This is an appeal from a judgment of conviction of income tax evasion on three counts of a five-count indictment which charged the defendant with evasion of his income tax for the calendar years 1947 through 1951, respectively. After a trial, in which the Government relied upon the so-called “net worth” theory of proof, he was convicted on the counts for the years 1948, 1949 and 1950, and acquitted on the count for 1951. The jury was unable to agree upon a verdict on the first count, for 1947, and following its stalemate the trial judge directed a verdict of acquittal on that count. Subsequently, on the defendant’s motion for acquittal on all counts or, in the alternative, for a new trial, but prior to any imposition of sentence, the judge vacated his previous directed verdict of acquittal as to Count 1.

The defendant was a Rochester policeman who had joined the Police Department in 1921 as a patrolman and was thereafter continually employed on the force throughout the pre-indictment and indictment years. His annual gross salary ranged from a low of $1125 in 1921 to a high of $4400 in 1951. The defendant was assigned to special duty investigating vice at various times beginning about 1934. He was so assigned throughout 1947 and thereafter until about July 16, 1948. On the latter date he was transferred to the detective bureau where he remained until about May 1949 when he was returned to the vice squad. There was no evidence that he was engaged specifically in vice investigation during 1950 but he was restored to his vice investigation duties on about December 15, 1951.

Prior to trial the Government conducted a thorough investigation of the defendant’s financial status and manner of life from the close of the calendar year 1941 through the close of the final indictment year 1951. The Government’s investigation was exhaustive. Every bank in Rochester was checked for accounts of the defendant. The Surrogate’s records in Monroe County were checked for possible legacies from relatives of the defendant and his wife, and a claimed legacy from the defendant’s sister was negated; the United States Treasury records pertaining to the issuance of United States Savings Bonds were checked for bond purchases year by year; the various contractors who worked on the defendant’s residence were interviewed to verify the cost figures given by the defendant; the department stores in Rochester were checked for purchases; the automobile dealers where the defendant had purchased cars were checked for the date of the purchase and the amount involved; the Rochester Gas and Electric Company was checked for payments by the defendant; the Clerk’s office of the Monroe County Court House was checked for the date that his old house was sold, the amount for which it was sold, and the date he purchased the property on which his new house was located. This list is by no means inclusive. It is set forth here only as some indication of the meticulousness of the Government’s investigation and to show that the specific items reflected in its net worth chart were based on undisputedly sufficient evidence.

In addition, Government agents questioned the defendant concerning his sources of income, inheritances, gifts, borrowed funds and the sources of his *60wife’s income and funds. The Government treated the defendant and his family as an economic unit and calculated the defendant’s assets, liabilities and expenditures on that basis. The Government purported to show the net worth (assets minus liabilities) of the defendant at the close of each calendar year beginning with 1941. Exhibit 80.1 It credited the defendant with an opening net worth of $57,490.40 as of January 1, 1947, which figure included no allowance for cash on hand. To the increase (or decrease) in the defendant’s net worth for each subsequent year in the indictment period, it added the defendant’s personal (nondeductible) expenditures for that year thus determining his gross income. It calculated the defendant’s taxable income for each year by subtracting from his gross income all known nontaxable income receipts and allowable nonbusiness deductions during that year. By subtracting his reported income from his taxable income figures thus arrived at, the Government calculated unreported taxable income during the indictment years 1947 through 1950 of $15,455.61, $7,892.20, $2,976.17, and $1,103.22, respectively. For 1951 the Government calculated an overestimation of taxable income of $180.62.

For the pre-indictment years 1942 through 1946 the Government, through the same basic scheme, calculated unreported taxable income of $4,996.32, $6,-167.55, $5,648.68, $9,009.12, and $17,122.-00, respectively. These figures were included in the net worth chart, Exhibit 80, which went to the jury.

It was the Government’s theory that the unreported taxable income as indicated by its net worth chart was graft received by the defendant for nonperformance of his duties as a policeman and as a member of the vice squad. Although there was no direct evidence that the defendant ever received a bribe, the evidence did disclose that he had “opportunities” flowing from his position in charge of the investigation of vice. There was evidence that there was wide open gambling in Rochester during the spring and early summer of 1948 when the defendant was on the vice squad; that on three or four occasions between April and June 1948 the defendant had been told by Officer Stanton about the sale of policy slips by one Passorelli, but no action was taken until Officers Faulkner and Van Auker were notified of similar sales and thereupon made an arrest; that statistics of gambling cases disposed of in the City Court of Rochester indicated an upswing in prosecution of gambling from 1948 to 1951 after other officers had been assigned to the squad; and that the defendant was on friendly terms with a professional gambler, Thomas L. O’Brien, whose restaurant he and his wife patronized and whose house he had visited on two occasions. There was evidence that O’Brien had called the defendant by telephone two or three times in 1954 to inquire about rumors that the defendant was going back to the police force, and that at an unidentified date the defendant and his wife had briefly visited O’Brien at his summer cottage. Police Lieutenant Irish testified that while he, Irish, was on the vice squad he had been offered bribes on several occasions. This evidence was first admitted and subsequently stricken out by the court with instructions to the jury to disregard it.

