[619]*619GILLETTE, J.
In this case involving criminal convictions for ex-convict in possession of a firearm and possession of a controlled substance, we are once again presented with an occasion to discuss an aspect of Oregon’s developing state law of search and seizure. We previously applied federal law, including Oregon cases interpreting federal law, to affirm these convictions. State v. Flores, 58 Or App 437, 648 P2d 1328 (1982). The Supreme Court accepted review and remanded the case to us for reconsideration in light of Oregon constitutional principles announced in State v. Caraher, 293 Or 741, 653 P2d 942 (1982). State v. Flores, 294 Or 77, 653 P2d 960 (1982). On remand, we look not only to Caraher but also to a later decision, State v. Lowry, 295 Or 337, 667 P2d 996 (1983), and to other relevant state constitutional principles and precedent, at least to the extent that we can divine them. We affirm.
OREGON CONSTITUTIONAL SEARCH AND SEIZURE LAW
The Oregon Supreme Court’s recent shift in direction on the use of the state constitution in criminal cases presents this court with difficult problems in its day-to-day decision-making, problems which we must face in this case. We are without recent state law precedents in many areas and are uncertain of the extent to which recent federal decisions may also express Oregon constitutional principles. We are therefore left to develop those principles with guidance only from a few recent Supreme Court cases and from that court’s traditional approach to the development of search and seizure law. That approach, as we will see, includes such fluctuations, indecisiveness and ambiguity that it is of little help to us, while the recent cases prove, on analysis, to stand for little, as yet.
Before the mid-1960’s, the Oregon Supreme Court construed only state constitutional provisions, because the federal Fourth Amendment did not apply to the states. Wolf v. Colorado, 338 US 25, 69 S Ct 1359, 93 L Ed 1782 (1949), overruled in Mapp v. Ohio, 367 US 643, 81 S Ct 1684, 6 L Ed 2d 1081 (1961), and in Ker v. California, 374 US 23, 83 S Ct 1623, 10 L Ed 2d 726 (1963); see Barron v. Baltimore, 32 US (7 Pet) 243, 8 L Ed 672 (1833). In determining what the Oregon Constitution required, the court frequently cited cases from [620]*620the federal system and from other states, finding precedents from other jurisdictions helpful but not binding. See, e.g., State v. Duffy, 135 Or 290, 295 P 953 (1931); State v. Lee, 120 Or 643, 253 P 533 (1927); State v. Laundy, 103 Or 443, 495-497, 204 P 958, 206 P 290 (1922). Although in a few instances it appeared to treat federal cases as controlling, see State v. DeFord, 120 Or 444, 250 P 220 (1926), the court generally kept the distinction between the two sovereigns and their separate constitutions clear, adopting federal principles when they seemed appropriate.
One example of the Oregon court’s approach before the 1960’s is its handling of the question of the exclusion of improperly obtained evidence from use in a criminal trial. The court first appeared to reject the exclusionary rule of Weeks v. United States, 232 US 383, 34 S Ct 341, 58 L Ed 652 (1914), in State v. Ware, 79 Or 367, 154 P 905, 155 P 364 (1916). Then, in State v. Laundy, supra, it appeared to adopt the rule, albeit by dictum, stating that it did so for the same reasons which commended it to the federal Supreme Court, not because it was bound to follow the federal lead. As the court undoubtedly knew, many state courts at that time had rejected Weeks. Yet, despite Laundy, the question of exclusion in fact — as opposed to an abstract solution — remained unresolved for another forty years. The original majority opinion in State v. McDaniel, 115 Or 187, 231 P 965, 237 P 373 (1925), expanded on the Laundy dictum and ordered suppression of evidence but, on rehearing, a different majority found the search good and treated suppression as an open question. The court refused to decide the question through 1959, see State v. Hoover, 219 Or 288, 347 P2d 69 (1959); State v. Flynn, 137 Or 8, 299 P 694, 300 P 1024 (1931), although trial courts had suppressed evidence as early as 1924 and continued to do so thereafter. State v. DeFord, supra; State v. Lanegan, 192 Or 691, 699, 236 P2d 438 (1951); see also Elkins v. United States, 364 US 206, 80 S Ct 1437, 4 L Ed 2d 1669 (1960) (evidence which Oregon state courts suppressed could not be used in federal prosecution if state officials obtained it in violation of Fourth Amendment standards); see Patterson, A Case for Admitting in Evidence Liquor Illegally Seized, 3 Or L Rev 334, 340 (1924); Note, Evidence — Admissibility of Illegally Obtained Evidence — The Law in Oregon, 39 Or L Rev 368 (1960). In [621]*621sum, the court avoided the issue by never finding a search invalid, at times straining hard to avoid doing so.
