Heights of Antimony and Bismuth Alloys. 227 



however that the wire and crystal were from the same sample 

 of metal. 



The effect of impurities can be seen in the table for bismuth 

 and tin given below, where it is seen that the presence of a 

 third of one per cent of tin is sufficient to reverse the direction 

 of the electromotive force of bismuth. Plainly then, if we ex- 

 pect to arrive at anything more than the general order of mag- 

 nitude of the quantities involved, the greatest care must be 

 taken to procure materials in a state of purity, and from these 

 materials the elements must be prepared in a uniform manner. 



Apparatus. 



A copper bar, l-5 cm square, 4 cm long, was shaped at one end 

 to a truncated wedge, and at the other inserted in a crossbar of 

 wood. A copper block was held upon either face of the wedge 

 by a light spring. Under these blocks, and pressed by them 

 against the faces of the wedge, pass the bars of the thermo- 

 couple, the bars meeting l em beyond the end of the wedge, 

 where they are soldered. The insulation between the bars, the 

 copper blocks and the wedge was effected by a thin film of 

 hard varnish only. It is thought that by the above means the 

 conduction of heat to the other junctions is prevented, and at 

 the same time the insulation was found to be complete. The 

 thermo-junction was connected with a sensitive galvanometer 

 and a thousand ohms additional resistance. It is assumed that 

 the galvanometer deflections produced by the junctions through 

 this circuit are proportional to the electromotive forces of the 

 junctions. 



The deflections were produced as follows. Two beakers of 

 water were provided, one at the temperature of the room, the 

 other at a higher temperature. The water in each was at such 

 a level that when the crossbar above spoken of rested upon the 

 rim of the beakers the junction would project a little below 

 the surface of the water. Then the water in each having been 

 stirred and its temperature read off, the junction was shifted 

 quickly from one beaker to the other, and, after the swing of 

 the galvanometer needle, back again. The deflection produced 

 by each couple under experiment was compared in each case 

 with the deflection by a couple of fine iron and copper wires 

 adopted as a standard. The electromotive force of the standard 

 junction was obtained by comparisons with a standard Daniels 

 cell, and found to be 1412 C. G. S. units at a temperature of 

 30° mean. 



Am. Jour. Sci — Third Series, Vol. XLVIII, No. 285.— Sept., 1894. 

 15 



