TRANSACTIONS OF THE SECTIONS. 47 



where h is the length of each tube. To prevent accidental breakage of the mano- 

 meter, it is fastened to the graduated brass plate, and with it screwed to a glass 

 cover J of an inch thick, capable of supporting a pressure of 20 atmospheres. 



Electbicitx and Magnetism. 



On the Influence of Clean and Unclean Surfaces in Voltaic Action. 

 By Thomas Bloxam, Lecturer on Cliemistnj, Cheltenham College. 



1. Gas was evolved by the contact of zinc and platinum surfaces, then an equal 

 amount from the same surfaces when the platinum had been cleaned by hot oil 

 of vitriol; the time was exactly half when the clean surfaces were used; contact of 

 the surfaces with the fingers or dipping them in solutions of various substances was 

 found to retard the evolution in a very marked degree. 



2. Heating the platinimi in a measure cleaned it, but not so satisfactorily as hot 

 oil of vitriol. Copper and other metals behaved similarly to platinum. 



•3. Platinized silver, from its method of manufacture, appeared to be already 

 clean, no advantage being obtained by chemically cleaning it. 



4. Mechanically roughened surfaces of platinum exhibited a decided advantage 

 over smooth ones. 



5. The cell of a Smee's battery, examined by a galvanometer, gave vastly better 

 results when the negative plate had been chemically cleaned. 



6. Voltameters, the plates of which had been chemically cleaned, exhibited a 

 marked superiority over those not so cleaned ; thus it appears that in all voltaic 

 action the results are superior where the surfaces of the negative metals, elec- 

 trodes, &c. have been chemically cleaned, and that mere contact with the finger is 

 sufficient to modify the evolution of gases from the surface. 



On a new Form of Constant Galvanic Battery. By Latimer Clabe, C.E. 

 (Extracted from a Letter to Sir William Thomson.) 



I have spoken to you several times about a form of battery which can be set up 

 under such conditions as to ensure uniformity of tension within limits of about '05 

 or "06 per cent., and that without any special precautions as to the purity of the 

 materials employed. I have not yet been able to make the necessary experiments 

 for determining its value in absolute imits, though I hope shortly to have made 

 an independent determination. I have, however, set up about 200 of the 

 elements in question, and have measured them on about 30 different days; 

 and from the mean of these experiments, taking the Daniell at 1'079 volts, I make 

 this element to be 1'403 volts. In obtaining this result I have had to make care- 

 ful measurements of electromotive force of more than 1000 different elements, 

 comprising some 40 or 50 different kinds ; in fact I have been working at it for 

 six years. 



The element in question varies about '07 per cent, for each degree Centigrade, 

 getting weaker with increased temperature : the temperature at which our com- 

 parison with the Daniell's cell is made is 18° Centigrade. 



The element consists of a cylinder of pure zinc resting on a paste of protosulphate 

 of mercury and satm-ated solution of sulphate of zinc, previously boiled to expel the 

 air, the other electrode being metallic mercury, connexion being made with the 

 latter by a platinum wire. It is desirable that the materials should be pure ; but 

 if commercial materials be employed the error does not exceed -06 per cent, at first, 

 and after three or four hours the value becomes sensibly the same as with pure 

 materials. 



The precautions necessary are that the protosulphate of mercury should be free 

 from persulphate, and that the solution of persulphate of zinc should be supersatu- 

 rated. The elements do not vary sensibly for two or three months, say "05 per cent. 



It is essential that the element should not be worked through small interpolar 



