Defendant appeals from his conviction, after a jury trial of driving while under the influence of intoxicants. We affirm.

Defendant makes two assignments of error, both of which relate to the court’s instructions to the jury.

First, defendant complains that the instructions twice mentioned that portion of ORS 487.5401 which forbids driving with a blood alcohol level of .10 percent or more, and thus unduly highlighted that portion of the law.

We find no error. In fact, the court’s instructions referred twice to all the elements of ORS 487.540— once to describe the statute under which the defendant was charged, and the second time to describe the state’s burden of proving each material element of the charge.2 Defendant acknowledges that both instructions correctly stated the law. This was an appropriate *854and balanced approach in defining for the jury the issues and the burden which the law places upon the state.

Defendant claims, however, that "* * * giving both instructions * * * resulted in the impression that the breath test was uncontrovertable [sic] evidence.” We are unable to perceive anything in the instructions as given which justifies defendant’s claim, and so reject it.

Defendant further notes, however, that the entire set of instructions was given again in response to a question received from the jury after deliberations had begun:

"Even though the law states that .10 means a person is 'under the influence,’ does this require a juror to go ahead and pronounce the defendant guilty? Can the breathalyzer say .10 or over and still I could come back with a verdict of not guilty? I just would like to have the law explained to me again so that it is perfectly clear to me.” (Emphasis added.)

The trial court’s response to this request was correct: it replayed the tape recording of its earlier instructions. As the emphasized portion of the jury inquiry shows, this was — at bottom — all the jury was requesting.

Defendant disagrees, claiming that the first two sentences of the jury inquiry required a specific response which the instructions could not provide. This brings us to the general theory involved in defendant’s second assignment of error, viz., that ORS 487.540(l)(a) embodies an impermissible conclusive presumption.

There are really two parts to this second contention of defendant:

(1) ORS 487.540(l)(a) constitutes a conclusive presumption, which is impermissible in the criminal law; and

*855(2) ORS 487.540(l)(a) and ORS 487.5453 must be read to permit consideration by the jury of other evidence which might tend to disprove the breathalyzer result.

Defendant’s first point was decided adversely to him in State v. Torrey, 32 Or App 439, 574 P2d 1138, rev den (1978). Driving with a blood alcohol level of .10 percent or greater is an alternative statutory definition of the offense of driving under the influence of intoxicants, not a conclusive presumption.

Defendant’s second point raises a question about a portion of State v. Torrey which, upon reconsideration, we are now convinced was unduly expansive. In Torrey, we held that, where a defendant is charged with driving under the influence of intoxicants under ORS 487.540(1)(a), (driving with a blood alcohol level of .10 or above) rather than (b) or (c), a trial judge could properly refuse to permit a defendant to offer evidence tending to show that defendant at the time *856exhibited no physical symptoms suggesting he was under the influence. Stated this way, the rule of Torrey is too broad. The correct analysis is as follows:

The gravamen of ORS 487.540(l)(a) is driving with a certain blood alcohol level. The legislature has seen fit to forbid this act, without more. The correctness of the evidence tending to establish the blood alcohol level is thus crucial. Equally crucial is defendant’s right to attack the evidence of blood alcohol level. Defendant can do this in one of two ways. First, he can show, by the testimony of those that performed it, that the chemical analysis used in the case was improperly conducted. Second, he can offer circumstantial evidence from other witnesses (including the defendant) to show that there is such a disparity between what the chemical test shows and other facts that one should infer that the test was in some way defective. See State v. Swarengin, 12 Or App 290, 506 P2d 729 (1973) (decided under analogous prior statute).

However, as with any such testimony, a proper foundation must be laid for it. The question then becomes: What is a proper foundation which will permit a jury to consider circumstantial evidence tending to contradict the state’s evidence of a particular blood alcohol reading? This is a question we did not specifically consider in State v. Swarengin, supra.

