456 



Mr. F. W. Jordan on the Direct 



The bismuth rod e was cut; from a crystallized* mass, 

 prepared by Dr. Lownds, so that its axis was parallel to the 

 principal cleavage plane of the metal. Two copper cylinders 

 c, d, were bored out to receive the heating coils, and their 



curved surfaces were planed down to within a millimetre of 

 the coils. The end of the bismuth rod was pressed against 

 and fused to the flat surface of the c}'linder. After solidifi- 

 cation the excess of fused bismuth around the rod was 

 removed. Each heating coil consisted of a spiral of fine 

 double-silk-covered eureka wire, which was soldered at 

 each end to thicker copper wire. The upper copper lead 

 was soldered to a circular copper disk a, and the lower 

 one was attached to the cylinder by electroplating with 

 copper. A copper wire h soldered to the upper lead served 

 as a potential lead for measuring the resistance of the coil, 

 and also as a lead for the heating current. The coil and 

 copper disk were insulated from the cylinder by paraffin wax. 

 The object of securing the coil in this way was to maintain 

 as far as possible an equality of temperature between the 

 copper cylinder and the terminals of the coil, and so minimise 

 the loss of heat bv conduction alono- the leads. 



