ARTHUR J. STANLEY, Jr., District Judge.

The appellant was convicted in the United States District Court for the District of Utah, of violation of 18 U.S.C.A. § 2312, commonly called the Dyer Act. He appeals, raising three points of claimed error.

The first point relied upon is that the trial judge erred in denying the motion to suppress evidence. The facts relating to this specification are as follows:

Late in the evening of April 11,1965, a Utah highway patrolman observed the appellant driving ninety miles an hour in a fifty-mile-an-hour zone. The officer stopped the appellant’s car. Upon stopping him, the officer ordered the appellant out of the car, and searched him. The officer then obtained from the appellant his driver’s license and a certificate of registration. The certificate purportedly showed the issuance to the appellant of a California license of the same number, ITF-754, as that on the car, but contained obvious erasures and alterations. Shortly thereafter the appellant was formally told that he was under arrest for speeding. Subsequently, a metal identification plate was taken from the doorpost of the car the appellant had been driving. The car was then towed to a garage and the appellant was taken into custody. The next day, two agents of the Federal Bureau of Investigation went to the garage and obtained the semi-confidential or “secret number” from the car. No search warrant was procured authorizing this search.

The appellant’s motion to suppress the certificate of registration, the metal identification plate, and the secret number was heard by the trial court before the trial.

As to the registration certificate and the metal identification plate, the motion was denied. The trial court, relying on Henry v. United States, 361 U.S. 98, 80 S.Ct. 168, 4 L.Ed.2d 134 (1959), and Kelley v. United States, 111 U.S.App.D.C. 396, 298 F.2d 310 (1961), held that the actual arrest preceded the formal words of arrest and became complete when the appellant was stopped and searched. Under the circumstances, this ruling was correct. In the Henry case, although *785 dealing with the problem in another context, the Court states:

“The prosecution conceded below, and adheres to the concession here, that the arrest took place when the federal agents stopped the car. That is our view on the facts of this particular case. When the officers interrupted the two men and restricted their liberty of movement, the arrest, for purposes of this case, was complete.” (361 U.S. p. 103, 80 S.Ct. p. 171).

The Supreme Court limited its holding to the case in question, but the reasoning is apposite here. In this case, the appellant was stopped by the officer, ordered from the car, and was searched. His liberty was restrained, and he undoubtedly felt that he was under arrest. The actual arrest took place at that time, although the formal words were not uttered until a few moments later.

Thus, the arrest preceded the search and seizure of the certificate of registration and the metal plate, and if the arresting officer had probable cause to make the arrest, the contemporaneous search incidental thereto was reasonable and therefore lawful. The officer testified that he had observed the appellant going ninety miles an hour. This was sufficient to give him probable cause to make an arrest for speeding, and it was for speeding that the arrest was made.

The arrest, probable cause existing therefor, served the function of a search warrant authorizing a search for things seizable in connection with the offense for which the appellant was arrested. Papani v. United States, 84 F.2d 160 (9th Cir. 1936). While it is true that the certificate and plate were unrelated to the offense, the search which, produced them was not invalidated by their discovery and seizure. Harris v. United States, 331 U.S. 145, 67 S.Ct. 1098, 91 L.Ed. 1399 (1947). Nor were the officers required to return them to the appellant because they were unrelated to the offense for which the appellant was arrested. Abel v. United States, 362 U.S. 217, 80 S.Ct. 683, 4 L.Ed.2d 668 (1960). In Cook v. United States, 346 F.2d 563 (10th Cir. 1965), this court has said:

“ * * * The offense premising the arrest was committed in the officer’s presence and routine and lawful police procedures were followed throughout. A bona fide arrest for one offense often reveals evidence of another and more serious offense and is not rendered illegal because of the prior suspicions or knowledge of the police concerning such other offenses.” (p. 565).

In that case the defendant had been arrested for the statutory misdemeanor of operating an automobile having a faulty muffler. A search of his car revealed a false Selective Service Notice of Classification. A judgment of conviction under 50 App.U.S.C. § 462(b) (2) was affirmed. This case is governed by the same rule. Insofar as the metal plate and the registration certificate are concerned, the motion to suppress was properly denied.

The “secret number” presents a somewhat different problem. The government conceded that under the rule of Simpson v. United States, 346 F.2d 291 (10th Cir. 1965), the search was illegal. Thus, direct evidence of the number was not to be introduced. However, the appellant contends that the only way the car could be identified at all was through the use of the secret number, and under the “poison fruits” doctrine, no evidence of identification of the car should have been allowed. The appellant also alleges error in the eventual introduction of the secret number into evidence.

It is true that the number was introduced, but this was brought about by the appellant himself. As part of his defense, the appellant established that the secret number had been obtained and that it differed from the number on the metal identification plate. At that point, it was proper for the government, in cross-examination, to establish that the number was substantially identical to the number on the tag, and that the *786 difference was not material. The appellant cannot complain, after having himself opened the door.

