1918] on Studies on Liquid Films 403 



sliding tube was adjusted so that its open end was kept within a 

 centimetre of the moving film. Small measured samples of the mixed 

 gases behind the films were withdrawn at intervals for analysis 

 through the two-way cock A. 



The maximum rate of diffusion of hydrogen above that of air was 

 found to be 42 c.c. per hour per cm-. When oxygen was used 

 instead of air the value obtained was 50 c.c. of hydrogen per hour 

 per cm.- greater than the oxygen going the other way ; and 1145 c.c. 

 of hydrogen passed out, while 632 c.c. of oxygen went in. When 

 the reverse process was tried 1173 c.c. of hydrogen went in through 

 the black film, while 635 c.c. of oxygen diffused out. The two 

 operations involved the passage of the film 7 cm. along the cylinder 

 in the first case, and the same back again for the reversal. Fig 38 

 gives the graphs of the movement of the film, and shows tbe 

 variation of the enclosed volume, with the time. The rate of 

 relative diffusion at any time is given by the slope of the tangent to 

 volume- time curves. This was found to decrease as the proportion 

 of oxygen under the film increased. 



The contraction of bubbles in air when blown up with from 

 lU per cent, to 30 per cent, of hydrogen was measured in order to 

 deduce the diffusion rates through coloured films. A much lower 

 value was obtained : thus a bubble coloured steely blue to pale amber 

 and containing 16 per cent, of hydrogen showed a rate of contrac- 

 tion of approximately f c.c. per hour per cm^. Measures were also 

 made with half bubbles blown on the roof of a glass chamber in 

 which a proportion of hydrogen was circulated ; and also by com- 

 paring the contractions of similar bubbles when hanging from a 

 nozzle and resting on a ring. This last was done to correct for 

 the distortion of figure produced by the buoyancy of the contained 

 hvdrosfen. 



Hydrogen Bubbles Blowx under Several 

 at3i0spheres. 



Bubbles that have thinned to blackness contract at an ever- 

 increasing rate by the continual transference of the enclosed gas 

 through the envelope. It was mentioned in last year's Discourse 

 that this contraction is greatly retarded when the bubbles are formed 

 in an enclosure in which the pressure has been raised, the excess- 

 pressure distending the bubble becoming then of relatively smaller 

 proportion to the total pressure. Further experiments have shown 

 that a black bubble in hydrogen under G'3 atmospheres pressure 

 took 100 days for its diameter to contract from 7 "7 to 4' cm., 

 whereas only 7 to 10 days is needed for the same contraction to take 

 place in a similar bubble at atmospheric pressure. When air was 

 used instead of hydrogen the difference was equally marked ; thus 



