The principle of the trial balance is obvious enough. Clearly
the total of all credits posted to the ledger must equal the total of
all debits posted to the ledger. A balance is simply the excess of
one side over the other, --that is to say, the two sides are equal
except for the balance. When, then, in taking a trial balance the
bookkeeper omits all accounts which have no balance, and takes
from other accounts only the excess of one side over the other, he
is simply omitting from his test of the ledger those debits and
credits which have already shown themselves to be equal and there-
fore beyond the need of test. If the parts tested show themselves
equal, and the parts not tested are already known to be equal,
obviously the totals must be equal.

In the ledger before us, since the accounts of Bills Payable and
Felix Holt show no balances, those accounts do not appear. No
ledger could in practice look quite like this one, for, as has been
already suggested, previous transactions must have taken place,--
else J. Straw and N. Nickleby could not owe the business, and
merchandise could not be sold before any had been bought; but
we may be sure that the previous transactions must also have had
a correspondence of debit and credit, and these equal balances
added to old equal balances must produce totals that are equal.
So the trial balance is correct for our purpose.

It is easy either to overstate or to understate the value of a trial
balance. The trial balance proves nothing; and yet by the law of
chance it is very good evidence that except in two particulars the
books are correct. If an error has been made in posting a debit to
John Jones's account when the posting should have been made to
John Smith's account, the trial balance will not indicate that any-
thing is wrong, for the trial balance shows simply whether a suffi-
cient amount has been debited somewhere; if, on the other hand,

-35-