230 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
“cold” readings of the differential thermometers, when in place in the calorimeter, 
for the different flows at different parts of the scale. 
The three “cold” readings at 60° are, for the largest flow 23"681, for the next 
23‘582, and for the lowest 23"544. A decrease in the bridge-wire reading means 
that the inflow thermometer is at a lower temperature than the outflow. The heat 
conduction at the inflow end, through the copper wires leading in the electric current, 
which depends only on the difference existing between the temperature of the 
calorimeter water and the temperature of the laboratory, can affect the temperature 
of the inflowing water less for the largest flow than for the smallest flow, and 
therefore explains the difference in the “ cold” readings for the different flows. For 
the flows used in the present work, the effect of the conduction at the outflow end on 
the temperature, as indicated by the outflow thermometer, must have been very 
small indeed, even at the highest points of the range. All of the “ cold” readings at 
the high points are slightly less than the interpolated reading from the differential 
ice and steam-points. For the “cold” readings at the ice-point, the effect was, as 
might be expected, reversed, and conduction of heat into the calorimeter from the 
laboratory took place. 
The readings given in Table XLIV., on November 18, 1899, are— 
For a flow of ‘59 gramme .... 23’330 
„ „ -39 „ .... 23-392 
on November 22, 1899, Table XLY.— 
For a flow of -62 gramme .... 23-329 
„ „ -37 „ .... 23-398 
The peculiar exception to this seems to be in the experiment made at the ice-point 
on March 24, 1899 (Table XV.), when the “cold” reading for the small flow was 
lower than for the high. I have not an exact record of the conditions under which 
this experiment was taken, but it is possible that the outflow end may not have 
been properly lagged, which would produce the effect indicated by the readings. As 
the observations, when reduced, give such a consistent measurement of the specific 
heat, even in the face of this apparent exception, additional evidence is given of the 
necessity of the “ cold ” readings to render the results independent of extraneous 
conditions. The “cold” readings for experiments with the jacket water at the 
temperature of the laboratory were, as a rule, all at the same point on the bridge- 
wire, and identical with the interpolated reading from the differential ice and steam- 
points. 
The radiation loss increases only very slightly as the temperatures of the calorimeter 
and jacket are raised, and this is of course on account of the temperature of the 
calorimeter flow-tube being always the same amount above the jacket water at all 