Only after Mapp v. Ohio, supra, which applied the federal exclusionary rule to the states, did the Oregon court, in State v. Chinn, 231 Or 259, 373 P2d 392 (1962), unequivocably state that unconstitutionally obtained evidence should be suppressed, and not until State v. Elkins, 245 Or 279, 442 P2d 250 (1966), did it ever find a search improper and reverse a conviction for failure to suppress. Throughout the entire period before Mapp and Ker v. California, supra, it normally treated this issue, and all other search and seizure questions, under the Oregon Constitution alone, looking to other jurisdictions for guidance but not accepting their decisions as controlling.1
During that period, the court dealt with a number of major issues in search and seizure law. They included searches incident to arrest, State v. McDaniel, supra; State v. Quartier, 114 Or 657, 236 P 746 (1925); State v. Laundy, supra; State v. McDaniel, 39 Or 161, 65 P 520 (1901), and its permissible scope, Keeler v. Myers, 119 Or 517, 249 P 637 (1926); the requirements for a search warrant, State v. Flynn, supra; Nally v. Richmond, 105 Or 462, 209 P 871 (1922); Smith v. McDuffie, 72 Or 276, 142 P 558, 143 P 929 (1914); and what constitutes probable cause. State v. Christensen, 151 Or 529, 51 P2d 835 (1935); State v. Duffy, 135 Or 290, 295 P 953 (1931). To be sure, the specific holdings in most, if not all, of these cases, while generally consistent with federal and state law of the period, are significantly more generous concerning police authority than are post-1960 decisions based on federal law; the changes in federal law alone in the interim make it questionable whether the Oregon Supreme Court would still consider them good law under the Oregon Constitution.
Whatever their present status, however, the pre-Mapp and pre-Ker Oregon cases necessarily established an analysis of Article I, section 9, of the Oregon Constitution that was independent of the federal constitution. That independent analysis began breaking down soon after Mapp and [622]*622Ker, as the Supreme Court, rather than continuing to develop the meaning of the Oregon Constitution, simply decided cases under the federal constitution. In State v. Chinn, supra; State v. Krogness, 238 Or 135, 388 P2d 120, cert den 377 US 992 (1964), and State v. Elkins, supra, the court relied primarily on state law, but thereafter it increasingly treated United States Supreme Court precedent not only as establishing the constitutional minima but also as stating the maximum extent of constitutional protection. Thus, the Oregon Constitution lost its independent significance at the very time of a major revamping of search and seizure law under the federal constitution.
There were occasional reminders in dissents that Oregon could act independently. In State v. Chinn, supra, which was clearly decided under the Oregon Constitution, Judge O’Connell opposed the broad scope the majority allowed for a search incident to arrest and urged that the Oregon court adopt the rule of United States v. Trupiano, 334 US 699, 68 S Ct 1229, 92 L Ed 1663 (1947), under the Oregon Constitution, despite its rejection in United States v. Rabinowitz, 339 US 56, 70 S Ct 430, 94 L Ed 653 (1953):
“[W]e have our own constitution to interpret and in interpreting it we are not bound by the cases decided by the United States Supreme Court if we propose to afford our citizens a greater protection against unlawful search than that required in those cases.” 231 Or at 288 (O’Connell, J., dissenting).
Judge O’Connell repeated that point in his dissent in State v. McCoy, 249 Or 160, 172, 437 P2d 734 (1968), a case which the majority decided primarily, if not exclusively, on federal grounds. Judge Sloan also argued in his dissent in State v. Cartwright, 246 Or 120, 147, 418 P2d 822 (1967), that Oregon could adopt different rules than the federal Supreme Court requires. It is not clear, however, whether he wished the court to construe the Oregon Constitution’s limitations on police authority more strictly than the federal or to adopt an independent construction of the federal constitution. (The Oregon Supreme Court several times indicated its belief that it was not bound by the federal Supreme Court’s construction of the Fourth Amendment, the last time in State v. Florance, 270 Or 169, 182, 527 P2d 1202 (1974). It apparently dropped this misconception only when the Supreme Court specifically [623]*623disapproved the Florance language in Oregon v. Haas, 420 US 714, 719 n 4, 95 S Ct 1215, 43 L Ed 2d 570 (1975).)
