OPINION

The first count of the two-count indictment charges ten defendants with a conspiracy to import marihuana illegally, and to “receive, conceal,” etc. such marihauna, “knowing said marihuana had been imported and brought into the United States contrary to law,” all in violation of 21 U.S.C. § 176a. 1 The second count charges one of the defendants with a substantive offense under the same statute.

The problem before the court at this time centers upon the second paragraph of § 176a, which provides:

“Whenever on trial for a violation of this subsection, the defendant is shown to have or to have had the marihuana in his possession, such possession shall be deemed sufficient evidence to authorize conviction unless the defendant explains his possession to the satisfaction of the jury.”

The defendants have moved to dismiss the indictment upon the asserted ground that the “presumptions” or “statutory inferences” thus authorized violate both the right to due process and the privilege against self-incrimination guaranteed by the Fifth Amendment. 2 The crux *778 of their position is quickly stated, if less easily adjudicated: it is that marihuana grows all over the United States as well as abroad, and that possession, therefore, is not a rational or reasonable, but an arbitrary, basis for inferring (1) unlawful importation and (2) the possessor’s knowledge of such importation. To demonstrate the alleged factual foundation for this thesis, the movants (more specifically, counsel for defendant Rojas, who pulled the laboring oar) requested an evidentiary hearing.

In its opposing arguments the Government substantially conceded the widespread growth at least of uncultivated marihuana. It urged, however, that the Congress, when it authorized the statutory inferences in question, had solid evidence that nearly all marihuana smoked in the United States comes from abroad (mainly from Mexico). Stressing this proposition, and citing cases closely approaching its goal, the prosecution urged denial of the motion without a hearing. If evidence should be received on the constitutional problem, the Government further argued, there would be time enough for that when the case went to trial.

The court concluded as a preliminary matter that defendants were entitled to the evidentiary hearing they sought. This is not by any means the first assault of its kind upon § 176a. In more than one case, this contention and others closely similar have been dismissed at the threshold because the defendants made no record thought sufficient to cast doubt upon the presumption “that Congress’ enactment of the presumption in § 176a * * * is * * * reasonable * * * ” United States v. Gibson, 310 F.2d 79, 82 (2d Cir. 1962); cf. United States v. Coke, 364 F.2d 484, 485-486 (2d Cir. 1966), cert. denied, 386 U.S. 918, 87 S.Ct. 877, 17 L.Ed.2d 789 (1967); Costello v. United States, 324 F.2d 260, 264 (9th Cir. 1963), cert. denied, 376 U.S. 930, 84 S.Ct. 699, 11 L.Ed.2d 650 (1964). Warned of that pitfall, defendants here have proceeded sensibly to avoid it. They seemed correct, too, in urging that the evidentiary record they desired be made separately and in ádvance, not as part of the trial record. The question is one of constitutional law; the factual issues are not particular ones as to the guilt or innocence of these defendants, but general ones affecting the reasonableness and the rationality of the inferences Congress enthroned as statutory guides or authorizations. Resolution of these issues, while it appeared unlikely to lead in any event to dismissal of the indictment, could be expected to affect the submission of proofs and the framing of instructions to the jury.

Accordingly, the court has held the factual hearing defendants sought. Two witnesses were heard: first, for the defense, Professor Richard Evans Schultes, Curator of Economic Botany, Lecturer in Economic Botany and Director of the Botanic Museum of Harvard University; then, for the Government, Narcotic Agent William J. Durkin, currently Director of the Federal Bureau of Narcotics Permissive Section, whose 17 years with the Bureau have encompassed a variety of assignments in most of Latin, Central, and South America, various Asian countries, and many parts of the United States. In addition to the testimony of these witnesses— considering that the problem concerns *779 the broad subject of “common experience” and the “circumstances of life as we know them,” Tot v. United States, 319 U.S. 463, 468, 63 S.Ct. 1241, 1245, 87 L.Ed. 1519 (1943) — the court received without objection references to books, legislative hearings, administrative reports, and newspaper stories. The objective, of course, was to test the reasonableness of two statutory inferences from possession of marihuana — (1) importation and (2) knowledge thereof —and so it was critical to appraise pertinent indicators of what the generality of men “know” or are likely to believe (including, of course, things we know which may not be so). The discussion in the first of the following sections of this opinion summarizes the court’s findings from .the evidence, broadly conceived, as well as other data of common experience which bear upon the ultimate questions for decision.

