172 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
cells X 3 -f- X 10 in series, the value 2'8684, pitmded they were equal to the true 
value. We have seen, however, that these two cells differ from the most probable 
true value of the Clark cells made in this laboratory by 00014 volt, which would 
give for the true value of X 3 + X 10 2-86826., or 1-43413 each. Reducing this value 
by ’066 per cent., we have, as the value of each of my Clark cells, 
1 "43318 iut. volt, 
and the most probable true Clark-cell value 
1-43325 int. volt at 15° C. 
which is in remarkable agreement with the absolute measurements of Carhart and 
Guthe for their Clark cells. 
From the ratio of the Clark to cadmium, the value of the cadmium cell is found 
to be 
1-01895 int. volt at 20 C. 
Sec. 3b.—Measurement of Resistance. 
Next in importance to the value of e for the Clark cells, which we assume for 
the calculation of the absolute value of the mechanical equivalent of heat, is the 
value to be assigned to R for the resistance used in these experiments. 
At the outset we were exceedingly fortunate in having the laboratory equipped 
with a large number of 1-ohm resistance standards certified by the Electrical 
Standards Committee of the British Association, which were obtained in 1893. The 
work, therefore, of standardizing the resistances which were made for the present 
series of experiments was reduced to a minimum by the facility with which they 
could be compared to these standards on a Nalder type of Carey-Foster commutator- 
bridge. This bridge was supplied with a set of ratio-coils and bridge-wires which 
could be interchanged at will, and selected to be comparable in size with the 
resistances compared. During the first experiments which we made on the specific 
heat of mercury and the early trial experiments with the water calorimeter, the 
electric heating current was passed through a 1-ohm manganin coil for standard, 
which was immersed in paraffin oil. The difference of potential across the terminals 
of this specially-constructed resistance was of the same order as that across the column 
of liquid in the fine flow-tube in the mercury-calorimeter, and also equal to that across 
the platinum heating-wire in the first water-calorimeter. This was arranged for 
convenience in balancing on the potentiometer. 
The coil was made from two manganin wires, 1 millim. in diameter, connected in 
parallel and wound on an ebonite frame. Connections were made to the coil at the 
bottom of the frame, which was held vertical, by two heavy copper-wires, ^ inch in 
diameter, so arranged as to have 3 or 4 inches immersed in the oil-bath with the coil. 
