94 
PROFESSOR HUGH L. CALLENDAR ON 
closed for the shortest possible time. This method should never be used in accurate 
work. Other writers have apparently found more serious difficulties, and appear to 
have considered that the heating effect was fatal to accurate work. This view has 
arisen merely from the use of unsuitable apparatus or faulty arrangements. 
(20.) Ice-point Apparatus. 
In accurate work the heating effect should never exceed a hundredth of a degree, 
and the correction can be readily applied if required, in case the current through the 
thermometer is not kept constant. It is possible, however, to obtain consistent 
results, even if the heating effect amounts to several tenths of a degree, provided 
that the conditions are steady, and that the heat gener¬ 
ated is not allowed to accumulate. This condition is 
generally satisfied in a bath of saturated steam or vapour, 
or in a well-stirred bath of liquid, but not at the freezing 
or melting point of a bad conductor. Errors due to 
variation of the heating effect are most common in 
observing the ice-point. The density of the water at 
this point is nearly constant, so that the convection 
currents are feeble, and the thermometer, if the current 
is excessive, or if there is considerable conduction of heat 
along the stem, as in the case of thick porcelain tubes, 
may become surrounded by a layer of water at a 
temperature appreciably above the freezing-point. Some 
advantage is gained in this case by employing a stirrer 
to make the water circulate vigorously through the ice. 
For this purpose I devised the following apparatus, 
which proved very useful for investigating the heating 
effects of large currents at the freezing-point, where 
accurate results could not be obtained by merely sur¬ 
rounding the thermometer with ice. The apparatus 
consists of two concentric cylindrical vessels. The ther¬ 
mometers are placed in the inner vessel. The whole 
apparatus is filled with melting ice, with enough water 
from previously melted ice to fill up the interspaces. 
The bottom and top of the inner vessel are fitted with 
gauze strainers to prevent circulation of the ice. The 
water is caused to circulate through the ice by means of a centrifugal stirrer below 
the middle of the inner vessel, worked by a shaft passing up through the centre of 
the apparatus. The thermometers should be deeply immersed, if they have thick 
rubes, or copper or silver leads, in order to minimise the effect of conduction along 
Fig. 3. Ice-point Apparatus. 
A, gauze-covered apertures ; 
B, outer vessel; C, inner 
vessel; D, shaft of stirrer ; 
E, gauze strainer; F, lagg¬ 
ing ; G, thermometer; H, ice; 
J, stirrer blades. 
