MERRITT, Circuit Judge.
This case raises issues of first impression in our court concerning the scope of the federal enterprise racketeering statute, 18 U.S.C. §§ 1961-68 (1976). The law was enacted as Title IX of the Organized Crime Control Act of 1970 and is popularly knov/n as RICO, an acronym for “Racketeer Influenced and Corrupt Organizations,” the heading under which it appears in the crim[263] inal code. RICO’s central aim is to prevent and punish the financial infiltration and corrupt operation, through patterns of racketeering activity, of “legitimate business operations affecting interstate commerce.” Iannelli v. United States, 420 U.S. 770, 787 n. 19, 95 S.Ct. 1284, 1294, 43 L.Ed.2d 616 (1975). The question in this case is whether the statute may also be applied to persons engaged in racketeering activity unrelated to any legitimate organization but in furtherance of something the government terms “a criminal enterprise.” We hold that it may not.
I.
After a jury trial in the United States District Court for the Southern District of Ohio, on an indictment containing 329 counts, the nine appellants — Carl Sutton, Jr., Joseph Elkins, Dyeatra Carter, Edwin Adams, Otis Hensley, Prince Albert Rankin, Samuel Harris, Charles Cravens, and Viola Holmes — were each convicted of conducting the affairs of an “enterprise” affecting interstate commerce through a pattern of racketeering activity, 18 U.S.C. § 1962(c), and of conspiracy to commit that offense, 18 U.S.C. § 1962(d). Each was also convicted of one or more counts of using the telephone to facilitate drug offenses, 21 U.S.C. § 843(b), and of various substantive drug offenses, primarily possession and distribution of heroin, 21 U.S.C. § 842(a)(1). In addition, Adams was convicted of seven counts of mail fraud, 18 U.S.C. § 1341, and of transporting and receiving stolen property in interstate commerce, 18 U.S.C. §§ 2314-15; and, Hensley was convicted of eight counts of mail fraud, thirteen counts of receipt by a convicted felon of firearms shipped in interstate commerce, 18 U.S.C. § 922(h), and of unlicensed dealing in firearms, 18 U.S.C. § 922(a).
The government’s evidence showed both a significant heroin distribution business and a large-volume stolen property fencing operation. They were centered in the Cincinnati, Ohio area, and involved many of the same participants.
The central figures in the narcotics distribution business were appellant Sutton and Herschel Weintrub, who was not tried in the instant prosecution. Sutton, with the aid of appellant Holmes, purchased heroin from Elkins and Carter in Cleveland with money supplied by Weintrub. The drugs were redistributed by appellants Rankin, Craven, Adams, Hensley and Harris.
Weintrub, Hensley and Adams comprised the fencing operation. Weintrub’s role again was apparently that of financier. Adams and Hensley actually marketed the stolen property, principally household goods. There was evidence that the goods were supplied by several burglary rings over which Hensley and Adams once claimed control to an undercover government agent.
It was the government’s theory of the case that these were not discrete criminal ventures but were merely separate departments of a unitary “criminal enterprise” under the management and control of Weintrub and Sutton. For example, there was evidence that Adams often sold both heroin and stolen property to a single customer in the same transaction. Adams told one such buyer, a government informant, that the stolen goods he had on hand— stoves in that instance — had been provided by Hensley and that the heroin he was selling was supplied by “Carl” (Sutton) and “Herschel” (Weintrub). Weintrub, Hensley, Harris, and Sutton were often observed by government surveillance agents visiting Adams’ place of business, a jewelry store, from which the heroin and stolen property were usually sold. During court-authorized electronic surveillance Adams and Hensley, and Adams and Weintrub, frequently were overheard discussing both the narcotics and the fencing operations in a single telephone conversation.
The telephone interceptions also revealed that Adams and Weintrub assisted Hensley in obtaining false receipts for jewelry, appliances, and other items of personal property which Hensley had reported stolen from his home in a burglary. Hensley used the receipts to collect insurance money on [264] the items, and the mailings made in connection with the insurance claims formed the basis of the mail fraud counts. The proceeds of the fraud apparently were applied to a debt Hensley owed Weintrub for narcotics.