The Government introduced proof of inconsistencies and fabrications in the defendant’s extrajudicial explanations of his finances. When the defendant was questioned by a superior in 1947 about the expensive home he was building on Gregory Hill Road, he claimed he could sell his Linden Street House for $10,000 and get a mortgage for the rest. Actually, the sale of the Linden Street residence netted him about $4800, and the new house, costing $37,800, with the land and all improvements, was not mortgaged. In April of 1952, when interrogated by Revenue Agents, the defendant stated that the construction of his new home cost him only $19,400. He claimed to have received in 1943 an inheritance from his brother, Clarence, of $500, with which *61the Government credited him, and a gift of $8800 from Clarence in 1987 which went into bond purchases in later years.2 3 He stated that he had no source of income from any other source than his salary. When asked how he could deposit over $13,000 in banks during 1946 and 1947 on his salary of about $3,000, the defendant replied, “That is a lot of money. It looks bad, doesn’t it?” At a conference with the Revenue Agents in July 1952 he stated that the house cost “around $25,000” and when confronted with his prior statement said, “I might have lied about that unwittingly. I have nothing to conceal.” At that time the defendant claimed the $500 inheritance and a gift of $8900 from Clarence in 1939 as his only gifts and inheritance. He stated that he used the gift from Clarence to make bank deposits in 1946 or 1947 and explained his bond purchases and other bank deposits by systematic savings and accumulations,3 an explanation which was completely destroyed by his wife’s testimony at the trial.4 In November 1952, at a meeting with Police Department offieials, the defendant claimed for the first time that in 1940 he and Clarence had pooled some money for a joint business venture and that when the venture failed to materialize Clarence before his death in 1941 turned over the pooled funds to the defendant. He also mentioned then for the first time a legacy of his wife’s from a Great Aunt Mary from the West which will be discussed later. The Government also presented evidence that the defendant had concealed bond purchases, expenditures, and bank accounts, and made misrepresentations relating to the purchase of a car for his daughter, Nancy, in his pre-indictment extrajudicial statements. In short, the evidence of fabrication, misrepresentation and concealment was strong and the jury may well have concluded that his constantly shifting explanations were incredible.

The defendant did not take the stand. The defense called several witnesses who were superiors and associates of the defendant on the police force. These witnesses testified to the defendant’s reputation for being an efficient policeman and *62to his activities in suppression of vice. Irma Ford, his wife, testified on his behalf and claimed the existence of a substantial cash hoard on hand at the opening date (January 1, 1947) derived principally as gifts from Mrs. Ford’s relatives, particularly a Great Aunt Mary of California, and from the defendant’s brother Clarence. Mrs. Ford testified that she had $2000 when she married the defendant. She claimed to have received $2000 from Great Aunt Mary in 1922 and another $6000 over a subsequent period of five or six years, before Mary died. Out of a sizeable estate which, she testified, Mary bequeathed to Mrs. Ford’s mother, she claimed to have further received approximately $5800 as gifts in three installments during 1937, 1938 and 1939. She also claimed gifts of $2400 from her daughter Jean in 1948 before Jean entered a convent. Mrs. Ford testified to the claimed gift of $8800 or $8900 from Clarence in about 1937 and also testified to finding a thick envelope containing money in the defendant’s dresser drawer in the summer of 1941, a few months before Clarence’s death, which contained the substantial sum (in an unspecified amount) allegedly left with the defendant by Clarence after their alleged business venture failed of consummation.

The Government introduced independent evidence tending to negate the claim of a substantial cash hoard on hand in 1947 and deriving from acquisitions in the 1920’s and 1930’s and to prove that its alleged existence was a fabrication. There was considerable evidence of small loans and debts throughout the 1930’s. As late as 1940 the defendant borrowed $100 from the Central Trust Company. He made a payment thereon of $25 on June 24, 1940, and renewed the note in the amount of $75 on August 8, 1940. On November 8, 1940, it was renewed again in the amount of $55, and was paid in full on February 8, 1941. In addition, the Government introduced evidence tending to show that even if the defendant’s explanations as to the accumulation of the hoard were true the hoard had been dissipated by the acquisition of non-cash assets prior to January 1, 1947. According to the pre-trial statements of the defendant and the trial testimony of his wife as to the alleged sources of the hoard it had been fully accumulated in an amount in excess of $24,700 prior to the death of Clarence in November 1941. During the period from January 1, 1942 until the opening date, January 1, 1947, the Government’s evidence purported to show approximately $43,000 of visible net worth increases and personal expenditures in excess of reported earnings. Therefore, even if the jury had believed the defendant’s explanation, there would have been ample evidence to indicate an exhaustion of the hoard before January 1, 1947.

Accumulation of Cash Hoard (per Defendant’s evidence) Dissipation of Cash Hoard (per Government’s evidence)

1919 Irma- Ford’s assets at marriage............ $ 2,000 1942 $ 4,996.32

1922-28 Gifts from Aunt Mary.. 8,000 1943 6,167.55

1937-39 Gifts from Irma Ford’s mother out of Aunt Mary’s bequest....... 5,800 1944 1945 5,648.68 9,008.12

1937-41 Gifts from Clarence5... 8,900 plus 6 1946 17,122.00

Total .............. 24,700 plus 42,942.67 7

*63The defendant does not challenge any of the specific items included in the net worth chart. Rather, he attacks the reliability of the opening net worth figufe (January 1, 1947), claiming omission therefrom of a substantial cash hoard in currency then on hand. The Government credited the defendant with an opening net worth of $57,490.40. This figure included a credit for $36,303.33 deposited in seven bank accounts which the Government’s investigation had uncovered. It did not, however, include any credit for currency on hand. The defendant asserts that this omission renders the Government’s evidence of net worth so unreliable as to lack probity as a foundation for its purported calculation of unreported net income.

