D4 REPORTS ON THE STATE OF SCIENCE. 
that point, and simultaneous measurements of the pressure at the begin- 
ning and end of compression are also made. A small correction for time- 
lag in the wire has to be applied; this amounts at a speed of 250 revs. 
per minute to about 6° C. at the completion of suction, and is negligible 
at the top of compression. The value of the volumetric heat calculated 
from the pressures and temperatures at these two points, on the assump- 
tion that the compression was adiabatic and that y was constant, is 
20°33 foot. lb. per cubic foot, the values in different experiments ranging 
from 19°76 to 20°95. The true value, according to Swann, for the range 
of temperature employed (20° to 270°) should be 20°1. Tt would appear, 
therefore, that the compression at the centre of the combustion space in 
an engine of this size (the diameter was 7 inches and the stroke 15 inches) 
is very nearly adiabatic, and there is reason to suppose that the method 
may yield good results when applied to higher temperatures. The ex- 
periments were done at various speeds, ranging from 60 to 250 r.p.m., 
and the fact that the values obtained at 60 r.p.m. are not systematically 
greater than those at higher speeds is further evidence that the 
loss of heat is small. The method has the advantage that it is in- 
dependent of leakage. Hopkinson found that in this engine at the top 
of compression the temperature of the air at a distance of half a centi- 
metre from the wall was about 30° less than in the centre. At points 
nearer to the wall, that is, within 1 mm., the temperature fell off very 
rapidly; it was still, however, decidedly above that of the wall at a 
distance of only 4mm. He is continuing his experiments with a view 
to giving a complete account of the temperature distribution and elucidat- 
ing more completely the phenomena of heat-flow. The general result, 
so far as he has gone, would appear to be that the layer of air in which 
the temperature gradients are considerable is extremely thin. It may, in 
fact, be difficult to decide precisely what is the nature of the film which 
determines the heat-flow. Once in solid metal it is quite certain that 
the temperature variations are extremely slight, and cannot be such as 
materially to affect heat-flow. On the other hand, the temperature of 
the air a fraction of a millimetre away from the wall is very much higher 
than that of the metal and approximates to the mean temperature of the 
ges. It is the temperature gradients in the composite layer of matter, 
partly air and partly solid or oil film, between these points which finally 
determine the heat-flow. It is possible that Clerk, who regards the 
solid and adherent film ! as the seat of the temperature changes, and Hop- 
kinson, who ascribes the action to a thin layer of air, are really dealing 
with the same thing from opposite points of view. 
(3) Explosion Haperiments. 
Much light has been thrown on the chemical processes in such ex- 
plosions as occur in the gas-engine by a very complete study made by 
Dr. Watson of the thermal and combustion efficiency in a petrol motor.” 
1 In later experiments on the compression and expansion of air in an ‘ R’ engine, 
with a very large flow of water through the jacket, Clerk finds that the heat-flow 
on compression and expansion becomes nearly proportional to relative temperatures. 
This appears in Clerk’s view to support the hypothesis of an adherent film which 
in some conditions of experiments accumulates heat and rises in temperature. 
2 Proc. Inst. Auto. Hng., May 1909. 
