MR. H. TOMLINSON ON THE COEFFICIENT OF VISCOSITY OF AIR, 
785 
Adopting, therefore, formula (12), we have the following equation for determining 
the value of /x at any temperature :— 
[x t = •00017155(14-'0027 5 U —•00000084^).(13) 
The differences between the observed and calculated values of ^ for the five 
different sets of experiments are given below :— 
Experiment. 
Observed value of jx t . 
Calculated value of fi t . 
Difference. 
I. 
■00017900 
•00017760 
+•00000140 
II. 
•00017680 
•00017850 
-•00000170 
III. 
•00017767 
•00017704 
+ •00000063 
IV. 
•00017581 
•00017653 
-■00000072 
V. 
•00017626 
•00017622 
+ -00000004 
The probable error is about ‘2 per cent., and, considering the manner in which the 
five sets of experiments varied as regards their conditions, it would seem that, even 
when all allowance has been made for aqueous vapour, &c., the number ‘00017155 
must represent the value of /x (J for dry air within at least ^ per cent. Now, this 
number agrees fairly with the values of /x 0 obtained by other observers with the 
transpiration method ; it is, however, more than 9 per cent, less than that obtained 
by Meyer with oscillating plates, and by Maxwell. The mathematical difficulties 
attending Professor Meyer’s method of oscillating plates have been already mentioned, 
but the method of Professor Maxwell does not seem open to these objections, and 
indeed appeared to me to be so good that I for some time attempted, though in vain, 
to account for the difference between Maxwell’s result and my own. Professor G. 
G. Stokes has, however, kindly interested himself in the matter, and has shown in 
the accompanying note the possibility of Maxwell’s result being too high. I may 
perhaps be allowed to add that, if we only take the first two of the five sets of 
Maxwell’s experiments, in which two the distances of the fixed from the oscillating 
plates are so great as to render any error such as suggested by Professor Stokes 
very small, we obtain a value for the coefficient which is nearly identical with that 
obtained by myself. 
5 H 
MDCCCLXXXVI. 
