294 
illuminating gas. The Departmental Committee on 
the Manufacture and Use of Water Gas, 1899, recom- 
mended that the Board of Trade should have the power 
to limit the proportion of carbon monoxide in illuminat- 
ing gas to 12 per cent. or such higher value as should 
be considered safe. This recommendation was not, 
however, made statutory. The Departmental Com- 
mittee on Carbon Monoxide, 1921, reported “ that it 
is not necessary or desirable to prescribe any limita- 
tions of the proportion of carbon monoxide which 
may be supplied in gas used for domestic purposes,” 
though a suggestion was considered that a limit of 
20 per cent. of carbon monoxide might be imposed. 
Pure coal gas contains 6-8 per cent. of carbon mon- 
oxide, water gas contains about 4o per cent., and 
carburetted water gas about 30 per cent. Water gas 
is often added for economic reasons to pure coal gas, 
and the illuminating gas supplied to the public not 
infrequently contains a quite considerable proportion 
of carbon monoxide. Occasionally so much as 50 
per cent. of water gas has been mixed with the coal 
gas, with the result that the illuminating gas has 
contained 20 per cent., or slightly more, of carbon 
monoxide. As a rule, however, the proportion of 
water gas is considerably lower than this, and some 
companies still continue to supply pure coal gas. 
An escape of gas is likely to be noticed and quickly 
remedied during the daytime, but far greater danger 
arises at night when a’ person may fall asleep in an 
ill-ventilated bedroom without noticing that the tap 
of a gas jet has been accidentally left turned on, or 
disregarding as trivial an escape of gas from some 
faulty fitting. He may then be disabled before he 
has any warning of the danger, and when once dis- 
abled he may not be found till many hours have 
elapsed and it is too late to save him. From the 
experimental data recorded by Dr. J. S. Haldane in 
the report of the Water Gas Committee it is possible 
to calculate the concentration of carbon monoxide 
which will finally be attained in a room if there is a 
leak*of gas into it, making the assumption that the 
carbon monoxide becomes uniformly mixed with the 
air. Even in the most unfavourable circumstances, 
when there are no special openings to admit of ventila- 
tion, there is always a considerable interchange of air 
through the walls, roof, and floor of any room. Even 
if the outside air is quite still a volume of fresh air 
equal to the cubic contents of the room will gain 
admission, in 1:8 hours in a room of 500 cubic feet, 
in.2°3 hours in one of 1000 cubic feet, and in 2-9 hours 
in one of 2000 cubic feet, and these rates may be 
doubled if a strong wind is blowing or if the room is 
furnished with a fireplace the chimney of which is 
not blocked. Taking the most unfavourable case 
the following table shows what will happen in three 
different-sized rooms if there is a leakage of gas into 
these rooms at the rates respectively of 4 and 10 cubic 
feet per hour : 

Carbon monoxide percentage eventually reached in the 
room when the illuminating gas contains the following 
percentages of carbon monoxide. 
A. LEAK or 4 Cusic Freer rer Hour. 
Capacity of room 
in cubic feet. 
5%: 10%. 15%. 20%. 30%- 
500. . 0°07 o'r4 0°22 0°29 0°43 
1000. . 0°05 O09) |) osrgiaeee O18: 0°28 
2000. . 0°03 0°06 0'09 O12 Some 
NO. 2783, VOL. 111] 
NATURE 
[Marcu 3, 1923 

Carbon monoxide percentage eventually reached in the 
Capacity of room room when the illuminating gas contains the following 
in cubic feet. percentages of carbon monoxide, 
B. Leak or 10 Cusic Fert per Hour. 
5%. 10%. 15%. 20%. 30%. 
500+, . or8 0°36" 0°54 0°72 I'04 
OOO ee, OFZ 0°23 0°34 0°46 0°69 
2an0 . . 0°07 O'I5 0°22 0°29 0°43 
The dotted lines mark the division into fatal and non- 
fatal percentages of carbon monoxide on the assumption 
that a concentration of o:2 per cent. of carbon mon- 
oxide may prove fatal if maintained for several hours. 
On testing the rate of escape of gas from an unlit 
gas-jet when the tap was turned on fully, four different 
burners chosen at random gave the following results. 
Under a gas pressure of 4 inches of water a properly 
regulated universal type of incandescent burner with 
inverted mantle passed about 4 cubic feet of gas per 
hour, and a Bijou burner of the same type about half 
that quantity. Under pressures of 14 and 3 inches 
of water a No. 3 Bray flat flame burner passed about 
6 cubic feet and 9 cubic feet per hour, respectively, 
and a No. 5 Bray burner about 8 and rr cubic feet. 
Under the most adverse conditions, therefore, the 
risk of fatal poisoning would appear not to be very 
great in the case of escape from a single well-regulated 
incandescent burner so long as the proportion of carbon 
monoxide in the illuminating gas does not exceed 
20 per cent., save in rooms of very small cubic capacity, 
though temporary severe symptoms might be caused. 
The real danger evidently lies in a leakage of gas con- 
siderably greater than that which might be obtained 
from such a burner. The pressure under which gas 
is supplied differs very greatly in different localities 
(under the Gas Regulation Act a minimum pressure 
of 2 inches water gauge has now to be maintained in 
the gas mains during the night), and an ordinary flat 
flame burner through which the rate of escape of gas 
will vary roughly as the square root of the pressure 
may clearly become a source of considerable danger 
if the tap should be accidentally turned full on. Some 
flat flame burners of improved type allow, however, 
much less gas to escape than the figures given above. 
Still greater risk attaches naturally to * the grosser 
forms of leakage from unlit gas-rings, fractured pipes, 
accidentally disconnected unions and the like, pro- 
vided that the escaping gas becomes sufficiently mixed 
with the air in the room. The Water Gas Committee 
found that with a large escape of coal gas the gas 
might collect mainly at the top of the room, and if 
the air in the room remained undisturbed it might 
be long before feeble convection currents could estab- 
lish a fatal atmosphere at the level of a bed, but the 
case is, of course, quite different when the gas is rich 
in carbon monoxide. 
It must be remembered that the table above in- 
tentionally depicts the most disadvantageous condi- 
tions for the occupant of the room. The majority 
of bedrooms possess a fireplace, and windows and 
doors often do not shut very tightly. The natural 
ventilation in a room might easily be double that on 
which the table is based without any special provision 
for ventilation, and the figures shown would then be 
halved. Under such conditions it would probably 
require a leak into a room of 1000 cubic feet capacity 
of 10 cubic feet per hour of gas containing 15 per cent. 


