196 



HoTborn and Day — Gas Thermometer at 



always colored black while pure copper bad no noticeable 

 effect upon it. 



Antimony. — The antimony with which our determinations 

 were made, as well as the following metals, were obtained from 

 C. A. F. Kahlbaum, chemist, in Berlin ; with the exception of 

 aluminium they may be regarded as pure. 



For melting the antimony only graphite crucibles were 

 employed, and even then a slight superficial oxidation was 

 observed after cooling. Eeducing gases were not introduced 

 on account of the thermo-elements. 



In solidifying, the temperature first sank far below (as much 

 as 20°) the melting point, to which it then rose suddenly (see 

 fig. 6, curve A). The time curves are otherwise normal. The 

 mean value for the melting temperature was Q l S0-6°. 



Minutes 10 20 SO hO 50 GO 70 SO 90 100 110 



Aluminium. — Aluminium was melted in both porcelain and 

 graphite crucibles. As has been noticed already by former 

 observers, this metal shows no sharply defined melting point. 

 Its time curves are similar in form to those obtained from 

 silver in air. We give the melting point as 657'3° — the mean 

 of the results obtained in the porcelain crucible. The melting 

 temperatures observed in graphite were slightly lower to be 

 sure, but the difference is more probably due to the smaller 

 quantity of metal used. 



The melting point of aluminium is not a sufficiently well- 

 defined point for use in calibrating thermometrical apparatus ; 



