698 Mr. J. H. Jeans on 
§ 9. Collected together for the sake of reference the 
numbers stand as follows :— 
Values of « x 108. 
: } 
Gas. Viscosity. bre ee, Diffusion. | ga ai _ Mean. 
eeseeale mene doen (PA TT see 2) aes eae nen Ga 
Hydrogen’. scree « 205) oi 1:99 | 203 | 2°05 | 2:03 
) Selina 2f<ecesee.2 18h. 4 oe ee . he 1-81 
| Water-vapour ... 339 | re | ne | = 3°39 
| Carbon monoxide 2-90 2-74. 2°92 2°86 
|, Eithylene J........ 3°77 3°88 ; 381 
Nitrogen ......... 2-90) 2-74 3:12 1 ae 
TATE ciesste tenet reece 2°86 2°72 2-90 +84 
Nitric oxide ...... 282 2°81... 2°82 
Oxysen (bw 2°81 2°58 & if 6 ae a 2°73 
Aroon. “ip sae nave 2°79 ue ce an | yoo 
Carbon dioxide... 3°47 3°58 S27 | ee 3°36 
Nitrous oxide ... 354 | 348 | : 3°52 
Ethyl! chloride ... 4-68 be ) ms | ee 4-68 
Chlorine ........, 40) fe | ae 2 oe 
In calculating the mean values of the last column, I have 
assigned double w eight to viscosity determinations, as these 
are probably more reliable than the rest. 
As regards the magnitude of error to be expected in the 
results, some indication is given by the amount of difference 
between the different determinations of ¢ for the same gas. 
It will be seen that there is no entry in which the error 
differs by as much as 10 per cent. from the mean value given 
in the last column, except in the case of one entry for carbon 
dioxide, 
It would, of course, be possible to arrive at this result in 
any table, by an appropriate choice of the calculations 
tabulated. It may, therefore, be well to mention that every 
calculation which I have started has been carried through, 
and the result tabulated, with one single exception to be 
explained later. The particular gases worked at have been 
selected partly on account of their importance, and partly on 
account of the amount of experimental evidence available. 
The single gas for which the results have not been tabulated 
is mercury-vapour. I found that viscosity experiments led 
to a value c=4'8 x 10-*; that this result differed widely from 
that obtained from the senda tian of heat; and finally that it 
was directly contradicted by calculations from the density of 
the solid. The density of solid mercury is given by Mendeleef 