A warrant-authorized search of Hensley’s residence during the closing days of the investigation uncovered several ledger books documenting transactions in firearms and stolen property. Entries in one of the ledger books formed the basis for Hensley’s convictions of receipt by a convicted felon of firearms that had been shipped in interstate commerce.
II.
Appellants’ main contention is that RICO was intended to proscribe only the infiltration and operation of legitimate enterprises through patterns of racketeering activity, something the government concedes was not involved in this case.1 Appellants argue that the statute does not reach a group of individuals like themselves, who “merely” have committed a series of racketeering offenses. To say the least, the argument lacks surface appeal, for it asks us to embrace the rather ironic proposition that racketeers should be immune from criminal liability under the statute so long as they keep their activities wholly illegitimate. We suspect that largely explains the hostile treatment the argument has received in the other courts of appeals that have considered it. United States v. Rone, 598 F.2d 564 (9th Cir. 1979); United States v. Elliott, 571 F.2d 880 (5th Cir.), cert. denied sub nom. Delph v. United States, 439 U.S. 953, 99 S.Ct. 349, 58 L.Ed.2d 344 (1978); United States v. Altese, 542 F.2d 104 (2d Cir. 1976), cert. denied, 429 U.S. 1039, 97 S.Ct. 736, 50 L.Ed.2d 750 (1977); United States v. Morris, 532 F.2d 436 (5th Cir. 1976); United States v. Hawes, 529 F.2d 472 (5th Cir. 1976); United States v. Cappetto, 502 F.2d 1351 (7th Cir. 1974), cert. denied, 420 U.S. 925, 95 S.Ct. 1121, 43 L.Ed.2d 395 (1975). Upon analysis, however, we do not find appellants’ proposed construction of RICO quite as unusual as it might seem at first blush. Indeed, we think there are compelling reasons to adopt it.
A.
Our analysis begins with the language of the statute. Section 1962(c), under which appellants were convicted, provides in pertinent part:
It shall be unlawful for any person employed by or associated with any enterprise * * * to conduct or participate, directly or indirectly, in the conduct of such enterprise’s affairs through a pattern of racketeering activity. . (emphasis added)
“Racketeering activity” is defined in section 1961(1) as including numerous federal offenses and, in addition, any offense involving murder, kidnapping, gambling, arson, robbery, extortion or drugs punishable under state law by imprisonment for more than one year. A “pattern of racketeering activity” is defined by section 1961(5) as requiring at least two acts of racketeering committed within ten years of each other.
The government’s argument is straightforward and relies entirely upon the text. The government first notes that the statute on its face does not distinguish between “legitimate” and “illegitimate” enterprises but instead, by its express terms, applies to “any enterprise.” “Enterprise,” the government then points out, is defined broadly in section 1961(4) to include “any individual, partnership, corporation, association, or other legal entity, and any union or group of individuals associated in fact although not a legal entity.” These provisions dispose of appellants’ claim, according to the government, since the evidence demonstrates that appellants were a “group of individuals associated in fact” and that each committed the required number of racke[265] teering offenses while in that association. To summarize the government’s theory of the case, the evidence showed the existence of a “single enterprise — operated for the purpose of making money from repeated criminal activity.”2
Of course, it is beyond dispute that the statute must be read to cover “any enterprise.” But what does that mean? The flaw we see in the government’s approach lies in its deceptively literal treatment of the statutory definition of the term “enterprise.” What parades under the guise of rigorous fidelity to the text turns out, upon examination, to read the “enterprise” element entirely out of the statute.
