Mr Miller's Description of a new Pyrometer. 129 



case, through this lid/and the one of iron under, are holes to 

 admit the thermometer to the mercury, the scale is exposed 

 from 30^ upwards. 



The weight of the platinum is about one-ninth of that of 

 the mercury and sheet-iron surrounding it. Mercury, which 

 forms the chief part of the weight, has nearly the same spe- 

 cific heat as platinum ; the temperature, therefore, when dif- 

 fused over, or diluted in, the mass is lowered to one-tenth, 

 not one-ninth, as the platinum retains its share ; thus, a tem- 

 perature of 2000° (on the supposition that the atmosphere is 

 at zero) would become 200°, and be measurable by thermo- 

 meters. The diffusion of all heats is completed in the same 

 time, viz., in six minutes when the thermometer becomes sta- 

 tionary. 



The scale, it will have been obseiwed, is determined by the 

 relative Aveights of the mercury and platinum, and may be 

 altered by increasing or diminishing either ; it may be found 

 by calculation, but more accurately and easily by experiment. 

 A scale of l°to 10° of Fahrenheit has been chosen, partly for 

 the sake of ready notation ; by the addition of a cipher to the 

 number of degrees through which the thermometer has risen, 

 and the addition afterwards of the atmospheric temperature, 

 the original heat is obtained.* 



As far as the thermometer served to check and carry out 

 the experiments, the results were found uniform and consis- 

 tent, or nearly so ; 100 clear degrees imparting 10° to the 

 mass, 200°, 20°, and so on. 



The boiling point of linseed oil, estimated at 600° was in- 

 dicated at 594°. The heat of an ordinary fire varied from 

 about 1200° to 1600°, according to intensity, which is depen- 

 dent on draught. The heat of a parlour fire is set down by 

 the late Professor Daniell at 1141°. The difference here ex- 

 hibited was unexpected, nor does it seem to be clearly ac- 

 counted for by attributing an excess in my determination to 



* Example. — With the atmosphere at 60°, let it be supposed the thermome- 

 ter has risen to 145°, the number of degrees passed over is 95 ; a cipher an- 

 nexed makes 950, and the weather temperature of 50° added, gives 1000° as 

 tlie original heat. 



VOT,. XTJV. NO. LXXXVII. — -JAN. 1848. I 



