436 Dr. J. A. Harker on the Specific 



It was found best for the experiment to use a somewhat 

 long and narrow sample, and that the iron should not be in 

 the form of a cylinder but of small pieces fitting not too 

 tightly into the tube. This necessitated the weight of the 

 porcelain being a rather large fraction of the whole — about 

 two-fifths of the total weight. It was obvious therefore that 

 an accurate determination of the specific heat of the kind of 

 porcelain used was necessary. Accordingly a set of experi- 

 ments was made with a specimen consisting of a porcelain tube 

 filled with fragments of the same kind of porcelain weighing 

 11 grams in all. The results of the experiments are given 

 in the following table : — 



Mean Specific Heat of Porcelain between 15° G. and 

 the temperature given. 



Temp. cent. 

 912 



Spec. heat. 

 •2582 



958 



•2563 



1075 



•2539 



These experiments indicate a small fall in the specific heat 

 over the range, through which the porcelain was used as 

 protecting sheath in the iron experiments. 



The results of the whole three sets of experiments on the 

 iron, made with an interval of more than a year and employ- 

 ing several different specimens, are tabulated together. The 

 most striking feature is the marked drop in the value of 

 S * above 900°. As the specific heat of both oxides of iron 

 is much greater than that of the metal, it is probable that in 

 those experiments where oxidation took place in the furnace, 

 the value found would be too high. The specimen was 

 cleaned from all but a thin firmly adhering film after each 

 experiment, and was weighed both immediately before and 

 after. The time of attainment of thermal equilibrium was of 

 course much shorter in the earlier observations, but even in 

 the series with the magnesia tube it is unlikely that the 

 calculation of the quantity of heat given out should be in 

 error more than about half a per cent, in most of the 

 experiments. 



The final result of the whole set of experiments is shown 

 in the table appended, arranged in order of ascending tem- 

 perature. Preliminary observations with the protected 