We think that was a question for the jury. Our function in passing on the sufficiency of the evidence is complete upon a determination that the opening net worth figure is supported by an adequate investigation, including the tracking down of leads furnished by the defendant, and is corroborated by some independent proof tending to negative the existence of a cash hoard on hand at the opening date.

Here, the opening figure was supported by an exhaustive investigation by the Government and was independently corroborated by evidence of the defendant’s precarious financial condition during the period when, according to the defense, he had a substantial cash hoard on hand.8 There was further evidence that even if the defendant had accumulated such a hoard as and when contended by the defense it would have been absorbed by his proved expenditures prior to the opening date.9 Finally, there was evidence of the defendant’s inconsistent, false and misleading extrajudicial statements which the jury might well have concluded were intentionally fabricated.10

In the context of this case, it is not necessary for us to decide whether any one of these lines of corroborative proof was sufficient. Their cumulative effect was to produce firm ground to infer the accuracy of the Government’s opening net worth figure. To reduce the totality of proofs to its component parts and examine each part for its independent probative efficacy would be to distort the overwhelming substantiality of the proofs as a whole into a congeries of irrelevant abstractions.

Since the Government thus furnished adequate evidence of the substantial accuracy of its opening net worth figure and since the other figures on its chart were independently supported and were never challenged, it follows that there was adequate support for a finding by the jury of unreported income during the indictment years from some source. The fact that the Government thoroughly investigated the defendant’s affairs and financial condition and yet failed to uncover any substantial nontaxable receipts beyond that shown on its chart, was sufficient, at least for purposes of its prima facie case,11 to nega*64tive' an inference that the receipts thus demonstrated may' have derived from a nontaxable source.

Somewhat overlapping the contention just above dealt with, the defendant claims that the Government’s proofs were insufficient because of its failure to track down leads furnished in his extrajudicial explanations as to the source of his net income and expenditures. We do not agree. His claim as to the low cost of his new house the Government did track down — and found substantially exaggerated. His claim of a legacy from his sister was checked and found untrue. His statement that he kept a large sum of cash derived from his brother in the Police Department safe was apparently tracked down as far as possible with a result by no means favorable to the defendant’s contention. His claims as to the receipt in pre-indictment years and his disposition of various cash gifts received from his brother Clarence and of gifts to his wife through her Great Aunt Mary were so vague that they were not susceptible of further investigation. The burden on the Government in this respect was only to provide “effective negation of reasonable explanations by the taxpayer inconsistent with guilt. Such refutation might fail when the Government does not track down relevant leads furnished by the taxpayer — leads reasonably susceptible of being cheeked, which, if true, would establish the taxpayer’s innocence.” Holland v. United States, 348 U.S. at pages 135, 136, 75 S.Ct. at page 135. Here the explanations were neither “reasonable” nor “susceptible of being checked.”

Moreover, here as in Holland, even if •the defendant’s leads were assumed to be true, the “incidents relied on by petitioners [the defendant here] were so remote in time and in their connection with subsequent events proved by the Government that” whatever the defendant’s net worth was in 1941 (when the last contribution from Clarence was allegedly received) it “appears by convincing evidence that” on January 1, 1947, he had only such ássets ás the Government credited to him in its opening net worth figure. Here, even if all the pre-indictment gifts claimed had been received as and when claimed, they did not prove cash on hand on January 1, 1947. At most, they would explain only part of the $43,000 by which his combined increase in visible net worth and personal expenditures, as shown by the Government’s chart, exceeded his aggregate income as reported for the years 1941 through 1946.

Nor do we agree with the contention made that in the circumstances of this case the Government, to prevail, was required to interview the defendant’s wife about her own assets, sources of exempt income and the receipt of the gifts sáid to have been received from her Aunt Mary. It is true that the financial status and history of Mrs. Ford was clearly relevant to the prosecution of the defendant, especially since the family was treated as a unit for purposes of calculating net worth increases and expenditures. But we think the Government adequately exhausted investigation of that possible source of income by its questioning of the defendant in July 1952 as to his wife’s assets and sources of income. He stated then that his wife did not work except occasionally in a hospital, that she had no stocks or securities, but that she "had a little money that her folks left her.” Moreover, it may not be assumed that a pre-trial interview with Mrs. Ford would have added anything to the case not contributed by her testimony on trial. And that testimony, even if true, did not impugn the Government’s opening net worth statement, as we have noted in the preceding paragraph.

In short, we think the amplitude of the investigation actually made, summarized elsewhere in this opinion, fully satisfied the investigatory burden on the Government and justified all the inferences, affirmative and negative, which the jury apparently based thereon.

The defendant also contends that the net worth proofs were not corroborated by proof of a likely source of additional *65income and hence were insufficient to support the verdict. The argument would have pertinence, we think, only if there were a lack of other evidence to prove that the defendant had unreported current receipts. Here, however, without proof of a likely source the Government’s proofs were fully adequate to support a finding that the defendant had substantial unreported receipts during the indictment years from some source. For the proofs, as we have seen, negatived the defendant’s claim of a cash hoard on hand during the indictment period and supported the accuracy of the Government’s opening net worth figure. The other figures in the Government’s net worth chart were affirmatively supported by the Government’s investigation and indeed were not challenged. The Government having thus established unreported current receipts from some source, all that was needed to complete a prima facie case was enough proof to negative the possibility that the unreported current receipts thus established were nontaxable.