The dictionary meaning of “enterprise” is any “undertaking” or “project,” i. e., some activity, and the term is also commonly used to describe a unit of organization established to perform any such undertaking or project. The statute defines “enterprise” only in the latter sense. Section 1961(4) catalogues the kinds of organizational units that may, for statutory purposes, qualify as an “enterprise” — anything from legal entities such as corporations or partnerships, to entities without formally recognized legal personalities such as “any union or group of individuals associated in fact,” even to “any individual.” However, the statutory definition is silent regarding what attributes or activities these units must assume or undertake before they may be deemed an “enterprise” in any meaningful sense. Obviously, every “individual” or “group of individuals,” considered in the abstract, is not an “enterprise.” Individuals and groups do not become “enterprises” except in relation to something they do. The statutory definition of “enterprise” contained in section 1961(4) is incomplete because it does not tell us what that “something” is.
The government would finesse the problem, and its theory of this case neatly illustrates the point: appellants were a “group of individuals associated in fact” around numerous patterns,of racketeering activity and therefore constituted a statutory “enterprise,” organized for the purpose of profiting from racketeering activity. Thus, in the government’s view, the “something” this group of individuals did to transform themselves into an “enterprise” is provided by their racketeering activity. In short, appellants’ enterprise was racketeering.
The government has successfully applied the statute in the same fashion on numerous other occasions. For example, in United States v. Cappetto, supra, the defendants were charged with engaging in a pattern of racketeering activity “consisting of participating on two or more occasions in an illegal gambling business” in the conduct of the affairs of an enterprise, “viz., an illegal gambling business.” 502 F.2d at 1355. In United States v. Morris, supra, the “enterprise” was described as “a group of individuals associated in fact to defraud in illegal card games persons who had travelled to Nevada,” and the racketeering activity through which the enterprise’s affairs was conducted consisted of running the fraudulent card games and recruiting victims to travel to Nevada. 532 F.2d at 442. In United States v. Hansen, 422 F.Supp. 430 (E.D.Wis.1976), aff’d, 583 F.2d 325 (7th Cir.), cert. denied, 439 U.S. 912, 99 S.Ct. 283, 58 L.Ed.2d 259 (1978), the enterprise was “for the purpose of defrauding insurance companies by the use of the mails and committing acts of arson,” and the racketeering activity consisted of defrauding insurance companies through the mails and arson. 422 F.Supp. at 433. And, in United States v. McLaurin, 557 F.2d 1064 (5th Cir. 1977), cert. denied, 434 U.S. 1020, 98 S.Ct. 743, 54 L.Ed.2d 767 (1978), the defendants were convicted of operating a “lucrative commercial enterprise specializing in prostitution” through a pattern of various prostitution offenses. 557 F.2d at 1066, 1073.
It requires no great insight to recognize that applying the statute in this fashion renders the “enterprise” element of the crime wholly redundant and transforms the statute into a simple proscription against “patterns of racketeering activity.” Under [266] the approach reflected in these cases, every “pattern of racketeering activity” becomes an “enterprise” whose affairs are conducted through the “pattern of racketeering activity.” Plainly, that is not the statute Congress has written.
The language Congress did use makes it unlawful “for any person employed by or associated with any enterprise ... to conduct . . . such enterprise’s affairs through a pattern of racketeering activity.” Surely, the draftsmen would not have opted for so complex a formulation if the legislative purpose had been merely to proscribe racketeering, without more. A straightforward prohibition against engaging in “patterns of racketeering activity” would have sufficed, and there would have been no need for a reference to “enterprises” of any sort. Although the government reminds us that the Organized Crime Control Act of 1970 “is a carefully crafted piece of legislation,” Iannelli v. United States, supra, 420 U.S. at 789, 95 S.Ct. at 1296, it would have us treat section 1962(c) as a purposeless circumlocution, written in terms of “enterprises,” and persons “employed” by them to conduct their “affairs,” but in reality directed at anyone who commits two acts of racketeering. Under this construction an individual or a group who robs two banks “for the purpose of making money” 3 commits a RICO offense.