I.

Marihauna (cannabis sativa), one of the oldest known plants, has been used through the ages for fibre, for oil, and for its narcotic principles. Until three decades or so ago, it was prescribed medicinally. Originally native to Central Asia, the plant was introduced here about the time of the American Revolution for the cultivation of hemp fibre, and it became, about 100 years or so ago, this country’s second largest non-food agricultural crop. With changing economic conditions, however, at least since as long ago as the end of World War II, the legitimate cultivation and use of the plant have ceased almost completely.

Whatever its actual or possible uses, there is botanieally but a single species of the plant. Despite the practically total end of licit cultivation, it continues to grow spontaneously or adventitiously in a substantial area of the world and. most interestingly for us, can grow in every State of the United States as well as Puerto Rico, Panama, and the Canal Zone. 3 It spreads, grows, and reproduces itself without human attention. One of the factors continuing to promote this lush growth is the former use, until recent times, of marihuana seeds in bird feed; and it appears until this day that the taste of winged creatures for such nourishment supplies continuing problems of noticeable magnitude for state and local health and drug-control officials.

It is difficult or impossible for all but the most expert analysts, using sophisticated techniques, to determine whether any given sample of marihuana comes from this or some foreign country. Nevertheless, upon the basis of enforcement experience both before and since the enactment of 21 U.S.C. § 176a in 1956, officials of the Bureau of Narcotics have concluded (and reported to Congress) that as much as 90%, or even more, of the smoking marihuana seized in this country over the years comes from abroad, primarily from Mexico, where cultivation is also illegal. These judgments are aided by the fact that Mexican producers employ characteristic methods of packaging the product. The belief that a preponderant part of the marihuana traffic here is comprised of imports is buttressed by the general impression and understanding, among narcotic agents and some popular observers, that smokers of the substance tend to prefer Mexican varieties. Not all students of the problem appear to share this view. 4

*780 On the other hand, both state and federal officials concerned with enforcement of the narcotic laws devote extensive time, energy, and publicity to the work, as part of their enforcement duties, of eradicating large areas of marihuana growth. 5 Year after year the Bureau of Narcotics reports acreages of growth destroyed. 6 The reports reflect that narcotics control must focus upon domestically grown marihuana as well as that smuggled in from abroad. In any single year, the amount of domestic growth thus destroyed could yield more marihuana than all that seized and destroyed by the Bureau and other enforcement agencies. And, of course, there are no figures for domestic growth which is not discovered and not destroyed except as this facet of the matter may be illuminated by the fact that the work continues annually with no apparent diminution over time.

Both “experts” of one kind or another and laymen are made conscious of the intensive concern of police officials over domestically grown marihuana as the “raw material” 7 for domestic use of this narcotic. In this and other ways there arises a widespread belief that the widely prevalent use of marihuana is supplied in substantial measure by domestic growth. Professor Schultes reported truthfully his judgment (whether correct or not, it is scarcely possible to know) that the marihuana so widely grown everywhere in the United States is the source of that smoked by most users in this country. In preliminary discussions some six months before passage of the bill which became the statute here in question Senator Langer of North Dakota observed that “it is very easy for a high-school boy or girl to get hold of it and thus form the habit.” 102 Cong. Rec. 274 (1956). A more deliberate and scholarly source, produced by well known authorities in its directly pertinent field, reported in 1954:

“The smoking of dried marihuana has become a fairly common drug habit during the past few years in the *781 United States. The ease with which the plant is cultivated in vacant lots and other places has made it difficult to eradicate the source of supply.” Gonzales, Vance, Helpern and Umberger, Legal Medicine: Pathology and Toxicology 835 (2d ed. 1954) (emphasis added).

