immersed in pure water and in oxygenated water. 351 



the water at the time of closing being Fig. 1. 



near the boiling temperature. On ^ ^ 



cooling, the space left vacant by the 



contraction of the fluid was estimated 



to be T^fth of a cubic inch ; the super- d 



ficies of each plate ^th of a superficial 

 inch. From previous trials, I knew that 

 when the above arrangement had a 

 communication with the atmosphere, a 

 flocculent deposit of the protoxide of 

 zinc was soon perceived, which steadily 

 increased. With the same hermetically 

 sealed there was no such deposit ; nei- 

 ther was there any perceptible change, 

 until the bursting of the vessel after 

 two years revealed another action of 

 the battery. Judging from the thick- 

 ness of the broken glass, I endea- 

 voured at the time to make an ap- 

 proximate estimate of the volume of 

 the gas generated, which in the vacant 

 space of y yth of a cubic inch, where it 

 could lodge, produced pressure suffi- 

 cient to burst the vessel : the result of 

 my estimate gave less than a cubic inch 

 of gas measured at the usual atmo- 

 spheric pressure ; for the development 

 of which six zinc surfaces of ^th of a U 11 



superficial inch each had been two 

 years in action. In a repetition of this 

 experiment, with zinc filings in lieu of plates, a small quan- 

 tity of gas was collected, and proved to be hydrogen. 



Afterwards examining the inner surfaces of the firagments 

 of the glass, the surface of the plates, and the fibres of the 

 thread with a powerful lens, I found all of them covered with 

 minute transparent crystals ; the largest crop of these was on 

 a copper surface opposite a spot on one of the zinc plates, to 

 w hich nearly the whole of the corrosion of the metal appeared 

 to have been confined. The red ground of the copper sur- 

 face showed them most distinctly. On heating the copper 

 the crystals parted with water of crystallization, and became 

 circular white spots, very much resembling the protoxide of 

 zinc. 



My friend Mr. Waldie examined the thread; his process 

 was, incinerating, dissolving the ash in hydrochloric acid, 

 adding excess of potash, filtering to separate a trace of oxide 



