Research on some Standards of Light. 



501 



Fig. 14. — Area of the aperture in the cover, 4*9 sq. cm. (or 145 per cent, above the 

 normal) ; intensity of the light at the temperature of solidification, 1005 (or 

 005 above the normal). 



% 0-5 



to 



to 



30 40 50 60 70 



Time in Seconds. 



altogether from the furnace, the constant part of the curve vanishes 

 and determinations become impossible. This case is shown in fig. 15. 



The recommendations made with regard to the rate at which the 

 gases are to be burnt, and to the time during which the blow-pipe is 

 to be alight, are only intended as an indication of the conditions 

 under which the observations will be most easily obtained. A varia- 

 tion of 10 or even 20 per cent, in these factors would leave the final 

 results practically unaffected, but a number of the observations would 

 probably have to be discarded according to one or other of the three 

 rules given on page 495. 



Fig. 15. — Showing the effect of removing the cover from the furnace. 



1-0 



"Co 



10 



BO 30 40 50 60 70 80 



Time in Seconds. 



The effect of contaminating the platinum with either silica or carbon is 

 very marked. Carbon forms the best illustration, as there is no 

 difficulty in subsequently getting rid of this impurity. On the 26th of 

 March, after a certain number of normal determinations had been 



