on the Theory of Voltaic Action. 371 



53 mm. diameter, and about 30 cm. long, with a leading tube attached 

 on one side. The measurement of the difference of potential was made 

 by a well-known zero method. The deflection given by a quadrant 

 electrometer, on separating the condenser plates, was annulled by 

 connecting the plates to points in a circuit of variable high resistance, 

 containing a large gravity Daniell cell ; the electromotive force required 

 thus to annul being taken as equal and opposite to the difference of 

 electrostatic potential at the plates. Three similar sets of measurements 

 were made with this apparatus, continuing after it had been sealed up 

 respectively six months, one and a half years, and seven and a half 

 years. In the intervals between observations the plates remained out 

 of metallic contact, and were kept, in Experiment I at their greatest 

 distance apart ; in III generally at their least. My notes of Experi- 

 ment II are not clear on this point. 



Experiment I, started December 12, 1888. The plates having been 

 cleaned with fine glass paper, the condenser was slipped into its tube. 

 A small porcelain cup of phosphorus pentoxide was introduced, in 

 order to dry out the interior, and the tube was then temporarily closed. 

 The difference of potential was then found to be 



0-74 volt. 



The end of the main tube was then sealed at the blowpipe. The 

 tube was exhausted through the side tube by a Sprengel pump, then 

 filled with nitrogen, again exhausted, and then refilled with nitrogen. 

 The condenser plates were now found to be in metallic contact, pre- 

 sumably due to the accidental sucking in of some minute globule of 

 mercury from the pump. In three days this contact ceased. About 

 3 grams of potassium and. 1 gram of sodium, in small pieces, were 

 now inserted by the side tube. (My notes do not state that the capsule 

 of phosphorus pentoxide was removed before closing the tube, but 

 probably it was.) The tube was again, December 17, exhausted to 

 3 mm., and refilled with nitrogen, when the difference of potential was 

 found to be about 



0-64 volt. 



The tube was finally exhausted to 4'5 mm., and the side tube was 

 sealed off, the difference of potential being 



0-61 volt. 



The following observations were then made at the intervals noted 

 in days after thus starting the experiment : 



Days... 13 25 27 30 61 106 173 181 

 Volt ... 0-56 0-52 0-55 0'51 0'47 0'34 0-32 0'33 



