556 Mr. G. Jaffé on the Conductivity of 
“These conditions — give for the rate of elongation 
dat A a : : 
Pp: af oe W, where P=pg sae -) mm being a viscous 
coefficient, p and q elastic coefficients, one for the thick and 
one for the thin band. We thus get for the elongation at 
any moment an expression w=athbe 4. 
This model then gives us an immediaté or primary strain 
followed by a slow movement in time—the secondary strain. 
To represent the final and uniform viscous flow, in order to 
make the model more complete, one could add say a piston, 
working in a cylinder of oil attached below, the weight itself 
being hung from beneath the cylinder. 
By supposing the bands to be able to support compression 
as well as tension, such a model can give, except in the one 
particular, a complete representation not only for the elon- 
gation curve under constant stress, but the curve of change of 
stress with time under constant strain. The one particular, 
as we have seen, is that the curve is an exponential of the 
wrong type. 
From these considerations we are forced to conclude that 
the explanation of the phenomenon must be sought for on 
altogether different lines from those based on a mixed viscous 
and elastic action in the way commonly attempted. 
LIT. On the Conductivity of Gases contained in Small Vessels. 
By GEORGE J Le Trinity College, Cambridge * 
T was shown by C. T. R. Wilson + that the spontaneous 
ionization of 2 ee is approximately proportional to the 
pressure, except at high pressures. At the same time the 
relative ionization of different gases is nearly proportional to 
the density of the gas, except in the case of hydrogen. 
As it seemed of interest to try whether this proportionality 
still holds for a gas of very high specific gravity and com- 
plicated structure, like nickel- carbonyl, I undertook, at the 
suggestion of Prof. J. J. Thomson, to measure the spon- 
taneous ionization of this gas. It was also to be hoped that the 
behaviour of the leak through Ni(CO), at different pressures 
in vessels of various sizes would throw some light on the 
part played by the walls of the vessel in ionising the gas. 
The apparatus used was a gold-leaf electroscope, and 
differed only in minor details from that used and described 
* Communicated by Prof. J. J. Thomson, F.R:S. 
Tt Roy. Soc. Proc. vol. lxviii. p. 155; vol. lxix. p. 277 
