Effusion of Argon, Helium, and some other Gases. 433 



little closer, for here the calculated time is also ll m 19 s, 6, 

 while the mean observed time is ll ra 17 s, 9. A similar result 

 is obtained by using von Obermaier's * values for the relative 

 viscosities. For the purpose of comparing argon and helium 

 with oxygen it was necessary to reduce the viscosity-cor- 

 rection as much as possible, so no further experiments were 

 carried out with this form of effusion-plug. Microscopic 

 examination showed that the hole produced by the spark was 

 irregular in shape and more of the nature of a series of cracks 

 or fissures than anything else. These are no doubt of ex- 

 cessively minute width and so, even with thin-walled glass 

 bulbs, relatively high viscosity-effects are unavoidable. 



The new effusion-plug was made by taking a circular disk 

 of platinum-foil (about ^ millim. thick) , backing it with a 

 flat piece of agate, and piercing it in the centre by gently 

 pressing a very fine needle against it. The disks so con- 

 structed were then tested by sealing them to the end of a 

 glass tube with Chatterton cement, and observing the rate 

 of 'effusion into the exhausted vessel of a Topler pump, by 

 noting the rate of fall of the mercury in the vertical capillary. 



A suitable disk having been obtained, it was melted off, 

 burnt clean, and soldered with gold to the end of a stout 

 platinum tube, which was finally sealed on to the end of the 

 glass tube w (see figure). The effusion-aperture employed in 

 the first series of measurements was about ^ millim. in 

 diameter. Preliminary experiments showed that the time 

 of effusion for any given gas varied somewhat from day to 

 day and — to a less degree — even during the course of a con- 

 secutive series of measurements. This source of error was 

 eliminated by comparing each gas directly with oxygen, and 

 usually the time of effusion for oxygen was measured imme- 

 diately before and immediately after that of the gas in 

 question, and the mean of these two values taken. 



The first effusion measurements were made with oxygen 

 and hydrogen to enable the viscosity-correction to be applied. 

 The results were as follows : — 



Oxygen 10 m 54 s - 7 

 Hydrogen 2 m 45 s . 



I0 m 54 s -8, 10 m 54 s -8. 



The time for hydrogen calculated from the densities alone 

 is 2 m 44 s, 4. The apparatus is therefore practically free from 

 viscosity and no correction is required. 



The apparatus was left filled with hydrogen overnight, 

 whereupon measurements of the effusion rate of hydrogen 

 * Wiener Berichte, Bd. lxxiii. ; Carl's Repertorium, Bd. xiii. 



Phil Mag. S. 5. Vol. 49. No. 300. May 1900. 2 H 



