JOHN R. BROWN, Circuit Judge.

This is an appeal by defendants, husband and wife, from convictions on a jury verdict of guilt for willful evasion of income taxes by filing false and fraudulent returns for the years 1950, 1951 and 1952. There is a frontal attack that for neither of the years is the evidence sufficient. Several less decisive attacks are made urging procedural errors requiring a new trial.

This is not the case of a wife, unwilling, reluctant, or acquiescent in her husband's demands, who finds herself criminally liable merely for the acts of her husband attributed to her. For here the Moores, owners and operators of the Pontiac automobile dealership in Freeport, Texas, were both active in the business and most, if not all, of the bookkeeping accounting activities were, as between the two of them, the immediate responsibility of Mrs. Moore. Nor is this case one of that kind so endemic in which the pressures of the net-worth method are brought to bear. For by indictment, bill of particulars and proof, the Government undertook to show willful error in specific items by which income was understated, cost of sales misrepresented, or operating expenses exaggerated, or combinations of two of them. Perhaps even more unique, the element of willful, fraudulent purpose, normally an inference from circumstances, was here, certainly as to 1950, proved directly by categorical admissions of culpability by defendants' accountant with much the same testimony coming from the other accountant who prepared the 1951, 1952 returns.

The substantive complaint of the Court's failure to grant defendants' motion for judgment of acquittal boils down to these points. As to 1950: the evidence did not adequately show that the extract copy of the 1950 return was a correct copy of the return as filed, and the element of incorrectness and willfulness rested wholly on testimony of the accountant, a self-confessed felon. As to 1951: admittedly the return, as filed, was not signed by defendants and there was no evidence to show that its contents were known to them. As to 1951 and 1952: since the accountant Danforth, who prepared the returns, testified that he thought they correctly reflected income, there was no evidence of fraudulent filing. As to all three years, 1950, 1951, 1952: the evidence merely showed that the income reported in the returns varied substantially from that shown in the account books of the business with no proof that the books were correct. In our view none of these contentions is sound.

The original 1950 return was not introduced. The evidence, however, demonstrated without qualification that it could not be because it had been destroyed in the fire of the warehouse in which returns had been stored by the District Director. Once the original was accounted for, there was likewise sufficient proof of its contents. The Revenue Agent who made the initial investigation in November 1953 testified that he requested, and Mrs. Moore furnished him, a penciled retained copy of the 1950 return. From this penciled Taxpayers' retained copy, he made extracts from which he prepared the typed "dummy" copy used on the trial. That this was fair and adequate proof that defendants had filed such a return was overwhelmingly established by impressive facts, both extrinsic and intrinsic.

There is even less to the objections covering the form of the 1951 return. The original (or agreed photostat) was introduced. While it was not signed by either Mr. or Mrs. Moore, the accountant, Danforth, testified that it appeared to be the one prepared by him and, in any case, extrinsic and intrinsic facts again overwhelmingly established it as the return submitted by defendants.

The fact of a filing by defendants of the return for 1950, and the return for 1951 showing the taxable income and income tax due and paid thereon, was thus adequately established. No such issues arise as to the 1952 return. We turn then to the questions whether, assuming this to have been established, there was sufficient evidence to show that defendants knew of the contents, knew that the true income was something else, and then willfully filed the false returns to defraud the Government?

Since the asserted incorrectness of reported income for all three years is based on the Taxpayers' books, it simplifies matters to outline this generally before discussing any specific complaints on the proof of 1950 or 1951-1952 violations.

The Moores began this business in late 1949. A Pontiac dealer is required to keep a standard set of books in a form and manner meticulously prescribed by General Motors. As neither of the Moores had training in this, they hired Vetterling, on a part time basis, for the period ending mid-1951 and Danforth (a one-time relative of Mrs. Moore) through 1952. If, as was the case here, the dealer obtained from GMAC "floor plan" financing of automobiles purchased from the manufacturer for resale, or made arrangements with it for the assignment of all or part of its conditional sales contract paper covering cars sold at retail, it was necessary for the dealer to submit, under an express warranty of correctness, a periodic and annual report on GMAC accounting forms showing the true condition of the business as reflected by the prescribed books of account. In addition, similar periodic and annual reports had to be made to the Pontiac Division of General Motors on the prescribed forms. These were substantially the equivalent of a trial balance, a detailed profit and loss statement and a balance sheet showing precisely for comparative analysis, increase in net worth of the business.

