in Gases to the Contact Potential Series. 395 
fall may be partly made up of that in the gas in a region 
beyond a space covered by the real anode-tall. Since the 
potential gradient in the gas increases with the pressure, an 
observed increase in the anode-fall with the gas-pressure may 
then be attributed to this extra quantity. Such a view was 
strengthened on comparing the potential gradient in hydrogen 
and oxygen with that for nitrogen. Both hydrogen and 
oxygen were found to possess near the anode a gradient only 
about one-third the value of that observed for nitrogen. 
Thus we find the disturbing element with these gases consi- 
derably smaller than it is for nitrogen. 
To obtain as nearly as possible the abrupt fall of potential 
between the anode and the gas, the test-wire W (fig. 1) was 
first adjusted parallel to and in contact with the face of the 
anode, giving a zero potential between them when the current 
was passing. Observing the electrometer, the wire was 
cautiously shifted out of contact, this being indicated by a 
sudden deflexion of the electrometer-needle. The smallest de- 
flexion thus obtained was taken as measuring the anode-fall. 
The observations for the three gases are given in Table I. The 
TasLe [.—Anode-Fall in Volts. Different Gas-pressures. 
ean a=) Pi | Ae. | Au. | Gu. | Fe. | Ni. | Sn.| Pb.| Cd. | Zn. | Al. | 
| press | ) | / | 
2-02 19-8 20-5 22-7 22-8 22-7 23-0)...... 228 237 236 
Hydrogen | 3-95 197) 20:5 22-9 23-9 22:5, 23-3)......| 23:6) 23-7 23-6 
je | ee ee |e es eee eee eee 
} mel ) 
= 1:37 | 18°5| 18-6)...... 19:0) 19-4! 19°4]:.....|...... 119-1) 18:5] 21-9 
se a Saget I 18:8 18°8)......| 19-6] 19-9) 19-9}......]...... | 20-2, 19-5] 22-5 
’ el 2 Ca ks ee een ee | Woke Ft) € e229 Fee Ses S 
Oxygen. | 1:20 | 229......| 243, 23-2) 23°8| 23°5) 249|......| 242 .....| 239 
Rvs ae 259 
| 
2°04 | 2353). QA-2, 23-4} 23°6| 23:6) 25°4)......| 25°6) ...... | - 
metals having been prepared as usual, their anode-fall for 
the lower gas-pressure was observed, then without repolishing 
them it was observed again in the fresh gas at the higher 
pressure. Several different metals were tested for the purpose 
of eliminating from the conclusions the effect of a change 
produced by use. The results with hydrogen show for seven 
metals the same value at the two pressures used; for copper 
and lead there is about 5 per cent. increase in fall for 
50 per cent. increase in gas-pressure, which is undoubtedly 
due to a change in the electrode surface. In oxygen, four 
metals give the same value at the two pressures observed, 
while four show an increase in value of the fail with pressure. 
From the fact that there is generally an increase in electrode 
