173 
BETWEEN THE FREEZING AND BOILING-POINTS. 
The wires were bent into an inverted U and made to fit into mercury-cups, either on 
the commutator-bridge or in the main calorimeter circuit from which the potential 
terminals were taken to the potentiometer. For the later experiments with the 
water calorimeter it was found advisable to alter the resistance of the heating-wire to 
- 5 ohm, so that another mangaiiin-resistance was made similar to the first one and 
connected in parallel with it in the calorimeter circuit. Our numerous comparisons of 
these ohms with the certified standards were far from satisfactory, but the cause was 
at first sight not apparent. 
The resistance of both the coils was found to increase, after carrying currents of 
from 4 to 8 amperes in a series of experiments, of the order of 2 or 3 parts in 10,000 
in two weeks. This was somewhat annoying, and necessitated repeated comparisons 
with the standards and numerous corrections. It was also a matter of doubt whether 
the resistance of the coils remained the same when the heavy currents were passing, 
seeing that they produced such a large permanent change in the resistance. We 
finally commenced to suspect the real cause of the trouble to be at the point where 
the manganin-wire was soldered on to the heavy copper-wire. In the face of this 
uncertainty it was decided to abandon these resistances altogether for others made of 
platinum-silver wire according to a different design. Both these new 1-ohm resistances 
have proved to be so reliable and constant since they were made, in May, 1898, that 
it is proposed to give a short description of them here. They were both made on an 
exactly similar design. 
The frame-work consisted of two heavy plates of mica, 4" X 2 W, placed side by 
side, and separated about § of an inch by ebonite strips at each narrow end. Both 
ebonite strips were split from end to end, parallel to the 
mica plates and half-way between them. The strips 
were fastened to the mica plates by ebonite washers and 
small screws, shown in fig. 4, which gives a general 
view of a resistance. The plates were arranged so that 
they could be separated or put together quickly by 
removing two screws at either end, clamping the ebonite 
strips together. Two jjtench copper-wires were passed 
through holes bored for them through the splits in the 
ebonite strips at each end, in such a way that they 
were clamped in place by the ebonite. These heavy 
wires, when in place, connected the space between the 
mica plates with the outside of the frame-work. At 
each end the two wires were bent at right angles so as 
nearly to meet, and were inserted and soldered into 
holes made for them on the opposite faces of a small 
copper block. Heavy copper-wires (^-inch) were soldered into holes in these copper 
blocks and bent into an inverted U for connecting to the commutator-bridge. 
