CONTINUOUS ELECTRIC CALORIMETRY. 
117 
arranged two in parallel, and 19 in series, giving a resistance of nearly 3 th of an 
ohm, capable of carrying a current of upwards of GO amperes without excessive 
heating. The maximum currents used in this experiment were about 8 amperes. 
The rise of temperature of the platinum conductor in each case was found to be 
nearly proportional to the watts expended. This was verified by varying the 
number of cells employed, and keeping the water-flow constant, for each arrange¬ 
ment of the conductor to be tested. The superheating of the conductor was 
estimated by deducting from the observed rise of temperature half the calculated 
rise of temperature of the water, making a suitable allowance for the heat-loss. The 
value of the water-flow was nearly half a gramme per second in all cases, and the 
results were reduced to a value of Q 0 = 5 calories per second to render the 
experiments with the different conductors strictly comparable. The actual heat-loss 
from the flow-tube could not be measured by this comparatively rough method, but I 
made some attempts to obtain comparative estimates of the temperatures of the 
outside of the flow-tube by winding a platinum wire round it, covering the spiral 
with flannel, and observing its resistance. These measurements could not lay claim 
to any accuracy, but were useful as an indication of effects to be expected. 
Table YI.—Superheating of Stranded Conductor. 
Form of Conductor employed. 
Conditions of Flow. 
Superheating. 
Five strands, irregular. 
. . Mixed 
0 
2-8 
Same, annealed and straightened . 
Linear 
6-5 
' ., twisted into a rope .... 
Linear 
8-5 
,, spiral fitting tube .... 
. . Mixed 
3-0 
In the first case the wire was taken as it came from the reel. The strands were 
well separated, and crossed each other irregularly, so that the colour-band was 
completely broken up and mixed to a uniform tint in a space of 10 or 15 centims. 
The flow was not precisely turbulent or eddying, but the stream-lines were so quickly 
sub-divided and mixed that the same effect was produced. When the wire was 
annealed and straightened, # the colour-band remained practically unbroken from end 
to end of the tube, but as the strands were still separate, the heat was more or less 
distributed over the cross-section. The twisting of the wire into a rope in the third 
experiment diminished the effective surface and increased the superheating, the 
value of which closely approached that calculated for a central conductor in a 
2 millim. tube assuming the conductivity of water ‘0016 C.G.S. The great 
diminution of the superheating in the fourth case, in which the wire was wound into 
* It is probable that the failure of Dr. Barnes to obtain consistent results with a stranded conductor 
may have been due to the use of annealed wire, which would inevitably become straightened in fitting up 
the apparatus unless special care were exercised. I was not aware of this mistake at the time. 
