(concurring specially in part).
While my doubt on the sufficiency of the evidence here to make out guilty knowledge is great, it is not sufficient to override the weight of the contrary view of my brothers. Consequently, while I am not moved to dissent there are, to me, some disturbing things about these cases made on “presumptions” or “inferences” from assumed possession which may warrant some comments. It is likely that their value, if they have any, will come entirely from the freshness, at this stage, of my approach — a naivete which, I suppose, many judges undergo in the early days of their career as they react with considerable surprise to supposedly entrenched principles in fields beyond their former competence. But the newcomer probably has a contribution to make — that is the essence of the Inspector General approach in the military and, of course, it is always the stranger who sees the spot on the parlor rug.
This case points up the clash of tensions between the demands of an orderly, law-abiding society as it shapes its rules to combat lawbreakers, whose stealth and methods conceal action from conventional testimonial proofs, and the continuous necessity that such society must provide, and use, adequate safeguards against punishment of a citizen except upon a charge that a specific law has been violated and proof showing that it is the law, and not others that has been offended. Here, the Appellant was convicted and sentenced to a substantial imprisonment for transporting a stolen automobile from Tallapoosa, Georgia, to Oxford, Alabama, knowing that the vehicle was stolen, 18 U.S.C.A. § 2312. The decisive issue of knowledge that it was stolen rests wholly upon the traditional presumption that possession of stolen property charges one with knowledge of its stolen character and, in turn, the indispensable ingredient of possession rests almost altogether on the thin fact that Appellant, as a hitchhiker, was momentarily driving the car in the place of the one having general custody of the vehicle.
The court’s opinion sketching in detail all of the revolting circumstances shows how unappealing must the accused have been in the eyes of the jury. I am convinced that, free as they were under the instructions of inference of guilty knowledge, the jury convicted Appellant, as he had convicted himself of being a liar, a social derelict, and an irresponsible drunkard. But he was charged not with these things indeed could not have been in that court. He was charged with transporting a vehicle knowing it to have been stolen. It is the serious task to make certain that he was convicted because he consciously transported a stolen vehicle, and not, disreputable as he paints himself to be, convicted because he is that sort of person.
But what is the proof that Appellant had knowledge that it was a stolen vehicle? On this phase, I certainly do not think that the record warrants any find*942ing that Appellant was the initial thief.1 Not being the thief and the record lacking any testimonial evidence that he knew it was stolen, the conviction stands alone on the inference from possession.2 The questions then are: Was there requisite possession ? Was the inference reasonably warranted here?
Separation into two parts aid analysis, although the problem is likely a single one. Here, the possession, if it exists, must be the manual driving of the automobile. While we conclude that here the evidence is adequate to constitute possession, this is not to lay down any inexorable fiat. ' I think it important to emphasize that what we are saying is. but another instance of the familiar but fundamental concept that a Federal Trial Judge at all times, in all stages of the trial, must make c'ertáin that a naked legal rule has factual relevance, and its application is surrounded with continuous safeguards. Whether the presumption can arise, whether it is permissible requires a determination- that the total circumstances give rational assurance to the feeling that in the abundant experience of mankind, the conclusion sought to be implied has the intrinsic mark of probability and reasonably convincing weight. Tot v. United States, 319 U.S. 463, 63 S.Ct. 1241, 87 L.Ed. 1519; Bollenbach v. United States,3 326 U.S. 607, 66 S.Ct. 402, 90 L.Ed. 350.
In this light, I think much more must be shown than mere driving of a vehicle by one temporarily in the car as a hitchhiker. First, it lacks the essential characteristics of possession.4
*943Driving can be and frequently is in the ordinary experience of men, merely the exercise of a temporary control over the physical movements of the vehicle. While the fact of driving by one alone in the vehicle might under many situations afford a basis for men to believe that such driver has, or is claiming to exercise, the right of dominion or full control over its disposition or use and thus has possession of it,5 this does not hold, without more, for one temporarily driving for another then in the vehicle.