Despite Judge O’Connell’s and Sloan’s statements, and possibly as a result of its misconception of its authority in construing the federal constitution, by the late 1960’s the basis of the Supreme Court’s search and seizure decisions was often hazy. At times, as in State v. McCoy, supra, it relied solely on federal law. At other times, as in State v. O’Neal, 251 Or 163, 444 P2d 951 (1969), and State v. Cloman, 254 Or 1, 456 P2d 67 (1969), it cited no constitutional provision whatsoever. It also, as in State v. Keller, 265 Or 622, 510 P2d 568 (1973), and State v. Blackburn/Barber, 266 Or 28, 511 P2d 381 (1973), cited both constitutions but made no effort to distinguish them or to provide a separate analysis of either. Cf. State v. Atkinson, 64 Or App 517, 669 P2d 343 (1983) (analyzing whether later federal precedents have undercut Keller and determining that it remains valid under the state constitution). In effect, the court treated searches and seizures as a generalized common law subject controlled by federal principles in which the state constitution has no independent role. The older state cases simply dropped from sight, along with any attempt to develop their principles in light of more recent constitutional formulations.
This movement away from the state constitution reached its climax in State v. Florance, supra, in which the court held that, in an attempt to provide one rule for law enforcement officers to follow, it would adopt the rule on searches incident to arrest of United States v. Robinson, 414 US 218, 94 S Ct 467, 38 L Ed 2d 427 (1973), as the law of Oregon. Thereafter, the court adopted other federal holdings, apparently as a matter of course. See cases cited in State v. Caraher, supra, 293 Or at 748; State v. Nettles, 287 Or 131, 135 n 2, 597 P2d 1243 (1979); State v. Flores, 280 Or 273, 570 P2d 965 (1977). While it continued in these cases to cite the state constitution, the court gave it no independent content and referred to it only as part of a rote incantation. At other times it cited only federal law in making its decisions, see cases cited in State v. Caraher, supra, 293 Or at 748-49 n 7. As Judges O’Connell and Sloan had opposed the movement away from the state constitution in the 1960’s, the court’s position in post -Florance cases provoked opposition from Judges Lent and Linde, see cases cited in State v. Caraher, supra, 293 Or at [624]*624750, but, until Caraher and State v. Lowry, supra, the Florance position appeared firm.2 Now, the court has reversed its course, but the result of its previous treatment of the state constitution leaves us with no ready way to determine the law we are now to apply.
As a practical matter, search and seizure law under Article I, section 9, of the Oregon Constitution stopped developing in the 1960’s, but search and seizure law in other jurisdictions has changed drastically since then. The Oregon court has referred to many of those changes in post-Florance cases, but it has not adopted them in what it apparently would now consider a principled fashion; it simply declared a certain federal rule to be the law of Oregon without further analysis. See, e.g., State v. Nettles, supra, 287 Or at 135 n 2. We might nevertheless assume that the rules that it adopted under the state constitution in the post -Florance cases remain the law of Oregon, because it claimed to construe the state constitution in adopting them. Cf. State v. Atkinson, supra, 64 Or App at 521-22 (State v. Keller, supra, a pre-Florance case, remains state law despite later United States Supreme Court cases casting doubt on its correctness under the federal constitution). However, the court in State v. Caraher, supra, and State v. Lowry, supra, did not simply abandon Florance’s position that the state constitution should follow the federal; it also overruled Florance’s specific holding on the issue in question and revived State v. O’Neal, supra, which Florance had overruled, as the law of Oregon. In so doing, it relied entirely on pre-Florance state constitutional cases. It thus appears that post -Florance cases in fact decided nothing under the Oregon Constitution and that the applicability in Oregon of the last 15 years’ growth of federal search and seizure remains an open question. The Oregon court always treated federal cases respectfully, even if it chose not to adopt them, but, in part because of the controversial nature of recent federal decisions [625]*625and in part because of its recent decisions, we cannot say that it will continue to do so rather than seeking an independent analysis in opinions of United States Supreme Court minorities, other state courts, or commentators. Cf. State v. Kennedy, 295 Or 260, 666 P2d 1316 (1983) (establishing an Oregon double jeopardy rule different from both the United States Supreme Court majority and minority).
The Supreme Court’s treatment of Florance, and its failure to cite post-Florance cases in Lowry or Caraher, convinces us that it will treat post -Florance but pr e-Caraher cases as federally-based and as deciding nothing under the Oregon Constitution, even when they purport to adopt a federal rule as the Oregon constitutional rule. We are thus returned to first principles, and what once appeared settled turns out not to be. For example, the impact of such cases as Katz v. United States, 389 US 347, 88 S Ct 507, 19 L Ed 2d 576 (1967) (the Fourth Amendment protects privacy interests, not property rights), and Chimel v. California, 395 US 752, 89 S Ct 2034, 23 L Ed 2d 685 (1969) (limiting the scope of search incident to arrest) on Oregon search and seizure law remains to be determined. It is quite possible, for instance, that the Oregon court will decide to stick closer to the constitutional language than has the Supreme Court and will not follow the federal lead in Katz and its progeny in totally abandoning a property rationale for Article I, section 9. See State v. Elkins, supra, 245 Or at 288-89 (Art I, § 9, protects property rights as well as privacy rights).
