254 



Sir William Thomson on a 



[Jan. 19, 



near the top of the jar. But he tells me that batteries on this plan, called 

 "gravity-batteries," were carefully tried in the late Electric and Inter- 

 national Telegraph Company's establishments, and found wanting in 

 economy. The waste of zinc and of sulphate of copper was found to be more 

 in them than in the ordinary porous-cell batteries. Daniell's batteries 

 without porous cells have also been tried in France, and found unsatisfactory 

 on account of the too free access of sulphate of copper to the zinc, which they 

 permit. Still, Graham's and Fick's measurements leave no room to doubt 

 but that the access of sulphate of copper to the zinc would be much less rapid 

 if by true diffusion alone, than it cannot but be in any form of porous-cell 

 battery with vertical plates of copper and zinc opposed to one another, as are 

 the ordinary telegraphic Daniell's batteries which Mr. Varley finds superior 

 to his own " gravity-battery." The comparative failure of the latter, there- 

 fore, must have arisen from mixing by currents of the liquids. All that seems 

 necessary, therefore, to make the gravity -battery much superior instead of 

 somewhat inferior to the porous-cell battery, is to secure that the lower 

 part of the liquid shall always remain denser than the upper part. In seek- 

 ing how to realize this condition, it first occurred to me to take advantage 

 of the fact that saturated solution of sulphate of zinc is much denser than 

 saturated solution of sulphate of copper. It seems* that, at 15° tempera- 

 ture, saturated aqueous solution of sulphate of copper is of 1*186 sp. gr., 

 and contains in every 100 parts of water 33*1 parts of the crystalline salt ; 

 and that at 15° the saturated solution of sulphate of zinc is of sp. gr. 

 1*44, and contains in every 100 parts of water 140*5 parts of sulphate of 

 zinc, both results being from Michel and Krafft's experiments f. Hence 

 I made an element with the zinc below ; next it saturated solution of sul- 

 phate of zinc, gradually diminishing to half strength through a few centi- 

 metres upwards ; saturated sulphate of copper resting on this ; and the 

 copper-plate fixed above in the sulphate-of-copper solution. In the be- 

 ginning, and for some time after, it is clear that the sulphate of copper can 

 have no access to the zinc otherwise than by true diffusion. I have found 

 this anticipation thoroughly realized in trials continued for several weeks ; 

 but the ultimate fate of such a battery is that the sulphate of zinc must 

 penetrate through the whole liquid, and then it will be impossible to keep 

 sulphate of copper separate in the upper part, because saturated solution 

 of sulphate of zinc certainly becomes denser on the introduction of sul- 

 phate of copper to it. To escape this chaotic termination I have intro- 

 duced a siphon of glass with a piece of cotton-wick along its length inside 

 it, so placed as to draw off liquor very gradually from a level somewhat 

 nearer the copper than the zinc ; and a glass funnel, also provided with a 

 core of cotton- wick, by which water semisaturated with sulphate of zinc 

 may be continually introduced at a somewhat lower level. A galvanic 



* Storer's Dictionary of Solubilities of Chemical Substances. Cambridge, Massa - 

 ehusetts : Sever and Francis, 1864. 



t Ann. Cb. et Phys. (3) vol. xli. pp. 478, 482: 1854. 