Nor, assuming that driving amounts to possession, do I think the transitory presence of a hitchhiker, without more gives him any reasonable basis in the light of human experience, to know that the vehicle has been stolen. If a hitchhiker merely by taking the wheel of his host’s automobile becomes infused with knowledge of the full history and genealogy of the car, the same result would follow for garage attendants, filling station employees, or others temporarily driving it. No distinction could be based on the status of hitchhiker in contrast to a vehicle-service employee unless courts are to say that all hitchhikers, as a class, are suspect — a fact which ignores another contemporary social phenomenon. The fact simply is that momentarily driving an automobile in the apparent custody, ownership, and control of another does not, cannot, communicate knowledge.
If that is so, it is not, in my view, enough in the superintendence of the administration of Federal Criminal Law for courts to be content that one so accused will find his vindication in the explanation of his innocence, Wilson v. United States, supra. Indeed, this case now serves as a classic example of the risk to liberty when a person may stand convicted unless he brings forward a convincing explanation. This unfortunate social misfit is not alone in that large group of people who, no matter how innocent in fact of the specific charge, from a confused and disorderly past, a chronic inability to tell the truth or relate a convincing story, can never hope to make a satisfactory explanation. Before a conviction is finally sustained, I think we must somehow be able to persuade ourselves that, conceding he had made no satisfactory explanation, had indeed created doubts, confusion, and much distrust by his conflicting and loose versions — when all of this was behind and put to rest — it all added up somehow to allow the jury genuinely to conclude that Appellant was aware that a week before, the car had been stolen in New Orleans and did not belong, as Smith said, to him or his sister. But the fact that he may have been with *944Smith continuously May 9 and 10, or on the 10th and 11th, may have engaged in riotous conduct with him through those days or hours does not, without more, supply what driving alone is inadequate to imply. I entertain substantial doubt that either driving on May 11 or his association with Smith or the automobile on the days immediately preceding will be adequate to afford a rational conclusion that he knew the car was stolen.6
It is the classic application of the constitutional presumption of innocence, made here and by all careful trial judges, that the. presumption of innocence attends the defendant at all times and at all stages — does not, and cannot, ever leave or abandon him.
This case has been my shock treatment into a sense of consciousness that where the crucial fact of (a) theft, (b) knowledge of stolen character, and (c) interstate transportation can all be established by “possession” this constitutional guarantee can become as sounding brass or a tinkling cymbal unless the judges, trial and appellate, are certain that possession really exists and furnishes a rational basis for the conclusions to be drawn.
. Many cases reason, that possession of recently stolen property gives rise to a presumption that the possessor (1) knew it was stolen, and (2) was the thief. Prince v. United States, 6 Cir., 217 F.2d 838; United States v. Stirsman, 7 Cir., 212 F.2d 900; Battaglia v. United States, 4 Cir., 205 F.2d 824; United States v. Guido, 2 Cir., 200 F.2d 105; Word v. United States, 10 Cir., 199 F.2d 625; Morandy v. United States, 9 Cir., 170 F.2d 5; United States v. Washington, D.C.Md., 69 P.Supp. 143.
In my view the total circumstances must furnish a legitimate basis for such an inference, and where not plausible in the setting of a particular record, the rule of law cannot supply the needed proof. Here, the theft takes place in New Orleans over 500 miles away, and whatever Appellant’s demonstrated moral shortcomings are, nothing permits me to believe that this boy ever got close to Louisiana. His trail was never a subtle or concealed one, and had he been there the PBI would undoubtedly have picked it up.
. The bare rule, plain in terms, is universally recognized: Wilson v. United States, 162 U.S. 613, 16 S.Ct. 895, 895, 40 L.Ed. 1090, “Possession of the fruits of crime, recently after its commission, justifies the inference that the possession is guilty possession, and, though only prima facie of guilt, may be of controlling weight, unless explained by the circumstances or accounted for in some way consistent with innocence.” Prom this court, Levi v. United States, 5 Cir., 71 F.2d 353, 354, “ * * * Possession of goods recently stolen justifies the inference that the possessor has guilty knowledge of the theft, in the absence of-explanatory facts or circumstances consistent with his innocence. * * * ”
See also: Dyer Act, 18 U.S.C.A. § 2311 et seq.; Niederluecke v. United States, 8 Cir., 47 F.2d 888; Bruce v. United States, 8 Cir., 73 F.2d 972; United States v. Di Carlo, 2 Cir., 64 F.2d 15; and Drew v. United States, 2 Cir., 27 F.2d 715, but probably modified, see Court of Appeals opinion United States v. Bollenbach, 2 Cir., 147 F.2d 199, reversed, 326 U.S. 607, 66 S.Ct. 402, 90 L.Ed. 350; and see cases footnote 1 supra.