We think such testimony, where it is specifically offered for the purpose of challenging, by inference, the accuracy of a test which showed a blood alcohol level of .10 percent or above, is relevant only if proper evidence of the relationship between physical size, blood alcohol content and reasonably expected behavior has been introduced so that the jury has some guidelines to follow in assessing the evidence. Such evidence will usually be in the form of expert testimony. Without such an evidentiary predicate, a jury of laymen would have insufficient data to assign any *857real probative value to observations of the physical behavior of the defendant.4

Applying this rule to the present case, and absent any expert testimony of the kind described, defendant would not have been entitled to have the jury initially instructed that they could consider any physical observations which had been made of defendant in determining whether or not they were persuaded as to the correctness of the blood alcohol test result. It follows that the trial judge correctly declined to answer the two intermediate questions posed to him by the jury, since he would then have been giving an instruction unsupported by evidence. His decision to play back the instruction he had previously given was correct.

Affirmed.

dissenting.

I respectfully submit that the majority’s reasoning is fallacious and renders ORS 487.540(l)(a), driving with a blood-alcohol content of .10 percent or more as measured by chemical analysis, a wholly superfluous statutory provision.

A juror made the following request of the trial court:

"Even though the law states that .10 means a person is 'under the influence’ does this require a juror to go ahead and pronounce the Defendant guilty. Can the breathalyzer say .10 or over and still I could come back with a verdict of not guilty?! just would like to have the law explained to me again so that it is perfectly clear to me.”* 1 (Emphasis supplied)

*858The trial court responded by playing back the audio recording of its previous instructions. The majority emphasizes the last sentence of the juror’s question and concludes that all the juror wanted was a repetition of the instructions. If the majority is correct in its analysis, then the remainder of the majority’s opinion is dictum. Unfortunately, the analysis is incorrect. The touchstone of the juror’s question is in the preceding sentence and the playback was not responsive to that question. A likely impression left on the jury is that the breathalyzer result introduced by the state, which was .13 percent blood-alcohol, was conclusive evidence of guilt. The proper answer to the juror’s question would have been that a finding of .10 percent or more blood-alcohol as shown by chemical analysis constitutes the offense of driving while under the influence of intoxicants (DUII), but that the jury must find that the chemical analysis, i.e., the breathalyzer, did in fact accurately measure defendant’s blood-alcohol content to be .10 percent or more.

The question remains, was the error prejudicial? I conclude that it was not because defendant did not offer any evidence that attacked the credibility of the breathalyzer result. It is at this point that my path of reasoning and the majority’s join, only shortly to detour in opposite directions. The only evidence introduced by defendant was testimony by eye witnesses that defendant was not visibly intoxicated. The majority and I agree that the evidence as presented was not competent to disprove the breathalyzer result. The majority reasons that such evidence is admissible under ORS 487.540(l)(a) and competent to prove or *859disprove blood-alcohol content provided that a proper foundation is made in the form of "scientific evidence of the relationship between physical size, blood-alcohol content and reasonably expected behavior.” Defendant failed to lay such foundation and consequently the majority concludes that defendant’s evidence was not competent. The foundation requirement imposed by the majority is an inappropriate judicial invention. My opinion is that defendant’s evidence of observable symptoms is not competent because the legislature has precluded consideration of such evidence under ORS 487.540(l)(a).

The majority states that "the gravamen of ORS 487.540(l)(a) is driving with a certain blood-alcohol level.” To the contrary, the gravamen of the offense is driving with a designated blood-alcohol level "as shown by chemical analysis of his breath, blood, urine or saliva * * ORS 487.540(l)(a). It is the quoted language that differentiates the elements of the offense defined in ORS 487.540(l)(a) from driving "under the influence of intoxicating liquor” under ORS 487.540(l)(b). Under the majority’s view, ORS 487.540(l)(a) becomes surplusage because proof of "not less than .10 percent by weight of alcohol in a person’s blood constitutes being under the influence of intoxicating liquor” in violation of ORS 487.540(l)(b). (Emphasis supplied.) ORS 487.545.2 The majority ignores that ORS 487.540(l)(a) and (l)(b) have entirely different origins. Prior to the 1975 revision of the *860Motor Vehicle Code the elements contained in ORS 487.540(l)(b) were contained in former ORS 483.992(2), originally enacted in 1931 as OCLA 115-318. Oregon Laws 1973, ch 360, § 18. In 1965, the legislature enacted a companion statute, former ORS 483.642, the predecessor of present ORS 487.545, establishing a rebuttable presumption that a blood-alcohol content of .15 percent or more constituted driving while under the influence of intoxicating liquor. Oregon Laws 1965, ch 574, § 9. In 1971 the legislature enacted a new statute, ORS 483.999(1) creating a separate offense identical to present ORS 487.540(l)(a), except that the prescribed level of blood-alcohol was .15 percent or more. Oregon Laws 1971, ch 564. The 1975 revision combined these two statutes into a single offense with alternate elements, lowered the prescribed blood-alcohol level to .10 percent and made the presumption under former ORS 483.642 conclusive in ORS 487.545. Oregon Laws 1975, ch 451, §§ 87, 88. The sole purpose of the consolidation was to avoid the possibility of a defendant being convicted of two offenses arising out of the same conduct. See Minutes, Interim Committee on Judiciary, August 30, 1974, pp. 44-45; Minutes, House Judiciary Committee, May 6, 1975, p. 3.

