n the distribution of power may be taken the
instance of air-drills. Although the work upon which the drill is
employed can be known, the power required for compression usually
comes from a common power-plant, so that the portion of power debited
to the air compressor is an approximation. The assumption of an
equal consumption of air by all drills is a further approximation.
In practice, therefore, many expenses are distributed on the theory
that they arise in proportion to the labor employed, or the machines
used in the various departments. The net result is to level down
expensive points and level up inexpensive ones.

The third sort of limitation of accounting difficulty referred
to, arises in determining into which department are actually to be
allocated the charges which lie in the borderland between various
primary classes of expenditure. For instance, in ore won from
development,--in some months three times as much development may
be in ore as in other months. If the total expense of development
work which yields ore be charged to stoping account, and if cost
be worked out on the total tonnage of ore hoisted, then the stoping
cost deduced will be erratic, and the true figures will be obscured.
On the other hand, if all development is charged to 'capital account'
and the stoping cost worked out on all ore hoisted, it will include
a fluctuating amount of ore not actually paid for by the revenue
departments or charged into costs. This fluctuation either way
vitiates the whole comparative value of the stoping costs. In the
following system a compromise is reached by crediting "development"
with an amount representing the ore won from development at the
average cost of stoping, and by charging this amount into "stoping."
A number of such questions arise where the proper division is simply
a matter of opinion.

The result of all these limitations is that a point in detail is
quickly reached where no further dissection of expenditure is justified,
since it becomes merely an approximation. The writer's own impression
is that without an unwarrantable number of accountants, no manager
can tell with any accuracy the cost of any particular stope, or
of any particular development heading. Therefore, aside from some
large items, such detailed statistics, if given, are to be taken
with great reserve.

WORKING COST SHEETS.--There are an infinite number of forms of
working cost sheets, practically every manager having a syste