e, these might always be brought within the pale of
the abstract science if it were worth while; and when we make the
necessary allowances for them in practice, if we are doing anything but
guess, we are following out the method of the abstract science into
minuter details; inserting among its hypotheses a fresh and still more
complex combination of circumstances, and so adding _pro hac vice_ a
supplementary chapter or appendix, or at least a supplementary theorem,
to the abstract science.

Having now shown that the method _a priori_ in Political Economy, and
in all the other branches of moral science, is the only certain or
scientific mode of investigation, and that the _a posteriori_ method,
or that of specific experience, as a means of arriving at truth, is
inapplicable to these subjects, we shall be able to show that the latter
method is notwithstanding of great value in the moral sciences; namely,
not as a means of discovering truth, but of verifying it, and reducing
to the lowest point that uncertainty before alluded to as arising from
the complexity of every particular case, and from the difficulty (not to
say impossibility) of our being assured _a priori_ that we have taken
into account all the material circumstances.

If we could be quite certain that we knew all the facts of the
particular case, we could derive little additional advantage from
specific experience. The causes being given, we may know what will be
their effect, without an actual trial of every possible combination;
since the causes are human feelings, and outward circumstances fitted to
excite them: and, as these for the most part are, or at least might be,
familiar to us, we can more surely judge of their combined effect from
that familiarity, than from any evidence which can be elicited from the
complicated and entangled circumstances of an actual experiment. If the
knowledge what are the particular causes operating in any given instance
were revealed to us by infallible authority, then, if our abstract
science were perfect, we should become prophets. But the causes are not
so revealed: they are to be collected by observation; and observation in
circumstances of complexity is apt to be imperfect. Some of the causes
may lie beyond observation; many are apt to escape it, unless we are on
the look-out for them; and it is only the habit of long and accurate
observation which can give us so correct a preconception what causes we
are likely to find