the Thermo-Electric Properties of Liquids. 



517 



glass vessel, when first constructed, was straight, and the narrow part 

 of it, A, afterwards bent by means of a row of Bunsen's burners. Each 

 of the cups contained a ribbon of sheet platinum, E, welded to the 

 lower end of a thin platinum tube, F, and coiled into a spiral form, the 

 coil resting upon a triangle of glass rod or platinum wire, Gr, placed 

 inside each cup ; each ribbon employed was 48 inches long and 1 inch 

 wide, and weighed two troy ounces; shorter and somewhat wider 

 ones would be more convenient, and for the cold electrode a much 

 smaller one would probably suffice, but only for use with such liquids 

 as offer but little electric conduction-resistance ; the hot electrode 

 must, however, be large, in all cases, in order to obtain conspicuous 

 results. Each of the platinum tubes was tightly fixed in a hole in the 

 centre of a bung of cork, which fitted the mouth of the cup, by which 

 the electrode might be lifted out ; each bung also had a hole in it, for 

 the insertion of a thermometer, and a small one, also, for escape of air. 

 The ends of thin platinum wires, M, N", for conveying the current to 

 the galvanometer, were also fixed securely between the tubes and the 

 bungs. The cup to be heated, C, was surrounded by a brass jacket or 

 cistern, H (viewed from the front), and I (viewed from the side), 

 provided with an exit- tube, J, and a projecting chamber, K (closed at 

 its outer end only), beneath which the flame of a lamp might be con- 

 veniently applied. The jacket was secured water-tight around C 

 (fig. 1), by fixing upon the narrow part of the glass tube, at L, a split 

 bung, of a diameter such as to tightly fit the lower opening of the 

 cistern, and securing the cistern to it water-tight by means of asphalt 

 varnish. The jacket was kept concentric with the cup C, by means of 

 three pieces of cork fixed between the two vessels at their upper end. 



The exit-tube, D, was provided with a piece of pure india-rubber 

 tubing, 15 inches long (fig. 2), provided with a narrow glass funnel, 

 0, at its extremity, by means of which both the filling and the 

 emptying of the cups, B and 0, was effected ; a glass stop-cock was 

 substituted for the tubing when liquids which acted upon india- 

 rubber were employed. The exit-pipe, J, of the cistern was also pro- 

 vided with a pinch tap, P, so that the hot water might be quickly 

 removed from the cistern after an experiment. Eig. 2 shows the com- 

 plete apparatus. The glass tube was supported in a vertical position 

 by a firm clip-support, which grasped it tightly at the point A, and 

 the funnel, 0, was held in an elevated position by means of a separate 

 clip-support, R, attached by a vertical metal rod to a metal hoop, Q, 

 which encircled the cup, B, T is a screen of cardboard, also supported 

 by the hoop Q ; U, and V, are thermometers. The glass vessel con- 

 tained about five ounces of solution, and the cistern required about 

 three ounces of water. All the sketches are about one-fourth the 

 actual linear dimensions. 



In using the apparatus the electrodes were first well cleaned, washed, 



