High Temperatures. 



191 



Callendar as well as Hey cock and Neville* have called atten- 

 tion to the fact that silver in an oxidizing atmosphere melts 

 and solidities at a lower temperature than in a reducing atmos- 

 phere, the difference being caused by the fact that silver 

 absorbs oxygen in melting which is given off again below the 

 melting temperature with considerable violence ("spitting"). 



Silver — Crucible Method. — In order to ascertain whether 

 the melting point as determined by the wire method differs 

 from the normal melting temperature we made determinations 

 of it by the crucible method also. 



(1) fn an Oxidising Atmosphere. — At first the silver was 

 melted in an open porcelain crucible without stirring. Under 

 these conditions the results differed from those with other 

 metals in that the silver showed no definite melting tempera- 

 ture. In melting as well as in solidifying the "time curve" 

 takes the form A (fig. 4), i. e. throughout its length it contains 



9300 



9200 



9100 



■5700 



8601 



Minutes 



no horizontal section but falls or rises slowly over an interval 

 of 6° to 8°, between which limits one point may be assumed as 

 the melting point nearly as well as another. After substitut- 

 ing mica sheets for the clay oven-covers so as to be able to see 

 the metal during the process, it was observed that it remained 

 liquid through the greater part of this interval. The melting 

 and solidifying occurred at about 954 , 5° (8940 microvolts). 

 It may be added that the substitution of three mica covers 

 separated by air layers, for the two clay covers, produced no 

 essential change in the conditions within the oven. 



If the thermo-element with its protecting tube were moved 

 about or the liquid metal otherwise stirred, the form of the 



*0. T. Heycock and F. H. Neville, Journ. Chem. Soc, 1895, 160 and 1024. 