That relatively authoritative pronouncement of 14 years ago is echoed repeatedly in the everyday news read by jurors (who must be told about “statutory inferences”) and judges alike in the current press. So, for example, even a casual reader of one unsensational newspaper is likely to notice a fair number of accounts within a short space of time telling about the domestic growth of marihuana, cultivated as well as uncultivated, its use by our native smokers, and the concern of both federal and state law enforcement officers with this evidently prevalent problem. 8

To summarize the general import for our purposes of the formal and informal record, we find:

(1) There is a respectable body of expert authority for the view that most marihuana found in this country has been smuggled in from a foreign country. There are respectable views to the contrary. Congress could reasonably have accepted the former view. But nobody can say with confidence how closely it accords with the facts.

(2) There is no reliable evidence showing the extent to which people “know” or believe that most illicit marihuana is imported. Even police officers and other interested government agents are by no means necessarily possessed of knowledge or a belief like the one declared to be inferable from possession under § 176a. See Costello v. United States, 324 F.2d 260, 263 (9th Cir. 1963), cert. denied, 376 U.S. 930 [84 S.Ct. 699, 11 L.Ed. 2d 650] (1964); United States v. Kapsalis, 313 F.2d 875, 876 (7th Cir.), cert. denied, 374 U.S. 856 [83 S.Ct. 1911, 10 L.Ed.2d 1077] (1963). Insofar as “common experience” may suggest an answer, it is that many people, perhaps a majority, users and non-users alike, believe that “pot” “is cultivated in vacant lots and other places” (Gonzales et al., op. cit. supra), which would account for its widespread use, regular and occasional, from the campus to the ghetto to the country club.

II.

While the enactment of an irrational or arbitrary “presumption” could not stand no matter what legislators might be thought to have believed, Tot v. United States, supra, we are obviously bound to pay respectful attention to the nature and sources of the information upon which Congress acted when it wrote § 176a. Nothing short of the closest study of this subject could entitle us even to approach, especially at nisi prius, “the exercise of the grave power of annulling an Act of Congress.” United States v. Gainey, 380 U.S. 63, 65, 85 S.Ct. 754, 756, 13 L.Ed.2d 658 (1965).

It is of interest, then, though by no means crucial, that the legislative history of the statutory inferences in § 176a, as described in a Government Memorandum of Law, “is indeed meager.” There is no explanation or support of any kind in committee reports. During Senate hearings preceding the enactment, however, Harry J. Anslinger, the Commissioner of Narcotics, reported that “most of the marihuana traffic in this country is fed from Mexico. * * * I would say,” he continued, “that over 90 percent of the cannabis that we have *782 seized has been smuggled from Mexico where they have a very considerable problem themselves and have done a lot in the destruction of clandestine fields.” 9 A Texas official estimated that two percent of the marihuana in his State was “homegrown” while the rest came from Mexico. 10 Other witnesses gave similar estimates relating to Texas, 11 which is, as the Government implicitly acknowledges, not necessarily representative of the entire country with respect to the problem of smuggling across the Mexican border. 12

