94 The Rev. Dr. Callan's Eosperiments on the 



bisulphate of potash, bichromate of potash, sulphate of soda, 

 borate of soda, and chloride of sodium mixed with carbonate of 

 soda. But they were all decomposed by the galvanic current, 

 as was e\ddent from the light which the gases produced. When 

 the current was sent from the battery through a solution con- 

 taining three ounces of bichromate of potash and five of carbonate 

 of soda in about five quarts of water, there was scarcely any foam, 

 but the hght produced by the gases was tinged with red, and 

 not so intense as the oxyhydrogen light. The heat of the flame 

 appeared not inferior to that of the oxyhydrogen flame. When 

 bisulphate of potash was mixed with carbonate of soda, the light 

 and heat produced were the same as when bichromate of potash 

 was used. The light was injured much less by mixing bichromate 

 of potash, or bisulphate of potash with the carbonate of soda, 

 than by the mixture of any of the other salts with it. 



To estimate the comparative expense of the coke light and the 

 lime light, it is necessary, first, to determine the size of the 

 batteiy which M'ill produce a brilliant coke light, with that of 

 one, which, by decomposing water, will produce a lime light of 

 equal illuminating power ; and secondly, to determine the time 

 each battery will work with a given charge. Although I made 

 a battery of a size which I thought would be most advantageous 

 for producing the coke light, I have not as yet succeeded in 

 determining satisfactorily either of these two points. The 

 battery which I prepared consisted of 60 cast-iron cells in which 

 each zinc plate was 4 inches by 2. In the battery which 



1 commonly used, the zinc plates were 4 inches square. I always 

 found that this battery soon destroyed the coke points. On one 

 occasion, about five years ago, I got a coke light sufficient for 

 the gas mici'oscope from a batteiy of 24 cells, which were 

 only half-filled with acid. Hence I inferred that zinc plates, 



2 inches by 4, would be large enough for a brilliant coke light. 

 "N^Tien the battery of 60 cells, in which the zinc plates were 

 2 inches by 4, was prepared, I sent the current, first, from thirty 

 of them through a pair of coke points. The light was sufficiently 

 brilliant for all illuminating pui-poses, but did not last very long. 

 I then changed the connexion that I might tiy the effect of 

 40 cells. The light was then both steady and brilliant. I 

 think that this is the smallest, and therefoi'c the least expensive 

 batteiy, by which, with the aid of a good apparatus for adjusting 

 the coke points, a continuous light of great illuminating power 

 can be obtained. If the plates be smaller, the illuminating 

 power* of the coke points will not be sufficient ; and if the 

 number of cells be less than 40, the electric current will not 

 have sufficient intensity to pass through the flame between the 

 coke points. With this battery a pair of coke points lasts a long 



