344 Proceedings of the Royal Society of Edinburgh. [Sess. 
division was "00950 mm., and in the left-hand, *00944 mm. It was possible 
to estimate a tenth of a division. 
The temperature coefficient of the electrical resistance of the wires was 
determined by coiling them on a glass tube covered with silk, and 
immersing them in a vessel of oil. The resistance was determined for 
temperatures ranging from near zero to 150°. Great care was taken to 
obtain accuracy in measuring the various temperatures at which the 
resistance was determined. The temperature was taken by a platinum 
resistance thermometer. It had been calibrated by finding its resistance 
at 0° and 100°, and drawing the graph on the assumption that its resistance 
was zero at — 240°, according to the results obtained by Professor Callendar. 
This graph was then tested by finding the resistance in the vapour of 
boiling ethyl-alcohol and of boiling amyl-alcohol, and the differences were 
within the limits of experimental error. 
The method adopted for finding the temperature coefficient of resistance 
was the following: — The platinum thermometer was immersed in the oil 
close to the wire whose resistance was being determined. Two ordinary 
mercury thermometers were also placed in the vessel, one near the top, 
the other near the bottom, their use being to ensure that the temperature 
did not vary while the resistances of the wire and of the platinum 
thermometer were being determined. The oil was kept well stirred ; when 
it reached a temperature at which a reading was desired the flame was 
carefully regulated ; and when the mercury thermometers had remained 
steady for not less than five minutes, the resistances of the thermometer 
and the wire were taken in the usual way by a metre-bridge. The 
resistance of the copper side-pieces of the bridge had been determined, the 
wire calibrated, and, to provide against any alteration in these through 
their temperature being raised by heat from the oil bath, the connecting 
wires were several metres long. The resistance of the connecting wires 
was allowed for by measuring that of similar wires lying parallel to them. 
The diameter of each wire was determined by weighing a measured 
length in air and in water, and calculating the diameter from the density. 
The diameter at different points was also measured by the micrometer 
gauge, and the average of the two methods was that used in the 
calculations. 
Arrangement of the Apparatus. 
AB is the wooden block, GH the glass tube resting on wooden blocks, and 
having openings at E and F, through which the readings were taken by 
the microscopes M and N. The wire was fastened at R, passed over a 
