CONTINUOUS ELECTRIC CALORIMETRY. 
73 
(7.) Values of the Resistance Standards. 
The currents employed in this investigation were measured on the Thomson-Varley 
potentiometer by comparing the difference of potential on the terminals of a specially 
constructed resistance, called the Current-Standard, with that of a pair of hermetically 
sealed Clark cells. This current-standard consisted of a single 1-ohm coil, or of two 
1-ohm coils in parallel, immersed in a well-stirred oil-bath. The resistance coils 
employed for this purpose in the earlier experiments were made of thick manganin 
wire of the best quality procured from Germany. The diameter of the wire was 
1 millim., and the maximum current carried by each single wire was 2 amperes. 
The wire was not materially heated above the temperature of the oil, but, as no cooler 
was used in the earlier experiments, the temperature of the oil generally rose some 
8 or 10° in the course of an hour. This was considered to be of no consequence 
with the manganin coils, as they had a temperature coefficient of only -f- '000020 for 
ranid changes. On re-testing these coils after some months’ work, it was found that 
their resistance had increased by two or three parts in 10,000, and that they 
continued to show small variations of this order. It was possible that these changes 
might have been due to the solder junctions, which would have explained certain 
anomalies observed in the earlier experimental results. It is equally likely, however, 
that they were caused by hysteresis in the wire as explained below. 
Although the variations of the manganin coils did not exceed a few parts in 10,000, 
it was felt that they were quite inadmissible, as the potential readings were taken to 
1 in 100,000. For this reason it was decided to make a pair of platinum-silver ohms 
wound on mica and annealed at a dull red heat after winding. These will be called 
the “ Mica Current-Standards.” They are fully described and illustrated by 
Dr. Barnes, p. 173. The method of construction was modelled on that of a platinum 
resistance thermometer, and had already been adopted for some years by the 
Instrument Company, Cambridge, for the manufacture of standard resistance coils. 
Standards constructed on this model possess great constancy, and their temperature 
can be determined with accuracy by immersion of the naked wire in an oil-bath. As 
compared with manganin coils, they have the disadvantage of a larger temperature 
coefficient, but the construction permits the determination of this coefficient with the 
greatest certainty, so that it is really no objection to their use for scientific purposes, 
for which coustancy is the primary desideratum. The adoption of platinum-silver 
necessitated the addition to the oil-bath of a cooling coil of composition tubing, with 
a water circulation to keep the temperature steady. 
For the determination of the variation of the specific heat of water, the absolute 
value of the current-standard was of no moment, but its constancy was of primary 
importance. It was also essential to be able to determine the temperature coefficient 
accurately, in order to reduce the observations at different temperatures in winter and 
summer to the same standard. This was determined under the actual conditions of 
VOL. CXCIX.—A. 
L 