It is also possible, as a result of the reliance on pre-Florance cases, that in some respects Oregon law may be less protective of an individual’s rights than is federal law. See State v. Kennedy, supra, 295 Or at 270-71. If so, federal law will control, although the Supreme Court has not always carried through on this analysis. For example, in State v. Caraher, supra, the court relied primarily on the approach of State v. Chinn, supra, in determining that a search was proper as a search incident to an arrest under the Oregon Constitution. Yet the specific search that the Chinn court approved would clearly be impermissibly broad under Chimel v. California, supra. In Caraher, the court also cited State v. Krogness, supra, even though the search in that case was held to violate the federal constitution in a federal habeas corpus action soon after the state decision. United States ex rel Krogness v. [626]*626Gladden, 242 F Supp 499 (D Or 1965). It is particularly unclear, in the light of the state cases on which it relied, why the Caraher court, after upholding the search under Oregon law, failed to go on and evaluate the search under federal law— part of the methodology it should always follow if it finds no state constitutional violation and a federal constitutional violation is also asserted. See State v. Kennedy, supra.
In short, the Supreme Court has now told us to place the Oregon Constitution in a hothouse and force it to make up for years of stunted growth which that court caused. Because of the changes in federal law in those years, because of the Oregon court’s history of looking to federal cases for guidance, and because the court in former years approved many searches it would not now approve, we cannot assume that pre-1970 Oregon cases remain the final word on the Oregon Constitution. Worse yet, we also cannot treat recent Oregon cases, even those which mention the state constitution, as controlling on the issue decided nor, given the court’s clear desire to cut itself free from federal decisions it considers undesirable, can we rely on federal law for guidance. As one member of the court recently said, we must “[b]e prepared to think from scratch. Liberating ourselves from the reactive approach of the [U.S.] Supreme Court means liberating ourselves from many of the ways things have been done.” Speech of Judge Linde to Conference of Chief Justices, quoted in 69 ABA Journal 1356 (1983). As another member of the court recently said, expressing the same thought, we now must make “a lonely journey in the dark of the moon and against the wind into the quagmire of the law of ‘search and seizure ***.’” State v. Caraher, supra, 293 Or at 760 (Campbell, J., concurring).
CARAHER, LOWERY, AND WARRANTLESS SEARCHES
Our previous opinion sets forth the facts of this case; we will only supplement them as necessary. The challenged items are a plastic baggie containing marijuana that officer Costanza took from defendant’s hand, a flowered cigarette purse and a leather change purse that he found in a brown paper bag on the seat next to the driver’s position in the car and a gun that officer Keever found in a lidless container under some clothing on the back seat.
[627]*627At the time of the seizure of the baggie, Costanza had only stopped defendant for a traffic violation. The warrantless search for and seizure of the baggie could not be incident to that stop or to any subsequent arrest related to the traffic violation; it required separate probable cause. State v. Elkins, supra. We therefore turn to a consideration of the probable cause which the state constitution and statutes require Costanza to have had in order to justify each step of this warrantless search.
We look first for guidance to the two cases in which the Oregon Supreme Court has recently addressed the subject of warrantless searches. In the first, State v. Caraher, supra, the court reaffirmed the doctrine that a person arrested for a crime may be searched incident to the arrest for evidence of that crime. 293 Or at 757-760; see also State v. O’Neal, supra; State v. Krogness, supra. In State v. Lowry, supra, it reaffirmed the requirement that a seizure of items found during a search incident to arrest have a justification independent of the arrest when the items seized are not related to the offense for which the defendant is under arrest. See State v. Elkins, supra. The court then went on to expand the warrant requirement in a way that we find difficult both to understand analytically and to apply practically.
In Lowry, the defendant, during a search after his arrest for driving under the influence of intoxicants, was found to possess a small, closed, transparent amber pill bottle containing a white powder that proved on analysis to be cocaine. The Supreme Court, in an extensive opinion, assumed that there was probable cause to seize the pill bottle but held that the subsequent testing was a separate search that was impermissible without a warrant. In doing so, it failed to note a well-established aspect of Oregon search and seizure law — one of the few — and appears to have extended the warrant requirement to situations where its only function will be to impede police investigations and to create busy work for judges.
The Supreme Court began its discussion in Lowry with a useful admonition:
“The constitutional text [of Or Const, Art I, § 9] itself ties the phrase ‘probable cause’ to warrants. It seems never to become superfluous to repeat that the requirement of a [628]*628judicial warrant for a search or seizure is the rule and that authority to act on an officer’s own assessment of probable cause without a warrant is justified only by one or another exception. * * *” 295 Or at 346.
