Gaseous Ammonia, PhospJiine, and Arsine. 



607 



been solidified, the gas v\as considered to be sufficiently dry. 

 This was confirmed by the smooth motion of the mercury 

 pellet in the fall tube. 



Observations were taken with the ammonia, first, of all at 

 atmospheric temperature, then at steam temperature, and 

 finally, after cooling, at atmospheric temperature again, 

 with the following results : — 



Table II.— A 



mmonia. 



Each recorded time is the mean of twelve observations, 

 mutually consistent to 0'4 per cent. 



Temperature 

 (deg. C.). 



Times of fall in seconds. 

 Whole pellet Two segments. 



Capillary 

 Correction 



Corrected 



Time 



(0- 



(a) 15-0 

 (c) 15-0 



70-10 72-91 

 70-22 72-56 



93 32 95-93 



00329 

 0-0308 



Mean 



00266 



68-09 

 68-05 





68-07 



(h) 100-0 



90 83 



This gives the ratio of times at 100° (J. and 15° 0. as 

 1*334, which corresponds to an abnormally large variation 

 of viscosity with temperature, the actual value of Sutherland's 

 Constant indicated being about 520. In addition the derived 

 value of the viscosity of ammonia at 0° 0, is noticeably 

 lower than those obtained in modern measurements by 

 Thomsen* and Vogel f. We believe this to be due to a 

 property of the ammonia which renders this apparatus 

 unreliable for viscosity measurements with this gas at 

 atmospheric temperature. We have observed that ammonia 

 is absorbed by glass to an extent which may be very con- 

 siderable, and depends upon the pressure and temperature 

 of the gas. The gas content of the glass of the apparatus 

 below the mercury pellet is greater than that above on 

 account of th 3 greater pressure, and as the pellet descends 

 free gas is transported from the space below the pellet to the 

 space above, thus affording a mode of escape for the gas in 

 parallel with the capillary. The time of fall is therefore 

 lower than that which would be obtained if the ammonia 

 had not this property. 



At steam temperature, however, there is reason to believe 



* Thomsen, Ann. der Physik, xxxvi. p. 825. 

 t Vogel, Berlin Dissertation, L914. 



