dissent, upon the ground that not only the administrator of the president of the bank but the other directors ought to be held liable to the extent to which they were held by the District Court, 229 Fed. Rep. 772.

delivered the opinion of the court.

This is a bill in equity brought by the receiver of a national bank to charge its former president and directors with the loss of a great part of its assets through the thefts of an employee of the bank while they were in power. The case was sent to a master who found for the defendants; but the District Court entered a decree against all of them. 229 Fed. Rep. 772. The Circuit Court of Appeals reversed this decree, dismissed the bill as against all except the administrator of Edwin Dresser, the president, cut down the amount with which he was charged and refused to add interest from the date of the decree of the District Court. 250 Fed. Rep. 525. 162 C. C. A. 541. Dresser’s administrator and the receiver both appeal, the latter contending that the decree of the District Court should be affirmed with interest and costs.

The bank was a little bank at Cambridge with a capital of $100,000 and average deposits of somewhere about $300,000. It had a cashier, á bookkeeper, a teller and a messenger. Before and during the time of the losses Dresser was its president and executive officer, a large stockholder, with an inactive deposit of from $35,000 to $50,000. From July, 1903, to the end, Frank L. Earl was cashier. Coleman, who made the trouble, entered the service of the bank as messenger in September, 1903. In January, 1904, he was promoted to be bookkeeper, being then not quite eighteen but having studied bookkeeping. In the previous August an auditor employed on the retirement of a cashier had reported that the daily balance book was very much behind, that it was impossible to *527prove the deposits, and that a competent bookkeeper should be employed upon the work immediately, Coleman kept the deposit ledger and this was the work that fell into. his hands. There was no cage in the bank, and in 1904 and 1905 there were some small shortages in the accounts of three successive tellers that were not accounted for, and the last of them, Cutting, was asked by Dresser to resign on that ground. Before doing so he told Dresser that someone had taken the money and that if he might be allowed to stay he would set a trap and catch the man, but Dresser did not care to do that and thought that there was nothing wrong. From Cutting’s resignation on October 7, 1905, Coleman acted as paying and receiving teller, in addition to his other duty, until November, 1907. During this time there were no shortages disclosed in the teller’s accounts. In May, 1906, Coleman took $2,000 cash from the vaults of the bank, but restored it the next morning. In November of the same year he began the thefts that come into question here. Perhaps in the beginning he took the money directly. But as he ceased to have charge of the cash in November, 1907, he invented another way. Having a small account at the bank, he would draw checks for the amount he wanted, exchange checks with a Boston broker, get cash for the broker’s check, and, when his own check came to the bank through the clearing house, would abstract it from the envelope, enter the others on his book and conceal the difference by a charge to some other account or a false addition in the column of drafts or deposits in the depositors’ ledger. He handed to the cashier only the slip from the clearing house that showed the totals. The cashier paid whatever appeared to be due and thus Coleman’s checks were honored. So far as Coleman thought it necessary, in view of the absolute trust in him on the part of all concerned, he took care that his balances should agree with those in the cashier’s book.

*528By May 1,1907, Coleman had abstracted $17,000, concealing the fact by false additions in the column of total checks, and false balances in the deposit ledger. Then for the moment a safer concealment was effected by charging the whole to Dresser’s account. Coleman adopted this method when a bank examiner was expected. Of course when the fraud was disguised by overcharging a depositor it could not be discovered except by calling in the pass-books, or taking all the deposit slips and comparing them with the depositors’ ledger in detail. By November, 1907, the amount taken by Coleman was $30,100. and the charge on Dresser’s account was $20,000. In 1908 the sum was raised from $33,000 to $49,671. In 1909 Coleman’s activity began to increase. In January he took $6,829.26; in March, $10,833.73; in June, his previous stealings amounting to $83,390.94, he took $5,152.06; in July, $18,050; in August, $6,250; in September, $17,350; in October, $47,277.08; in November,- $51,847; in December, $46,956.44; in January, 1910, $27,395.53; in February, $6,473.97; making a total of $310,143.02, when the bank closed on February 21, 1910. As a result of this the amount of the monthly deposits seemed to decline noticeably and the directors considered the matter in September, 1909, but concluded that the falling off was due in part to the springing up of rivals, whose deposits were increasing, but was parallel to a similar decrease in New York. An examination by a bank examiner in December, 1909, disclosed nothing wrong to him.

In this connection it should be mentioned that in the previous semi-annual examinations by national bank examiners nothing was discovered pointing to malfeasance. The cashier was honest and everybody believed that they could rely upon him, although in fact he relied too much upon Coleman, who also was unsuspected by all. If Earl had opened the envelopes from the clearing house, and had seen the checks, or had examined the deposit *529ledger with any care he would have found out what was going on. The scrutiny of anyone accustomed to such details would have discovered the false additions and other indicia, of fraud that were on the face of the book. But it may be doubted whether anything less than a continuous pursuit of the figures through pages would have done so except by a lucky chance.

