1921.] Marsden and Fenton.—Coal, Gas, and Electricity. 
267 
It will be seen that, neglecting the question of labour, coal is three and 
a half times more economical than either gas or electricity. The economic 
advantage is still further increased if, by draught-regulation and a suitably 
placed fire, advantage can be taken of the heat conducted to other rooms 
from the flue. 
As regards gas and electricity, if we consider the greater convenience 
and labour-saving of the latter it is probably, on the whole, slightly more 
economical. Sufficient power, however, must be available to heat a room 
rapidly, and, in general, 1 to 1-5 watts per cubic foot of space must be 
provided, depending on climatic conditions and on the number of outside 
walls of the particular room. 
VII. Efficiency of various Methods of Lighting. 
Tests were made of the cost of lighting by gas, electricity, &c., and, 
although these tests were not so exhaustive as might be desirable, yet the 
results are conclusive indication of the superiority of electricity for lighting 
purposes, apart altogether from the economy of the latter as regards labour. 
The results are as follows :— 
Source. 
Price. 
Cost per Mean Hori¬ 
zontal Candle-power 
per Hour. 
Electricity 
ljd. per unit 
d. 
0*0015 
Electricity 
6d. per unit 
0-006 
Gas (Welsbach) 
8s. 4d. per 1,000 cubic feet . . 
0-015 
Kerosene 
3s. per gallon 
0-04 
Candles 
Is. 4d. per pound 
0-195 
In the case of electricity the figures are based on an efficiency of 1 watt 
per candle-power. So-called “ half-watt ” lamps of less than 100 watts 
generally average about 0-85 watt per candle-power, while vacuum-type 
lamps are generally about 1*6 watts per candle-power. 
In the case of gas, a fair average upright Welsbach mantle was used. 
The cost per candle-power hour for a batswing burner is about twenty-five 
times that for a Welsbach mantle, depending on the amount of air in the 
gas. It is worthy of note that a report of the Institution of Gas Engineers 
states that incandescent gas-mantles with low-pressure gas require under 
best conditions 27*5 B.Th.U. per candle-power hour.* This would corre- 
27.5 
spond to cubic feet per candle-power hour, or 0-006d. It appears, 
therefore, that we could have obtained a more efficient burner. 
The candle-power measurements in the above experiments were referred 
to a special incandescent electric-lamp standard compared by the National 
Physical Laboratory with the International standard. The figure for 
kerosene was obtained from tests on two medium-sized kerosene-lamps of 
5 candle-power. No doubt a more extensive and careful selection may have 
revealed a more efficient lamp, but the figure given is a fair indication of 
average conditions. 
* See D. Clerk and Professors Smithells and Cobb, Report to the Institute of Gas 
Engineers on Goal-gas and Electric-supply Industries of the United Kingdom, 1919. 
