THEORY OF THE STEAM ENGINE. CXLIII, 
3, cup heale 
d At 
pe ES 
4 Q=¢ avs 
Ke 24 
5, 0 Wim py Wi T. 
2 C (r—1)™ 
ad N* 
The want of harmony in these various formulz is astonishing. 
Cylinder condensation is made proportional, as regards speed to 
-. 1 ‘ : 
Y wr wnt and wr 38 regards ratio of expansion to log, 7, rt 
and (r—1),™ and similarly as regards the other conditions, 
Such disagreement would be brought about by the neglect of 
the third of the above mentioned requirements, even if the first 
two were satisfactorily complied with. It by no means follows 
as has been assumed by the authors of most of these formule that 
the relationship between initial condensation and any one work- 
ing condition, say the speed, is even approximately the same when, 
say, the boiler pressure is at 80 Ibs. per square inch as when it is 
at 160 ibs. per square inch. This makes it necessary in order to 
determine the relationship between cylinder condensation and one 
particular working condition, not merely to carry out one series 
of tests, but to carry out such a series again and again, each time 
altering the value of one of the other working conditions, in order 
to determine the effect which such a change would have on the 
form of the function connecting cylinder condensation and the 
first mentioned condition. If we suppose r, W and P to be the. 
ratio of expansion, speed and initial pressure respectively, then it 
is not true as is often assumed that 
L=f (r)f' (N)Ff" (P) 
where Z is the cylinder loss, and f, f’ and f” are constant func- 
tions of their respective conditions, but that 
L=F(r,N,P,) 
where Fis a function involving all the variable working conditions. 
