334 PROCEEDINGS OF THE AMERICAN ACADEMY. 



tions of the second were made as nearly as possible like those of the 

 first. The palladium wires were cut from the same piece and were sub- 

 jected to jDracticallj the same treatment in preparation for their respec- 

 tive experiments. The curves are plotted with increase of resistance 

 (above the resistance at start) for ordinates, and cubic centimeters of 

 hydrogen occluded for abscissae. These curves should, then, be similar 

 in form to curves whose coordinates are " proportional increase of resist- 

 ance " and " proportional increase of mass " (see P"ig. 1). The curves 

 exhibited here show a decided increase in the resistance of the palladium 

 from the very beginning of the experiment. In fact the first few 

 volumes of occluded hydrogen seem to be more effective, in increasing 

 resistance, than similar quantities added later in the chai'ging process. 

 The two curves are noticeably similar, and show a marked departure 

 from the straight line. The writer hopes to repeat these experiments, 

 varying the conditions somewhat, as regards rapidity of charging, 

 temperature, etc., and, perhaps, using apparatus of different design, 

 in the hope of fixing, more conclusively, the exact sha^ie of the 

 curve. 



The maximum quantity of hydrogen occluded is, in the case illustrated 

 by Fig. 8, 130 cu.cm. or about 1030 volumes. This increases the re- 

 sistance from 0.0927 ohm to 0.1558 ohm, an advance of 68 per cent. 

 It is believed that the last point of this curve corresponds, practically, to 

 saturation of the wire; after the data for this point had been taken, the 

 charging current was allowed to continue for twelve hours, and the 

 resistance was measured at the end of the first five hours and again 

 at the end of the twelve hours, but the results were practically those 

 obtained for the last point of the curve. No attempt was made to 

 measure the volumes of escaping hydrogen during this twelve-hour 

 run. Knott found a maximum increase of resistance of about fifty-two 

 per cent (0.1865 ohm to 0.285 ohm), corresponding to an occlusion 

 of from 900 to 1000 volumes. Graham, in Poggendorff"s Annalen 

 for 1869, states that the conductivities of pure palladium and the 

 hydrogen-charged palladium are as 5.99 to 8.10. He does not state 

 the exact quantity of gas occluded in the hydrogen-charged wire, but 

 elsewhere in the same paper describes experiments in which 900 volumes 

 were absorbed. 



The method employed in these experiments seems to be capable of 

 results more accurate than those obtained by the investigators named 

 above, and these results would seem to indicate that the resistance of the 

 palladium wire, as hydrogen is absorbed, increases rapidly from the very 