The court then went on to note, at the conclusion of the same paragraph:
“* * * The reasons for the exceptions from the rule are always one form or another of practical necessity to act before a warrant can be obtained.” 295 Or at 347. (Emphasis supplied.)
This last sentence contributes to our puzzlement. It is not only “practical necessity” in the obvious sense of the phrase that justifies a warrantless search; arrest itself justifies such a search, so long as the search is for the protection of the officer or the preservation of evidence or is related to the crime for which the defendant was arrested and is reasonable in time, scope and manner. State v. Caraher, supra, and State v. Chinn, supra, establish this rule ineluctably. The first two purposes are matters of necessity, but the third is not. What Lowry fails to face squarely is that the search authorized in Caraher extends beyond the “necessary” to the merely reasonable. See State v. Caraher, supra, 293 Or at 759:
“These cases, Chinn, Krogness and O’Neal, have expanded the justification for a search incident to arrest beyond considerations of the officer’s safety and destruction of evidence. They permit a search when it is relevant to the crime for which defendant is being arrested and so long as it is reasonable in light of all the facts. * * *”
In Caraher, the court approved a search of the defendant’s purse that occurred while she was under arrest and locked in the back seat of a police car and the officer, with the purse, was sitting in the front seat. There was no “practical necessity” for that search. We cannot believe that Lowry was intended to overrule Caraher sub silentio; the opinion cites Caraher with approval several times and even, as noted, reiterates Caraher’s holding. We therefore conclude that, by suggesting in Lowry that warrantless searches can be justified only by “practical necessity,” the Supreme Court used the phrase in a technical sense, which we must now try to discern.
The phrase the Supreme Court used to describe the justification for a warrantless search was “practical necessity” [629]*629rather than the more common “exigent circumstances.” We assume that it did so primarily to free itself from any technical glosses that may have been put on the latter term and to emphasize the need to look afresh at the practical situation facing the officer in deciding whether the officer, in the light of all the circumstances, should have sought a warrant rather than conduct an immediate search. Because two of the three justifications for a search incident to arrest (protection of the officer and preservation of evidence) are also matters of practical necessity, the phrase seems to apply there also. Inasmuch as the court apparently did not intend to alter its acceptance of the third justification (relatedness and reasonableness) either, we assume that the court also subsumed that justification under the “practical necessity” heading. That is, “practical necessity” in its technical meaning includes any search incident to arrest which meets the criteria of State v. Caraher, supra, and State v. Chinn, supra. While this conclusion may strain the Supreme Court’s language, it is the only one which makes its holdings consistent.
This brings us to our second difficulty with Lowry. In order for a search to be incident to an arrest, the officer must have statutory authority to arrest the defendant, which means that the officer must have an arrest warrant or probable cause to believe that the arrestee has committed an offense. ORS 133.235; ORS 133.310. The crime for which the officer has probable cause need not be the crime for which the arrestee is ultimately charged or even the one for which the officer makes the arrest. See State v. Cloman, supra, 254 Or at 10-13; State v. Krogness, supra, 238 Or at 145-149. If a warrantless arrest is for a felony or a Class A misdemeanor, it is sufficient for the officer to have probable cause to believe that the arrestee committed it. If it is for a lesser offense, the officer must have probable cause to believe that the arrestee committed it in the officer’s presence. ORS 133.310(1).
As the Supreme Court explained in State v. Krogness, supra:
“Where the officer, without trespassing, sees contraband or other evidence reasonably causing him to believe that contraband is being transported or that some other crime is being committed in his presence, he may have probable cause to make an arrest for the newly discovered offense as well as for the traffic offense which initially brought the subject to [630]*630the attention of the officer. In such a case, while there may be no distinct demarcation between the first and second arrests, there does exist, prior to any extensive search, a probable-cause foundation for an arrest for an offense more serious than a traffic violation. The officer then is justified in making such a search as may be commensurate with the gravity of the newly discovered situation. Probable cause to arrest for the more serious offense, when present, will answer constitutional objections to the rigor of the ensuing search. * * *
a* * * * *
“* * * If there is sufficient cause, as a matter of law, to justify whatever arrest is necessary physically to make a search, then a reasonable search is a lawful incident of such an arrest. * * *” 238 Or at 145-47. (Emphasis supplied.)
This concept of escalating probable cause is a principle familiar to virtually anyone with a speaking acquaintance with criminal constitutional law and has been part of Oregon law for almost 60 years. State v. McDaniel, supra, 115 Or at 240-242. Yet here, too, Lowry — if taken literally — raises havoc with the familiar.
