CXLIV, 8. H. BARRACLOUGH. 
The amount of labour necessary for a complete determination 
of this function is very considerable, as may be seen by consider- 
ing a particular case. Let it be supposed that it is desired to 
make an investigation of the effect which changes of pressure, 
speed and ratio of expansion have upon the heat loss by initial 
condensation in some particular engine, and that five tests are 
considered sufficient for each series. The following table may 
then be taken as giving the particular values chosen. 
Pressure Ibs. abs. Speed R.P.M. Ratio of Expansion. 
20 200 10 
100 160 8 
80 120 6 
60 80 4 
40 40 2 
Tt is at once evident that the number of engine tests necessary 
te carry out this investigation completely, is 125, and if a fourth 
working condition were to be introduced into the problem the 
number would be increased five times. As the carrying out and 
working up of an engine test is a matter involving a considerable 
expenditure of both time and labour, we have doubtless in the 
large number of such tests necessary, an explanation of the 
unsatisfactory nature of the experimental investigations of the 
heat losses hitherto attempted. In conjunction with Mr. L. 8. 
Marks, the writer initiated such a series of tests,! as is here Sug- 
gested, on a Corliss engine at Sibley College, U.S.A. The time 
at their disposal only allowed of the complete investigation of the 
effect of two conditions—speed and pressure, but the result of 
the investigation was to support the above opinion entirely. The 
heat loss was found to vary as the reciprocal of some power of N, 
the number of revolutions per minute, but the index of this power 
depended on the pressure, so that the formula for “ cylinder con- 
densation” when both speed and pressure were subject to variation 
was of the general form 
: ~ Wie) 
1 Proc. Inst. C.E., Vol. cxx., p. 323. 
L 