Beyond these “meager” expressions there is nothing. Nobody appears to have paid the slightest attention to the separate element and additional inference respecting knowledge of importation. It was known then, as it is now, see e. g., Butler v. United States, 273 F.2d 436, 438 (9th Cir. 1959), that marihuana, whatever its source, is liable to pass through many hands from grower to smoker. It could readily have been seen that beyond the smuggler and perhaps an early middleman, those found in “possession” of marihuana were likely to be at a far remove from, and without knowledge of, the source. Yet it was and is possession alone which allows the whole gamut of statutory inferences, including the inference that the possessor knew of the illegal importation. And, of course, “the jury must be satisfied, on knowledge as on all other elements, beyond a reasonable doubt.” United States v. Llanes, 374 F.2d 712, 716 n. 4 (2d Cir.), cert. denied, 388 U.S. 917, 88 S.Ct. 24, 19 L.Ed.2d 208 (1967). It is here, in this court’s view, that the statutory inferences become “strained” and “arbi trary,” and thus invalid under the Due Process Clause. Tot v. United States, 319 U.S. 463, 467-468, 63 S.Ct. 1241, 87 L.Ed. 1519 (1943).

The familiar statement of the general principle governing “presumptions” or “statutory inferences” is that “constitutionality * * * depends upon the rationality of the connection ‘between the facts proved and the ultimate fact presumed.’ ” United States v. Gainey, 380 U.S. 63, 66, 85 S.Ct. 754, 757, 13 L.Ed.2d 658 (1965), quoting from Tot v. United States, supra. The legislatively authorized inference must be “rational” and “reasonable,” not “strained” or “arbitrary.” United States v. Romano, 382 U.S. 136, 139, 86 S.Ct. 279, 15 L.Ed.2d 210 (1965); United States v. Gainey, supra 380 U.S. at 66-67, 85 S.Ct. 754; Tot v. United States, supra 319 U.S. at 467-468, 63 S.Ct. 1241. While these words are general, and while the decisions have (perhaps surpisingly) not often distinguished explicitly between the civil and criminal contexts, see Note, Constitutionality of Rebuttable Statutory Presumptions, 55 Colum.L.Rev. 527, 531 n. 31 (1955), it is obvious from the most modestly attentive analysis of the precedents that a good deal more is required than a mere relationship sufficient to be called “relevancy” between the proved fact and the ultimate one declared to be inferable or presumable. Before the inferred relationship may serve as a basic element establishing criminal guilt, the two facts must be “very probably connected; ” the inference must be one “very likely” to be correct. See United States v. Romano, supra, 382 U.S. at 141, 86 S.Ct. 279. *783 While the point has not been, and probably cannot be, put statistically, it ought “rarely” to be the case that the inference will be contradicted by the actual experience of relationships in the real world. See United States v. Gainey, supra, 380 U.S. at 67, 85 S.Ct. 754. There must be a “broad ground of experience which supports the conclusion” to be reached by the route of the statutory inference; the experience must sustain the inference “in such a large proportion of * * * instances that a jury may be instructed that it may draw [the authorized] conclusion in the absence of any explanation.” United States v. Romano, 330 F.2d 566, 571 (2d Cir. 1964), aff’d, 382 U.S. 136, 86 S.Ct. 279, 15 L.Ed.2d 210 (1965). 13

The subject is in its nature “empirical,” and so, inevitably, must be the “process of making the determination of rationality * * *.” Gainey, supra 380 U.S. at 67, 85 S.Ct. at 757. Because the question is of this character, and because it relates ultimately to what lay juries or equally (in this respect) lay judges may be permitted to find as facts warranting criminal convictions, the test of rationality or reasonableness has been stated and applied in terms of “common experience” and “the circumstances of life as we [judges, apparently, and other non-expert adults generally] know them * * Tot v. United States, supra, 319 U.S. at 467, 468, 63 S.Ct. at 1245. Room has been left, however, of some uncertain dimensions for legislative creation of “a valid presumption * * * upon a view of relation broader than that a jury might take in a specific case.” Id. at 468, 63 S.Ct. at 1245. Evidently reaffirming that thought, the Supreme Court has recently declared that “significant weight should ne accorded the capacity of Congress to amass the stuff of actual experience and cull conclusions from it.” United States v. Gainey, supra 380 U.S. at 67, 85 S.Ct. at 757. But the test in the end is “experience” — reality. Nature and the Constitution alike bar Congress from legislating the existence of facts that do not exist. Congress cannot declare, and thus make it generally true, that felons in possession of guns have “very probably” received them from across state lines, because that is not known — and has not been shown — to be generally true. Tot v. United States, supra. For the same reason, Congress cannot authorize a jury to find from a man’s presence near a still, without more, that the man has “possession, custody and control” of the still. United States v. Romano, supra.