Lowry was a case that escalated from a stop of the defendant’s car on account of a faulty headlight to an arrest for driving under the influence of intoxicants to the discovery, seizure and subsequent analysis of a small, closed, transparent amber pill bottle containing a white powder. In that opinion, the Supreme Court said, touching on the subject of the inadvertent discovery of suspected contraband:
“In the course of a valid traffic stop of a vehicle or a permissible frisk incident to a stop or an arrest, officers sometimes may come upon other suspicious items. But these may not be seized on suspicion alone; probable cause is required. State v. Elkins, supra, found a constitutional violation in the seizure of pills in a pill bottle incident to an arrest for public drunkenness, because mere suspicion did not constitute probable cause for the officer to believe ‘that the article he has discovered is contraband and therefore a crime is being committed.’ 245 Or at 284. The state’s brief in the present case .proceeded on the premise that State v. Elkins had been overruled by State v. Florance, 270 Or 169, 527 P2d 1202 (1974), an assumption that the Court of Appeals properly rejected. The majority of that court nevertheless sustained the officer’s action in retaining and opening the bottle and testing its contents, on the theory that the characteristics of the [631]*631bottle could cause a ‘hypothetical reasonable officer’ or ‘an officer trained with respect to illegal drugs’ reasonably to believe that the bottle contained contraband. 59 Or App at 344.
“We need not follow the Court of Appeals into judicial speculations on the extent to which contemporary culture has made pill bottles and containers of white powdery substances prima facie evidence of criminál possession of their contents and therefore subject to seizure. That question is of potential concern to people who carry a supply of baby powder, or table salt, or legitimate medicines, because if observation of such a substance in an unconventional container suffices as probable cause to suspect possession of contraband, it also suffices as probable cause to arrest the person in whose possession the container is observed. In the present case, however, the seizure of the bottle incident to a valid arrest and the subsequent testing of its contents are properly analyzed not as one but as two events. For ‘probable cause’ is not alone dispositive of both steps. The question is not simply whether probable cause to investigate, that is to say, to ‘search,’ the contents of the bottle did or did not exist, but whether there was any need to do so without a warrant.” 295 Or at 345-46. (Footnote omitted; emphasis supplied.)
The Supreme Court then went on to decide that, assuming the officer had probable cause to believe that the amber bottle contained contraband and that the bottle’s seizure was therefore proper, the subsequent testing of the bottle’s contents was a separate, impermissible search, because the officers did not obtain a warrant authorizing the analysis. 295 Or at 346-349. The predicate for this conclusion is unclear. It may be that the officer who found the amber bottle had no special training or other articulated basis for believing he had discovered contraband and, thus, had no probable cause to seize the bottle. State v. Elkins, supra. That is the basis for the concurring opinion, but it does not explain the thrust of the court majority’s analysis.3 The court’s language suggests, rather, that probable cause “is not alone dispositive of both [the seizure of the bottle and the testing of its contents].” 295 Or at 346.
[632]*632It is at this point that the majority opinion in Lowry becomes inexplicable. With due respect to the Supreme Court, if the latter statement was meant to be a recitation of existing law, nothing could be more patently wrong. If the officer had probable cause to believe that what he had found was contraband, then he had probable cause to arrest defendant for possession of that contraband and the seizure was incident to the arrest for that offense. State v. Caraher, supra. The court danced around the probable cause issue throughout the opinion; it is a shame that it did not directly confront it.4 It noted that there was a dispute whether the officer had “reasonable cause” to believe that the bottle contained contraband and then stated that, if he did, he would be entitled to retain the bottle long enough for a magistrate to determine if there was “probable cause” to seize and test the contents. The court made no effort to explain the distinction between “reasonable cause” and “probable cause,” assuming that there is one. It ultimately held that the distinction between the case before it and Caraher is that, in Caraher, the item seized was related to the crime for which the defendant was arrested, while in Lowry it was not. 295 Or at 347. In order to make this distinction as to relatedness, however, the court had, at least, to assume that there was probable cause because, unless there was, the seizure was unlawful and the court had no justification for going on to the relatedness point it seems so desperately to have wanted to make. But there is the rub: If there was probable cause, there also was probable cause to arrest and the seizure and subsequent search were as much incident to the arrest as were those in Caraher.5
[633]*633The only possible distinctions we can find between Lowry and Caraher are that, in Caraher, the defendant challenged only the warrantless discovery of the contraband, not its subsequent testing, and that in Lowry there may not have been probable cause to justify the seizure. The majority refused to rely on the second possibility; we are left with the first, a weak reed indeed — an analytically untenable one, in fact, in view of the fact that this issue (among others) was not raised in Lowry, either.