Common sense, not to mention the first principle of statutory construction, leads us to reject the government’s reading and to seek a construction that gives some content to each element of the crime set forth in the text. The plain meaning of the words in context indicates that the reference to “enterprise” was included to denote an entity larger than, and conceptually distinct from, any “pattern of racketeering activity” through which the enterprise’s “affairs” ■might be conducted. If the “enterprise” element of the crime is to have independent meaning, but is still to encompass “criminal enterprises” as the government contends, then a “criminal enterprise” must involve something more than simply an individual or group engaged in a pattern of racketeering activity.
The problem is thus to discover what the distinction might be for statutory purposes between simple patterns of racketeering activity, which we think are outside RICO’s purview, and a “criminal enterprise,” which the government insists is within the ambit of the statute. Here the text is no help, for it does not even hint at what the standard should be for determining when racketeers have crossed the line to become a “criminal enterprise.”
In a passage which the government urges us to follow, the Fifth Circuit describes a “criminal enterprise” as “an amoeba-like infra-structure that controls a secret criminal network.” United States v. Elliott, supra, 571 F.2d at 897-98. With all due respect, we think greater precision than that is required if the statute is not to violate “the first essential of due process of law” by forbidding “the'doing of an act in terms so vague that [persons] of common intelligence [would] necessarily [have to] guess at its meaning and differ as to its application.” United States v. Culbert, 435 U.S. 371, 374, 98 S.Ct. 1112, 1114, 55 L.Ed.2d 349 (1978) quoting Connally v. General Construction Co., 269 U.S. 385, 391, 46 S.Ct. 126, 70 L.Ed. 322 (1926). Although government prosecutors may be trained nowadays to recognize an “amoeba-like infra-structure” when they see one, our instincts are not so keenly developed, and we think even racketeers are entitled to know before the fact at what point their criminal activities will be deemed sufficiently “amoeba-like” to transgress the statute.
Some reasonably definite standards are ■ needed if the statute is to apply to so-called “criminal enterprises.” We turn to the legislative history to see whether it offers any guidance concerning what those standards should be.
B.
RICO’s legislative history is remarkable for the clarity with which it speaks to the [267] issue of the intended scope of the “enterprise” element of the crime. Unfortunately, these materials do not solve the problem that we think is created by the government’s proposed construction — that of finding a workable definition of “criminal enterprise.” But the materials do point the way to a construction that will insure to the “enterprise” element some content independent of the racketeering element in all potential applications of the statute. The construction unmistakably endorsed by the legislative history is the one appellants have urged — limiting section 1962(c) to the conduct of a “legitimate” enterprise’s affairs through racketeering activity.
RICO in its present form is the product of two bills introduced separately in the Senate in 1969. The first, S. 1623, the “Criminal Activities Profits -Act,” was sponsored by Senator Hruska and prohibited the investment of income derived from criminal activities into any legitimate business enterprise affecting interstate or foreign commerce. Senator McClellan introduced the second, S. 1861, the “Corrupt Organizations Act of 1969,” which proscribed the infiltration or management of legitimate organizations by racketeers. Senator McClellan succinctly explained the purpose of the measures as follows: “ . . . What we are trying to do . . . [is] to keep them [racketeers] from using that [racketeering and income derived therefrom] to infiltrate legitimate businesses and pollute the stream of commerce.”4 And, in 1969 at least, the Department of Justice apparently shared Senator McClellan’s understanding of the thrust of the proposals. Delivering the Department’s endorsement, Assistant Attorney General Wilson described the McClellan bill as one “to prohibit the infiltration or management of legitimate organizations by racketeering activity or the proceeds thereof,” which, “[l]ike [the Hruska bill], is designed to attack the infiltration of legitimate business by organized crime.”5 Wilson went on to praise Senator McClellan for drafting an “innovative approach to the problem of racketeering infiltration of legitimate business.”6
Nothing happened that reflects a departure from this original understanding during any subsequent stage of the process by which the two bills were merged into Title IX of the Organized Crime Control Act of 1970. The Report of the Senate Judiciary Committee declares that Title IX “has as its purpose the elimination of the infiltration of organized crime and racketeering into legitimate organizations operating in interstate commerce.” 7 That general statement of purpose is followed by a lengthy discussion of various examples of the subversion of business and labor by racketeers. The House report on the measure is to like effect.8 Before the House Judiciary Committee, the Department of Justice described the measure it was endorsing in the following terms:
Title IX is designed to inhibit the infiltration of legitimate business by organized crime, and, ... to reach the criminal syndicates’ major sources of revenue.