While it has not commonly been featured in the standard formulations (but see Mr. Justice Black, dissenting in United States v. Gainey, supra 380 U.S. at 82, 83, 85-86, 85 S.Ct. 754), it is fitting to stress that our problem is in the setting of a criminal prosecution. No authority in any relevant context has ever suggested that the statutory inference was intended, or may properly serve, to lighten the prosecution’s burden of proof beyond a reasonable doubt. On the contrary, it is routine and standard *784 practice under 21 U.S.C. § 176a, as under § 174, to instruct’the jury (or to hold in a bench trial) that no conviction may be had unless the prosecution has proved each dement of the offense — including importation and knowledge thereof — beyond a reasonable doubt. See, e. g., United States v. Gibson, 310 F.2d 79, 81 (2d Cir. 1962); United States v. Peeples, 377 F.2d 205, 211 (2d Cir. 1967); United States v. Llanes, 374 F.2d 712, 714, 716 (2d Cir. 1967). And so when we allow the drawing of an inference like the type herein issue, we must be telling juries (or ourselves as triers of fact) that the inference is one sustained by substantially more than a preponderance of the evidence, that the probability is far in excess of 50-50, that it is sufficient to warrant a reasonable man’s being convinced “to a moral certainty” of the correctness of the inference. See, e. g., United States v. Bilotti, 380 F.2d 649, 654 (2d Cir.), cert. denied, 389 U.S. 944, 88 S.Ct. 308, 19 L.Ed.2d 300 (1967).

The obvious importance of this consideration does not depend upon the premise that the requirement of proof beyond a reasonable doubt has become a part of due process — though few things are more central, for lawyers and laymen alike, among pertinent notions of fundamental fairness. See Leland v. Oregon, 343 U.S. 790, 802-803, 72 S.Ct. 1002, 96 L.Ed. 1302 (1952) (Mr. Justice Frankfurter, joined by Mr. Justice Black, dissenting); State v. Edwards, 269 Minn. 343, 130 N.W.2d 623, 626 (1964); cf. In re Whittington, 391 U.S. 341, 342, 88 S.Ct. 1507, 20 L.Ed.2d 625 (1968). It is enough that Congress and the courts have proceeded steadily with the unquestioned and repeatedly expressed view that § 176a and statutes like it leave the burden of proof unchanged. At the very least, Congress has never so much as hinted — let alone given the kind of clarion announcement suited to so unsettling a departure — that any change was contemplated. So it stands as the known and controlling rule that the burden of proof is the familiar one, announced to juries every day across the land. Surely, courts and legislatures, not less than administrators, are bound to faithful adherence to such procedural basics until or unless some valid departure is promulgated in a lawful way. Cf. Yellin v. United States, 374 U.S. 109, 120-121, 83 S.Ct. 1828, 10 L.Ed. 2d 778 (1963); United States ex rel. Accardi v. Shaughnessy, 347 U.S. 260, 74 S.Ct. 499, 98 L.Ed. 681 (1954). The force of this conclusion should be evident quite apart from the profound subversiveness of a view that would countenance our telling juries in one breath that proof of all elements must be beyond a reasonable doubt while effectively saying in the next breath that the burden on some or all elements may actually be a good deal less substantial. 14

