retorts I made as much gas as would have required 42 retorts to produce
on the old system. But I know that even this can be excelled; and I am
aware that there are works where, by the introduction of retorts
measuring 21 by 15 inches, instead of 18 by 13 inches--and which, I may
say, can be put quite easily into the same arch--a production of 12,000
cubic feet per mouthpiece can be obtained. This will, of course, still
further reduce the cost of production.

With such an experience, gentlemen, I think it is almost needless for me
to add that I am a strong advocate of the regenerative system. I have
often heard it asked, "But can the system be profitably adapted to small
works?" In answer to this, I will say I have proved that it can. During
last summer the manager of a small gas works in my neighborhood called
on me regarding the working of this system, and expressed a desire, if
it was at all possible to adapt it to his present settings without much
expense, to try it. I must say I admired his progressive spirit and
pluck; and, after a somewhat lengthy conversation with him, during which
I gathered the full details of his working and his requirements, I
determined to encourage him in his desire to prove if it could be
successfully applied to a works of the size mentioned. The present
setting consisted of three [semicircle] retorts in one arch; and one of
his stipulations to me was: "You must so contrive the setting that if it
should prove a failure I can reconvert it into the old system in a few
hours." I at once saw that the stipulation was reasonable, or he might
be caught in a fix in midwinter. But, with true "Scotch caution" and
forethought, he was, while anxious to experiment, determined not to be
"caught napping." After some consideration, I prepared a sketch for him
of how I thought it could be done, and at the same time comply with his
stipulation; and having received full explanations, he set about it, and
has had it working now for something like six months. His experience has