89 VISCOSITY OF GASES 233 



in those of Schumann 1 on benzol and the esters in 

 most of his experiments at least. Schumann, however, 

 mentions that with methyl formate a slight increase of the 

 coefficient of friction, amounting to about 1 per cent., was 

 noted when the pressure at 21*8 was raised from 37*47 to 

 46-37 cm. of mercury, or nearly to the pressure 50 -1 cm. 

 corresponding to the point of saturation. Perhaps in this 

 small variability we may see a confirmation of our theoretical 

 formula. 



For a sure resolution of the question whether the 

 formula correctly represents the actual behaviour of vapours 

 we may, with a better prospect of success, employ in our 

 calculations the observations, already mentioned in 77, 

 which Warburg and von Babo 2 have made on the 

 friction of carbonic acid under high pressures. For these 

 experiments confirm the theoretical formula in that they 

 both agree in showing that the viscosity depends on the 

 pressure only in so far as it varies with the density. 3 Prob- 

 ably, too, the observations of Houdaille 4 on the varia- 

 bility of the viscosity of water-vapour may help to good 

 results. 



90. Comparison of the Theory with the Behaviour 

 of Carbonic Acid 



In order to submit the numerical results of the observa- 

 tions to the theoretical calculation I have put the formula 

 for the coefficient of viscosity 77 in the simpler form 



by making the following substitutions, for the sake of short- 



1 Wied. Ann. 1884, xxiii. p. 394. 



2 Ber. ub. d. Verh. d. naturf. Ges. zu Freiburg i. B. 1882, viii. ; Wied. Ann. 

 1882, xvii. p. 390 ; abstract in Berl. Sitzungsber. 1882, p. 509. 



3 Warburg and von Babo, Results, 12, i. 3. Hence the viscosity 

 seems to be connected with the density much more simply than with the 

 pressure. 



1 Fortschr. d. Phys. 1896, 52. Jahrg. I. p. 442. 