The gravamen of the "driving while under the influence of intoxicating liquor” offense now incorporated in ORS 487.540(l)(b) was proof that defendant was sufficiently under the influence of alcohol that his physical and mental condition were affected to some ’’perceptible”degree. E.g., State v. Gaylor, 19 Or App 154, 527 P2d 4 (1974). The method of proof depended on evidence of perceptible, i.e., observable, symptoms such as odor of breath, flushed appearance, lack of muscular control, speech difficulty, disorderly conduct, mental disturbance, visual disorders, sleepiness, muscular tremors, dizziness and nausea. See American Medical Association, Alcohol and The Impaired Driver 143-44 (1970) (hereinafter "AMA”). As a result of the enactment in 1965 of former ORS 483.642, evidence of *861observable symptoms could be buttressed with evidence of defendant’s blood-alcohol content and the attendant rebuttable presumptions that arose from such evidence.

The enactment in 1971 of former ORS 483.999(1), now ORS 487.540(l)(a), was based upon an entirely different premise and gravamen. The new statute was premised upon a body of accepted medical knowledge concerning the effects of alcohol on driving of which we can take judicial notice. There are four conclusions that can be drawn from that knowledge: (1) the traditionally accepted observable symptoms of intoxication are unreliable in determining in fact whether a person is or is not alcoholically impaired, AMA at 143-44; (2) there is little correlation between traditionally observable symptoms of intoxication and the concentration of blood-alcohol in a person’s body because some individuals, for example, will display "gross impairment” with a blood-alcohol content of .05 percent or even less, while others will not display such symptoms with readings as high as .35 percent, AMA at 10-12; (3) all persons’ mental and physical capabilities are in fact "definitely impaired” at blood-alcohol concentrations in the body of .10 percent or more, AMA at 146; (4) the more reliable and scientific method for determining whether a person’s physical and mental capabilities are impaired because of the consumption of alcohol is through chemical analysis of his blood, breath, etc., AMA at 162-67. The legislative history of former ORS 483.999(1) indicates that the legislature intended that liability under that statute was to be premised on chemical analysis and not on observable symptoms. See Minutes, Senate Select Committee on Traffic Safety, May 3, 1971; Minutes, House Judiciary Committee, May 28, 1971. Former ORS 483.999(1) itself made this clear. It provided in pertinent part:

"Any person who drives any vehicle upon any highway of this state when that person has .15 percent or more by weight of alcohol in his blood as shown by *862chemical analysis of the person’s breath, blood, urine or saliva made pursuant to ORS 483.634 to 483.646 shall be punished, upon conviction * * (Emphasis supplied.)

The majority ignores the quoted statutory language citing State v. Swarengin, 12 Or App 290, 506 P2d 729 (1973). I would overrule the reasoning in that case because it likewise ignores the statutory language. Swarengin involved a prosecution under former ORS 483.999(1) wherein defendant contended that it was prejudicial error for the trial court to admit evidence offered by the state of observable intoxication as shown by field sobriety tests administered to defendant by the police at the time of the arrest. We held the evidence admissible as having "some tendency to support an inference that defendant’s blood-alcohol level at the time of his arrest was not substantially different from his blood-alcohol level when he took a breathalyzer test about an hour later.” 12 Or App at 291. The evidence should not have been admitted because ORS 483.999(1) like ORS 487.540(l)(a) requires that the state prove blood-alcohol content by chemical analysis performed in accordance with the implied consent law, and not by field sobriety tests.3