* $ # % * sk
The statute would also proscribe the acquisition, maintenance, or control of any interest in business enterprise through a pattern of racketeering activity or the collection of unlawful debts. Here the emphasis is against illegally acquired ownership or control of businesses by [268] members and associates of the Mafia 9
The floor debates on the measure are filled with similar descriptions of Title IX’s scope,10 And the formal “Statement of Findings and Purpose” preceding the substantive provisions of the Organized Crime Control Act as finally adopted declares that Congress was acting to stop the “billions of dollars” obtained “from such illegal endeavors as syndicated gambling, loan sharking, the theft and fencing of property, the importation and distribution of narcotics and other dangerous drugs and other forms of social exploitation" from being used “to infiltrate and corrupt legitimate business and labor unions and to subvert and corrupt our democratic processes,” a process which had “weaken[ed] the stability of the Nation’s economic system, harm[ed] innocent investors and competing organizations, interfered] with free competition, seriously burdened] interstate and foreign commerce, threaten[ed] the domestic security and undermine[d] the general welfare of the Nation and its citizens.”11
Legislative history is sometimes equivocal, and arguments from it may not always be dispositive of concrete issues of statutory construction. But, on this issue, we feel confident in concluding that the Congress was of one mind. Senator McClellan summed up the legislative consensus nicely during the floor debates when he characterized the special class of criminals at whom RICO was aimed as those who “operate illegitimately in legitimate channels.”12
C.
Against this background, appellants’ defense that they were “mere" racketeers and therefore not properly charged under section 1962(c) begins to make a great deal of sense. The legislative history conclusively demonstrates that RICO was enacted in response to the growing subversion of our society’s legitimate institutions of business and labor by organized crime, a relatively recent development that Congress deemed a significantly more dangerous threat to the nation’s social and economic stability than the age-old problem of crime for crime’s sake. It thus seriously misconceives this legislative purpose to argue, as some have, that limiting RICO to the corruption of legitimate enterprises “does not make sense since it leaves a loophole for illegitimate business to escape its coverage.” United States v. Altese, supra, 542 F.2d at 106-07. “Illegitimate business,” so-called, is already comprehensively proscribed and severely punished by the many provisions of state and federal law listed under RICO’s definition of “racketeering activity.” RICO’s evident purpose was to single out racketeering activity undertaken in connection with the subversion of legitimate institutions as a special case, deserving of even harsher penal sanctions. Under the view we take of the congressional scheme, appellants are not so much attempting to slip through a “loophole” in the law as they are quite properly pointing out that they did not engage in the aggravated form of racketeering activity for which RICO was exclusively designed.
Appellants' proposed construction of section 1962(c) also makes sense in terms of the logic and structure of the statute- as a whole. Subsection (a) of section 1962 prohibits the investment of income derived from racketeering activity “in acquisition of any interest in, or the establishment or operation of, any enterprise” and goes on to state that a “purchase of securities on the open market for purposes of investment” with tainted money is not unlawful so long as the holdings of the purchaser, his family, and accomplices in crime “do not amount in [269] the aggregate to one percent of the outstanding securities of any one class, and do not confer . . . the power to elect one or more directors of the issuer.” Clearly, subsection (a) is talking about legitimate enterprises only. Logic dictates that the provisions which follow — subsection (b), prohibiting the use of racketeering to “acquire or maintain . . . any interest in or control of any enterprise,” and subsection (c), prohibiting “the conduct of [an] enterprise’s affairs through a pattern of racketeering activity” — should be read in pari materia.
