Appellant was convicted by jury verdict on two counts alleging theft from the United States mails in violation of 18 U.S.C. § 1708 (1964). He was sentenced by a District Judge in the United States District Court for the Southern District of Ohio, Eastern Division, to two three-year terms to be served consecutively.

On appeal he contends that admission of evidence concerning his finances, the joinder for trial of the two counts of the indictment, and certain comment by the prosecutor constitute reversible error. He also claims that the evidence as a whole does not support the jury verdict. Finally, he argues that the District Judge should have disqualified himself because of previous contact with the case as United States District Attorney.

As to all of these issues, except the last one, we are convinced that there was no reversible error. As to the last issue, however, we believe that the express command of a federal statute requires reversal for new trial.

We comment first on the two issues which are bound to recur at retrial.

Appellant at the time of the events involved in the indictment was a truck driver for a company which had a "star route"1 contract with the United States Post Office Department. He drove a truck which was used in mail delivery between Columbus, Ohio and Logan, Ohio. The indictment charged in Count I that on December 14, 1965, appellant stole $36,000 in currency from the United States mails, and in Count II that on January 13, 1966, he stole $400 in coin. Both of these amounts were in mail shipments which were required to be carried by appellant's truck. The government proofs tended to show that both of these shipments were placed on appellant's truck and were not received at the station to which he should have delivered them. The $36,000 was not recovered; the $400 was.

The prosecution offered no proofs which established the exact time or circumstances of the theft of the $36,000 charged in Count I. It did, however, offer proofs of appellant's opportunity to commit the theft and it did seek to negate (mostly, of course, by self-serving testimony) the possibility that any other person had stolen the money.

But as to Count I, there was significant testimony of appellant's embarrassed financial circumstances immediately prior to December 14, 1965, and his expenditure of large sums of money immediately thereafter. This testimony, if believed, established that prior to December 14, 1965, appellant who had a wife and three children earned $70 a week, that he was heavily in debt and had been for a long time, and that he had recently and unsuccessfully tried to borrow money from his employer. In the immediate aftermath of the disappearance of the $36,000 mail shipment, on December 14 and 15 appellant paid off a substantial number of old debts, and he made a $2,000 down payment on a new car in $20 bills. A major portion of the $36,000 shipment had been in $20 bills.

The admission of evidence of appellant's impecunious state before December 14 and his sudden affluence immediately thereafter is the subject of appellant's first two appellate issues. We believe that under the total facts of this case, where there was much other evidence of guilt, this testimony was relevant and admissible. E. g., United States v. Daniels, 377 F.2d 255 (6th Cir. 1967); Young v. United States, 358 F.2d 429 (9th Cir. 1966). If the jury believed it, the jury had a right (as the District Judge charged) to draw an inference that the funds employed by appellant on December 14 and 15 to pay debts and to buy a car came from the missing $36,000.

There was, in our opinion, evidence in this case from which the jury could have inferred a "natural connection" between the missing $36,000 and the funds appellant spent immediately after December 14, 1965. Cf. Williams v. United States, 168 U.S. 382, 397, 18 S. Ct. 92, 42 L. Ed. 509 (1897).

Count II of the indictment charges appellant with theft of $400 of "bait" money. After the disappearance of the $36,000 shipment, suspicion centered upon appellant. He was questioned about his handling of the shipment, but was not arrested. On January 12, 1966, postal inspectors made up a "test sack" of $400 in coin with an address which would require it to be delivered via the star route truck which appellant drove. After the truck was loaded and dispatched, postal inspectors went to the Lancaster, Ohio, post office which should have received the package and found that it had not been delivered. They also found that it had not been dropped off at the next of appellant's regular stops at his home town of Logan. By this time appellant had parked the trailer at a parking lot and two postal inspectors approached him at a nearby service station to which he had driven the tractor. One told him, "We are looking for a mail sack" to which, according to the inspector, appellant replied, "I don't know anything about any mail sack." The inspectors then drove back to the parking lot where the trailer was parked and found the missing mail sack on the ground half way between the trailer and appellant's own car.

We find no reversible error in the joinder for trial of the two counts of this indictment. These were similar offenses. Fed.R.Crim.P. 8. Moreover, Rule 14 of the Federal Rules of Criminal Procedure places a great deal of discretion in the District Judge to grant or deny motions for severance. United States v. Vida, 370 F.2d 759 (6th Cir. 1966), cert. denied, 387 U.S. 910, 87 S. Ct. 1695, 18 L. Ed. 2d 630 (1967); United States v. Carter, 311 F.2d 934 (6th Cir. 1963). On this record we find no abuse of judicial discretion in denial of appellant's belated motion for severance. See Bayless v. United States, 381 F.2d 67 (9th Cir. 1967); Baker v. United States, 401 F.2d 958 (D.C. Cir. 1968). Cf. Drew v. United States, 118 U.S.App.D.C. 11, 331 F.2d 85 (1964).

We cannot, of course, foresee completely what a motion for severance before the required retrial will produce. What we can say is that in this case the District Judge acted well within his judicial discretion.

