Portable Agricultural Steam Engines at Newcastle. 715 
The last column in the balance-sheet shows the percentage 
which each source of loss bears to the total amount of heat 
generated. Heating the fuel, using the word in its ordinary 
sense, and the air necessary for its combustion, and displacing 
the atmosphere (items 4 and 5) take 6^ per cent. ; while the cost 
of dealing with the excess air amounts to 6^ per cent. The 
loss by cooling is, however, the most serious of all, and, although 
this engine was, as regards the usual parts, lagged with 
exceptional care, amounted to 9^ per cent. 
The losses which cannot be certainly accounted for amount 
to less than 3j per cent. A portion of these were probably due to 
the increased rate of cooling while the engine was at work, for 
the cylinders, piston rods, valve, spindles, and the working 
parts generally, were hotter, and therefore emitted more heat 
than when at rest. 
The process by which this balance sheet has been obtained 
is sufficiently complicated. There is a simpler method of 
judging of the efficiency of an engine based upon the prin- 
ciple to which we have already alluded, namely, that the 
proportion of useful work to be got out of a heat-engine 
depends upon the proportion which the fall of temperature 
bears to the original absolute temperature. In the case of a 
boiler, the difference between the temperature of the furnace 
and that of the smoke-box represents the fall of temperature, 
while the initial absolute temperature is that of the furnace. 
Unfortunately, we have, as yet, no trustworthy pyrometer 
for ascertaining the temperature of a furnace. Attempts were 
made to determine this by means of a Wilson's pyrometer. A 
ring of iron, of known weight, was buried in the glowing coal, 
and, when thoroughly heated, was quenched in a given weight 
of water ; the rise in temperature of the water was oberved, and 
in this manner, if the specific heat of iron at various tempera- 
tures were accurately known, it would have been easy to calcu- 
late its temperature at the time of quenching. Could this heat 
have been ascertained, it is by no means clear it would have 
represented that of the fire, as this was so very thin that it was 
impossible to be sure that the iron ring was perfectly covered, 
and had really attained the temperature of the fuel, and, in 
addition, as we have just said, there is much uncertainty as to 
the specific heat of metals, and especially of iron at high 
temperatures. Under these circumstances, the attempt was 
abandoned. An approximation may however be arrived at in 
the following way. Setting aside the wood used for kindling 
the fire, the total weight of matter heated in the furnace per 1 lb. 
of coal was 24*45 lb. The specific heat of the coal, the air, 
