764 REPORT—1885. 
The dielectric strength of air is ordinarily far below the maximum charge, 
but if it varies with pressure more rapidly than the 2 power, say as p— 
it follows that at some high pressure it will overtake the maximum 
charge and pass above it; and in that case an air condenser at some very 
high pressure would be unbnurstable, i.e. the dielectric strength of suffi- 
ciently high-pressure air would be infinite. 
On the other hand if dielectric strength were to vary say as p}, then 
very low-pressure air would possess this property. 
Experimental evidence tends to show that, over a moderate range, 
dielectric strength varies roughly as p; but that at very low pressures 
this law is largely departed from, dielectric strength rapidly falling 
to a minimum and then apparently increasing again as the exhaustion 
proceeds. 
Cailletet found that air at 50 atmospheres was tremendously strong, 
a millimetre of it being able to resist a powerful induction coil. 
But it is well known that high vacua are apparently very strong too; 
though the latter is open to a doubt suggested by Schuster, who does 
not feel sure that the potential applied to small electrodes is really com- 
municated by them to the gas. 
Another mode of attacking the question, whether an upper limit to 
charge exists, would be by seeking for a diminution in the capacity of a 
condenser when highly charged. It is generally assumed that the capa- 
city of a condenser is a constant quantity ; but perhaps our evidence 
for this is at present insufficient. The electrostatic permeability of a 
medium (1.e. its specific inductive capacity) may be found to diminish 
with high strain, and tend toward zero, in other words, a dielectric may 
become ‘ saturated with electricity,’ just as magnetic permeability tends 
towards zero in a substance very highly magnetised or nearly ‘ saturated 
with magnetism.’ Unless it has been already done, condensers of various 
material (and air is the easiest) ought to be examined over a very great 
range of absolute potential; for a departure from the law of simple pro- 
portion between charge and E.M.F. could not fail to have an important 
signification, whether it be the signification I have hinted at or no. 
In conclusion, I must confess that the latter portion of this communi- 
cation has been perhaps too widely speculative ; and it behoves me to ex- 
plain that only a small portion of the whole paper was spoken by me at 
the meeting of the Association in Aberdeen; and that in ordering it to 
be printed in extenso the Committee have not made themselves respon- 
sible for the contents of the whole. I judged, however, that a little 
latitude would naturally be allowed in the opening of a discussion, and 
have therefore set down all the matters which, had time and convenience 
served, I was prepared to bring forward. The preliminary notes published 
in ‘Nature,’ September 10, 1885, willaccordingly now serve as atable of con- 
tents. I expect my suggestions to receive a severe and salutary criticism, 
but trust that my admitted ignorance of much that has been done may be 
pardoned, for it is to be remembered that I was not called upon to draw up 
a conclusive Report, but to furnish food for reflection and discussion. 
