in Gases to the Contact Potential Series. 391 
several days in connexion with a drying-chamber containing 
phosphorus pentoxide. Pressure readings taken from day to 
day during this time indicated the rate at which occluded 
gas was given off. In these flasks mercury valves were used, 
so that while stored the gas came in contact with only glass 
and mercury. Observations made with freshly prepared 
hydrogen and that which had been stored for weeks were at no 
greater variance than those obtained from different tests with 
the stored gas. 
Hydrogen was generated by electrolysis of dilute phos- 
phoric acid ; it was purified by passing through a solution of 
potassium permanganate, a tube containing pieces of caustic 
potash, another filled with phosphorus pentoxide, and finally 
through a chamber containing pure sodium to the storage- 
chamber. 
Nitrogen was obtained by passing air over heated phos- 
phorus and through the same series as hydrogen. 
Oxygen was obtained by heating a mixture of potassium 
chlorate and manganese dioxide. It was purified by passing 
through a solution of caustic potash, a tube containing 
sticks of the same, and finally a drying-tube of phosphorus 
pentoxide, 
Gas-pressures were measured by means of a Macleod 
gauge. ‘The electric current was furnished by a battery of 
storage-cells and measured by a Weston milliammeter. <A 
Kelvin quadrant-electrometer was used for the potential-mea- 
surements—the anode-fall being placed between the quadrants 
and a definite potential on the needle; the cathode-fall being 
placed on the needle and a definite potential between the 
quadrants. The anode-fall could be observed accurately to 
one-tenth volt ; the cathode to less than one volt. 
Preliminary Tests. 
In hydrogen and oxygen a preliminary study. of both 
cathode and anode fall was made, the controlling conditions 
in nitrogen* having been previously determined. The metals 
and discharge-tube were prepared and left over-night as 
described above. On the next day the tube was again 
flushed several times with the gas, great care being taken to 
avoid entrance of moisture through the pumping apparatus. 
Gas was finally let into the tube at the desired pressure, and 
after standing half an hour or more the fall of potential at the 
cathode (or anode) for the metals quickly observed—no metal 
* Wied. Ann, lxviii. p. 752 (1899); Phil, Mag. J. ¢. 
2D2 