Merely to supply this necessary link in the Government’s case, it was not necessary to prove directly the receipt of graft or that graft was a “likely source” in the sense that it was a source which the defendant had actually drawn upon. “Likely source” is but one method to negative all the possible nontaxable sources of the alleged net worth increases: it is not always an indispensable element in a net worth case. United States v. Adonis, 3 Cir., 221 F.2d 717. This conclusion, we think, is supported by passages in Smith v. United States, 348 U.S. 147, 158, 75 S.Ct. 194, 99 L.Ed. 192, and United States v. Calderon, 348 U.S. 160, 164 and 165, 75 S.Ct. 186, 99 L.Ed. 202, and is wholly consistent with Holland v. United States, 348 U.S. 121, 137 and 138, 75 S.Ct. 127, 99 L.Ed. 150. Direct proof of a likely source here was unnecessary because the Government sufficiently proved unreported current receipts by directly substantiating its opening net worth figure by independent evidence and because the results of its exhaustive investigation of the defendant’s affairs warranted an inference that he had no nontaxable income during the indictment years not credited to him on the Government’s net worth chart.12

The defendant urges that the evidence relating to graft opportunities should have been excluded as prejudicial to the defendant since the Government failed to link him directly with graft. We do not agree. The evidence showing the possibility of graft, although not sufficient in itself to prove unreported current receipts, was relevant and material for its tendency to show that the unreported receipts, independently established, derived from a taxable source. The evidence therefore was properly received as corroborative. It added strength to the inference that the unreported receipts were taxable.

The finding of the defendant’s willfulness was supported by an abundance of evidence. Here, as in Holland v. United States, 348 U.S. at page 139, 75 S.Ct. at page 137, there was evidence of a consistent pattern of underreporting comparatively substantial amounts of income both within and before the indictment years. And the use of fabrications to deny the underreporting of income, rather than a defense of non-willful negligence in failing to report it, provides some additional evidence of a willful attempt to evade income taxes through concealment of taxable sources. Gariepy v. United States, 6 Cir., 189 F.2d 459, 463; Sasser v. United States, 5 Cir., 208 F.2d 535, 539.

On appeal, it has been suggested in behalf of the defendant that despite the absence of admissions or other proof that he received graft, it is possible that he did do so prior to the indictment years *66and thereby built up a cash hoard sufficient to account for the apparent unreported income in the indictment years. On this theory the defendant’s fabrications are explained as an effort to conceal guilt in receiving graft rather than guilt of tax evasion. In consequence, it is argued, the fabrications may not be treated as corroborative of the Government’s opening net worth figure. Although this contention runs counter to the defendant’s pre-trial assertions and the theory of the defense at the trial, it has some plausibility and deserves consideration to the extent that it may have bearing on the prima facie sufficiency of the Government’s proof.

But we cannot indulge in the assumption that the defendant took graft in large amounts prior to the opening indictment date and shut our eyes to the likelihood that, if so, he continued to do so during the indictment years when the source remained available. If we assume, as we are now asked to do, that prior to 1947 the defendant actually profited through graft, by the defendant’s own hypothesis accompanied by proof of opportunities for graft there existed a likely source of unreported taxable income during the indictment years which constituted adequate corroboration of the Government’s proofs of additional taxable income, under the Holland doctrine.

In addition to the major contentions already discussed, the defendant presses several other claims of error all of which we find without merit. He suggests the possibility that some of his excess funds might have come from his son, James Ford. Not only was such a claim not made prior to or at the trial, but James Ford testified that prior to the indictment years he was successively at college, in the army, and at law school, and during the indictment years until April 1949 he was successively employed by a law firm for approximately $25 a week, and as a Post Office clerk for about $60 a week. Thereafter, he was employed as a claims examiner with the State Labor Department. On the evidence the jury might reasonably have concluded that none of the defendant’s expenditures derived from James. Moreover, the Government need not negate every possible source of nontaxable income where leads are not forthcoming. Holland v. United States, supra, 348 U.S. at page 138, 75 S.Ct. at page 136; Rossi v. United States, 289 U.S. 89, 91-92, 53 S.Ct. 532, 77 L.Ed. 1051; Warring v. United States, 4 Cir., 222 F.2d 906, 911-912.

The defendant contends that the conviction must be reversed because the entire case rests on the accuracy of the Government’s opening net worth figure and that the jury’s inability to reach a verdict for 1947 imports a finding that the figure was inaccurate. That argument is specious. On appeal we are concerned not with the impact of the evidence on the jury but with its legal sufficiency. We think the evidence sufficiently supported the Government’s opening net worth figure for 1947 even though the jury for some reason failed to convict on the 1947 count. Even if the jury had acquitted the defendant on the 1947 count that acquittal would not destroy the validity of its verdict for subsequent years. United States v. Costello, 2 Cir., 221 F.2d 668, 673; United States v. Coplon, 2 Cir., 185 F.2d 629, 633, 28 A.L.R.2d 1041; Horne v. United States, 5 Cir., 193 F.2d 175. That is so quite apart from the fact that in this case the opening figure for 1948, although calculated by the scheme of that for 1947, was based on independent data.

