BURGESS AND WALTENBERG: MELTING POINTS 373 



in the Arsem furnace, by the method of optical heating and cool- 

 ing curves,^ values which are close to Day and Sosman's value, 

 1452°. Four observations of the melting point of nickel on 

 iridium with the micropyrometer gave us 1452°. 



With the long filament pyrometer lamps used, the equation® 

 log c = a + 6 log r (in which c = current and T = absolute temper- 

 ature and h is very nearly unity) is sufficient for very considerable 

 temperature ranges, and permits a calibration in terms of two 

 temperatures only, i.e., nickel and palladium. Applying this 

 equation by extrapolation to the determination of the platinum 

 melting point by observing the melting of platinum strips mounted 

 as usual for taking metal melts, we obtained 1755° ± 5 from six 

 observations. The values found for diopside (1391°) and anor- 

 thite (1549°) by Day and Sosman were also exactly reproduced 

 by us with the micropyrometer. 



The metals actually used for calibration in hydrogen were 

 "Kahlbaum" electrolytic nickel {MP -= 1452°) and pure Heraeus 

 palladium {MP = 1549°), both of the same grade as used by Day 

 and Sosman.^ A sample of their palladium gave exactly the same 

 melting point as ours. In controlling the constancy of the pyrom- 

 eter lamps and studying the reproducibility of the melts some 

 35 observations were taken of nickel and 20 of palladium. The 

 various metals were not studied in regular order but were mixed 

 as indiscriminately as possible, with observations on the standard 

 points, nickel and palladium. 



Iron. Pure electrolytic iron from three sources and as many 

 different methods of preparation was used, namely, from Prof. 

 C. F. Burgess (99.97 per cent Fe) from Langheim-Pfanliauser 

 Werke A. G. (99.98 per cent Fe) and from a sample (99.99 per 

 cent Fe) prepared by Mr. J. R. Cain of this Bureau. "Kahl- 

 baum" iron in powdered form and iron reduced on the platinum 

 strip in hydrogen from Kahlbaum iron oxide were also used. 



5 C. W. Kanolt, Tech. Paper No. 10, Bureau of Standards, 1912. 



6 Pirani, Verh. Phys. Ges., 12: 323. 1910. 



^ Day, Sosman and Allen, 1. c. The analysis by Dr. Allen of these products 

 gave the nickel as 99.835 pure and the palladium 99.975. 



