can be, and should be, covered by use of a controlling account, as the items are of such a nature that they are grouped in the original records, and where the posting of a thousand entries in one can be proved by one operation, time is saved. This method assumes, of course, that there will be a verified trial balance of the customers' ledgers to support the controlling account. The question of accounts which have been collected, but which do not show the collections, will be considered later.

As to the footings, verification is somewhat more important than with purchase records. In a large concern, prove, say, every eighth page, and in a small concern, say, every third or fourth page, always including the last and sometimes the next to last page of each month.

Cheques Received must be Deposited: In those corporations where the capital stock is all, or nearly all, owned by its officers or directors, the latter frequently handle the funds of the corporation as if no corporate responsibility existed. Usually they act honestly and no harm is done, but an auditor cannot afford to pass over illegal acts which come to his knowledge, although so far as can be foreseen no one will be injured thereby.

The practice of an officer using cheques drawn to the order of the corporation for his private use cannot be condemned too severely, and yet it is not uncommon. All cheques should be deposited, and the officer can withdraw a similar sum if the corporation is indebted to him.

In a New York case, decided by the Court of Appeals in 1908, where the president of a corporation (the stock of which was all owned by himself and the secretary-treasurer) used a cheque drawn to the order of the corporation and indorsed by it to pay his personal debt, the creditors of the corporation upon its being declared insolvent compelled the trust company that cashed the cheque to refund the money.

Periodical Verification of Bank Balances: The auditor should ascertain by personal investigation whether the bank balances have been verified at frequent intervals throughout the period covered by the audit. It frequently happens that clerks are careless about reconciling the bank balances with those shown by the books. In some cases the pass books are not left for settlement for many months; in other cases the pass books are balanced by

the banks and delivered to the depositors, but are not verified by the latter.

In the middle and far West, the banks have quite generally adopted the plan of sending monthly statements to all depositors, accompanied by the paid cheques. There is a growing tendency on the part of the banks in the East to follow this plan. This is to be commended, as it leads naturally to more frequent reconciliations between the parties.

But the bank statements may be filed away without examination unless some one in authority insists on prompt attention. Instances have been known where junior clerks have forged signatures to cheques, and owing to carelessness in inspecting the bank settlements, the fraud was not discovered for a long time.

The law is that the writing up of a bank pass book, with a return of paid cheques, or a statement of account, is not conclusive upon the parties so as to preclude an ascertainment of the true state of the accounts in a case where cheques are included in the balance which are subsequently discovered to be forged. But where the depositor is under a duty from the usages of business or otherwise to examine the account within a reasonable time and to give timely notice of any objections he may have, an omission to perform this duty, leaving the bank to rely upon the presumption that the account is acquiesced in, whereby it is misled to its prejudice, will make the account conclusive. The law is not unreasonable in holding that where bank settlements lie around untouched for long periods those in control of the business should know that a risk is being run for which there is no

excuse.

The auditor may sympathize with the clerk who is overworked and who omits the reconciliation of the bank balances for what he considers more important duties, but nevertheless, such neglect may be costly and should be reported.

(3) Cash Receipts: In all well-regulated concerns all cash receipts are deposited in bank and all payments, therefore, must be made from the bank account. This almost disposes of the question of verifying the footings of the cash book. Where the bank account is proved and the cash receipts and payments are traced into and out of bank, it would seem logical that the footings of the cash book could be automatically proved at the same

time. If this does not seem to be complete verification, in a large business, the proving of every third or fourth page will be a complete check.

In many small concerns the cashier handles the receipts, writes up the cash book, and makes the deposits. He might retain a portion of the receipts, force the footings of the column to agree with the actual deposit, and increase the footings of the discount column. The cash book would then balance across, the cash balance agree with the amount shown by the pass book, and as the discount column would be posted in total by the bookkeeper, the ledger would be in balance without any falsification being made therein. If there were special columns provided for cash sales or similar earnings, the footings of such columns could be reduced by the amount of the shortage and thus afford additional facilities for covering up the amounts taken.