It is urged by the defendant that the evidence of opportunities from graft went beyond the limits of the order of preclusion entered by the judge before trial which precluded the Government “from introducing evidence at the trial as to the claimed possible or probable sources of the defendant’s alleged undisclosed income, except to show that the defendant was employed as a police officer, the nature of his duties as such * * * and that the defendant’s increase in net worth did not come from a non-taxable source or sources.” This order substantially followed the language *67used by the Government in its bill of particulars stating what it intended to prove. Wo think the proofs properly fell within the confines of the order. In addition there is no claim that the defendant was surprised by the evidence.

Error is assigned for failure of the trial court to inform defense counsel as to its rulings on the defendant’s requests to charge in advance of summation. Rule 30 of the Federal Rules of Criminal Procedure, 18 U.S.C.A., provides that, “The court shall inform counsel of its proposed action upon the requests prior to their arguments to the jury * * We think that in the setting of this case strict compliance with the Rule was waived when counsel, without request or objection, proceeded with his summation. But even if there were technical error it was harmless. The judge charged in substantial compliance with all the defendant’s requests save one and there is neither assertion nor showing of harm resulting from noncompliance with the Rule.

The defendant also assigns as error the admission of the items on the net worth chart purporting to show Additional Unreported Net Income for the pre-indictment years 1942 through 1946. This evidence was highly relevant both on the issue of the defendant’s intent13 and to refute the defendant’s claim as to a substantial cash hoard on hand at the opening indictment date consisting of receipts from Clarence or Aunt Mary prior to 1942. We hold that this evidence was properly admitted.

The defendant claims prejudice from the admission of Irish’s testimony that on occasions he had been approached with offers of graft money. Subsequent to its admission this testimony was stricken on the authority of Ford v. United States, 5 Cir., 210 F.2d 313, and in the charge the jury was admonished to disregard it. If its admission was erroneous, we think the subsequent action taken must be deemed to have sufficiently safeguarded the defendant from resulting prejudice. We incline to think, however, that in view of the independent evidence of unreported income during the indictment period the testimony was relevant and admissible for its bearing on the source of that income. It tended to show that Rochester police officials had opportunities to obtain taxable income in the nature of graft and thus reinforced an inference, permissible on other evidence, that the defendant’s unreported income was taxable. We hold that nothing connected with this item is cause for reversal.

Finally, there is a claim that the trial court erred in vacating its directed verdict of acquittal on Count 1 of the indictment. The argument is that if the defendant is retried on that count he will be placed in double jeopardy. Since the order is interlocutory and the defendant has not yet been placed in jeopardy thereunder, the issue is not currently appealable and the pending appeal as to Count 1 must accordingly be dismissed. Lewis v. United States, 216 U.S. 611, 30 S.Ct. 438, 54 L.Ed. 637; United States v. Swidler, 3 Cir., 207 F.2d 47, certiorari denied 346 U.S. 915, 74 S.Ct. 274, 98 L.Ed. 411. Cf. Berman v. United States, 302 U.S. 211, 58 S.Ct. 164, 82 L.Ed. 204.

Affirmed on Counts 2, 3, and 4. Appeal dismissed on Count 1.

*68

*69

(dissenting).

Holland v. United States, 348 U.S. 121, 75 S.Ct. 127, 99 L.Ed. 150, and its companion decisions warn us that the net-worth method is a dangerous instrument, to be used with caution, lest it may have the effect of unconstitutionally putting on an accused the burden of proving his innocence.1 Because of that- warning, I have such doubts about affirming this conviction that I feel constrained to dissent.

The trial judge, in the plainest terms, told the jury that it could not find defendant guilty, unless it found (1) that he had a “probable source of income” (other than his salary or interest on his bank account and bonds) and (2) that this probable source was graft. His charge, in this respect, reads: “To justify the use by the Government of this net-worth-expenditure method you must be satisfied on all the evidence in the case that the defendant has a probable source of income other than his salary and the interest on his bank accounts and the bond interest. You must be satisfied that the probable source of his income was in some way connected with his duties as head of the Vice Squad, with receiving money or things of value either in exchange for services rendered to people whom he came in contact with as an officer of the Vice Squad, or exchange for forbearing to do something he was supposed to do as an officer of the Vice Squad.”

According, then, to the theory on which the case went to the jury, probable graft was an essential element of the government’s case; and Holland v. United States teaches that every essential element must be proved beyond a reasonable doubt. Of course, such proof of any essential may be by circumstantial (“indirect”) evidence. But I think that here the circumstantial evidence of probable graft — which Judge HlNCKS significantly calls proof of “the possibility of graft” — was much too meager to justify a jury finding of the existence of that element. For look at that evidence as Judge HlNCKS describes it:

(a) When defendant was on the vice squad, there was wide open illegal gambling.

(b) While on this squad, in 1948 defendant was told, by another policeman, of the sale of policy slips by one Passorelli, but “no action was taken” until two other officers, learning of this fact, made an arrest.

(c) The number of prosecutions for gambling increased when other policemen were assigned to the vice squad.

(d) Defendant was on “friendly terms” with O’Brien, a professional gambler; defendant and his wife patronized O’Brien’s restaurant ; twice, they visited at O’Brien’s house and once at O’Brien’s summer cottage. After defendant had left the police force, O’Brien phoned defendant two or three times to ask whether defendant was going back on the force.

