242 PROCEEDINGS OF THE AMERICAN ACADEMY 



Standard Candle. 

 This was found one of the hardest lights to manage. It was 

 necessary to snuff the wick continually, otherwise the flame would 

 become too brilliant, besides which too much red would be intro- 

 duced. After a little practice, however, better results were obtained, 

 and when calculated, the curve followed very closely that of the 

 gas-flame. (See Fig. 1.) 



Mean Readings. 



r Y G v 



20.5 17.9 16.5 19.2 



22.5 19.0 19.6 12.6 



21.5 18.4 18.0 15.9 



73 



R 



677 

 691 



14 



Lime Light. 



This was the next flame measured, and has already been referred 

 to. It is very steady and uniform, and comparatively easy to 

 measure. 



Magnesium Light. 



This was obtained by burning two coils of wire simultaneously in a 

 lamp adapted for that purpose. The coils weighed together 56 gins., 

 and burned at the rate of .37 gins, per minute, and would therefore last 

 without renewal for about two hours and a half. Three bright lines 

 were visible in the spectrum, namely - D, b', and a line which would 

 come about half way between b' and F. These lines fortunately did 

 not come into the field of view in either of my measurements, but 

 would be represented on the curves in Fig. 1. by long vertical lines 

 drawn at these points. The light itself had a very curious appear- 

 ance when viewed through colored glass. It was the shape of a 

 broad, inverted candle-flaine, wavering from side to side, and some- 

 times splitting in two for nearly its whole length. There seemed to 



