372 BURGESS AND waltenberg: melting points 



knowledge of these melting points. The melts were made for the 

 most part in pure hydrogen, and except for titanium, were taken 

 on platinum strips 6 to 8 cm. long, 4 mm. wide, and 0.01 or 0.02 

 mm. thick. Manganese was also melted on nickel strips and 

 titanium was melted on iridium strips. 



Calibration with nickel and palladium. For melts on platinum 

 the pyrometer was calibrated in terms of the melting points of 

 nickel and palladium. The melting of these metals is extremely 

 sharp on platinum and they appear to show no alloying until 

 they melt. It is assumed that they melt on platinum at the same 

 temperatures as in crucibles; this we have found to be the case 

 for nickel, palladium, platinum, and iron. Numerous observa- 

 tions were made with various metals upon the possible effects on 

 the melting point determinations of alloying with the platinum 

 strip. We are convinced, that except possibly in some cases which 

 we shall not consider here in which there may be a chemical 

 reaction among the materials in the furnace, as for example when 

 silicon, platinum and hydrogen are together, there is no signifi- 

 cant alloying effect with the apparatus and materials as we use 

 them. 



The melting points of both nickel and palladium may be con- 

 sidered as well established by several series of measurements.^ 

 We have made an independent optical determination of the nickel 

 melting point in hydrogen with an Arsem furnace.* The nickel 

 was melted in a magnesia crucible contained within an out-glazed, 

 closed-end porcelain or kaolin tube. Temperatures were meas- 

 ured with a Holborn-Kurlbaum form of Morse pyrometer which 

 had been calibrated by Dr. Kanolt at the melting points of anti- 

 mony (630°), AgaCuo (779°), silver (960?5), copper (1083°), and 

 diopside (1391°). Five observations by Dr. Kanolt on the melt- 

 ing point of platinum with this pyrometer gave 1755° =t 5 in 

 an iridium furnace, or in exact agreement with the generally 

 accepted value. ^ 



For nickel, four observations gave us 1449°, 1450°, 1449°, 1449° 



' See Day and Sosman, Carnegie Institution of Washington, Pub. 157, 1911. 

 Waidner and Burgess, Bull. Bureau of Standards, 3: 163. 1907. 

 « Jl. Am. Chem. Soc, 28: 921. 1906. 