(Another policeman, Irish, testified that while on the vice squad he (Irish) had “been offered bribes” several times. But this item of evidence was stricken.)

I think a jury could not reasonably infer from this evidence that defendant had probably received graft. Observe that defendant was but a subordinate of the Chief of Police. Even assuming that defendant allowed too much gambling to go on unmolested when he was on the squad, it may well be that he did so under orders from the Chief; judicial notice informs us that experience in many cities shows the great likelihood of such a situation. The statistics as to the number of arrests are misleading: Win-field, the government witness who supplied these figures, testified-that one single raid accounted for a very large number of arrests; that in a single raid in 1950 about 100 persons were arrested and that the figures did not show the *71number of raids as distinguished from the number of arrests.

The “Texas Ford case,” Ford v. United States, 5 Cir., 210 F.2d 313, 318 — cited by Judge HINCKS in another connection —was a pre-Holland decision. The defendant was ■'Chief of Police. On its face, the court’s opinion might seem to support Judge HINCKS as to the proof of probable graft. But, in the instant case, the government in its brief itself says of that case: “While it does not appear in the court’s opinion, the record in that case discloses that the defendant there had received gifts of clothing from the operators of some of the gambling establishments.”2 3 In the case at bar, there is not a syllable of such evidence.

It is this circuit’s doctrine that, in a criminal case, there is no difference, when it comes to granting a judgment of acquittal, between a “preponderance” and “beyond a reasonable doubt,” that the difference is for the jury only. I have recently said why I disagree'with that doctrine. United States v. Masiello, 2 Cir., 235 F.2d 279. But in the instant case, I think the difference unimportant because I think that, if probable grafting was an indispensable element of the government’s case, the government’s proof of that fact was so flimsy that it did not approach anything remotely like a preponderance.

As I read Judge HINCKS’ opinion, he is not willing to hold that, if graft were an essential element, the evidence on that issue would be sufficient. For otherwise he would surely have simplified his opinion by affirming on the quite simple theory presented by the judge to the jury (which Judge HINCKS mentions only in passing and then in a footnote). Judge HINCKS sedulously avoids affirmance on that ground. Instead, he elaborately, over several pages, works out a different theory which is as follows:

(1) There was no need whatever here for the government to prove a “likely source,” because the evidence was entirely adequate to support a finding that defendant had unreported receipts during the indictment years which must have derived from “some source” or other, without specifying any.

(2) Consequently, the government made a case for the jury without the evidence of probable graft, i.e., the jury could reasonably have found defendant guilty if no evidence of probable graft had been received.

(3) However, evidence of “possibility of graft” was properly received as “corroborative” of the evidence that the unreported taxable income derived from some unidentified source or other, corroborative in the sense that it added what was not essential, i. e., proof of a particular likely source.

I shall, for the moment, assume that Judge HINCKS correctly says that no proof of a “likely source” was needed here, that it sufficed to prove “some source” without proof of any particular prohable source.3 If so, the evidence of possible grafting was admissible only if it tended to show (although unnecessarily) a particular likely source. But, as noted above, that evidence was so exceedingly slim that it could not tend to prove that fact.

Therefore, the reception of that evidence should have led to a mistrial or, at the very least, that evidence should have been stricken from the record. For it was most prejudicial since, on the basis of that flimsy evidence, a jury of Rochester citizens were allowed to hear arguments by the prosecutor that the accused had been a grafting Rochester policeman, and the judge’s charge heavily underscored the importance of that evidence.

*72Judge HINCKS attempts to get rid of this error by relegating to a footnote that part of the trial judge’s charge which made probable graft a central issue, and by then saying in that footnote that that charge “was too favorable to the defendant." Surely the phrase “too favorable” is a quaint way of characterizing a charge centering on prejudicial evidence that should not have been in the case. Had the case gone to the jury without the evidence concerning graft and under a charge contrived in accord with Judge HINCKS’ theory, the jury might or might not have found defendant guilty. But I think we cannot properly affirm a conviction resulting from a conjectural verdict based on an imaginary charge with reference to an imaginary record, i. e., a record not containing the prejudicial evidence about graft. Judge HINCKS seems to regard this error as “harmless,” for he apparently seeks to exorcise it by stating that we must look to “the totality of proofs," their “overwhelming substantiality,” and must avoid “irrelevant abstractions.” I think the markedly prejudicial error in the reception of the graft evidence, emphasized by the judge’s charge, is no “irrelevant abstraction.” It is the very kind of error which could easily acedunt for the verdict. Kotteakos v. United States, 328 U.S. 750, 66 S.Ct. 1239, 90 L.Ed. 1557, admonishes us that such an error is not “harmless,” that we may not affirm a conviction merely because, reading a cold record, we believe the defendant would have been convicted if the error had not occurred.

Moreover, I gravely doubt the validity of a basic postulate of Judge HINCKS’ theory. That is, I question whether Holland v. United States, 348 U.S. 121, 75 S.Ct. 127, 99 L.Ed. 150, ever permits á conviction in a net-worth case without proof of a “likely source,” which (to . quote Judge HINCKS) “the defendant had actually drawn upon.” Cf. United States v. Costello, 2 Cir., 221 F.2d 668, 670-671. True, in Holland, the Supreme Court said, 348 U.S. 137-138, 75 S.Ct. 136, that proof of such a source is “sufficient” ; and one can argue that there is a difference between what is an essential and what is “sufficient.”

