240 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
water already in the bottles. The correction to the fundamental interval of Thermo¬ 
meter E is the same as for the last series. 
Series 4. Nos. XXXIII. to XXXV. 
This series includes experiments with Calorimeter C, using the flattened platinum 
strip for heating wire, instead of the central wire conductor and elastic strip. These 
observations are summarized in Section 5. The use of the flat wire was found to 
produce more irregularity in the heat-loss between the different flows, especially 
apparent on the small flows, and no doubt occasioned by the fact that the wire 
touched the sides of the tube. The correction for the fundamental interval of 
Thermometer E is that given under Group III., in Section 3, c. 
The agreement of the value of J with the other measurements is very satisfactory, 
and the more so because the heating conductor was changed not only in form but in 
resistance. 
Series 5. Nos. XXXVI. and XXXVII. 
Ex]i>eriments with Calorimeter E with flat heating strip. The fine-bore tube was 
slightly over 3 millims. in diameter, and was the largest tried in these experiments. 
The first experiments with this calorimeter were made with the central wire conductor 
and large rubber cord, but were neither satisfactory nor consistent. The effect of the 
streamdine motion apparently began to come in, with the helical motion in the water, 
probably from the size of the flow-tube compared to the size of the heating-wire. 
The thermometer was the same in every respect to that used in Series 4. 
Series 6. Nos. XXXVIII. to XLVIII. 
In this series, Calorimeter C was refitted with central solid platinum heating-wire 
with silk-covered rubber cord. Thermometer E, Group III., was still used. Measure¬ 
ments from 20° to 0° were made, and the values obtained under Series 1 and 2 
completely verified. The observations were extended below 4°, and the lowest point 
obtained was for a rise of temperature of 1° above 0°. In this experiment only one 
flow was obtained, but the value of d may be calculated with some degree of 
approximation by assuming the value of the heat-loss for the two other determina¬ 
tions with inflow-water at 0°, and correcting for the temperature coefficient of the 
radiation. This experiment was done principally to test the rapid increase of the 
specific heat at the freezing-point. An attempt was also made to obtain the high 
points, but with no more success than in the previous attempts. 
Series 7. Nos. XLIX and L. 
Between Series 6 and the present series, various devices were tried to obviate the 
effect of dissolved air in the boiled water. The plan was finally adopted of preparing 
