measured by means of a photometer. Then from the law of in- 



* 



verse squares, viz., that the intensity of li<*ht per unit 

 surface varies inversely as the aquare of t^ ft. distance from -B 

 the source , the intensity at any distance from the arc can 

 be computed. 



r 



The calibration oT~nTe\screens was repeated seven times 

 or until satisfactory checks of the distances were obtained. 



some thermopiles of less rapid action than the one used 

 here, it ^/ in impossible to get results oy waiting for 

 the galvanometer indicator to come to a steady fftate.In such 

 a case the deflections produced by exposure to the ligh^'""' 

 for equal intervals of time can be compared. A series of 

 measurements fom five second exposures agreed very well 

 with those obtained by the other mejjthod, 



TABLE SITC'VINr, 



LIGHT FILTERS^) 



?ROM AN 



TKAlOaTT2DARB EQUAL. 



Filter No. Wave- lengths Color 



70 6600-^-7000 A.U. Red 



71 6200-6800 A.U. Red 



Distance from light 

 320 cm. 

 275 cm. 



;e the same" effect it migTTE oe tne resort or a general inten- 

 iy gradiant; if only the blue ligh<; is effective, the problem 

 one of a chemical effect; if only the red it might be assumed 



