Fhydcal Cuiisfants of Thallium. 199 



observed value of Copper- Lead. The results with the silver 

 are not very consistent. Various diagrams can be con- 

 structed from the above observations, but the nearest to the 

 true one will ])robably be obtained from the first, third and 

 fourth, from which the figure is drawn (see Plate XVII), tlie 

 dotted lines on the figure being Professor Tait's results for 

 silver and copper. 



I have been for some months engaged on a series of 

 observations, of which I hope to give an account to this 

 Society shortly, from which it appears that thermo electric 

 values cannot be absolutely constant, and which explains the 

 above inconsistent results of the observations on these 

 metals. 



Summary. 



The constants obtained are thus : — 



(1) Coefticient of expansion - "0000225 



(2) Specific resistance - - 20200 at 20° C. 



(3) Range of resistance with 



temperature - - -00392 at 20° C. 



(4) Thermo electric height - 216— '79^ 



The thallium was obtained from Schuchardt, and Professor 

 Masson has kindly anal^^sed it and supplied the following 

 statement of his results : — 



" The small sample of thallium wire submitted to me, 

 was found to contain as impurities, lead, arsenic and copper. 

 An estimation of the lead showed it to be present to the 

 extent of 1-50 per cent. The arsenic and copper were 

 present in too small amount to be estimated in so small a 

 sample. A direct estimation of the thallium itself sliowed 

 the wire to contain 9790 per cent, of that metal." 