However the details of the governing principles are stated, the inferences in § 176a must fail the constitutional test, at least in substantially pertinent part. The difficult question is whether importation should be inferable from possession alone. The less difficult question is whether possession alone should warrant an inference of knowledge of importation. This court has resolved grave doubts on the first question in favor of deferring to the always weighty judgment of the Congress. But it would pass deference and become abdication to sustain the remaining aspect. There was no basis in 1956, and there is none today, for thinking it “very *785 likely” that people found in possession of marihuana “know” it to have been imported. Whether or not most marihuana is imported, there is powerful evidence that enormous numbers of people, experts as well as laymen, “know” a good deal (or even most) marihuana found in the United States is grown and harvested here. The statutory inference is written to embrace possession and possessors generally. It is not confined in place— e. g., to Texas or other border States. It is not limited as to prospective defendants — to users, or wholesalers, or even retailers. It reaches students, housewives, slum-dwellers, the schooled and unschooled, the naive and the sophisticated, everywhere and all alike. There is no need— and no solid basis — to suggest that a more confined version would be more nearly sustainable. It is enough that Congress has crossed a forbidden line by inadvertently announcing an “arbitrary” and “strained" and highly unlikely proposition about a supposedly general matter of fact.

Insofar as § 176a proclaims a statutory inference that one found in possession of marihuana knows it to have been illegally imported, the statute is unconstitutional because it denies due process.

III.

The conclusion just stated is a grave one, not lightly to be reached by a court of first instance. The pressures against such a decision are not diminished by the fact that the statute in question is a dozen years old and has heretofore withstood similar assaults. While there is no controlling authority rested upon an evidentiary record like the one now made, see United States v. Gibson, 310 F.2d 79, 82 (2d Cir. 1962), there are strong intimations in the books of a drift of judicial sentiment opposed to the result herein. It should be observed, therefore, that the precedents have been studied with respectful care and found to supply no justification for a course other than the one now followed.

An early, and therefore “leading,” decision is Caudillo v. United States, 253 F.2d 513 (9th Cir.), cert. denied sub nom. Romero v. United States, 357 U.S. 931, 78 S.Ct. 1375, 2 L.Ed.2d 1373 (1958). There, recognizing that § 176a leaves undiminished the Government’s burden of proof beyond a reasonable doubt, the Court said (p. 517):

“If we agree that by far the larger part of all marihuana found within the United States is imported then it is a reasonable inference that it is probable that one in possession of marihuana is in possession of imported marihuana. And, a fortiori, it can be inferred that the person in possession knows of the illegal importation.”

With all respect, the inference from possession to knowledge of importation does not follow, “a fortiori” or otherwise. It does not follow “logically,” of course, because the question is one of fact, not logic. It does not follow empirically because, as has been indicated, the empirical data are at best equivocal, and a long way from the kind of high probbility required to sustain a finding of proof beyond a reasonable doubt.

This vital point is buttressed, in an indirect but revealing way, by a later opinion of the Ninth Circuit, Butler v. United States, 273 F.2d 436 (1959), where the unsuccessful appellant asked that Caudillo be overruled. In stating its reasons for rejecting the appeal, the Court began by pointing out the important prosecution need supplied by § 176a (p. 438):

“The government has no practical method to trace back through one or a dozen hands to the person who originally grew the weed. If it had the means to so trace the paths of commerce to the plant’s origin, there would be no need of any rule of evidence presuming importation, for importation could either by proved by the government, or the government would establish the marijuana as home-grown, and the government’s case would fail.”

*786 But that unquestionably correct analysis, again with deference, proves far too much. Congress had (at least so far as it disclosed) no pragmatic basis whatsoever for supposing that a possessor of marihuana is any better equipped, or even as well equipped, as the government “to trace back through one or a dozen hands to the person who originally grew the weed.” This court — which may perhaps be said to deal more intimately, week in and week out, with narcotics problems than most members of the Congress — cannot pretend not to know, judicially or otherwise, what it knows of “connections” and “pushers” and the other miscellaneous figures in the trade. There is powerful reason to disbelieve, rather than to believe, that random people found with marihuana could know (or, a fortiori, could prove) its ultimate source.

