684 
PHYSICS: O. E. BUCKLEY 
preferably situated between the other two electrodes and of such form 
as not to entirely block the electron current to the anode. A milliam- 
meter is used to measure the current to the anode and a sensitive gal- 
vanometer to measure the current from the collector which is main- 
tained negative with respect to the cathode so as to pick up only the 
positive ions. 
If there were no gas at all in the space between the electrodes a pure 
electron current would flow from cathode to anode and no current 
would flow to the collector. However, if gas is present positive ions 
are formed by collision in amount proportional to the electron current 
and the number of gas molecules in the space. Since the collector is 
negative with respect to the cathode a certain proportion of the positive 
ions, depending on the form, dimensions, and potentials of the electrodes, 
will flow to the collector. Hence the ratio of the collector current to 
the anode current is proportional to the pressure and may be used to 
measure the pressure when the constant of proportionality has been 
determined. 
This relation has been tested experimentally with air over a pressure 
range from 10"^ to 4 X 10~^ mm. of mercury by comparison with Mc- 
Leod and Knudsen manometers. The actual apparatus used consisted 
of a glass bulb 6 cm. in diameter enclosing three parallel, V shaped fila- 
ments of thin platinum strip, each about 3.5 cm. long, placed 5 mm. 
apart, the collector being between the other two. Leads from both 
ends of each filament were brought through the glass. This arrange- 
ment permits glowing the electrodes to free them from occluded gases. 
An oxide coated filament was used for the cathode. The bulb was 
sealed to a large glass reservoir which was connected to a high vacuum 
pump and either the Knudsen or McLeod manometers. When the 
latter was used a liquid-air trap served to keep the mercury vapor of 
the McLeod manometer out of the ionization manometer. 
Currents from 0.2 to 2.0 milliamperes were used with from 100 to 
250 volts between cathode and anode. The collector was held at 10 
volts negative with respect to the cathode. The resulting current to 
the collector at a pressure of 10"^ mm. was about one-thousandth the 
current to the anode and at lower pressures was proportionately less. 
Hence at a pressure of 10~^ mm. with a current of 2.0 milliamperes 
to the anode a collector current of 2 X 10"^ amperes could be obtained. 
With a sensitive galvanometer much lower pressures could easily be 
measured. 
Experiments with hydrogen and with mercury vapor in place of air 
gave constants of proportionality nearly the same as with air. 
