1906-7.] Change in Lead Wires by Permanent Stretching. 17 
jack which could be adjusted to the required height. In order to prevent 
jerking, the first weights, which were made of brass and contained a slot 
for the copper rod, were put on while the disc rested on the screw jack ; the 
last one or two were put on after stretching had begun, and were carefully 
adjusted by hand. By previous tests we knew how many weights were 
required to produce any considerable stretch. After the removal of the 
weights, the wire was always allowed to hang freely for several hours 
before any measurements were made. 
The instruments used to determine the resistance of the wire were a 
Carey Foster Bridge by Nalder (N.C.S. pattern) and a Nalder’s Four Coil 
Astatic Galvanometer of Kelvin type (N.C.S. pattern), which was provided 
with a “ scissors ” magnet control. The bridge wires were detachable, and 
of different resistances, varying from very thin to about J inch in thickness. 
In the experiments, the resistances forming the arms of the bridge were 
the lead wire that was stretched, and a second lead wire of the same 
diameter and length approximately. The wires hung close together, and 
the second wire carried no weights. From the mercury cups with which 
the ends of the lead wires were in contact, thick insulated copper cables, 
of the same lengths and diameters, led to the bridge. 
Several of the detachable bridge wires were used, and these were 
calibrated by Strouhal and Barus’ method. The bridge wire was divided 
into parts of equal resistance by comparing a length on the bridge wire 
with the same one of several approximately equal resistances, joined in 
series by mercury cups, this resistance being moved forward one place after 
each determination. Then, by a graph, we computed the relative resist- 
ances of any two lengths on the bridge wire. We then determined the 
absolute resistance of the bridge wire between two points, one near each 
end, and, from the graph, got the resistance of the part between any two 
points lying in this interval. In making the absolute determination, one 
arm of the bridge was a thick copper piece of negligible resistance, and the 
other arm was, first, a definite resistance, and, secondly, this definite 
resistance in parallel with a standard ohm. 
The resistance of the unstretched wire was measured on one of the 
Carey Foster wires before the other lead wire was stretched, and also 
between the stretches. The difference of the resistances of the two lead 
wires was measured after each stretch, and the whole resistance of the 
wire under investigation determined. By this means we insured that 
the change in specific resistance, if any, should be due to the stretching. 
The wires were hung in a long cupboard to prevent temperature changes, and 
the whole of the apparatus guarded from air currents by means of screens. 
VOL. xxvit. 2 