Possession and Receipt Stolen Goods: Najjar v. United States, 5 Cir., 152 F.2d 965; Janow v. United States, 5 Cir., 141 F.2d 1017.
Miscellaneous: Tractenberg v. United States, 53 App.D.C. 396, 293 F. 476; McNamara v. Henkel, 226 U.S. 520, 33 S.Ct. 146, 57 L.Ed. 330; Andrews v. United States, 5 Cir., 157 F.2d 723; Roberts v. United States, 5 Cir., 151 F.2d 664.
. Prince v. United States, supra; Battaglia v. United States, supra; Morandy v. United States, supra; and United States v. Washington, supra, seem to doubt that Bollenbach would be significant. I think it, at least, challenges the supposed presumption and inference to the closest scrutiny.
. 72 C.J.S., Possession, p. 233, “ * * * denoting custody coupled with a right or interest of proprietorship; and ‘possession’ is inclusive of ‘custody,’ although *943‘custody’ is not tantamount to ‘possession’. In its full significance, ‘possession’ connotes domination or supremacy of authority. It implies a right and a fact; the right to enjoy annexed to the right of property, and the fact of the real detention of a thing which would be in the hands of a master or of another for him. It also implies a right to deal with property at pleasure and to exclude other persons from meddling with it. Possession involves power of control and intent to control, and all the definitions contained in recognized law dictionaries indicate that the element of custody and control is involved in the term ‘possession.’ * * * ” See especially Pearson v. United States, 6 Cir., 192 F.2d 681, mere driving not possession of truck’s cargo; mere sitting in truck not sufficient. United States v. O’Brien, 7 Cir., 174 F.2d 341; United States v. Hodorowicz, 7 Cir., 165 F.2d 218, 220, “ * * * possession is the exercise of such a power over a thing as attaches to lawful ownership * * * ”; Toney v. United States, 62 App.D.C. 307, 67 F.2d 573, 574; United States v. Gardner, 7 Cir., 171 F.2d 753; Johnson v. United States, 8 Cir., 195 F.2d 673; see § 2513 footnotes 2, 5, Wigmore on Evidence; Camilla v. United States, 6 Cir., 207 F.2d 339; Caringella v. United States, 7 Cir., 78 F.2d 563.
. Nor can this deficiency be remedied on the theory that Appellant acquires a “possession” from aiding and abetting Smith’s possession. To be aiding and abetting possession, it must be knowing it to have been stolen. And assistance, assuming that to be so, cannot be turned into a vicarious possession to prove that as a fact on which to base the subsequent inference of knowledge. Johnson v. United States, supra; Pearson v. United States, supra.
. Despite the sweeping language frequently used, I do not think courts will convict or sustain conviction where all that is available is the presumption and in*945ferenee. If so, a defendant could be convicted on simple proof (a) vehicle stolen in one state and, (b) defendant temporarily present in or near vehicle in second state, since the crucial issues of (1) theft, (2) transportation, (3) guilty knowledge would arise by inference from (b).
The following cases reflect action where, in totality, the application of inferences made something to be that which was not probable in fact: Johnson v. United States, supra; United States v. Gardner, supra; Ezzard v. United States, 8 Cir., 7 F.2d 808; United States v. Ruffino, 2 Cir., 67 F.2d 440, cited and approved, Andrews v. United States, supra, and all others, including those cited elsewhere herein, show on eareful examination some additional circumstances such as altered or forged title papers, motor numbers, grossly reduced sales price, use of assumed names, and, most significant, admitted or adequately proved possession of a real and demonstrable dominion over the car or thing.