1904— o.] Mr J. Fraser on Electricity based on Bubble Atom. 713 
figments of the imagination,” they ought to answer to this test ; 
and I was disappointed that they did not answer more fully to it. 
It ultimately became clear to me that there was some factor in the 
problem of which I took no account ; and that this factor, what- 
ever it was, acted equally on silver, copper and aluminium ; and 
also equally on zinc and platinum, though differently from its 
action on the three former ; and also equally on lead and antimony, 
though differently on these from all the others. After casting 
about in every direction I could think of for some feasible solution 
of the problem, the most feasible — and, indeed, one which ought to 
be accounted for if this theory is not the mere ravings of a dis- 
ordered imagination — is self-compression, as I have termed it. It 
cannot be supposed that in solid bodies, constituted of atoms such 
as are postulated in the theory under discussion, that the ether 
should have the same freedom of circulation through them all. 
It must be that in some of them, according to the sizes and arrange- 
ment of the molecules, the ether must be absolutely trapped in a 
corner or pocket, so that the pressure of the vibrating molecules on 
it, and its reaction on them, must produce a speed in the constituents 
of the atoms which would prevent their acceptance of new motion, 
in the way already set forth (see par. 33), and could only find relief 
in the direction of least resistance — namely, outwardly, or by 
radiation of heat. 
38. I append a list of the resistances found by the above rule, 
along with the observed resistances for comparison; also the 
melting points * of the different metals used in the calculations, and 
the numbers used as multipliers to bring the calculated resistances 
up to that observed. It will be seen that the latter are not mere 
empirical numbers, for they hold good in their proportions 
throughout the calculation. They are empirical in so far as that 
they are the nearest ivhole numbers required to bring the calculated 
up to the observed resistances, but when once established they 
hold their ground. For instance, the resistance found for gold 
when paired with aluminium, silver or copper, had to be multiplied 
by 2 to bring it up to the observed resistance, as already said, 
seeming to show that some factor acted with twice the influence on 
gold as on the three former metals. Similarly the resistance found 
* 273° to be added for reduction to absolute temperature. 
