€10 Prof. Rankine and Mr. Smith on the Viscosity of 



necessary because early attempts at measurement were un- 

 successful, owing to a chemical action between the mercury 

 and arsine which takes place slowly at atmospheric temper- 

 ature. Even with these precautions we have not succeeded 

 in eliminating this action entirely. From a purely chemical 

 point of view the decomposition is negligible, although the 

 surface tension of the mercury altered profoundly, an 1 traces, 

 apparently of arsenic, could be s^en. Further, when the gas 

 was so'idified after the viscosity measurements had been 

 made, no measurable quantity of hydrogen could be obtained 

 by fractional distillation at — 183° C. The abnormally large 

 surface tension effect, however, particularly at steam tem- 

 perature, made it necessary to curtail the number of readings 

 in order to secure results while the mercury pellet was 

 running fairly smoothly. The result is that the data for 

 arsine eannot be regarded as reliable as for the other oases — 

 not on account of the impurity of the arsine, but because of 

 some residual uncertainty of the capillary correction. It 

 will be noticed how much larger this correction is at steam 

 temperature, presumably because the chemical action is 

 accelerated with rising temperature. There was evidence 

 that when the gas was at constant temperature the capillary 

 correction slowly increased, and in order to allow for this 

 the observations were taken in the following order : — whole 

 pellet, two segments, and then whole pellet again, the mean 

 of the first and last being used in conjunction with the 

 intermediate observations. 

 Results : — 



Table IV. — Arsine. 



The numbers in parentheses indicate the number of observations 

 for which the figure quoted is the mean value. 



Temperature 

 (.leg. C). 



Time of fall in seconds. n .,, 



Capillary 



Corrected 

 Time 



(0. 



108-7 

 139-3 



AT hole pellet. 



iwo segments. 



160 

 1000 



(4) 113 9 

 (4) 150-3 



(2) 119-8 0-047 • 

 (2) 1631 0073 



Time of fall at 16 o, 0. = 108-7 sec, whence time at 

 15°-0 C. = 108-8 sec. 



Th refore . ?? 10 o\ _ 139-3 i. 9S7 



