438 
MB. J. T. BOTTOMLEY ON THERMAL 
PtEGNAULT # in determining the boiling point of mercury. BuxsExt has also used an 
air-thermometer which was closed with a stopper when the high temperature was 
reached. The quantity of air left in the thermometer at the high temperature was 
determined voiumetrically instead of by weighing. The form of thermometer here 
described, with a capillary tube at each end, will, I think, be found an improvement 
in point of practical convenience. 
Fig. 4, a, shows the electric connections used in determining the resistance of the 
platinum wire at various temperatures. The potential method has been chiefly 
employed, and has been found very convenient and satisfactory. For this purpose there 
are connected in series a single gravity Daxiell’s cell, a Thomsox rheostat, the 
platinum wire to be tested, and a standard coil of platinoid wire about double the 
resistance of the platinum wire. At low temperatures the difference of potentials 
between the two ends of the standard coil is half, or considerably less than, that between 
the ends of the platinum wire, and when the temperature has been raised to 500° C. or 
600° C. this relation is reversed. The standard coil is silk-coverecl wire wound in a 
single layer on a brass can which is kept filled with cold water ; and, being of platinoid, 
its variation of resistance for two or three degrees of temperature is quite unimportant. 
A potential galvanometer having a resistance of 5000 ohms, and with an interposed 
resistance of 10,000 ohms or more, was used for determining the differences of potential 
on the standard coil and on the platinum wire ; and the rheostat enabled me to obtain 
convenient deflections of the potential galvanometer. 
In recent experiments on this subject I have also used a Wheatstone’s bridge 
instead of the potential method ; the four resistances of the bridge being two standard 
coils of platinoid, the platinum wire to be tested, and a calibrated rheostat. As the 
resistance of the platinum wire rises with temperature in one branch, increased resistance 
is thrown in in the rheostat branch so as to maintain the balance in the bridge. 
I come now to give an account of the main results which have been obtained up to 
the present time.J 
Three modes of experimenting have been used. 
1. A constant current was kept flowing through the wire, and the Sprengel pump 
was worked so as gradually to improve the vacuum. The resistance of the wire at 
different vacuums was determined, and thus the temperature which the given current 
would maintain in the wire at a measured vacuum was determined. 
* ‘ Paris, Acad. Mem.,’ vol. 21, p. 230. 
t ‘Wiedemann, Annalen,’ vol. 24, 1884. 
X In doing this, I may perhaps he permitted to remark that, owing to the very limited amount of time 
at my disposal, these results have been obtained in the course of experiments scattered overlong intervals 
of time. They have also been obtained with appai-atus which I have incessantly altered wherever I could 
see a possibility of improving. Moreover, it was impossible to foreseethe bearing of any one result in the 
long series of experiments, and few of them could even be calculated till long after they were obtained and 
recorded. I have, therefore, reason to be satisfied with their agreement among themselves. 
