218 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
graduated to ‘1° and reading to Youths. Its error was determined by comparison 
with the standard thermometer, and its readings were found to be "03° too large. The 
standard thermometer was a Geissler reading to Tooth between — 2° and 50° C., 
which has been in the possession of the laboratory for several years, as before stated. 
In 1896 both Professor C allendar and I separately determined its error, and reduced 
its readings to the nitrogen scale over the entire range by comparison with a platinum 
thermometer. The error was very consistent, and showed its readings too high by 
•11° to ‘12° from 0° to 50°. This thermometer was used in the Clark-cell bath and 
lias already been referred to in that section. For jacket-temperatures between 
50° and 90° C., a second Muller thermometer, reading between 50° and 150°, was 
used, which was graduated to -g-th of a degree. 
Specimen Tables of Observations. 
\_Added April '20th, 1901.—H.T.B. Owing to the necessity of condensing the 
tables since the communication of this paper, I have considered that it would be of 
advantage to give specimen tables of observations as recorded during a complete 
experiment. I have therefore included here two sets, made as typical as possible, 
which illustrate more clearly the remarkable steadiness of the conditions. The 
experiments selected are those given under date October 27, 1899, and March 10, 
1900. They include two calorimeters and different-sized flow-tubes (Calorimeters C 
and E), as well as flat and round heating wires. One of the sets includes the 
observations taken for the measurement at the highest point of the range. In the 
first set box 1 was used, on the second set box 2. The order in which the readings 
were obtained has been described in Section 4. In all cases the time of taking the 
readings was as closely as possible on the even minute. The time for the start and 
finish of an interval, during which the flow was measured, is of course given, as 
recorded on the chronograph, to ’01 second. The reading of the contact point on the 
bridge-wire (b.-w.) is given in centims. (10 centims. = 1 C C.), and the deflections of 
the galvanometer noted, it being accurately set to the nearest millimetre by means 
of the vernier reading to hundredths of a millimetre. The sensitiveness of the 
galvanometer remained very constant, but was repeatedly checked during a set of 
observations. The balance-point on the bridge-wire was calculated by interpolation 
from the observations of the deflection. In the electrical readings of the potential, 
S stands for the balance-point for the difference of potential across the standardized 
resistance, and P for the same across the calorimeter heating-wire. These readings 
are of course uncorrected for the errors in the Thomson-Varley slide coils. The 
temperature of S is that of the oil in which the standardized resistance was immersed, 
and the temperature of the Clark cells is here uncorrected for the thermometer error. 
The inflow temperature is the same as the temperature of the jacket. In some cases 
