by this solu- 

 tion are in the ultra-violet," and further, " . . . , I do not 

 consider that there is any good reason for classifying the sun's 

 rays into infra-red or heat rays, visible rays, and ultra-violet 

 or chemical rays, as there are just as many chemical reactions 

 affected by light corresponding to the visible and even infra-red 

 parts of the spectrum as there are by the ultra-violet portion." 

 In his Table V, however, the decomposition in a solution exposed 

 in a quartz beaker was somewhat less than that in a similar 

 solution exposed at the same time in a glass flask. Since the 

 reaction does take place in a glass container, it seems probable 

 that rays other than the ultra-violet are active upon the solu- 

 tion. This seems even more certain when it is considered that 

 of the total range of ultra-violet light (3920 to 1000 Angstrom 

 units) only those over 2910 units in length reach the earth and 

 that none less than 3000 units are capable of passing through 

 glass. It should be stated further that the rays over 3000 

 units in length are not bactericidal and, hence, are probably 

 the least active chemically of the ultra-violet portion of the 

 spectrum so far as living organisms are concerned. "^ 



Granting that the chemical photometer shows the chemical 

 effect of light* belonging chiefly to the violet end of the spectrum 

 and that the pyrheliometer records the heating effect of the 

 entire spectrum, with its maximum in the infra-red, it seems, 

 from the comparisons made, that the proportion of the two 

 kinds of rays commonly obtaining in ordinary sunlight is respon- 

 sible for the apparent agreement of the two methods of measure- 

 ment in a majority of the tests. The cases of marked departure 

 of the ratio of the number of calories to the number of grams 

 from the mean of all the observations are thought to be due in 

 most instances to the occurrence of clouds or haze w^hich affected 



•» Ayers, H. S. and Johnson, W. T., The destruction of bacteria in milk by 

 ultra-violet light. Centbl. Bakt. (etc.) Abt. 2 : Bd. 40; 1914, No. 1-8, .pp. 109-131. 



