BETWEEN THE FREEZING AND BOILING-POINTS. 225 
Sec. 5.— Experimental Proof of the Theory of the Method. 
In Section 2 it was shown that the conditions to be studied in the general 
difference equation of the method were the relations of the heat-loss to the rise of 
temperature and to the flow. In the present section I wish to summarize the 
different experiments which have a more particular bearing on the theory of the 
method. In the first place in regard to the question, which arises in all experiments 
where a quantity of water is heated by an electric current conveyed in a wire, of the 
excess temperature of the wire over the water, it may be said that in the present 
method the measurement of the electrical energy is completely independent of any 
value to be assumed for the resistance of the heating-wire, and not only that, but 
owing to the steady temperature conditions inside the apparatus, no uncertainty of a 
change in resistance in the wire with a change in temperature is introduced. When 
the temperature inside the calorimeter has arrived at a steady state, only such 
energy is used in warming the water as is supplied to the calorimeter by the electric¬ 
heating current. The fact that the results were completely independent of the 
resistance of the heating-wire was shown by using heat-wires of very different 
resistances. 
In regard to the insulation of the platinum heating-wire and of polarization and 
similar effects, it was considered that these played no part in the results. The 
resistance of the water column through which the heating-wire passed was 
enormously high and equal to a column of water 50 centims. long and 2 millims. in 
diameter, hence in comparison with the resistance of the central heating-wire, which 
varied from '4 to '8 ohm, was quite negligible. This is true even if it is admitted 
that the conveyance of the electric current by the water itself could have produced 
any error on the final result. 
Polarization by the naked wire in the water, I am satisfied, did not take place. 
Not the slightest trace of gas was ever generated in the calorimeter which could not 
be referred to the liberation of air in the water, and this was verified by watching 
the column of water in the fine tube when the calorimeter was removed from the 
water-jacket and a large electric current passed through. The effect of reversing the 
electric current in the apparatus, and making it flow either with or against the water 
flow, was tried in some of the earlier experiments, but it was found to produce no 
effect on the heat-loss as measured by the difference between the electrical and 
thermal measurements. The effect, if any, on the electrical readings was entirely 
negligible. 
The first experiments which were tried to test the method, were on the relation of 
the heat-loss to the flow, and were made by varying the flow over a wide range and 
at the same time keeping the inflow temperature and rise of temperature constant. 
The following list of calorimeters, used in the present work, will aid in describing 
these and subsequent experiments :— 
VOL. CXCIX.—A. 2 G 
