340 DECOMPOSITION AND RECOMPOSITION OF WATEIl. 



female screw is wrought so as to allow a large iron valve cock (C) to 

 be screwed to it. 



Upon the tubiilures also iron caps are cemented, which are so 

 wrought as, with the aid of appropriate screws, to constitute stuffing 

 boxes. 



Through each of these a platina rod (D, d) is introduced, and 

 fastened to plates of platina, to act as ^'- electrodes,'^ agreeably to the 

 language of the celebrated Faraday. 



The tube being supported over the mercurial cistern, by means of 

 a communication with an air pump, through the valve cock and flexi- 

 ble leaden pipe, the bore of the tube is exhausted of air, so as to cause 

 the mercury to take its place. 



The mercury is so far displaced by a solution of borax, consisting of 

 equal parts of water and saturated solution of that salt, as to sink the 

 surface of the column of metal in the tube about an inch or more 

 below the "electrodes." The projecting end of one of the rods (D, rf). 

 to the other ends of which the " electrodes" are severally attached, is 

 bent at right angles outside of the tube, so as to enter some mercury in 

 an iron capsule, supported purposely at a proper height, and communi- 

 cating with one end of my deflagrator of an hundred pairs of Cruick- 

 shank plates of about eight inches by fourteen. Of course the rod of 

 the other electrode must have a commimication with the other end of 

 the deflagrator. Under these circumstances, if the circuit be com- 

 pleted by throwing the acid on the plates of the deflagrator, a most 

 rapid evolution of hydrogen and oxygen will ensue in consequence of 

 the decomposition of the water, so that within a few seconds, several 

 cubic inches of gas will be collected. 



The action being now suspended by throwing the acid off* the plates, 

 and the foam being allowed to subside, the resulting gaseous mixture 

 may be ignited, and of course condensed, by completing the circuit 

 again as at first, and at the same time causing the ends of the " elec- 

 trodes" to come into contact with each other, and thus to produce a 

 spark. 



This contact is effected by causing a very slight movement in the 

 rod, bent at right angles, and entering the mercury in the iron capsule. 

 Of course the process may be repeated as often as can be reasonably 

 desired. 