The fact, so readily discovered by the Revenue Agent in the initial routine audit, that there was a substantial discrepancy in the taxable income reported in the three returns and that shown by these elaborate books and reports, has not, nor can it ever be, denied. To escape this awful predicament, Taxpayers asserted a plea of good faith ignorance, but in refutation their main trust was put, not on facts, but on a legal theory. The legal theory is that while there was proof that there was a discrepancy as such, there was no proof that the accounts, as reflected in the books, rather than the accounts as reflected in the returns, were correct. Elaborating further, it was that since books of account are normally evidential only, Sitterding v. Commissioner, 4 Cir., 80 F.2d 939; United States v. Berman, D.C.Ga., 75 F. Supp. 789, 790, a taxpayer can be convicted, not for failing to pay tax on what the books show, but only on what the real income was. In translating that further into tangible terms, the Taxpayers, by urging the plea of good faith ignorance of what the accountants had done, took positions which may well have been considered by the jury as self-defeating, irreconcilably inconsistent ones: (1) the books were incorrect so it was proper not to follow then in preparing the returns; (2) the books were correct, but the accountants did not advise Taxpayers that the books were not followed in preparing the returns.

But this legal theory collapsed in the face of a record which overwhelmingly supported the contrary conclusions implied by the verdicts of guilt. A major part of this refuting evidence serves double harness and satisfies as well the element that the false returns were knowingly submitted.

There was first the fact that the books recorded all of the transactions showing income and expenses of the business. And, it is another one of the ironic distinctions which set this case off as a unique one that, instead of proving that the books were in error, defendants' own expert witness, a Certified Public Accountant, testified that, except for two items in 1952, these books were correct and taxes, substantially in the amounts calculated by the Government to be due, were in fact due. The jury could also credit the strong suggestion that another set of accountants to whom the books were first submitted when the Special Agent entered the case, likewise could find nothing wrong with the books or the tax computations based on them.

The Revenue Agent checked and tested the books for the three years and found them to substantiate that which was reflected in the periodic and annual reports submitted to GMAC and General Motors. Both Vetterling and Danforth affirmed that the books correctly reflected the state of the business.

Where taxpayers obtain essential credit and procure the very inventory of merchandise which is the main stock in trade on the basis of books and records regularly kept in accordance with accepted accounting principles, the jury is entitled to conclude that such books are an accurate reflection of the business. It is not required, as defendants seem to assert, that the Government go back and reconstruct the books item by item, sale by sale, check by check, to establish anew that the books and records are correct.

Moreover, there is in the circumstance credited by the jury from Vetterling's testimony further corroboration. He (and his wife corroborated this in essential detail) stated that in January of 1951 he had prepared the 1950 return. As the filing deadline date was approaching (see note 3, supra), Mr. and Mrs. Moore came to his home on a Sunday morning to sign the return which Mrs. Vetterling had typed up. The return was based on the books and records which had been kept for the Moores by Vetterling. When Mr. Moore saw the proposed return, he was indignant and outspoken. He refused to sign the return and said the tax should be about $1300 to $1400. When Vetterling told him that that could not be done without falsifying the return, Moore, in effect, told Vetterling that that was Moore's worry, not his. A new return was then prepared by Vetterling making arbitrary adjustments in the Moores' presence, signed by the two of them, the initial return and all work papers torn up and the remnants taken away by Mr. Moore.

Danforth, the accountant who succeeded Vetterling, was not so strong. But for 1951 and 1952, he testified that Moore told him that the business had not earned the profits shown by the periodic reports, and in preparing the returns, he should use figures substantially the same as for the preceding year.

In addition, the jury had the right to impute to the defendants an admission-agreement that the books were correct. On the completion of the Agent's audit, a statement of a proposed deficiency was submitted for the assessment of additional taxes in almost the identical aggregate of the three counts of the indictment. Taxpayers, while insisting at that time that they had not made that much money, nevertheless submitted a special formal offer to pay the proposed deficiency which was subsequently declined by the Commissioner.

Likewise the jury had the opportunity to determine for itself whether the reason continually pressed by Taxpayers as the major cause for the books being incorrect had either substance in fact or was asserted in good faith. As justification for the instructions which Danforth testified they gave to him, and as an explanation why they had not followed the books in having the returns prepared, the Moores personally and through Danforth testified that the books did not correctly reflect income because of excessive allowances on the trade-in of used cars in the sale of new automobiles.