*863The rules of statutory construction, legislative history and common sense dictate that we construe ORS 487.540 in a manner which does not render part of that statute, i.e. ORS 487.540(l)(a) superfluous. The legislative intent under ORS 487.540(l)(a) was to establish a factual basis for liability independent of ORS 487.540(l)(b), utilizing only evidence having scientific validity and precluding evidence of observable symptoms. The state must prove liability by chemical analysis of the defendant’s breath, blood, urine or saliva performed in accordance with the statutory procedures of the implied consent law, ORS 487.805 to 487.815 and 487.825 to 487.835. The defendant likewise is limited to introducing only that evidence which has scientific validity and cannot rely on evidence of observable symptoms. Logic further dictates that if such evidence is to be excluded under ORS 487.540(l)(a) for the purpose of determining whether or not a defendant’s blood-alcohol content was .10 percent or more, it likewise was to be excluded in determining the accuracy of the breathalyzer.

Of course, in the legal sense, evidence of observable physical symptoms is relevant in that it may establish some degree of probability concerning the level of blood-alcohol content. However, even under the common law rules of evidence, relevance alone does not determine admissibility. Legislative exclusion under ORS 487.540(l)(a) is wholly consistent with the common law. McCormick states:

"* * * relevance is not always enough. There may remain the question, is its value worth what it costs? There are several counter-balancing factors which may move the court to exclude relevant evidence if they outweigh its probative value. In order of their importance, they are these. First, the danger that the facts offered may unduly arouse the jury’s emotions of prejudice, hostility or sympathy. Second, the probability that the proof and the answering evidence that it provokes may create a side issue that will unduly distract the jury from the main issues. Third, the likelihood that the evidence offered and the counter proof will consume an *864undue amount of time.” (Footnotes omitted) McCormick, Evidence 438-40, § 185 (2d ed, E. Cleary, 1972).

The second and third considerations mentioned by McCormick are particularly applicable here. There is a likelihood that an extensive inquiry into defendant’s observable mental and physical condition will not only prolong the trial, but focus the jury’s attention on that side issue and thus distract it from the principal issue, i.e., whether or not defendant had a blood-alcohol concentration of .10 percent or more as measured by chemical analysis. Weighed against these considerations is the fact that evidence concerning defendant’s observable mental and physical condition has only a slight probative relationship to a person’s blood-alcohol content, and consequently to the accuracy of the breathalyzer.

Defendant argues that exclusion of consideration of such evidence means that evidence of breathalyzer results of .10 percent or more establishes a conclusive presumption of guilt which is violative of due process. I would agree that if the presumption is conclusive it may be violative of due process considering the fallible nature of the breathalyzer machine and the persons who operate it. See State v. Michener, 25 Or App 523, 550 P2d 449 rev den (1976); AMA at 164-66. However, the presumption is not conclusive. There is a variety of readily available and reliable methods of proof by which defendant can effectively impeach the breathalyzer. Defendant may require that the state provide him with the breathalyzer ampule which can be used to retest the breathalyzer. See State v. Michener, supra, at 528. An equally if not more reliable method of disproving the breathalyzer is an independent blood test. ORS 487.810 provides that the defendant who is administered a breathalyzer test must upon request be afforded reasonable opportunity to have "any licensed physician * * * or other qualified person of his own choosing administer a chemical test or tests for the purpose of determining the alcoholic content of his blood.” Significantly, the state *865must prove that it has complied with ORS 487.810 in order to establish that the chemical analysis comes within ORS 487.540(l)(a). The legislature thus contemplated, and indeed insisted, that defendants have available a readily accessible and reliable means for impeaching the breathalyzer.

Another concern of the legislature in enacting the new vehicle code was efficient judicial administration of traffic offenses. See Commentary, Oregon Vehicle Code, Foreword pp. XII-XV (1975). It was thus not unreasonable nor violative of due process for the legislature to want to disallow evidence which is inherently unreliable and unduly prolongs and confuses the trial in DUII cases. The rule adopted by the majority will accomplish the opposite result. The trial will not only be prolonged, as with this case, by lengthy eye witness testimony concerning the defendant’s observable behavior and demeanor, but the trier of fact will be confronted with a confusing array of expert testimony concerning the relationship between blood-alcohol content and observable symptoms of alcoholic impairment.

