1823.] occurring in the Thermoelectric Series. 323 



It is obvious that the abovementioned thermoelectric pecu- 

 liarity of iron v/ill affect its place in the series, which at a red 

 heat will be at least above brass, though at low temperatures it 

 is decidedly below plumbago. 



In regard to the series itself, it is not, perhaps, very material, 

 provided the order be correct, whether we consider bismuth as 

 the most positive, and antimony as the most negative metal, or 

 the contrary ; but analogy with the galvanic series seems to 

 make the last the preferable arrangement. In this case, which 

 I have now adopted, antimony, heat, and bismuth, form a circuit 

 similar to silver, acid, and zinc ; the silver and antimony being 

 the positive, and zinc and bismuth the negative poles. The 

 corresponding thermoelectric and galvanic series will, there- 

 fore, be : 



Positive, 

 In this thermoelectric series. 



Positive. 



I have corrected an error as to 



the place of silver, arising from the wire formerly used, and 



which had been purchased as pure silver, being, as I have since 



J found, alloyed with copper. The other changes are unimportant, 



^ excepting the in&ertion of galena, which in whatever way I have 



■p tried it, appears more strongly negative even than bismuth. 



Believe me, my dear Sir, most truly yours, 



J. CUMMING, 



y2 