The question of the liability of the directors in this case is the question whether they neglected their duty by accepting the cashier’s statement of liabilities and failing to -inspect the depositors’ lédger. The statements of assets always were correct. A by-law that had been allowed to become obsolete or nearly so is invoked as establishing their own standard of conduct. By that a committee was to be appointed every six months "to examine into the affairs of the bank, to count its cash, and compare' its assets and liabilities with the balances on the general ledger, for the purpose of ascertaining whether or not the books are correctly kept, and the condition of the bank is in a sound and solvent condition.” Of course liabilities as well as assets must be known to know the condition and, as this case shows, peculations may be concealed as well by a false understatement of liabilities as by a false show, of assets. But the former is not the direction in which fraud would have been looked for, especially on the part of one who at the time of his principal abstractions was not in contact with the funds. A debtor hardly expects to have his liability understated. Some animals must have given at least one exhibition of dangerous propensities before the owner can be held. This fraud was a novelty in the way of swindling a bank so far as the knowledge of any experience had reached Cambridge before 1910. We are not prepared to reverse the finding of the master and the Circuit Court of Appeals that the directors should not be held answerable for taking the cashier’s statement of' liabilities tó be as correct as the *530statement of assets always was. If he had not been negligent without their knowledge it would have been. Their confidence seemed warranted by the semi-annual examinations by the government examiner and they were encouraged in their belief that all was well by the president, whose responsibility, as executive officer; interest, as large stockholder and depositor; and knowledge, from long daily presence in the bank, were greater than theirs. They were not bound by virtue of the office gratuitously assumed by them to call in the pass-books and compare them with the ledger, and until the event showed the possibility they hardly could have seen that their failure to look at the ledger opened a way to fraud. See Briggs v. Spaulding, 141 U. S. 132; Warner v. Penoyer, 91 Fed. Rep. 587. We are not laying down general principles, however, but confine our decision to the circumstances of the particular case.

The position of the president is different. Practically he was the master of the situation. He was daily at the bank for horns, he had the deposit ledger in his hands at times and might have had it at any time. He had had hints and warnings in addition to those that we have mentioned, warnings that should not be magnified unduly, but still that taken with the auditor’s report of 1903, the unexplained shortages, the suggestion of the teller, Cutting, in 1905, and the final seeming rapid decline in deposits, would have induced scrutiny but for an invincible repose upon the status quo. In 1908 one Fillmore learned that a package containing $150 left with the bank for safe keeping was not to be found, told Dresser of the loss, wrote to him that he could but conclude that the package had been destroyed or removed by someone connected with the bank, and in later conversation said that it was evident that there was a thief in the bank. He added that he would advise, the president to look after Coleman, that he believed he was living at a pretty fast pace, and that he *531had pretty good authority for thinking that he was supporting a woman. In the same year or the year before, Coleman, whose pay was never more than twelve dollars a week, set up an automobile, as was known to Dresser and commented on unfavorably, to him. -There was also some evidence of notice to Dresser that Coleman was dealing in copper stocks. In 1909 came the great and inadequately explained seeming shrinkage in the deposits. No doubt plausible explanations of his conduct came from Coleman and the notice as to speculations may have been slight, but taking the whole story of the relations of the parties, we are not ready to say that the two courts below erred in finding that Dresser had been put upon his guard. However little the warnings may have pointed to the specific facts, had they been accepted they would have led to an examination of the depositors’ ledger, a discovery of past and a prevention of future thefts.

We do not perceive any ground for applying to this case the limitations of liability ex contractu adverted to in Globe Refining Co. v. Landa Cotton Oil Co., 190 U. S. 540. In accepting the presidency Dresser must be taken to have contemplated responsibility for losses to the bank, whatever they were, if chargeable to his fault. Those that happened were chargeable to his fault, after he had warnings that should have led to steps that would have made fraud impossible, even though the precise form that the fraud would take hardly could have been foreseen. We accept with hesitation the date of December 1, 1908, as the beginning of Dresser’s liability, but think it reasonable that interest should be charged against his estate upon the sum found by the Circuit Court of Appeals to be due. It is a question of discretion, not of right, Lincoln v. Claflin, 7 Wall. 132; Drumm-Flato Commission Co. v. Edmisson, 208 U. S. 534, 539, but to the extent that the decree of the District Court was affirmed, Kneeland v. American Loan & Trust Co., 138 U. S. 509; De La Rama *532v. De La Rama, 241 U. S. 154, 159, it seems to us just Upon all the circumstances that it should run until the receiver interposed a delay by his appeal to this Court. The Scotland, 118 U. S. 507, 520. Upon this as upon the other points our decision is confined to the specific facts.

. Decree modified by charging the estate of Dresser with interest from February 1, 1916, to June 1, 1918, upon the sum found to be due, and affirmed.