Where, then, does Lowry leave us? We are unsure. However, the case’s insistent reliance on both Caraher and Elkins, and its emphasis on the requirement of procuring a warrant as soon as possible, lead us to conclude that the court may have had two intentions: to reinstate, as a fixture of Oregon constitutional law, the principle of State v. Elkins, supra, and to place a gloss on that part of State v. Chinn, supra, which tests the validity of searches incident to arrest by the officer’s opportunity to obtain a warrant.
The basic principle of Elkins is that an arrest allows only a limited search and that one of the limitations is that' material seized must be related to the crime for which there is probable cause to arrest. In Lowry, the Supreme Court and both the majority and dissent in this court quoted Elkins on this point:
“If the rule were otherwise, an officer who desired to inculpate an arrested person in another crime, could seize everything in such person’s immediate possession and control upon the prospect that on further investigation some of it might prove to have been stolen or to be contraband. It would open the door to complete temporary confiscation of all an arrested person’s property which was in his immediate possession and control at the time of his arrest for the purpose of minute examination of it in an effort to connect him with another crime. Such a practice would be as much an exploratory seizure as one made upon an arrest for which no probable cause existed. Intolerable invasions of a person’s property rights would be invited by an ex post facto authorization of a seizure made on groundless suspicion.” State v. Elkins, supra, 245 Or at 287-288; quoted in State v. Lowry, supra, 295 Or at [634]*634348 and in State v. Lowry, 59 Or App 338, 341, 650 P2d 1062 (1982), and 59 Or App at 347 (Buttler, J., dissenting).
In State v. Chinn, supra, the Supreme Court stated criteria for evaluating the reasonableness of a search incident to an arrest; it reaffirmed those criteria in State v. Caraher, supra, 293 Or at 758. In doing so, however, the court in Caraher failed to mention the Chinn criterion that the officer’s opportunity to obtain a warrant is relevant to the determination of the reasonableness of proceeding without a warrant. 231 Or at 268, 272-73. Chinn gave the criterion an uncertain weight; Lowry could be read to make it decisive, but the absence of analytically defensible underpinnings and the tremendous ambiguity of its language leaves the matter in doubt.
We shall not read Lowry as making an abstract opportunity to obtain a warrant dispositive under the present circumstances, for two reasons: First, we think it clear that, if that is what was intended, the majority could have said it. It did not. Second, we think that, had Lowry been meant to stand for such a proposition — that the seizure of contraband incident to lawful arrest nonetheless separately requires a warrant before that contraband can be tested — its incredible departure from traditional practice would have been made with greater candor and would have provoked a firestorm of dissent, not a concurrence. Third, such an approach is both impractical and of little protection to the person being searched. We recognize that the opportunity to seek a warrant now has greater significance in a search incident to arrest than it did previously and that, when a search reaches a logical stopping point the police must seek a warrant before proceeding further.6 However, requiring officers to stop midway in a search like the present one, leave everyone in limbo, and seek a warrant would tend to benefit only the stationers who sell blank search warrant forms.
Reading Caraher and Lowry together in the manner described — which, given Lowry’s anomalies, admittedly takes [635]*635some nerve — and keeping in mind pre-Florance cases, we conclude that there are two major exceptions to the warrant requirement of Article I, Section 9, of the Oregon Constitution: searches incident to an arrest, in which the relevant criteria are those of Chinn and Caraher as modified by Lowry, and searches supported by probable cause and “practical necessity” in its obvious meaning.
Searches incident to arrest are the primary justification for warrantless searches of the person. If the officer is without probable cause to arrest a person, the officer is without probable cause to search him, for a search of the person necessarily implies an arrest. The Supreme Court made that clear 20 years ago:
“We believe, however, that the spirit of the constitutional proscription against unreasonable searches requires the same degree of good-faith belief in the guilt of a suspect to justify a search of his person or of his effects as would be required to support his arrest or an application for a warrant to search his home. As a practical matter, it is difficult to explain how the police can search an individual without arresting him, since any substantial detention without his consent would fit the definition of an arrest found in such criminal cases as State v. Christensen, 151 Or 529, [51 P2d 835 (1935)], and such civil cases as Lukas v. J. C. Penney Co., 233 Or 345, 378 P2d 717 (1963). If there is sufficient cause, as a matter of law, to justify whatever arrest is necessary physically to make a search, then a reasonable search is a lawful incident of such an arrest. If there is not probable cause to arrest, there is no logical basis for saying that one may nevertheless be arrested on suspicion and detained long enough for the officers to search him to satisfy themselves that he is indeed as innocent as the law presumes him to be.” State v. Krogness, supra, 238 Or at 146-47.
