Heat of Iron at High Temperatures. 431 



point of silver by this method 907° C. We know from more 

 recent determinations that the correct value for this melting- 

 point lies between 955° 0. and 962° C. From this it is 

 obvious that before the results of Pionchon's experiments 

 can be compared with those of modern work, his temperature- 

 scale must be corrected, at all events at the higher ranges. 

 For this purpose I have assumed that the scale is accurate up 

 to 400° C, and from that point upwards have added a gradually 

 increasing correction to reduce the value of the silver point 

 to its proper figure. The results obtained are discussed at 

 the end of the paper. 



The experiments now to be detailed were undertaken to 

 test the accuracy of the results obtained with the iron ball 

 pyrometer still employed in certain industries for the 

 measurement of temperature. No protecting envelope was 

 employed for the earlier experiments at lower ranges. The 

 specimens, of which several of different sizes were used, were 

 cut to the shape of a cold chisel, and were made of a sample 

 of exceptionally pure metal containing only '01 per cent, of 

 carbon*, supplied to the laboratory by Mr. Had field of Sheffield. 

 The largest weighed 197 and the smallest 7 grams. They 

 were heated to the required temperature in a small vertical 

 electric furnace conveniently mounted on a swinging bracket, 

 which permitted the whole to be rapidly brought to an exact 

 position over the calorimeter when required. The furnace 

 was closed at the top and bottom by closely-fitting hinged 

 lids. In the earlier experiments the temperature was 

 measured by a platinum thermometer, enclosed in a thin 

 porcelain tube, the bulb of the thermometer being in close 

 contact with the specimen. To ensure equality of temperature 

 between the specimen and the thermometer, a thick sheath 

 of iron tubing was placed round both, and in all cases ample 

 time was allowed for the establishment of a sufficiently steady 

 temperature in the furnace to render errors in the measure- 

 ment of the temperature of the specimen as small as possible. 



The calorimeters employed were of the type formerly used 

 by the author for investigations on the latent heat of steam, 

 and were of the form of flattened cylinders made of thin 

 brass, nickel-plated heavily on the exterior. The stirring- 

 was continuous and was effected by a system of screw-blades 



* The complete analysis of the specimen is as follows : — 



Carbon = 



0-01 



Silicon = 



002 



Sulphur = 



003 



Phosphorus = 



0-04 



Manganese = 



trace 