The postings of the nominal accounts should usually be verified, not because there is any great danger of fraud lurking therein, but for the purpose of locating any possible posting to a wrong account. For instance, it frequently happens that part of a plant or old machinery may be sold. Sometimes such items are posted to an earning account instead of to a capital or a reserve account. These postings are, as a rule, few in number and are important enough to be verified in extenso.

The postings to the credit of customers should be proved in totals through the controlling accounts. If there is no controlling account and one cannot be constructed readily, a fair test should be made of the individual ledger credits, working, of course, from the ledger back to the cash book and not vice versa. The reason for this is obvious; if a customer has been credited with an amount which purports to have been posted from the cash book, but which, as a matter of fact, is not entered there at all, the discrepancy would not be discovered by using the cash book as a basis, and it would not be safe to depend on looking through the ledger subsequently to see that all items are ticked. It is sometimes suggested that the chief danger in such practice lies in the possibility that the ledger clerk can, where the work is not finished at a sitting, supply the ledger tick marks himself. There is not much basis for this fear, but it would be foolish to expose one's self to it where no necessity exists.

If it is thought wise to verify the individual postings to the customers' accounts, do not check every one unless some very good ground for suspicion exists. If the audit is a periodical one, say, for six months, cover about half the letters of the alphabet only. Six months later cover the other half, or one-fourth only at each audit, and take two years to the entire list. It is not an infrequent fraud wherein ledger credits do not appear as cash debits, but it is hard to imagine one case where a good test would not have disclosed the fraud. Very few men would confine their peculations to customers whose names commenced with X Y Z only. The auditor can afford to take the chance that the defaulter will inadvertently manipulate the account of an A customer, in which case he would not escape the first time, and if he used only one or two letters, he would still be detected in a reasonable time.

(4) Cash Payments: On the payment side there is also a possible opportunity for fraud being covered by erroneous footings. In most cash books there are columns for different expenses which are posted in total at the end of the month. If the cashier were to take an unnumbered cheque from the back of the book, or one of a style similar to the ones in current use, and have it drawn to his order or to some name representing himself and did not enter the amount in the cash book, the footings could be falsified to that extent. The cheque could be numbered to correspond with those in use at the time. When the bank settlement was received and the balance verified, the cheque could be destroyed.

Instances are known where cheques were marked as void on the stubs and the two halves of one cheque pasted on two stubs (one of the portions, of course, being without a number), thus providing a cheque which was used by the cashier to obtain money fraudulently.

These manipulations are not likely to occur more than once in an auditor's experience, but all of the possibilities mentioned are based on actual experience, so that an auditor cannot afford to neglect all reasonable precautions to ascertain if such fraud exists.

The auditor should insist, wherever feasible, on having all payments represented by cheques. It reduces the possibility of

manipulation of cash-book footings to a minimum, and for this reason alone is worth all the trouble which may be occasioned by the necessity of depositing all currency receipts. If the footings cannot be proved by the bank account, verify, say, every third or fourth page.

If the cash book is properly columned and a controlling account is kept with accounts payable, most of the postings will consist of monthly aggregates, which should be checked to see that they do not get into the wrong ledger account. Here, however, it is important to avoid duplication. In many audits it is desirable to make full analyses of the various expense and purchase accounts for use in the reports. If feasible and convenient this work should be done at the time of the verification of the postings, for it may be found that the details in the ledger are not sufficient and the cash book pages must be consulted again.

Postings to the individual accounts need not be verified unless some special reason appears. The payments are supposed to be vouched to establish their authenticity, and it is not necessary to trace the payments to the debits of the accounts. A controlling account supported by a trial balance of the subsidiary ledger is a good proof, but even where this is not in evidence the checking of the debit postings is usually superfluous.

Summary: As against the practice-fairly common-of checking all postings and footings, the above course will seem radical. It is not radical, however, if it is approved after full discussion and thought, and if it will stand the additional test of each particular audit. Where the slightest cause for suspicion exists, there must be a careful study of every phase of the situation.

But if suspicion has been aroused and there is a probability of something being wrong, the most foolish thing an auditor can do is to jump in blindly and tick every entry in the books. This has been done more than once, but the practice cannot be condemned too strongly.

Other Records: In many lines of business the books of account bear distinctive titles, and it may be that in the foregoing pages these books have not been called by their technical names. For instance, in a magazine publishing business a sales