590 Prof. Kia-Lok Yen on the Coefficients of 



connected in series and filled with charcoal of coconut-shell, 

 which were immersed in liquid air. 



Having passed through the purifying agents but once, 

 the gas was found to be spectroscopically pure upon the 

 examination of the spectra produced by discharge in the 

 tubes V 1 and V 2 , and thus further purification was proved 

 unnecessary. 



Before the gas was introduced into the apparatus the 

 stopcock between the chamber and the purifying tubes was 

 opened, and with the stopcocks S 2 and S 9 closed, the whole 

 apparatus was evacuated until no discharge could pass between 

 the terminals of the tubes Y 1 and V 2 . While thus being 

 evacuated, the tubes I\ and T 2 were heated by electric heaters 

 up to about 500 degrees centigrade in order that the charcoal 

 might give up all of its gas content. 



To render the pressure gradient in the apparatus as small 

 as possible while it was being evacuated, shunt passages were 

 provided from various parts of the apparatus to the outlet to 

 the pump. These passages were provided with stopcocks S 2 , 

 S 3 , S 4 , and S 5 , which closed them up when the process of 

 purification was taking place. 



The apparatus was partly rilled and then evacuated, and 

 then refilled and re-evacuated three or four times before the 

 chamber was finally filled up to the desired pressure for 

 observation. 



A compression-and-suction apparatus ('was provided in 

 order to fill the chamber up to a pressure higher than that 

 in the generating and purifying part of the apparatus. With 

 this device it was possible to make all the observations under 

 a pressure of exactly 76 cm. of mercury. 



Table I. contains the results of all the individual obser- 

 vations. 



Observations 1 to 10 were made with electrolytic hydrogen, 

 while the rest were made with hvdrooen general ed from zinc 

 and hydrochloric acid. Observations 11 to 11 were made 

 with one content of the chamber, while the rest were made 

 with another. As may be seen from the table, the greatest 

 difference between individual observations and the mean 

 value is not more than 0*25 of one per cent. 



This value of 77 for hydrogen agrees with that found by 

 Professor Millikan some years ago from the fall of droplets 

 in hydrogen ; taking the value for e from the fall of droplets 

 in air as the basis of calculation. The value thus found was 

 881-0 xl0~ 7 *. 



* Phys. Review, n. s. vol. viii. No. 6, Dec. 1916, p. 618. 



