on the Electrical Resistance of Bismuth. 195 



undertake to make qualitative analyses of various specimens 

 for us. "We wish again to thank him here for the indispensable 

 assistance which he has given to our work. 

 Eesults of the qualitative analyses : — 



Bismuth. Impurities. 



Monheim . . . Copper*, lead, iron. 



TrommsdorfF I. . Iron, copper. 



Trommsdorff II. . Lead, copper, nickel. 



Absolutely pure \ T j n. \ • ^ 



T A W l ^ m ^j copper (traces), iron, carbon. 



Schering . . . Copper, lead ?, nickel. 



It is very difficult to prepare pure bismuth. Even the 

 method indicated above for preparing the metal which has 

 been compressed, does not yield a pure product. A small 

 quantity of lead and copper is always precipitated with the 

 bismuth ; and it is only by repeating this operation several 

 times that the metal can be obtained very pure. 



M. A. Classen is occupied at the present time in pre- 

 paring for us pure bismuth, which is to be used in our later 

 researches. 



It results from all this that the various specimens of 

 so-called pure bismuth are very different from a chemical 

 point of view. Their points of fusion and their specific 

 weights have been likewise determined ; they vary very 

 sensibly from one sample to another. 



IV. Measurement of Electrical Resistances. 



The threads of bismuth were placed in a water-bath, which 

 was heated all along its length by a linear gas-burner, forming 

 a series of small flames : a thermometer indicated the tem- 

 perature. 



The whole was placed between the poles of a RuhmkorfFs 

 electromagnet of large size, worked by an Otto gas-engine 

 and a dynamo-electric machine of Siemens and Halskef. 



The intensity of the electric current in the spirals of the 

 electromagnet was sensibly 28 amperes during the whole 

 course of the experiments. 



* The metals whose names are in italic are those which were found in 

 considerable quantities in the bismuth analysed. 



t The electromagnet was set in such a direction as to have no influence 

 on the galvanometer. 