Most importantly, however, appellants’ proposed construction is to be preferred over the government’s because it infuses some content into-each element of the crime. All of the words of section 1962(c) take on some independent significance when the statute is applied, for example, to a shop steward who conducts the affairs of his labor union through a pattern of extortion, bribery and fraud.13 The same cannot be said for a construction that would permit the prosecution of illegal gamblers for conducting illegal gambling through a pattern of illegal gambling14 or of prostitutes for conducting prostitution through a pattern of prostitution.15
In our view, the only alternative we have to accepting appellants’ position on the scope of section 1962(c) is to rewrite the statute completely. To reiterate, the government’s approach is unacceptable because it reads the “enterprise” element out of the crime. In order to extend the statute to illicit enterprises of some description, and yet preserve some content for the “enterprise” element, we would be required to engraft upon the definition of “enterprise” contained in section 1961(4) some set of standards that would serve to warn any person or group engaged in racketeering activity when they will be deemed to have embarked upon an “enterprise” to that end.
Several considerations discourage us from choosing that course. In the first place, there is no mandate in the legislative history for us to indulge in such uneasy and speculative construction. Indeed, the unambiguous thrust of that history supports appellants’ construction and no other. Without such a mandate, we would truly be inventing from our own notions of sound legislative policy the additional elements we believe are needed to save the statute from fatal vagueness and at the same time have it apply to “criminal enterprises” in any meaningful sense. Although Congress has declared that RICO’s provisions should be “liberally construed to effectuate its remedial purpose,” we do not read that directive as authorizing us to write a new and substantially different law. Appellants’ construction fully serves the statute’s remedial purpose as revealed by all of the guides to legislative intent. It is for Congress, not the courts, to amend the statute to expand its coverage, if that is what an effective policy against organized crime requires.
Two of the canons courts traditionally follow in construing criminal statutes also counsel us not to stray from the path marked out by the legislative history. The first of these is that “ambiguity concerning the ambit of criminal statutes should be resolved in favor of lenity,” Rewis v. United States, 401 U.S. 808, 812, 91 S.Ct. 1056, 1059, 28 L.Ed.2d 493 (1971), unless the legislature has spoken “plainly and unmistakably” to the contrary. United States v. Gradwell, 243 U.S. 476, 485, 37 S.Ct. 407, 61 L.Ed. 857 (1917). When as in this case the legislative history speaks “plainly and unmistakably,” but in support of resolving the statute’s ambiguity in favor of the construction argued by defendants, the maxim applies with special force and tells us that [270] we ought not to strain to accommodate the government’s desire for a broader construction.
The other canon that guides us is that, “unless Congress conveys its purpose clearly, it will not be deemed to have significantly changed the federal-state balance.” United States v. Bass, 404 U.S. 336, 349, 92 S.Ct. 515, 523, 30 L.Ed.2d 488 (1971). To be sure, even under appellants’ view, RICO still represents a substantial incursion into state criminal jurisdiction. But the government’s construction would take us much further into areas traditionally left to state regulation by making a federal felon out of “any individual” or any member of a “group” who has committed any two of the broad range of state offenses denominated “racketeering activity” under section 1961(1). We do not seriously doubt the power of Congress to undertake such a bold expansion of federal criminal jurisdiction. But we will not simply assume that Congress has done so with this statute, when the text, at best, is ambiguous on the matter and the legislative history suggests a different construction that would not alter the traditional division of responsibilities between federal and state governments quite so radically.
For all of these reasons, we are persuaded to construe section 1962(c) in the manner appellants have suggested. We therefore hold that an “enterprise” within the meaning of the statute is “any individual, partnership, corporation, association . and any union or group of individuals associated in fact,” that is organized and acting for some ostensibly lawful purpose, either formally declared or informally recognized. Section 1962(c) is violated whenever any person associated with such an enterprise conducts its “affairs,” i. e., undertakes any activity on behalf of or relating to the purposes of the enterprise, by committing at least two criminal acts constituting a “pattern of racketeering” as defined in section 1961(5). Since appellants’ numerous acts of racketeering were not shown to have been related in any way to the affairs of such an enterprise, we reverse their convictions under section 1962(c) for conducting the affairs of an enterprise through a pattern of racketeering activity and their convictions under section 1962(d) for conspiracy to commit that offense.
