454 A. L. Day and J. K. Clement — Gas Thermometer. 



interpolation of this character even when used with the best 

 experimental data which we have obtained. The error amounts 

 to 2° in the region 600-700°. 



The solution of this difficulty is obviously to obtain additional 

 fixed points and thereby to reduce the interval for interpola- 

 tion but we have so far found difficulty in obtaining suitable 

 substances. The metals which melt in the desired region either 

 are not obtainable in uniform purity or easily become oxidized 

 or otherwise contaminated during the manipulation necessary 

 for a melting point determination. Suitable eutectic mixtures 

 may eventually offer a solution of the difficulty. 



The Metals Used* — After some investigation, it was found 

 that we could obtain in sufficient quantity, silver, copper and 

 zinc which ranged in purity from about 99*94 per cent to 99*997 

 per cent, and gold which was probably still purer. These 

 figures do not include oxygen nor carbon (except in case of 

 the silver), for, since the melting points of the metals had to 

 be determined in carbon crucibles, it is evident that the pres- 

 ence of these impurities would have no significance for present 

 purposes. Of course it would have been possible to prepare 

 these metals, or at least the zinc and copper, in still purer 

 condition, but it was not thought to be worth while, since it is 

 improbable that the most refined of present-day methods could 

 safely determine any difference between the melting points 

 of the chemically pure metals and those actually used. 



The Gold. — About 250 grams of " proof gold " were obtained 

 from the Philadelphia Mint. It was prepared by Mr. Jacob 

 Eckfeldt. A sample of gold prepared in a similar manner by 

 Mr. Eckfeldt was used by Prof. Mallet in his determination of 

 the atomic weight of this metal. The method of purification 

 is given in the Am. Chem. Jour., vii, 73, 1889. Prof. Mallet 

 found no systematic difference between this gold and two 

 other samples, one of which was obtained from the Mint of 

 England, and the other of which was prepared by himself. In 

 view of these facts, it was evidently unnecessary to analyze the 

 gold. 



Regarding the methods which were used in the analysis of 

 the silver, copper and zinc, there will be no need of giving all 

 the details, especially where accurate methods of procedure 

 are well known, but in view of the very small quantity of 

 impurities estimated, some explanation and some conclusions as 

 to the accuracy of the data will be presented. Most of the 

 work was done in a new laboratory under exceptionally favor- 

 able conditions of cleanliness ; large samples, generally 100 

 grams, were taken for analysis and the reagents were subjected 

 to rigid examination. Separations were always repeated, in 



* By E. T. Allen. 



