782 MR. H. TOMLINSON ON THE COEFFICIENT OF VISCOSITY OF AIR, 
where M' is the mass of air inside, and Q is the real part of the imaginary expression 
2 4 6 
2 . 4(“) 5 + 2 S 6 ( “) 4 + 2 Mt 6 t 8 + * ' ‘ 
11 1 ’ 
1 + 2M ( ma ) 2 + 2Mth W 4+ 2M8 ( mft ) 6+ • • • 
which is of the form 
A + y/—IB 
C+v/-ll) 
and of which the real part is 
AC + BD 
C 3 + D 2 ' 
In the following table will be found the corrections necessary to be made for the 
rotations of the spheres and cylinders about their axes :— 
Cylinders. 
Number of 
experiment. 
Logarithmic 
decrement uncorrected 
for rotation about 
axes. 
Effect of air outside. 
Effect of air inside. 
Corrected logarithmic | 
decrement. 
I. 
■0027411 
•0000313 
•0000011 
•0027087 
II. 
•0012546 
•0000030 
•ooooooo 
•0012516 
III. 
•0016905 
•0000173 
■0000019 
•0016713 
IV. 
•0014638 
•0000150 
•0000017 
•0014471 
Spheres. 
V. 
•0003483 
•0000159 
•0003324 
1 
There is still a further slight correction to make, inasmuch as the mercury of the 
barometer was not at 0° C. when the aneroid was compared with the mercury baro¬ 
meter, whereas the density of the air was calculated on the assumption of the mercury 
being at 0° C. The correction is very slight, but the closeness of agreement ol the 
different experiments justifies us in making it. It will be sufficient for this purpose 
to multiply each value of /r as determined from the above table by (1 + ‘00018£), 
where t is the temperature at which the aneroid was compared with the mercury 
barometer. Applying all the corrections, the final results are as follows :— 
* In making these corrections, an approximate value of /< was used. 