This theory, though expressly submitted to, was presumably rejected by the jury. Not the least reason may well have been the fact that whether excessive or not, whatever was allowed was entered in the books and was taken into account in all subsequent transactions. While an inflated value might temporarily swell assets, it was, on the records, the cost of that used car. If a car, so valued, was sold, there was an automatic loss to the extent that the resale price was less than the initial trade-in. And, in any case, the books showed, and Moore readily acknowledged, that each December the inventory of used cars was reappraised, and inventory value written off so far that December, in contrast to the other months of substantial profits, invariably showed a marked loss.

The jury was entitled, of course, in its everyday cumulative wisdom to think that a businessman, asserting the doubtful correctness of his books, would have the means of demonstrating it, and if the explanation failed in content, that that inferentially went far toward establishing correctness, and certainly in indicating a lack of good faith.

To this the jury could add a further circumstance, plain and simple in its obvious existence, and having profound relevance in the business world. The uncontradicted fact was that instead of this being a business which remained static (as would have been the case had net profits remained the same each year as that reported for 1950), it was a business of marked growth with an increase of over 400% in net worth and 800% in cash.

Once substantial correctness of the books was established, there was ample basis for the finding that returns incorrectly showing different taxable income were willfully filed with fraudulent intent. Vetterling's testimony, while categorically refuted by the Moores, showed, if accepted, a deliberate unlawful purpose. The jury could read Danforth the same way, for it was not required to accept the self-serving protestations that he was not intentionally falsifying the returns and thought that the Moores were correct when they stated to him each year that they had not made the profits indicated on the periodic reports. What he did may have drowned out the sound of what he said, and in so doing, it established as well specific omissions and misstatements of income and expense.

When it comes to procedural errors, none require reversal. It was not error to refuse the requested instruction that the jury, in weighing the testimony of the Vetterlings and Danforth should "take into consideration the very keen interest that said witnesses have in giving the testimony which they gave." Full instructions treating these witnesses as accomplices and giving the jury the usual precautions for receiving such evidence, as well as general and specific charges concerning bias or hostility of any witness were given. We cannot discern how Vetterling had a "keen interest" in categorically confessing to a deliberate falsification of a return. Nor was Danforth aided or his interest advanced in any way by his testimony.

The instruction on "reasonable doubt" while subject precisely to the infirmities of the one criticized in Holland v. United States, 348 U.S. 121, 141, 75 S.Ct. 127, 95 L.Ed. 150, 167, does not, again for the reasons pointed out in that very case, present a situation, on this record, of harmful effect. Fed. Rules Crim.Proc. rule 52(a), 18 U.S.C.A.

The final point concerns the objections to the charge as given and the refusal of the Court to grant a requested charge on evidence of good reputation. The Court gave one in substantial accord with our prior decisions, Le More v. United States, 5 Cir., 253 F. 887, certiorari denied 248 U.S. 586, 39 S.Ct. 184, 63 L.Ed. 434, and Grace v. United States, 5 Cir., 4 F.2d 658, certiorari denied 268 U.S. 702, 45 S.Ct. 637, 69 L. Ed. 1165, unaware at the time of the trial of our decision in Holland v. United States, 5 Cir., 245 F.2d 341, which was not published until after the trial. In this record with its devastating facts, both direct and circumstantial, we would not have reached a conclusion that any such difference in the wording between the requested and given instructions could have had any harmful effect. United States v. Kushner, 2 Cir., 135 F.2d 668, certiorari denied 320 U.S. 212, 63 S.Ct. 1449, 87 L.Ed. 1850. Nevertheless, since trial courts are entitled to guides as clear as can be fashioned, we think it unsound to undertake any comparative analysis of the instructions here requested in contrast to those given by the Court or those criticized in Holland. Rather we should state plainly that in Holland neither by briefs nor argument were our prior decisions in Le More and Grace called to our attention. Those cases represented the law in this Circuit, see Kreiner v. United States, 2 Cir., 11 F.2d 722, at page 726, certiorari denied 271 U.S. 688, 46 S.Ct. 639, 70 L.Ed. 1152. The contrary holding in Holland is disapproved so that these two prior decisions continue their vitality.

Affirmed.