As to the identification of the car through the secret number, the trial court deferred ruling until it could be established that the number either was or was not used in identifying the car. The evidence showed that the only number used in identifying the car was the number on the metal plate, that the secret number was not used, and that the secret number would only have confused the issue. In view of that evidence, the motion to suppress was properly denied. There can be no “poison fruits” where there are no fruits.

The second specification of error concerns the identification of the car appellant was driving in Utah as being the same car as that stolen in California. Mr. Franco, the owner, testified that his car was stolen from a parking lot in San Francisco. That car was a 1963 Pontiac Lemans with the identification number 263S6850. The California license number on the car when stolen was BKW-813. He further testified that when he regained possession of his car he recognized it by several identifying features (e. g., color, six-inch scratch on front hood, faulty turn indicator, cigarette burn in the front seat) and that it was “exactly the same as when it was stolen.”

The car Massey was driving when arrested was also a 1963 Pontiac Lemans, but bore California license plates numbered ITF-754. The testimony established that the identification number on the metal plate was 263S6850, although it originally had been 163S6850 (the first digit apparently having been altered at the factory). There was further testimony by the police officers that the metal plate on the ear had holes burned at both ends, and was attached with a soft adhesive, while such plates usually are attached by riveting. The secret number was 163S6850, but the secret number on Franco’s automobile was not shown. Thus, there was evidence that both cars were 1963 Pontiac Lemans cars, each bearing the identification number 263S6850.

The appellant relies upon Tyler v. United States, 323 F.2d 711 (10th Cir. 1963), in which this court held that proof that a car was stolen in one state and a car of the same make and type was found in another, is not sufficient identification to sustain a conviction. In the Tyler case, there was no evidence of identical serial numbers. In this case, there was direct evidence of identity. It was for the jury to weigh that evidence in the light of the surrounding circumstances, including the temporary nature of the fastening of the metal plate. In commenting on the evidence, the trial court pointed out that a key question for the jury to consider was whether the car being driven by the appellant was the car stolen in California. The jury by its verdict answered that question adversely to the appellant.

The trial judge twice allowed the government to reopen its case. The trial court has a large discretion with respect to order of proof and permitting a party to reopen after it has rested. See Lucas v. United States, 343 F.2d 1 (8th Cir. 1965). The testimony allowed on both occasions related to identification of the automobile. There is no suggestion that this surprised the appellant, or that any further preparation was necessary to meet this testimony. The appellant did not request a continuance. Under these circumstances, the trial court did not abuse its discretion in allowing the government to reopen.

The appellant contends that the trial judge participated too actively in the conduct of the trial, especially in suggesting necessary proof to the prosecuting attorney. This court has recently stated, in Ayash v. United States, 352 F.2d 1009 (10th Cir. 1965), that:

“ * * * The trial Judge is not a mere moderator or umpire in the trial of a case in federal court, and, within reasonable bounds, he has the right *787 to participate in eliciting the truth. He should, however, be careful not to become an advocate for any of the parties.” (p. 1010).

In this case, the trial judge suggested that a Pontiac dealer should be able to shed more light on the subject of identification, and the government, heeding the suggestion, produced a factory representative of the manufacturer as a rebuttal witness. We note that the judge also assisted the appellant by reminding his counsel that he had failed to move for acquittal at the close of the appellant’s case. The activity of the trial judge was not that of an advocate, but clearly was designed to get all the available facts fully and fairly before the jury. In so acting, he did not abuse his discretion.

The final specification of error relates to the instructions and the trial court’s comments on the evidence. The appellant contends that the court erred in not giving an instruction to disregard statements made by one of the witnesses. This instruction was not requested, and if it had been requested, refusal to give it would not require reversal. The statements were stricken from the record when made and the jury admonished to disregard them. A further instruction was not necessary.

The appellant contends that in the comments on the evidence the court erred in saying, “it would only be a softheaded person that would say there wasn’t a car stolen in California, because the owner testified it was stolen.” A federal judge may comment on the evidence, if he leaves the issues of facts to the jury’s decision. United States v. Waldman, 240 F.2d 449 (2d Cir. 1957). The trial court properly charged the jury that the jurors were the sole fact finders and that the court’s comments on the evidence were not binding on them. Here there was no abuse of discretion, and no error.

The appellant further complains that the court instructed repetitiously on the subject of the permissible inference that could be drawn from possession of recently stolen property. An examination of the record discloses that the subject was mentioned in different contexts and was not prejudicially repetitious.

We have carefully reviewed the record, and find no errors or defects affecting the appellant’s substantial rights.

The judgment of conviction is affirmed.