BELL. — ULTRAVIOLET COMPONENT IN ARTIFICIAL LIGHT. 29 



Snow is a good reflector of ultra violet radiations, at least throughout 

 the limits of the solar spectrum. At two meters distance a square 

 meter of snow surface may reflect to the eye as much as 10 ■* ergs 

 per second per square cm. If even one tenth of this is in the ultra 

 violet then a square meter of snow in the field of vision at two meters 

 distance would deliver about 1000 ergs per second of ultra violet per 

 square cm., which is in excess of the greatest amount which would be 

 given at this distance by any of the artificial sources of light here 

 investigated. 



Fortunately the sun is weak in the extreme ultra violet, but the 

 very large amount of radiation which can be reflected to the eye from 

 a snow covered surface is quite sufficient to account for all the phe- 

 nomena observed, even although the ultra violet per foot-candle in 

 the sunbeam is rather exceptionally low. 



Two sources of light, not here measured for reasons already stated, 

 should not be forgotten. One of these is the iron arc used for thera- 

 peutic purposes, of which the spectrum is shown along side of the 

 mercury spectrum in Plate, 2, g. It will be observed that it is enor- 

 mously rich in lines, even to the extreme ultra violet, and as the light 

 giving power between iron terminals is not high, this source would 

 stand very near the bottom of Table III. The yellow calcium fluoride 

 arc, of which the spectrum is similarly shown in Plate 2, i, would un- 

 questionably stand near the quartz arc at the head of the list, owing 

 both to its very high luminous efficiency and to the comparatively 

 weak lines in the extreme ultra violet. 



In conclusion it ma}' be confidently stated that no commercial 

 illuminant radiates for any ordinary working value of the illumination 

 enough ultra violet energy to be at all harmful, provided one exercises 

 ordinary discretion is keeping unpleasantly bright visible light out 

 of the eves. 