The “comparative convenience of producing evidence of the ultimate fact” is at best “but a corollary” of the “controlling test,” the demonstration of “a rational connection between the facts proved and the fact presumed * * Tot v. United States, supra 319 U.S. at 467, 63 S.Ct. at 1245. But here the corollary test, like the controlling one, serves to condemn the enactment. Suitable cases for “presumptions” are those where the truth of the fact in question, and the existence of evidence about it, is almost certain to be within the defendant’s ken — as, for example, what he himself was doing near a still, United States v. Gainey, supra, how he himself came to be in possession of something, Morrison v. People of the State of California, 291 U.S. 82, 91 n. 4, 54 S.Ct. 281, 78 L.Ed. 664 (1934), or whether he himself had a license or permit of a specified kind, ibid. Such instances are utterly unlike a requirement of “going forward” concerning a chain of transfers linking unknown people at distant removes in a world of stealth and shadows. And it bears reemphasis that even relative ease of proof (the “corollary”) — as, what the defendant was doing near a still — is not even reached where the controlling test of high probability (rationality) is unsatisfied. United States v. Romano, supra. 15

The decisions sustaining the provisions here in issue rely commonly upon Yee Hem v. United States, 268 U.S. 178, 45 S.Ct. 470, 69 L.Ed. 904 (1925), which upheld, as applied to opium, the much older statute (now 21 U.S.C. § 174) from which § 176a was copied. An obvious and sufficient distinction is that opium has never been produced in the United States at all, and has long been associated, in popular literature and “folklore,” with faraway places. Cf. Erwing v. United States, 323 F.2d 674, 679-682 (9th Cir. 1963); but cf. United States v. Coke, 364 F.2d 484 (2d Cir. 1966), cert. denied, 386 U.S. 918, 87 S.Ct. 877, 17 L.Ed.2d 789 (1967). It is a far cry from that case, allowing an inference that defendant “knew” what every relevant source would teach him, to this one, where there is a large and steady stream of information teaching the contrary of what is presumed.

*787 IV.

It remains to state briefly the limited consequences of what the court now decides. The partial invalidation of the inferences in § 176a does not lead, as plaintiffs contend, to dismissal of the indictment. To begin with, the continued vitality under today’s decision of the inference of importation actually leaves, in strictest theory, “some” evidence relevant to every element before the grand jury. Furthermore, there are many cases, see, e. g., Caudillo v. United States, supra — and this case is responsibly represented to be one of them— where the Government in fact has evidence beyond the statutory inferences for the elements in question. Finally, and most importantly, there is no later learning that effects any pertinent limitation upon the rule of Costello v. United States, 350 U.S. 359, 363, 76 S.Ct. 406, 408, 100 L.Ed. 397 (1956), that indictments are not “to be held open to challenge on the ground that there was inadequate or incompetent evidence before the grand jury * * See also United States v. Weber, 197 F.2d 237, 238 (2d Cir.), cert. denied, 344 U.S. 834, 73 S.Ct. 42, 97 L.Ed. 649 (1952); Carrado v. United States, 93 U.S.App.D.C. 183, 210 F.2d 712, 717-718 (1953), cert. denied, Atkins v. United States, 347 U.S. 1018, 74 S.Ct. 874, 98 L.Ed. 1140 (1954).

What does follow, however, from the decision of defendants’ motion is a departure from the proofs normally required and the standard instructions given in the cases of this kind. Subject to revisions that may be ordered by higher authority, 16 the contents of a prima facie case and the court’s charge to the jury will reflect the requirement that there be substantial evidence beyond the statutory inference to show that a defendant found in possession of marihuana had knowledge of its importation. It is not useful to predict or speculate as to the form such evidence may or should take. It is sufficient for now to hold, and the court does, that in the absence of such independent evidence, the prosecution will have failed to make out a case for the jury.

Subject to that anticipatory ruling, and having allowed the making of the record defendants sought for the preservation of their asserted rights, the court denies the motion to dismiss the indictment. It is so ordered.