III.
It remains to consider the effect of our reversal of the RICO and RICO conspiracy counts upon the validity of appellants’ numerous other convictions. These remaining counts can be divided into roughly four categories:
(1) offenses involving controlled substances, principally possession and distribution of heroin and using the telephone to facilitate those offenses;
(2) receipt and transportation of stolen property;
(3) mail fraud;
(4) unlicensed dealing in firearms.
All nine appellants were convicted of one or more narcotics offenses. Only appellants Adams and Hensley were charged with or convicted of offenses in the second and third categories, involving stolen property and mail fraud. Only appellant Hensley was charged with or convicted of firearms offenses.
We reluctantly conclude that the convictions on all of these counts must also be reversed, on grounds of misjoinder. Trying all of these counts against all of these appellants in a single proceeding was proper only if appellants were participants “in the same series of acts or transactions constituting an offense or offenses.” Rule 8(b), Fed.R.Crim.P. The government sought to supply the requisite nexus among defendants and offenses by alleging, in counts I and II of the 329 count indictment, that appellants were associated in a single “criminal enterprise,” the affairs of which were conducted by appellants through the commission of the many disparate crimes charged in the remaining counts. We have just determined, however, that the RICO and RICO conspiracy counts were fatally [271] defective, in that they failed to state offenses under 18 U.S.C. §§ 1962(c) and (d). The legal insufficiency of those two counts removes any conceivable justification for appellants’ joint trial on the other counts.
Without benefit of the “criminal enterprise” concept, rejected in Part II supra, there is simply no basis for lumping the four categories of offenses with which appellants were variously charged into a single “series of acts or transactions” within the meaning of Rule 8(b). It may be that each category of offenses (narcotics, stolen property, mail fraud and firearms), considered in isolation, would meet the “same series” test and that, accordingly, all of the defendants and offenses charged within a single, category could be jointly tried. However, the sole connection among the various categories, compared one with another, is the presence of some common participants. And even that connection is tenuous. None of the appellants was charged with offenses in all four categories, and seven of the nine were charged in only a single category, the one involving narcotics. Only Hensley and Adams were charged with offenses in more than one category: Adams, with narcotics, mail fraud and stolen property offenses, and Hensley, with narcotics, mail fraud, and firearms offenses.
Although “all of the defendants need not be charged in each count” under Rule 8(b), we would be making a mockery of the Rule were we to find Hensley’s and Adams’ participation in three of the four categories enough to transform the others, who had no connection with mail fraud, stolen property, or firearms, into participants in a single criminal transaction encompassing all of these offenses. It is true, as the government argues, that Hensley used the proceeds of his mail fraud to pay off a debt he owed Herschel Weintrub for heroin. But we fail to see how that makes the mail fraud a part of Weintrub’s heroin distribution ring or makes those guilty of operating the heroin business participants in the same “series of acts” which constituted the mail fraud. It is also true that Hensley dealt in unlicensed firearms as well as stolen property, and that he apparently documented sales of both in the same ledger books. But that fact hardly justifies characterizing appellant Adams’ trafficking in stolen property as part of the same transaction as Hensley’s illicit gun sales, and it affords no basis whatsoever for connecting the participants in the narcotics distribution ring to either the stolen property or firearms offenses committed by others. Nor could all of these offenses have been tried together on the theory that appellants were members of a single conspiracy with multiple criminal objects under 18 U.S.C. § 371. Analyzed under conventional principles of the law of conspiracy, the allegations of the indictment, and the proofs at trial, made out, at most, multiple, unrelated conspiracies,16 We think Rule 8(b) exists “to prohibit exactly what was done here, namely, allowing evidence in a case against one defendant to be presented in the case against another charged with a completely disassociated offense, with the danger that the jury might feel that the evidence*/ against one supported the charge against the other.” Ingram v. United States, 272 F.2d 567, 570 (4th Cir. 1959).
