
Principles of Aerodynamics. 677 
By similarity ( a=B=m=h=1Ls b= =) it is evident that 
this law, if valid for a certain tube, can be applied to a tube 
n times wider and longer only if the pressures be diminished 
in the same proportion. 
Then velocity will remain unchanged, the outflowing 
volume will be increased n? times. This last result, how- 
ever, is not limited to capillary tubes, not even to stationary 
flow; it can be applied just as well, for example, to effusion 
of gas from a closed reservoir through a small aperture. 
This law of Poiseuille is the starting-point for the usual 
method of measuring viscosity of gases. Butas its theoretical 
deduction implies neglect of inertia terms, of longitudinal 
friction, of thermic effects, forming serious obstacles to its 
accuracy and wider applicability, it may be worth drawing 
attention to the fact that relative measures, nevertheless, will 
give quite accurate results if effected in a suitable way. 
For if we use pressures not such as we like, but such as 
are proportional to the value of the ratio - i Me for different 
gases, the motions will be similar (by h=n=1; m=—_; 
a 
— ue _), and the ratio of effective difference of pressure to 
a 
the outflowing volume —_ El 
viscosity. it | 
Again, the validity of this result does not depend on 
the use of capillary tubes; it holds good, just as well, for the 
opposite case, of effusion through a hole in a thin wall; but 
evidently the use of capillary tubes is more convenient, since 
errors arising from incorrect application of pressures are of 
no importance there. 
Likewise exact measurements of the thermic variability of 

will be the exact measure of 
easy So 
viscosity may be performed by observing the value —, if 
such pressures 7, p2 be chosen at different temperatures as 
are proportional to the (e+ 4) power of temperature 
[by ae=B=n=1; m=/h; b= hers). 
Maxwell’s method of oscillating disks could be improved in 
a similar but somewhat more complicated manner. It would 
be interesting to take these researches up again with regard 
to those improvements, as investigations on thermic varia- 
bility of viscosity have made no decisive progress since the 
time when Schumann (Wied. Ann. xxill. p. 353, 1884) 
