102 
PROFESSOR HUGH L. CALLENDAR ON 
values of the coefficients assumed, namely, 1‘50 and 1'54, cannot be entirety explained 
by the difference of the scales at the boiling-point of sulphur. # 
Part IV.—Calorimetry. 
(24.) Temperature Regulation. 
The question of temperature regulation was particularly important in this method 
of calorimetry. It also presented exceptional difficulties on account of the form of 
the calorimeter, and the large range of temperature to he covered. The apparatus 
enqiloyed for this purpose was made in the summer of 1895, and I spent a good deal 
of my leisure time during the session 1895-96 experimenting with various forms of 
regulator before I was able to obtain a satisfactory arrangement. 
On account of the extreme length of the calorimeter (over 1 metre), and since it 
Fig. 5. Heater, Circulator, and Regulator. 
AB, oil bulb; C, stirrer blades; D, gland; E, inflow; F, F, outflow. 
was necessary to have both the ends accessible for inserting leads and thermometers, 
I decided to employ a tube form of water-jacket open at both ends in place of the 
more usual bath and stirrer. The most convenient method of maintaining- such 
a jacket at a constant temperature appeared to be by means of a vigorous water 
circulation maintained by a centrifugal pump. The apparatus constructed for this 
purpose is shown diagrammatically in fig. 5. It was intended to serve as heater, 
regulator and circulating pump simultaneously. The annular bulb AB was filled 
with oil, the expansion of which actuated a gas regulator in the usual manner. The 
revolving blades CC of the centrifugal pump were connected through the gland D 
with a water motor by means of a short piece of stiff rubber tubing. The circulating 
* The Joule-Thomson equation, extrapolated to 445°, would make the constant-volume thermometer 
read two or three-tenths of a degree higher than the constant-pressure thermometer, which would account 
for the difference between Chappuis’ corrected result, 444° - 7 C., for the sulphur boiling-point, and the 
value, 444°-53 C., obtained by Callendar and Griffiths. —Added March 11, 1902. 
