454) TRANSACTIONS OF SECTION A. 
very different. A comparative experiment was made in which two rods of glass 
were laid from earth to the knobs of two electroscopes, passed up for the purpose 
into a vessel in which the percentage of moisture could be changed at will. One 
of these rods was dried by heating, while the other was not. As the percentage 
of moisture was increased the air-dried rod was observed gradually to acquire con- 
siderable conducting powers, but the other rod did not conduct until the critical 
pressure was reached, when it did so relatively suddenly. The critical pressure 
for glass is roughly 60 per cent. saturation. 
In the case of shellac the critical pressure was found to be above 90 per cent. 
A simple experiment illustrating the two modes of condensation was exhibited. 
An electroscope with shellac insulation was covered by a bell jar which had been 
previously held for a moment over a Bunsen flame to acquire a damp atmosphere. 
No discharge occurred. The experiment was now repeated, but the shellac was 
first moistened and then dried with a cloth, so as to again insulate. On the elec- 
troscope being covered by the bell jar immediate discharge took place, even though 
in the meantime the air in the jar must have somewhat dried. This was the b or 
conducting type of condensation, On now melting the surface of the shellac so as 
completely to dry it, the surface was found to have reverted to the condition in 
which the @ type of condensation is deposited. 
6. On a Theoretical Method of attempting to detect Relative Motion between 
the Ether and the Earth. By A. O. Rankine. 
The explanation which has been suggested to account for the negative results 
in experiments so far made is that the distance between two points in a body is 
shorter when the line joining them is parallel to the drift than when the line is 
perpendicular to it, and this by just the necessary amount to produce exact 
compensation. If 6/ is the increase in a length / in changing from a direction 
parallel to a direction perpendicular to the drift, then for exact compensation 
Z = B*, where B is the ratio of the velocity of the drift to the velocity of light. 
The method about to be described is based on the assumption that this change 
does exist. 
Imagine two particles, each of mass m, one at each end of a massless bar of 
length 27, Let it be suspended in a horizontal position by a vertical wire through 
its mid-point. The moment of inertia of the system about a vertical axis through 
the mid-point is 2m*, and the time period of an oscillation is given by 
T? = Km", where K is a constant depending only on the elastic properties of the 
wire. Let this time period be determined for a small oscillation, the equilibrium 
position of the length of the bar being parallel to the drift of the ether which is 
supposed horizontal. ; 
Now attach the bar to the wire perpendicularly to its original direction, The 
wire itself will be in precisely the same position relative to the drift as before, 
but the bar is now perpendicular to the drift instead of parallel to it. We there- 
fore expect an increase in length, and, supposing no other factor to change, a 
corresponding increase in the time period. From the above equation 
T6T = Kmrér 
whence 
Unfortunately this calculated change of time period is far too small to be detected 
practically, but it is interesting to regard the experiment from another point of 
view. It has been aflirmed that no experiment is devisable which will yield a 
positive result owing to the entry of a compensating factor. If this be true, 
then no change of time period would take place in the above experiment, and it 
becomes necessary to seek for the compensating factor. The only possibilities 
are that K or m, or both, change. Of these K is excluded because, as we have 
