TRANSACTIONS OF SECTION A. 711 
meter to measure the difference of potential between the terminals of the tube ; 
(5) a five-fall Sprengel pump, by means of which the density of the air inside the 
tube could be reduced to any desired extent. The galvanometer was placed on a 
block of paraffin between one terminal of the electric machine and one terminal of 
the glass tube. Its deflections were read by a telescope’ and its sensibility was 
arranged by external magnets, so that one division of ‘deflection corresponded to 
0:3 mikroampere. Our method of experimenting was to keep the ‘density of the 
air constant while we varied the difference of potential between the terminals of 
the tube, and taking simultaneous readings on the voltmeter and‘ on the galvano- 
meter. The electric potential was varied either by varying ‘the speed of rotation 
of the machine or by varying the distance between the needle-point terminals of 
the machine, or by a combination of both. ol 
“We found that at ordinary atmospheric density it’ requires a difference of 
‘potential of between 2,000 and 3,000 volts at the terminals of the tube before the 
galvanometer indicates any current. As the difference of potential is now increased, 
the current through the galvanometer increases at a greater ratio, so that if a curve 
be drawn with differences of potential as absciss and galvanometer readings or 
currents as ordinates, the curve is always concave towards’ the axis of current. 
Through this particular tube the currents at 3,000, 5,000; and 8,000 volts difference 
of potential were 7:2, 17-6, and 63:2 mikroamperes respectively. As the density of 
the air was diminished, the difference of potential necessary to start a current, as 
‘indicated by the galvanometer, gradually diminished also, till, at'a density of about 
zoo Of the ordinary density, a few score volts were suflicient to start a current. 
For the same difference of potential the current increased as the density of the air 
diminished ; or, otherwise, the same current was obtained by smaller differences of 
potential as the density of the air was reduced. Thus a current of about 56 
mikroamperes was obtained by differences of potential of 7,400, 1,090, 700, 370, 
‘405, 570 volts, when the densities of the air were 1 (ordinary density), 0-058, 
0:0095, 0:0007, 0:00006, 0:000024 respectively; or, otherwise, when the air 
‘pressures were 750, 44, 7, 4, 35, =; millimetres of mercury respectively. 
As the air density was still further reduced, the difference of potential necessary 
to start a current increased, and the current for the same difference of potential 
diminished. Thus, when the density of the air was reduced to one five-millionth 
of the density of air at ordinary atmospheric pressure and temperature, differences 
of potential of 3,000, 5,000, and 8,000 volts gave currents of 1:3, 4:4, and 14:6 
mikroamperes respectively. 
\. If a curve be drawn for a constant difference of potential, with air densities as 
abscissze and currents as ordinates, we find the curve rising as the air density is 
diminished to about z,/55 Or zs'q_Of ordinary density; then falling again as the 
density is still further reduced to about a five-millionth of ordinary density. This 
is the lowest density we have experimented with, but we have no reason to doubt 
that at very much lower densities we would still be able to get measurable currents 
through the tube. 
_ We are now experimenting with a tube 15 cms. long and 1} em. diameter, 
having ball terminals of } cm. diameter and about 2 mm. apart. The investigation 
is not complete enough for publishing any results. 
5. The Duration of X-Radiation at each Spark. 
By Frep. T. Trouton, IA., D.Se. 
The object of these experiments was to ascertain how long a Crookes’ tube 
continued at each spark to give out Roéntgen radiation. 
The method adopted was to rotate a metallic-toothed wheel (cut out of sheet 
zinc) interposed between a tube and a sensitive photographic plate. Only one spark 
is allowed to pass by making one brake of the primary of the inductive coil used. 
The departure from sharpness of outline of the image of the moving teeth on 
