608 Prof. Rankine and Mr. Smith on the Viscosity of 



that the effect becomes negligible : certainly much gas is 

 driven from the glass by the increase of temperature, 

 numerous visible bubbles appearing between the pellet and 

 the glass. In these circumstances we have refrained from 

 estimating the viscosity of ammonia at atmospheric temper- 

 ature from the results given in Table II., but we think that 

 the calculation is justified at steam temperature. 

 We have at 100° C, 



'ammonia t/U OO d'^QKA 



t. dir ~ 152*56 ~~ l ' 



which, when corrected for the different slipping coefficients 

 of the two gases, becomes 0*5947. 



The viscosity of air at 100° C. is 2*191 X 10~ 4 C.G.S. units, 

 whence the viscosity of ammonia at 100° C. = 1*303 x 10~ 4 

 C.G.S. units. This is the first value obtained for ammonia 

 at 100° C, and cannot therefore be compared with previous 

 work. It may be somewhat low on account of a small 

 persistence of the absorption effect at this temperature, but 

 the error is probably not more than 0*3 per cent. 



Phosphine. 



This gas was made by allowing a concentrated solution of 

 caustic potash to drop on phosphonium iodide. The air was 

 displaced from the apparatus by means of a current of 

 hydrogen. The gas was dried by passing it through a 

 U-tube surrounded by a freezing mixture, and then over 

 tubes containing phosphorus pentoxide. The phosphine was 

 then collected over mercury as in the case of ammonia. It 

 was found that the gas did not inflame when exposed to the 

 atmosphere, — a fact which showed that the gas was free 

 from liquid phosphorus hydride. 



The gas was finally purified by being fractionated at liquid 

 air temperature. In order to avoid complications arising 

 from the fact that the liquid hydride may be formed when 

 the phosphine is being transferred to the exhausted visco- 

 meter, the condenser containing the solidified phosphine was 

 connected direct!}* to the apparatus, and the liquid allowed 

 to evaporate slowly until the viscometer was filled with 

 phosphine at atmospheric pressure, the connecting tap being- 

 closed before evaporation was complete. Since phosphine 

 boils at -85° 0., the P 2 H 4 (B.P. 26° C.) which might be 

 formed and any trace of water would certainly be retained 

 in the condenser. 



To test whether any P 2 H 4 had been formed a sample of 

 the gas was allowed to come into contact with the air after' 



