332 PROCEEDINGS OF THE AMERICAN ACADEMY. 



By turning the commutator M through 45°, after the measurement of 

 resistance before charging, the resistance measuring circuit was broken 

 and the charging current from B^ was started. This was allowed to flow 

 for a carefully measured length of time, — fifteen minutes, — and ammeter 

 readings were taken at the end of every minute. A turn of the commu- 

 tator key, M, stopped the charging current, and at the same time made 

 the proper connections for measuring resistance ; but before the latter 

 work was undertaken the burette was examined with a view to record- 

 ing the volume of hydrogen collected. No hydrogen was found in the 

 burette after this or any other of the earlier chargiugs ; and it was there- 

 fore assumed, in such cases, that the total volume liberated during the 

 charging run, as calculated from ammeter readings, had been absorbed 

 by the wire. This was usually, for each fifteen minutes' run, about 

 twenty times the volume of the wire. The resistance of the wire, 

 thus charged, was then carefully measured as before, just fifteen minutes 

 being allowed for this work. Again the wire was charged with an addi- 

 tional twenty volumes (approximately) and its new resistance immediately 

 determined. This process, of alternately charging the wire and measur- 

 ing its resistance, was continued until the wire appeared to be saturated, 

 the entire experiment requiring about thirty hours of uninterrupted 

 observations. This slow charging rate was employed because prelimi- 

 nary experiments had shown that time is a very important element in the 

 absorption of hydrogen by palladium; if a strong current is used, most of 

 the liberated hydrogen bubbles up to the surface of the liquid, even at 

 the beginning of the run with a fresh wire. But with the weak currents 

 used in these experiments no hydrogen bubbles were seen until after the 

 palladium had absorbed more than six hundred volumes (75 to 80 cu. cm.); 

 oxygen bubbles, of course, rose from the platinum wires from the very 

 start. The first appearance of hydrogen in the burette was carefully 

 watched for, and thereafter, at the close of each charging run, the 

 volume of hydrogen in the burette was carefully recorded and reduced to 

 standard conditions of pressure and temperature. In the subsequent cal- 

 culations the new volume of hydrogen, collected in the burette during 

 each charging period, was substracted from the total volume liberated by 

 the current during that period, as calculated from ammeter readings, and 

 it was assumed that the difference represented the volume of hydrogen 

 (at 0° C. and 760 mm. pressure) absorbed by the wire during that run. 

 In the earlier part of the work tl>e charging time was, in each case, just 

 fifteen minutes ; later, when the hydrogen was escaping very freely, this 

 time was lengthened. In every case just fifteen minutes was allowed, 