A risk of prejudice, either from evidentiary spillover or transference of guilt, inheres in any joinder of offenses or defendants. In the interests of judicial economy, our system of justice nevertheless tolerates that risk so long as the requirements of Rule 8 are satisfied, subject of course to the remedy of severance should the risk become a reality. Rule 8 “set[s] the limits of toler[272] anee,” however. King v. United States, 355 F.2d 700, 703 (1st Cir. 1966). Misjoinder under the Rule is deemed prejudicial per se: “[WJhere multiple defendants are charged with offenses in no way connected, and are tried together, they are prejudiced by that very fact . . . .” United States v. Reynolds, 489 F.2d 4, 6 (6th Cir. 1973), cert. denied, 416 U.S. 988, 94 S.Ct. 2395, 40 L.Ed.2d 766 (1974), quoting Ingram v. United States, supra, 272 F.2d at 570. And, the remedy for misjoinder “is reversal . with orders for . . . new and separate trial[s].” United States v. Reynolds, supra, 489 F.2d at 6.
We think appellants are entitled to that remedy. This is not merely a case of “retroactive misjoinder,” such as confronted the Supreme Court in Schaffer v. United States, 362 U.S. 511, 80 S.Ct. 945, 4 L.Ed.2d 921 (1960). Schaffer involved an indictment charging seven defendants with three counts of transporting stolen property in interstate commerce and one count of conspiracy to commit those offenses. The conspiracy count, alleging a connection among the three substantive counts, provided a basis for joinder under Rule 8(b). At the conclusion of the government’s case, however, the trial court dismissed the conspiracy count for insufficient evidence, and submitted the remaining counts to the jury.
The Court ruled, 5-4, that, despite the mid-trial dismissal of the count that had initially justified joinder, there had been no misjoinder under Rule 8(b). The dismissed conspiracy count had been alleged in the good faith belief that it could be proved and therefore provided a valid basis for joinder at the point when compliance with the Rule must be assessed, i. e,, the beginning of the trial. Since joinder had been proper, the question was not one of misjoinder but rather whether the trial court had abused its discretion in denying the defendants’ motions for relief from prejudicial joinder under Rule 14. Finding no abuse, the Court affirmed the convictions on the three substantive counts.
We think this case stands on an entirely different footing. Here, the counts alleged to justify appellants’ joinder — the RICO and RICO conspiracy counts — failed not on grounds of insufficient proof, but rather because they rested upon an erroneous construction of the statute under which they were brought. Had the indictment been tested against the correct construction of 18 U.S.C. §§ 1962(c) and (d), the RICO counts would have been dismissed, and, as we have already, explained, joinder of the remaining counts could not have been sustained under Rule 8. Thus, unlike the situation in Schaffer, where joinder was fully justified at the relevant time by a legally sufficient conspiracy count, appellants’ joinder was never proper under the Rule. In these circumstances, the only sensible approach is to view appellants’ joinder as having been void ab initio, and to afford them the remedy to which all victims of misjoinder are entitled.
We therefore reverse the judgments of conviction as to all of the appellants on all of the remaining counts17 and remand the cause for new and separate trials. At this point, we do not think it either desirable or possible to issue precise instructions as to how these offenses and defendants ought ultimately to be subdivided for purposes of retrial. That is properly a decision for the prosecutor in the sound exercise of his discretion, subject to scrutiny by the District Court under Rule 8. We would observe, however, that, in the event the prosecutor opts for multiple defendant retrials, prudence would counsel strongly against mixing the four categories of offenses we have identified in this opinion.
In the interest of expediting the proceedings on remand, we also note that we have reviewed, and we reject, appellants’ claims that the District Court should [273] have suppressed: (1) the fruits of judicially-authorized electronic surveillance of a telephone at appellant Holmes’ residence, because there was no probable cause to believe that Holmes herself was engaged in criminal activity18; and (2) the fruits of a search of an apartment' located at 3619 Clarion Avenue because the warrant was not supported by probable cause.19
Reversed and remanded for further proceedings consistent with this opinion.