316 kanolt: melting points of refractoey oxides 



tically downward thru a glass window in the top of the furnace. 

 The carbon-filament pyrometer lamps were calibrated against the 

 following melting points : 



Material Melting point 



degrees 



Antimony 630 



Copper-silver eutetic 779 



Silver 960.5 



Copper 1083 



Diopside 1391 



These materials were melted in graphite crucibles, the pyrome- 

 ter being sighted into a thin-walled graphite tube inserted in the 

 metal. Heating curves or cooling curves were plotted. 



As the temperatures to be measured were far above the working 

 limit of the pyrometer lamps, absorption glasses were interposed 

 between the pyrometer and the furnace. The true temperatures 

 were then found from the apparent temperatures measured thru 

 the glasses, by means of the equation: 



where T^ is the absolute temperature of the furnace, T^ is the 

 apparent temperature observed thru the glasses, and A is a con- 

 stant. The value of A for each glass or combination of glasses 

 used was determined by sighting the pyrometer at a black body 

 held at a constant temperature within the working limit of the 

 pyrometer lamps, and taking readings alternately with and with- 

 out the interposition of the glass. It was found that a small 

 error might be introduced by the lag of the pyrometer lamp when 

 its temperature was thus changed at short intervals. This was 

 avoided by using two pyrometer lamps, each maintained approxi- 

 mately at one of the temperatures to be measured. A correction 

 was applied for the absorption and reflection of the glass window 

 of the furnace. 



It was found that measurements of the same high temperature 

 made with different pyrometer lamps or different absorption 

 glasses were always in satisfactory agreement. As a check upon 

 the work five determinations of the melting point of platinum 



