258 _ REPORTS ON THE STATE OF SCIENCE. 
petrol as 1 part by weight to 19 parts by weight of air up to 1 part petrol 
tq 10 parts air. : 
The efficiency rises from A to B and falls from B'to D. The line 
C to E starting from the point C on BD gives the efficiencies calculated 
on the basis of the heat liberated during the chemical changes which. 
actually occur. The efficiency rises from point A to point B, the lower 
efficiency at point A being due to slow inflammation of the mixture for 
the speed of the engine, so that burning is continued after the constant 
volume phase is passed. B is the point of maximum efficiency, but it 
does not appear to be the point of maximum chemical combination ; the 
most complete chemical combination is found at the point C. The fall 
of efficiency is doubtless due to the continually increasing flame tempera- 
ture between B and C, that is between a mixture of 1 part by weight 
of petrol to 17°3 parts of air and 1 part of petrol to 14 parts of air. The 
increasing flame temperature with the richer mixture increases the heat 
THERMAL EFFICIENCY 
™ 
os 
me 
be 
f { 
Be IN 
Reb fes a0 hae 
Ratio of Air to Petrol. 
ww, 
FIG.4. 
loss and also increases the mean specific heat of the gases over the range 
of the temperature used, and so efficiency is diminished from both causes. 
From C to D the petrol is in excess, and hence the flame temperature 
must on the whole be falling. At the point D a very large quantity, 
however, of the fuel is being discharged incompletely burned, as only 
64 per cent. of the heat of the fuel is liberated. Calculating, however, 
the line CE, which is the line giving the thermal efficiency of the 
engine for the chemical action actually completed, it is somewhat sur- 
prising and interesting to find how greatly efficiency increases. With 
1 of petrol to 10 of air it rises as high as 0°289. For these tests the 
air standard efficiency was 0.46, so that the efficiency ratio was 0°64—an 
extremely high efficiency for so small an engine.* 
1 The portion CE of the curve is calculated by deducting from the calorific 
value of the petrol the heat which would be liberated if the CO, H, and CH, in the 
exhaust were burnt to CO, and H,O. There is some uncertainty in the result 
because there is undoubtedly some combustible matter in the exhaust which is not 
accounted for in the analyses. Moreover, in Dr. Watson's experiments the H and 
. 
