632 REPORT—1899. 
his formula for the higher points, with such modification only as is necessary to 
make it fit with the observations at lower temperatures. It happens that the 
rate of variation of the specific heat given by Regnault’s formula agrees with that 
given by the formula (CB) between 55° and 60°. The two formule can therefore 
be very accurately fitted at this point by the simple expedient of subtracting a 
constant quantity from the values given by Regnault’s formula at temperatures 
above 60°C. ‘This apparently arbitrary method would not be suggested if it 
were not that it leads to results which are intrinsically most probable, and which 
require the simplest modification of existing tables. 
If the formula (CB) is adopted for the range 0° to 60°, over which it has been 
accurately verified, and the formula of Regnault (corrected as above explained) 
from 60° to 100°, the ratio of the meau specific heat between 0° and 100° to the 
specific heat at 20° is 1:0014. Taking Rowland’s value as 4:181 joules at 20°, 
this ratio would give 4:1868 for the mean specific heat, which exceeds the value 
found by Reynolds and Moorby by less than one part in a thousand—a discrepancy 
so small as to be within the limits of possible error even in the case of these two 
extremely accurate determinations. Since it is a work of great labour and difficulty 
to redetermine the specific heat at temperatures above 100°, and since it is 
extremely unlikely that more accurate results over this part of the range will be 
forthcoming in the near future, it has appeared desirable to adopt this basis for the 
construction of the annexed table of the variation of the specific heat of water 
over the whole range 0° to 220°C. The general effect of the table is to diminish 
the extent of the variation hitherto assumed, but it is believed that the results here 
tabulated are within the limits of possible error of all the best experiments. The 
order of agreement may be inferred from a comparison of the values of A, the total heat 
of the liquid, given in the lasttwocolumns. The agreement with Rowland is within 
1 in 3,000 between 10° and 40°, and with Regnault within 1 in 1,000 at 160°C. 
The variations of Regnault’s individual observations exceed 5 parts in 1,000. 
The values of the total heat 4 are found by integrating the specific heat from 
0 to ¢, according to the formula. The formula does not represent the rapid change 
of s near the freezing-point, but accurate account may be taken of this, when 
desired, at any point above 10°, by adding the constant quantity 0:020 to the 
value of 4 as givenin the table by the formula. This correction, however, is seldom 
of importance. 
3. On the Expansion of Porcelain with Rise of Temperature. 
By T, G. Beprorp.—See Reports, p. 245. 
4, Interim Report on Methods of Determining Magnetic Force at Sea. 
See Reports, p. 64. 
FRIDAY, SEPTEMBER 15. 
The following Reports and Papers were read :— 
1. Report on Electrolysis and Hlectro-Chemistry. —See Reports, p. 160. 
2. On the Energy per Cubic Centimetre in a Turbulent Liquid when 
Transmitting Laminar Waves. By Professor G. F. FrtzGEratp, 
F.RS., Trinity College, Dublin. 
In the ‘Phil. Mag.’ vol. xxiv. p. 342, October 1887, Lord Kelvin has given 
equations for the transmission of laminar waves through a turbulent liquid. He 
expresses doubt as to the possibility of any turbulency being possible to which 