The last issue in this case requires the statement of some additional facts. District Judge Kinneary, who started trial of this case on June 24, 1968, had been United States Attorney for the Southern District of Ohio until his resignation on August 3, 1966, in order to be sworn in as a United States District Judge on August 5, 1966. The indictment under which appellant was tried in this case was returned on August 25, 1967, a year after the date Judge Kinneary took the bench.

During the course of the trial, apparently at a conference in chambers, Judge Kinneary disclosed to counsel for appellant that the original complaint against appellant in relation to Count II had been filed on January 13, 1966, while he was still United States Attorney. A preliminary hearing had also been held on that complaint on January 17, 1966. Judge Kinneary subsequently stated on the record that prior to trial of this case he had no knowledge of the prosecution or other contact than the purely formal contact referred to above. At the conclusion of all the proofs in this case, appellant made a formal motion for mistrial based upon the disqualification of the District Judge. On appeal appellant does not dispute the facts referred to above, but does assert that the District Judge's formal relationship to the prosecution at the time of the complaint made him "of counsel" within the meaning of 28 U.S.C. § 455 (1964).

"Any justice or judge of the United States shall disqualify himself in any case in which he has a substantial interest, has been of counsel, is or has been a material witness, or is so related to or connected with any party or his attorney as to render it improper, in his opinion, for him to sit on the trial, appeal, or other proceeding therein." 28 U.S.C. § 455 (1964).

This trial was actually based on a grand jury indictment containing two counts. Previously appellant had been the subject of a formal complaint on the facts alleged by Count II of the indictment and had been present at a preliminary hearing on that complaint. Both of these proceedings took place while Judge Kinneary was United States District Attorney. We think it must follow that such proceedings were a part of this "case."

Further, the federal statute prescribing the duties of the United States Attorney provides in pertinent part:

"Except as otherwise provided by law, each United States attorney, within his district, shall —

"(1) prosecute for all offenses against the United States; * * *" 28 U.S.C. § 547 (Supp. II, 1965-1966).

This statute has previously been interpreted (properly we think) as making the United States Attorney "of counsel" in any criminal prosecution in his district. United States v. Vasilick, 160 F.2d 631 (3d Cir. 1947); United States v. Maher, 88 F. Supp. 1007 (N.D.Me. 1950).

There are two plausible reasons for rejecting appellant's disqualification issue. The first is that in actuality appellant had a fair trial before a judge who knew nothing about his case except as the trial record developed before him. We recognize that this jury case was well tried by an able and impartial judge. The trouble with this argument is that the language of the statute is mandatory: "Any * * * judge of the United States shall disqualify himself in any case in which he * * * has been of counsel * * *." (Emphasis added.) As to this feature of the statute, prejudice is presumed whether actually present or not. We believe that the language employed by Congress compels the conclusion that Congress was concerned with avoiding the appearance of partiality as well as avoiding the fact. See Rapp v. Van Dusen, 350 F.2d 806 (3d Cir. 1965); Texaco, Inc. v. Chandler, 354 F.2d 655 (10th Cir. 1965), cert. denied, 383 U.S. 936, 86 S. Ct. 1066, 15 L. Ed. 2d 853 (1966). See also American Cyanamid Co. v. F.T.C., 363 F.2d 757, 763-768 (6th Cir. 1966).

Judge Biggs, writing for the Third Circuit, has held in a similar case:

"[W]here, as here, a judge of a district court has been of counsel for a party in the case pending before him, his disqualification is not a matter for the exercise of his own discretion but is unconditional and absolute." United States v. Vasilick, 160 F.2d 631, 632 (3d Cir. 1947).

We agree.

The second reason argued to us for not reversing this case on the issue of disqualification is that appellant through counsel waived any disqualification claim. It is clear that appellant never suggested disqualification until the issue had been raised by the District Judge himself, and then the formal motion for disqualification was presented belatedly at the end of the trial.

As we have already pointed out, however, the language of the statute is mandatory and it is addressed to the District Judge. United States v. Vasilick, supra.

We believe Adams v. United States, 302 F.2d 307 (5th Cir. 1962), is distinguishable from this case. In Adams the District Judge had resigned as United States Attorney four months before the indictment which commenced the case which he tried. The court's opinion held that he could not have been "of counsel" in this case. Hence, we do not consider the waiver discussion in Adams applicable to our instant case.

Further, whatever logical force the waiver argument might have otherwise, we believe it is completely negated by the fact that this statute originally required one of the parties to move for disqualification (see 28 U.S.C. § 24 (1946)), but in 1948 Congress amended the statute, removing this feature entirely. Judicial Code § 455, 28 U.S.C. § 455 (1964).

As the statute now stands, the duty to disqualify is placed solely upon the District Judge. Burdensome as this duty may occasionally prove to be (particularly where a District Judge has moved from the prosecutorial to the judicial function) we think the statutory mandate must be observed.

The judgment of the District Court is vacated and the case is remanded for further proceedings consistent with this opinion.