BETWEEN THE FREEZING AND BOILING-POINTS. 
231 
points of the scale (when the water is heated by the electric current). The increase 
is from the temperature coefficient of the radiation, and appears to be almost exactly 
linear over the range 0° to 100°. 
To show the relation between the heat-loss and the rise of temperature, I have 
summarized the observations in Tables XXXIII., XXXIV., XXXV., for rises of 8°, 5 C , 
and 2° respectively. These were made with Calorimeter C, but the flat heating-wire 
was used in place of the central heating conductor with rubber elastic. 
Relation of heat-loss to rise of temperature. 
Calorimeter C.—Mean Temperature, 28 =, G C. October 14, 18, and 19, 1899. 
dd. 
Q. 
(EC-4-2Q d0)ldd. 
Diflerence from mean 
value, '04535. 
Larg 
3 How. 
8-3069 
•626436 
•04521 
- -00014 
8-3212 
•625128 
•04533 
- -00002 
5-1009 
• 636545 
•04518 
- -00017 
5-1086 
•635186 
•04573 
+ -00038 
2-2054 
•620890 
•04544 
+ -00009 
2-2096 
•619353 
•04525 
- -oooio 
< 
Small How. 
Mean value, • 05047 
8-2446 
•381577 
•05044 
- -00003 
8-2446 
•381454 
•05021 
- -00026 
5-0887 
•388460 
•05058 
+ -00011 
5-0894 
•388232 
•05066 
+ -00019 
2-2417 
•376414 
•04951 
- -00096 
2-2433 
•375879 
•04992 
- -00055 
The observations were taken for different inflow temperatures so as to give approxi¬ 
mately the same mean temperature for the different values of dO, in consequence of 
which I have used the value 4'2 joules in obtaining the heat watts, the value of S 
being the same for the different values of dd. For the large flows, the agreement of 
the values of the heat-loss is good, and much better than might be expected, having 
obtained the values from experiments made on different dates, although they were 
within a day or two of each other, and the calorimeter did not vary more than a few 
degrees between each experiment. 
The values for the small flow do not agree so well for the 2° rise, but the values for 
