power went up, past 4, 8 and 12, to 25 or 50 horse power; and in the
exhibition of 1889 there were gas engines seen in which the explosion
chamber had a diameter of as much as 23 inches.

But the price of coal gas seemed to be too high for use in these large
engines, in which sizes steam is comparatively cheap; and so poorer gas,
which, though possessing only about 28 per cent. of the heating power,
is still cheaper in proportion than coal gas, when it is made on the
spot, was introduced to tide over the difficulty. Difficulties have been
successively overcome, with the result which we have just seen, namely,
1.37 pounds of anthracite per effective horse power, or about half the
carbon which a steam engine of the same power of excellent design, and
well kept up, would consume. A 50 horse simplex at Marseilles, in
Barataud's flour mill, is said to have run for the last 2 years on 1.12
pounds of English anthracite per effective horse power; and thus M. Witz
says his predictions of 10 years ago, that the gas producer would some
day replace the boiler, are being verified in such a way as to surprise
even himself.

But the objection is stated, and it is a serious one: the weight of fuel
is not the only thing to be considered. The steam engine uses coal, the
producer requires English anthracite, which is dearer; the gas motor
uses a great deal of water and a great deal of oil, which cost money;
and gas motors are dear, while gas producers and their adjuncts cost a
tidy bit of money, and wear out pretty fast. Is not steam, after all,
more economical in the long run? Besides, producers are bulky and take
up a great deal of space; the weight of fuel is only one element in a
complicated problem.

In order to study the grounds of this objection, M. Witz has instituted