294 The N.Z. Journal of Science and Technology. [Jan. 
and rails. Locomotives having this maximum axle-load can easily be 
built for the 3 ft. 6 in. gauge. In fact, the outside cylinders that form the 
only possible arrangement for such large engines can be more easily accom¬ 
modated on this gauge than on wider ones. It is generally recognized that 
the largest American locomotives, the enormous Mallet compound type, 
could be better designed on the 3 ft. 6 in. gauge, since it is impossible to get 
sufficiently large low-pressure cylinders between the wheels and the loading- 
gauge limits on the 4 ft. 8J in. gauge. In England the narrower loading- 
gauge restricts the size of the outside cylinders to such an extent that 
designers have been forced into the regrettable expedient of adopting the 
complicated and expensive three- or four-cylinder type, whereas if the 
track-gauge there were 3 ft. 6 in. the simple two-cylinder type could be 
built. When the other extreme is reached we find that on the lightest 
lines locomotives of any reasonable power must have as many driving- 
axles as possible to get the necessary adhesive weight. This leads to fire¬ 
boxes between the coupled wheels, and if the gauge is less than 3 ft. the 
width of grate is reduced below 2 ft., and the resulting grate-area is too 
small for any reasonable efficiency. Taking all these points into consider¬ 
ation, I have no hesitation in saying that a gauge between 3 ft. and 3 ft. 6 in. 
is suitable for the heaviest locomotives and largest rolling-stock, while a 
gauge narrower than 3 ft. offers no advantages whatever for light railways, 
the slight saving in first cost being under 10 per cent, of the total capital 
expenditure, and this is more than offset by the increased working-charges 
if the traffic handled is anywhere near sufficient to return interest on the 
capital involved. 
Having decided that no possible advantage can be secured by a change 
of gauge, the next point at issue is the question of steam versus some other 
source of power. The varied power and speed requirements of railway¬ 
working are peculiarly unfavourable to the oil-engine, but, in any case, the 
fact that motor-oils are so suitable for internal-combustion engines will 
always keep their cost above that of coal when computed on a calorific 
basis. Oil can therefore be dismissed without further reference for loco¬ 
motive work, the more so as electricity is able to offer all the advantages 
of oil at a cheaper rate. In proceeding to compare the economic possi¬ 
bilities of steam and electricity, it is necessary to declare at once that the 
modern superheated steam-locomotive is a much more economical machine 
than is generally appreciated. As a source of power for relatively large 
units (say, over 300 h.p.) the locomotive engine and boiler stands alone 
for first cost, as even at the present time it can be put in commission for 
£8 per horse-power, as against, say, £30 for steam turbines, and £80 for 
electric generators driven by water-power. This is a fact that is not suffi¬ 
ciently appreciated, and any engineer can call to mind power plants where 
the load-factor is so low that the locomotive boiler and engine is a very 
attractive proposition. 
There is no doubt that the difficulty of obtaining reliable indicator- 
diagrams from a locomotive in service, and the lack of adequate testing 
plants, have caused the power and efficiency of the locomotive to be 
grossly underestimated. The usual method of “ begging the question ” is to 
calculate the horse-power by expressing it as a linear function of the 
heating-surface—say, the heating-surface in square feet divided by three, 
four, or five, as the calculator pleases—and this is the favourite method of 
those who trace the amusing curves comparing the torque-speed charac¬ 
teristic of the steam and electric locomotives. There is the less excuse for 
