.SYA' WILLIAM SII-.Ml-:\S, 1<\R.S. 149 



Plate 14, with a constant resistance of 1 unit for each metal at the 

 /m. Centigrade, and in diagram No. 2, Plate 15, with each metal 

 represented by ita own specific resistance at zero Centigrade, and 

 the close coincidence of the calculated resistances with the experi- 

 mental resistances as shown in the tables, excepting a certain 

 number of evidently erroneous observations, proves the entire 

 applicability of the law of increase expressed by the formula to 

 vnrious metals at temperatures between and 350 Centigrade. 

 It remained to be proved, however, whether the same law would 

 apply to higher degrees of temperature. 



PLATINUM BALL PYROMETER. For this purpose I had recourse 

 to a pyrometer, constructed upon the supposition that the specific 

 heat of solids and liquids is the same at all temperatures. 



An instrument of this description was designed by me some years 

 since, and is used by ironmasters in determining the temperature 

 of their hot blast. It is represented at Fig. 2, Plate 1 2, and consists 

 of a cylindrical vessel of thin sheet copper capable of containing 

 an imperial pint of water. The inner vessel is surrounded by the 

 two external vessels of thin metal plate, the narro\v space between 

 the first and second being filled with air ; and the space between 

 the second and third, or the outer vessel, with cow-hair or other 

 non-conductor of heat. A delicate thermometer is fixed against 

 the side of the innermost vessel, being protected from injury by a 

 perforated plate. It is provided with a sliding scale having 

 divisions equal in breadth to the degrees on the thermometer, but 

 each division counting as the equivalent of 50 degrees. A copper 

 or platinum ball is provided, the weight of which is so adjusted 

 that the heat capacity of 50 balls is equal to that of an imperial 

 pint of water at ordinary temperature. This is dropped into the 

 vessel and the sliding scale thereupon fixed so that its zero index 

 shall coincide with the position of the mercury level in the ther- 

 mometer tube. The copper or platinum ball is perforated, in 

 oilier that it may be placed at the end of a rod to be exposed to 

 the heat which is intended to be measured. 



Upon being fully heated, the ball is dropped into the water, and 

 the reading indicated upon the sliding scale, added to that of the 

 mercury thermometer, gives the temperature of the ball. 



Although a high degree of accuracy cannot be claimed for this 

 instrument, its indications are, nevertheless, useful for obtaining 