But in Holland, the Court, 348 U.S. at page 137, 75 S.Ct. at page 136, carefully pointed out that there was a “likely source” in the income from -the hotel. In discussing (348 U.S. at page 138, 75 S.Ct. at page 137) United States v. Johnson, 319 U.S. 503, 63 S.Ct. 1233, 87 L.Ed. 1546, the court said that there “the taxpayer was the owner of an undisclosed business capable of producing taxable income”, and that in Holland there was such a “disclosed business” (i. e., the hotel). Also, 348 U.S. at page 126, 75 S.Ct. at page 131, the Court said: “In each of the four cases decided today the allegedly unreported income comes from the same disclosed sources as produced the taxpayer’s reported income * *

Judge HINCKS, I think, misinterprets Smith v. United States, 348 U.S. 147, 75 S.Ct. 194, 99 L.Ed. 192. I think it makes clear (1) that proof of a “likely source” is necessary and (2) that even a defendant’s affirmative admissions, during a pre-indictment investigation, must be corroborated. In the Smith case, the evidence showed the following: The year before the indictment years, defendant had acquired a racing-news service, described by the Court as “a business, producing unrecorded amounts of income.” See 348 U.S. at pages 149, 159, 75 S.Ct. at page 200. In the course of a preindictment investigation, defendant had made a written statement affirmatively admitting what hé called his “true net worth.” The government relied, in large part, on this written admission. The Court said that, “whether we consider the statement an admission of one of the formal ‘elements’ of the crime or of a fact subsidiary to the proof of these ‘elements' ", it was “crucial” to, “an element vital to,” the government’s case. As the admission was made after the commission of the alleged crime, the Court held it required corroboration because, though not “involuntary,” still “its reliability” was “suspect,” since it was “extracted from one * * * under presure of” an *73investigation, “whose words may reflect the strain and confusion attending his predicament rather than a clear reflection of his past.” In discussing, 348 U.S. at pages 156-157, 75 S.Ct. at pages 198, 199, the “quantum of corroboration,” the Court said, “all elements of the offense must be established by independent evidence or corroborated admissions,” adding at once, “but one available mode of corroboration is for the independent evidence to bolster the confession itself and thereby prove the offense ‘through’ the statements of the accused. * * * Under the above standard the Government may provide the necessary corroboration by introducing substantial evidence, apart from petitioner’s admissions, tending to show that petitioner willfully understated his taxable income.” This entire discussion is, in its context, concerned exclusively with proof of corroboration of an affirmative admission. The same is true of the court’s subsequent discussion, 348 U.S. at pages 158-159, 75 S.Ct. at page 200. There the Court said, “But substantiating the opening net worth is just one method of corroborating these extrajudicial statements”, and went on to state that the “admissions may also be corroborated by an entirely different line of proof' — by independent evidence concerning petitioner’s conduct during the prosecution period, which tends to establish the crime of tax evasion without resort to the net worth computations.” The Court then summarized the evidence of his conduct “coincident with petitioner's opening of the racing-news service.” This conduct, the Court concluded, “does corroborate the net worth statement by tending to show that the petitioner was understating his income during the prosecution years.” Here, again, the Court spoke of the necessity, in one way or another, of corroborating the defendant’s admissions, and again referred to a business which was a “likely source.”

However, let us assume, arguendo, that in an appropriate case, it is not essential to prove a “likely source.” Even so, I have these doubts about Judge HINCKS’ reasoning, which, as I understand it runs thus:

For proof of a likely source, the’ government could substitute evidence, in the form of statements emanating from the defendant, which negated the possible existence of enough non-taxable sources to account for the apparently unreported taxable income. This substituted essential element of the government’s case was not proved by defendant’s testimony, since he did not take the stand. It was, says Judge HINCKS, sufficiently proved by evidence of defendant’s lies during the course of the pre-indictment investigation.

If, then, I understand Judge HINCKS, these out-of-court lies became an essential part of the government’s case. Judge HINCKS’ reasoning therefore disturbs me: Assuming arguendo, that proof of a likely source was not necessary, then, if defendant, out of court, had made affirmative admissions of facts tending to show unreported taxable income (or lack of a sufficient “cash hoard” or the like), such admissions, if corroborated, would have sufficed. But defendant’s out-of-court statements were, at most, lies, by way of denying such affirmative facts. Do such negative statements measure up to admissions? To put it differently, assuming again that proof of a likely source was not needed, do these negative statements (falsehoods) suffice, even if corroborated, to support the verdict?

I think not. As above noted, the Smith case calls for corroboration of even an affirmative admission. The reason for such a requirement is also stated, in Holland v. United States, 348 U.S. 121, at pages 128-129, 75 S.Ct. 127, at page 132, where the Court observed that, in the course of a pre-indictment investigation, the taxpayer “may be more concerned with a quick settlement than an honest search for the truth”, and said, “The problem of corroboration, dealt with in the companion case of Smith v. United States, 348 U.S. 147, 75 S.Ct. 194 [99 L.*74Ed. 192], and United States v. Calderon, 348 U.S. 160, 75 S.Ct. 186 [99 L.Ed. 202], therefore becomes crucial.” Note, then that (as in Smith) in United States v. Calderon, 348 U.S. at page 164, 75 S.Ct. at page 188, the defendant had made pre-indictment affirmative admissions, and that, 348 U.S. at page 165ff, 75 S.Ct. at page 188, there was evidence corroborating these admissions.