It thus appears that, under Oregon law, the search of a person or of property closely associated with the person at the time of the search generally is supportable only if there is probable cause to arrest. While a formal arrest may occur after the search, if at all, Krogness points out that the search itself implies an arrest.7 See also State v. Caraher, supra. [636]*636Conversely, a search based on probable cause and practical necessity applies to situations such as that in State v. Elk, 249 Or 614, 439 P2d 1011 (1968), in which the officers found a deserted car in circumstances which gave them probable cause to believe that it contained stolen property. There was no one to arrest, but there were grounds for the search.8
THE OREGON CONSTITUTION APPLIED TO THESE SEARCHES
We turn — at last — to the facts of this case. Costanza saw defendant run a stop sign at about 4:15 a.m. As soon as he turned on his overhead lights to stop the car, defendant quickly reached over to the passenger’s side and then straightened up. Defendant continued driving for another 100 yards before stopping. During that time, he was constantly bobbing up and down, moving his head and arms in and out of the officer’s sight. This was such unusual behavior for a driver in a traffic stop that Costanza felt concern for his own safety and called for a cover unit before leaving his vehicle to approach defendant.
Defendant continued making quick movements toward various places in the front seat of the vehicle while Costanza approached him and even after he had asked defendant for his driver’s license and registration. Defendant shifted around, talked nervously and reached into a brown paper bag with his right hand, pulling out a small, folded, clear plastic baggie. Defendant attempted to hide it in his hand, but a small portion protruded and Costanza could see it. When [637]*637defendant first tried to leave his car, he left the transmission in drive and the vehicle moved. Costanza had to tell him to stop and to put the vehicle in park. Defendant attempted to do so but only moved the transmission lever to reverse. When he got out, the car began backing up and Keever, who had just arrived in response to Costanza’s request, jumped in and stopped it.
In short, defendant acted nervous, worried and not fully in control of himself. He was obviously trying to hide something that he had taken out of the brown paper bag and that something was, as Costanza could see, a plastic baggie. Costanza knew from his training and his previous experience that people often store marijuana in plastic baggies. The totality of these circumstances gave probable cause to believe that what defendant was hiding was a controlled substance and that it was most likely marijuana. That was sufficient probable cause to believe that defendant was committing the offense of possession of a controlled substance in Costanza’s presence and, whether any formal arrest occurred then or later, defendant was in fact under arrest for that offense.9 ORS 133.005(1). The search involved in Costanza taking the baggie from defendant was incident to that arrest. The trial court correctly refused to suppress the baggie.
The baggie contained a substance which Costanza recognized on sight as marijuana. He therefore continued to have probable cause to hold defendant.10 In addition, he noticed that defendant’s behavior appeared “strangely [638]*638altered” from the normal and suspected that defendant was under the influence of some intoxicant. He then searched the front seat, including the brown paper bag from which he had seen defendant take the baggie; it was within easy reach. We hold that that search was reasonably related in time, scope and intensity to one of the offenses for which defendant was under arrest — possession of the marijuana. See State v. O’Neal, supra; State v. Chinn, supra.
In the bag, Costanza found a flowered cigarette purse and a leather change purse. After feeling them from the outside, he opened them and found containers with a white powder in them and drug user paraphernalia.11 His activity in opening the two closed containers would probably not be permissible as an inventory of defendant’s property resulting from the arrest. See State v. Keller, supra. However, when an arrest is for possession of contraband, a search of the person and the area within the immediate control of the person for evidence of that offense — including the opening of closed containers — is permissible. State v. Caraher, supra; State v. Chinn, supra. That is all that occurred here. The trial court did not err in refusing to suppress the bag, the flowered cigarette purse and its contents, or the leather change purse and its contents.12
Having found the powder, which they had probable cause to believe was a drug possession of which is a felony, ORS 475.992(4) (a), it was reasonable for the officers to believe that they might find more contraband in the rest of the [639]*639passenger compartment. We hold that defendant’s arrest for that felony was a sufficient basis for this further search. As noted, searches incident to a lawful arrest must be scrutinized as to time, scope and intensity. State v. Chinn, supra. Here, we find nothing unreasonable in searching the passenger compartment for evidence of the felony for which defendant had just been arrested. Such a search was essentially contemporaneous to the arrest, was no more intrusive than was justified by the nature of the offense, and was no more extensive than the area within defendant’s immediate control. See State v. Caraher, supra. The gun that led to the firearm charge was found during that search. Its seizure was appropriate.
Turning to consideration of defendant’s claims under the federal constitution, we adhere to our former opinion.
Affirmed.