The majority assumes a medical relationship between "physical size, blood-alcohol content and reasonably expected behavior” that does not exist. Physical size has a definite relationship to a person’s alcoholic absorption and dissipation rate, AMA at 21, but slight relationship to the observable symptoms of alcoholic impairment. Other factors such as environment and metabolism may have a greater degree of relationship although there is medical disagreement as to the degree and quality of the relationship. AMA at 12, 20. The most important factor is a person’s drinking history. AMA at 10. A more experienced drinker is often able to conceal many of the otherwise observable symptoms of impairment. The medical facts upon which ORS 487.540(l)(a) is premised, which the majority ignores, is that there is no reliable correlation between blood-alcohol content and observable symptoms and that all persons are suffering *866impairment at .10 percent or more. The fact that some persons may manifest observable symptoms of alcoholic impairment at blood-alcohol levels of less than .10 or may not manifest such symptoms except at much higher levels is irrelevant.

I conclude that the evidence concerning defendant’s observable mental and physical condition should not be considered either as evidence of his blood-alcohol content or as to the credibility of the chemical analysis offered as evidence under ORS 487.540(l)(a). Since defendant makes no other contentions concerning the accuracy of the breathalyzer, the trial court’s instruction was harmless error.

Because of the volume of DUII cases being tried, I believe we should also address in this case the problem posed by the consolidation of the elements of former ORS 483.992(2) and 487.999(1) into a single offense with alternate elements under ORS 487.540(l)(a) and (l)(b). The trial courts require guidance under either the majority’s or my view in treating these two former offenses as alternative methods of proof. As I have already stated, the only purpose of the consolidation was to avoid the problem of multiple prosecutions arising out of the same conduct. The legislature evidently did not consider the problem of proof presented by treating these as alternate elements of a single offense. Under ORS 487.540(l)(b), evidence of observable symptoms as well as evidence concerning blood-alcohol levels is admissible and may be considered by the jury. See ORS 487.545. I would hold that evidence of observable symptoms is not admissible under ORS 487.540(l)(a). The majority would hold that evidence of observable symptoms is only admissible under ORS 487.540(l)(a) if a proper foundation is laid. Thus we have an inherently confusing situation for the trier of fact. The conventional solution to this problem would be to require that the trial court carefully instruct the jury as to what evidence it may or may not consider under each of the alternative elements. The probable effect of such instructions will *867be to confuse rather than enlighten. More important, where the state wants to buttress its case under ORS 487.540(l)(a), evidence such as that introduced in Swarengin will be allowable under ORS 487.540(l)(b), thus creating a real probability of prejudice to the defendant.

The logical solution is to require the state to elect prior to trial whether it will prosecute under subsection (l)(a) or (l)(b) of ORS 487.540. Admittedly, this would be breaking new ground in requiring such an election. There is precedent to the effect that the state can be required to elect if it appears that simultaneous trial of all the offenses charged would be prejudicial to the defense. See, e.g., State v. Reyes, 209 Or 595, 303 P2d 519, 304 P2d 446, 308 P2d 182 (1957); State v. Davis, 13 Or App 225, 508 P2d 471 rev den (1973). However, it has been generally held that this rule is inapplicable where there is only one statutory offense charged even though there may be alternate methods of proof. The rationale for the latter rule is that "there is nothing to elect.” Reyes at 623. But that rationale cannot withstand analysis. Due process would certainly require an election if, for example, the legislature decided to consolidate all crimes into a single statutory offense. Here, the legislature has consolidated into a single offense alternate methods of proof, one premised on observable symptoms and the other on chemical analysis, which are incompatible with each other. Furthermore, a trial in which both methods of proof are utilized by either the state or the defense seemingly defeats the legislative purpose and makes absurd the provisions of ORS 487.540(l)(a).

In view of these circumstances, I would require that the state make an election prior to trial as to whether it intends to prosecute under ORS 487.540(l)(a) or 487.540(l)(b). More important, however, I think it is incumbent upon this court under either the majority’s view or mine, to provide some guidance to trial courts in how to handle this difficult problem.