252 
Mr Wilson, On the Hall Effect in Gases 
gauge besides a bulb containing P 2 0 5 and a three-way tap which 
enabled fresh air to be admitted when required. The air was 
kept in a CaCl 2 tube for a long time before admitting it. The 
whole apparatus was perfectly air-tight, no leak being detectable 
by the gauge even in a whole week, although a change of pressure 
of 001 mm. could be detected easily. 
The discharge between the electrodes A and B was produced 
by means of a battery of small secondary cells, of which one 
thousand were available when required. The circuit included an 
Ayrton-Mather galvanometer and an adjustable liquid resistance, 
besides a telephone to detect oscillations in the current. The 
P. D. being used was measured by means of a Braun’s electro- 
static voltmeter reading from 100 to 1500 volts. 
The magnetic field for producing the Hall effect was produced 
by means of a circular coil of 143 turns and 37 '4 cms. outside 
diameter, which was placed so that the axis of A B coincided with 
a diameter of the coil, the plane of the coil being horizontal. 
The axis of the tube K was then perpendicular to the plane of 
the coil, and the centre of the coil midway between the small 
electrodes E and E'. The current sent through the coil was 
supplied by six large secondary cells, and measured by a Weston 
ammeter. Its value could be changed by means of an adjustable 
resistance. The field due to this coil was vertical and nearly 
uniform near the small electrodes EE'. One ampere sent through 
the coil gave a field at EE' of 5‘20 c.G.S. 
When measuring the Hall effect the electrodes EE' were first 
turned round until the electrometer indicated that they were at 
the same potential. A magnetic field was then applied, when a 
deflection of the electrometer was obtained which was reversed 
on reversing the direction of the field ; its direction showed that 
the negative ions have the greater velocity. 
The present paper only deals with results obtained when E 
and E' were immersed in an apparently uniform positive column 
produced by a sensibly steady current. 
If the positive column is striated the Hall effect depends on 
the position of the electrodes with reference to the striae, and if 
the current is not steady (as indicated by the telephone), the 
results obtained are irregular in character. 
Before making measurements therefore at any pressure the 
current and the number of cells used were adjusted until a 
uniform column and steady current were obtained, or at least 
until as near an approximation to these desired conditions as 
possible was obtained. At pressures of several millimetres there 
is usually a considerable range of current for which the discharge 
is steady and uniform, but as the pressure is diminished this 
range of current rapidly diminishes, until at about 0’2 mm. a 
