582 PROCEEDINGS OP THE AMERICAN ACADEMY. 



moments due to pressures for the silver and glass sides of the vanes re- 

 spectively toward the light. The next two columns contain these mo- 

 ments corrected for a period of 24 seconds of the torsion balance. The 



P v C P v C 



columns headed and — — are the corresponding forces re- 

 duced to standard sensitiveness, G = 1000. The final column contains 

 the averages of the two columns which precede it. Table IV exhibits 

 corresponding data for " red glass " and " water cell." The air pressure, 

 period of the balance, lever arm and galvanometer constants are those 

 given in Table III for the same date. 



In these ballistic measurements the lamp reading was the throw due 

 to an exposure of the light upon the bolometer for six seconds, but in 

 the energy measurements the lamp reading was a stationary deflection 

 due to prolonged exposure. To bring the pressure values into compari- 

 son with the energy measurements it is necessary to reduce the average 

 of the quantities in the last column to pressures in dynes by multiplying 

 by 0.363 X 10 -6 , the torsion coefficient of the quartz fibre, and to re- 

 duce not only to a static deflection of the torsion balance but also to a 

 static deflection of the lamp galvanometer G 2 . The ratio of a ballistic 

 to a static deflection of the galvanometer G 2 was obtained from a long 

 series of lamp exposures. This ratio was found " through air " to be 

 = 1.55 ; " through red glass " = 1.535 ; " through water cell " = 1.502. 

 These differences are probably due not solely to the damping constant of 

 the galvanometer but to the peculiar manner in which the bolometer 

 was warmed up to its stationary conditions by the beam from the lamp. 

 Applying these reduction factors to the averages in Tables III and IV, 

 we obtain the following results. The pressure of the standard light 

 beam which has passed 



1 55 



(a) through air = 16.91 X -^= X 0.363 X 10~ 5 = 



(7.01 ± 0.023) x 10- 5 dynes; 



1 5^5 



(b) through red glass = 16.91 X — ^ X 0.363 X 10~ 6 = 



(6.94 ± 0.024) X 10~ 5 dynes; 



1 50 9 



(c) through water cell = 16.20 X y^_ X 0.363 X lO" 5 = 



(6.52 ± 0.028) X 10~ 6 dynes. 



