338 kanolt: melting points of fire bricks 



proportions to form sillimanite, to protect them from the small 

 amount of reducing gas in the furnace, altho the action of this 

 gas was slight. The samples were observed thru a glass window 

 in the top of the furnace. 



The temperatures were determined by means of a Morse optical 

 pyrometer of the Holborn-Kurlbaum type, which was sighted 

 vertically downward thru the glass window. The carbon-fila- 

 ment pyrometer lamp was calibrated by two methods. In the 

 first calibration it was sighted into a platinum resistance furnace 

 in which black-body conditions were obtained, and the tempera- 

 ture of which was measured by platinum platinum-rhodium ther- 

 mocouples. These thermocouples had been calibrated against 

 the freezing points of pure metals. In the second calibration the 

 lamp was calibrated against the freezing points of metals directly, 

 without the intermediation of thermocouples. The metals used 

 were copper, silver and the copper-silver eutectic, which freeze 

 at 1083°, 961°, and 779° respectively. The metals were melted 

 in the vacuum furnace in graphite crucibles, the pyrometer being 

 sighted into a thin-walled graphite tube inserted in the metal. 

 The pyrometer readings corresponding to the freezing points were 

 determined by means of cooling curves. With silver and copper, 

 heating curves were also obtained. 



As the melting points to be measured were above the working 

 limit of the pyrometer lamp, an absorption glass was interposed 

 between the pyrometer and the furnace. 



The true temperatures were then found from the apparent 

 temperatures measured thru the glass, by means of the equation 



where T\ is the absolute temperature of the furnace, T 2 is the 

 apparent temperature observed thru the glass, and A is a constant. 

 The value of A was determined by calibrations at various temper- 

 atures. A small correction was also applied for the absorption 

 and reflection of the glass window of the furnace. 



The samples, which were from 1 to 2 cm. in diameter, were 

 heated at the rate of about 10° per minute when near the melting 



