204 PHENOMENA DEPENDENT ON MOLECULAR PATHS 80 



To explain this remarkable phenomenon by the formula, 

 we note that the numerator of the fraction may be put 



PL + &?!*) = vjl - (l - " 

 p p m^J | \ M!/ p 



= l_Vl_*V)^_. . = 1-0-48?'--. 

 m/p p 



in the case under consideration, where w 2 = 2 andmj = 43-67 

 if p 2 is the pressure of the hydrogen, while the denominator 

 becomes 



fa + p*(ii\*(*\*\* - fi h ^iVr^V^' 

 \p pVV ;J | " L 1 " I 1 Uv W }p 



. 1 _|fl - (aVf^)*!*' - . . = 1 - 0-52?? - ... 



3{ ^J \rn.J } p p 



if the values of the coefficients of viscosity are put in from 

 79. For small values of p z , therefore, the numerator 

 diminishes less than the denominator as p 2 increases, and 

 the value of 77 must therefore rise in magnitude, and not fall, 

 with increase of p v so long as this remains small. For 

 larger values of p 2 the relation alters. 1 



We may not, therefore, as this instance shows, without 

 further consideration conclude that, because the viscosity of 

 air is greater than that of carbonic acid, the coefficient of 

 viscosity of carbonic acid is increased by the mixture with 

 it of some atmospheric air. But if we put in the formula 

 w 2 = 28*69 for air (which we assume as a mean value from 

 N 2 = 27-88 and 2 = 31-76), and also 77 2 = 0-000172 for air, 

 and 77, = 0-000145 for carbonic acid, the formula becomes 



TJ = Vl (l - 0-17 p 2 /p -...)/(!- 0-29 p,lp -...), 



and from this it follows that 97 increases with p 2 , and 

 therefore with the amount of added air, even for small 

 values of p 2 , as Warburg and von Babo 2 have actually 

 observed. 3 



1 [The numerator of the fraction is greater in this case than the denomi- 

 nator so long as the ratio p^p does not exceed 0*615. TB.] 



2 Ber. ilber d. Verh. d. naturf. Ges. in Freiburg i. B. 1882, viii. p. 117 ; 

 Wied. Ann. 1882, xvii. p. 422. 



3 [This behaviour is independent of the relative amounts of the two gases ; 