To be sure, if there were enough independent evidence, aside from pre-indictment lies (negative statements), then, of course, evidence of those lies would be admissible as cumulative evidence. Like flight, or a false alibi, or spoliation of documents, such lies are proof of a guilty mind. But, absent independent evidence otherwise sufficient to prove facts essential to a verdict of guilt, I doubt whether evidence of lies, flight, false alibis or spoliation will support such a verdict, especially in a net-worth case. The reason for not accepting such evidence as proof of an essential element was stated by Judge Shaw in Com. v. Webster, 5 Cush., Mass., 295, 316: An “innocent man, when placed in a condition of suspicion and danger, may resort to deception.” See also, Ayala v. United States, 1 Cir., 268 F. 296, 300; Vick v. United States, 5 Cir., 216 F.2d 228, 232.

Judge HINCKS apparently disagrees with what I said in the preceding paragraph. He cites, inter alia, Holland v. United States, 348 U.S. 121, 75 S.Ct. 127, 99 L.Ed. 150; Friedberg v. United States, 348 U.S. 142, 75 S.Ct. 138, 99 L.Ed. 188, and United States v. Calderon, 348 U.S. 160, 167, 75 S.Ct. 186, 99 L.Ed. 202. I think my discussion of Holland and Smith, supra, shows that they do not bear him out. I see nothing on the subject in Friedberg: there the defendant took the stand and “vacillated” in his testimony, see 348 U.S. at page 143, 75 S.Ct. at page 139. Speaking of the defendant’s “false and misleading extrajudicial statements”, Judge HINCKS surprisingly remarks that the Supreme Court, “with regard to similar evidence”, stated in United States v. Calderon, 348 U.S. 160, 167, 75 S.Ct. 186, 190, “There could hardly be more conclusive independent evidence of the crime.” Judge HINCKS has overlooked the fact that this statement of the Supreme Court refers to “respondent’s testimony at the trial” conflicting with his “testimony at a former trial.” 4

Judge HINCKS also cites several circuit court cases. All but one (i. e., United States v. Adonis, 3 Cir., 221 F.2d 717) are (1) unrelated to a net-worth problem and/or (2) ante-date the Supreme Court decision in Holland and its companion decisions:

(a) Sasser v. United States, 5 Cir., 208 F.2d 535 was pre-Holland. Moreover, the evidence of the out-of-court lies was but one among several factors enumerated by the court, and not a factor essential to support the verdict.

(b) Gariepy v. United States, 6 Cir., 189 F.2d 459, 463 was preHolland. Moreover, the court said the defendant had made affirmative “admissions”; the court also said they were “admissible” because they “reinforced permissible inferences from evidence at the trial.”

(c) United States v. Simone, 2 Cir., 205 F.2d 480, 482, was not a net-worth case; also, it was pre*75Holland. It relied on Smolin, United States v. Smolin, 2 Cir., 182 F.2d 782, 786, “and cases there cited.” So turn to Smolin, 2 Cir., 182 F.2d 782, 786. It, too, was not a net-worth case, and was pre-Holland. It cited and relied on Tucker v. United States, 151 U.S. 164, 14 S.Ct. 299, 301, 38 L.Ed. 112. But in Tucker the defendant testified at the trial, and the Supreme Court said that his out-of-court statement was used to “contradict his testimony upon the stand.”

The one post-Holland net-worth case which seems to support Judge HINCKS is United States v. Adonis, 3 Cir., 221 F.2d 717. If it does, I think it wrong. See McCormick, Evidence (1954) 538-539. See also Morgan, 59 Harvard L. Rev. (1946) 480 at 557 to the effect that the “greater number of cases hold that spoliation does not amount”' to “evidence of necessary facts not otherwise appearing in the case”; and see Maguire and Vincent, Admissions from Spoliation or Related Conduct, 45 Yale L.J. (1935) 227.

So I think, that, even on Judge HINCKS’ theory, the judge should have granted defendant’s motion for a judgment of acquittal at the close of the government’s evidence. Defendant, however, did not stand on the denial of this motion, but introduced evidence in his defense. If this evidence had contained anything affirmative which supplied the missing essential of the government’s case, the jury’s verdict would have been proper, on Judge HINCKS’ theory. But there was no such evidence introduced by the defense. The testimony of Mrs. Ford, a witness for the defense, did not do the trick. To be sure, the jury may well have believed that she lied. However, as she was but a witness, not a defendant, her lies constituted a negative only. They cannot, I think, supply affirmative proof, for the government, of a missing essential. Even in a civil case in most jurisdictions, “disbelief of testimony is not the equivalent of proof of facts contrary to that testimony”; Boice-Perrine Co. v. Kelley, 243 Mass. 327, 137 N.E. 731, 733; Zarrillo. v. Stone, 317 Mass. 510, 58 N.E.2d 848, 849; Cruzan v. New York Cent. & H. R. R. Co., 227 Mass. 594, 116 N.E. 879, 880; Pariso v. Towse, 2 Cir., 45 F.2d 962, 964. Consequently, although the defense put in evidence, that evidence did not strengthen the government’s case. Wherefore, even on Judge HINCKS.’ theory, we should reverse.5