of coal, taking good years together with bad, must cover the costs at
which these mines can produce. If the price rules lower than this,
sooner or later they will close down, and we will be left with a
smaller number of mines, among which great variations of conditions
will still prevail. Once more, the price must cover the cost incurred
by the least profitable of these remaining mines, unless their number
is still further to be diminished. Thus we can conceive of a "margin
of production" which will shift backwards to more profitable or
forwards to include less profitable mines, according as the demand for
coal contracts or expands. But, wherever this margin may be, there is
no escaping the conclusion that it is the cost of production of the
"marginal mines," of those that is to say which it is only just worth
while to work, to which the price of coal will approximate.

It follows that there is no real connection between price and cost of
production throughout the industry as a whole. It follows incidentally
that those concerns which can market their coal at an appreciably
lower cost than the marginal concerns, are likely to reap more than an
ordinary rate of profit, though royalties may absorb part of the
excess.


§2. _The Various Aspects of Marginal Cost_. This relation cuts much
deeper than the particular system under which the mines are at present
owned and worked. If, for instance, we supposed that the various
mines were amalgamated together in a few giant concerns, each of which
comprised some of the richer and some of the poorer mines, the
preceding argument would need to be recast in form, but its substance
would be unaffected. For though a great coal trust could in a sense
_afford_ to sell at a price lower than the marginal cost, setting its