Among the items of evidence placed before the jury of six which convicted William Cochran of driving under the influence of intoxicating liquor (G. L. c. 90, § 24[1] [a][l]), were the results of a breathalyzer test which placed Cochran at the outer limits of drunkenness. 1 Cochran’s primary contention on appeal is that the breathalyzer test results should not have been received in evidence because, demonstrably, the test device was malfunctioning. We think the point is a sound one and reverse the conviction.
*261 These are the facts the jury could have found. On March 10, 1986, the defendant, Cochran, precipitated a three-car collision by slamming into the rear of a car stopped at a red light. That car was propelled, in turn, into the car in front of it. Nobody was seriously injured, but the defendant demonstrated a certain incapacity of speech, gait, and thought. His reactions to the accident were passive and detached. By reason of the defendant’s conduct, his physical unsteadiness, his withdrawn attitude, and the smell of alcohol on his breath, the driver of the car struck by the defendant’s vehicle and a passerby concluded that the defendant was quite drunk.
Such was also the conclusion of Officer DeFrancesco of the Cambridge police. DeFrancesco, who knew the defendant reasonably well, and, more to the point, knew him sober, thought the occasion called for field sobriety tests. Cochran could not walk a straight line but did prove able to recite the alphabet.
Although the field sobriety tests had not been devastating to Cochran, Officer DeFrancesco thought what he had observed of the defendant warranted placing him under arrest. He was taken to the Cambridge police station and booked. There, duly advised of his rights regarding chemical testing, Cochran elected to submit to a blood alcohol examination. See G. L. c. 90, § 24(1)0), as amended through St. 1980, c. 383, § 1. Austin Maxwell, a Cambridge police officer trained to administer breathalyzer tests, fed two samples of the defendant’s exhalations into a Smith & Wesson Breathalyzer, model 2000. Those samples registered an alcohol level at .24 and .25, respectively, a stunning result considering that a reading of .10 or higher triggers a statutory presumption of intoxication. G. L. c. 90, § 24(1)0?), as amended by St. 1980, c. 383, § 1. The defendant waived his right to an independent alcohol level blood test.
In addition to describing the incriminating results which Cochran’s two breath samples produced, Officer Maxwell testified about how the breathalyzer device worked, what standard operating procedures were to be followed (e.g., purging" the machine before causing it to inhale a new breath sample), and how he had faithfully followed those procedures. Maxwell *262 described, among other things, that following each test of a subject, the operator fed a control sample into the machine. This was called the simulator test.
Cross-examination of Maxwell produced the following exchange:
Mr. Macdonald: “What’s the importance of the simulator test?”
Officer Maxwell: “To verify that the machine is working.”
(• • •)
Mr. Macdonald: “What kind of simulator reading would it take for you to see in order to realize the machine wasn’t operating properly?”
Officer Maxwell: “Something more than one-thousandths off the .15.”
Mr. Macdonald: “So it’s your testimony that if the reading on the simulator is . . . less than a . 14 or greater than a .16, that would indicate the machine is not operating properly?”
Officer Maxwell: “That’s correct.”
Mr. Macdonald: “Do you happen to know what the simulator reading was in this case?” (...)
Officer Maxwell: “It was .13.”
Mr. Macdonald: “Did you notice that at the time?” Officer Maxwell: “No, I didn’t.” (. . .)
Mr. Macdonald: “If you had notice (sic) the .13 at the time of Mr. Cochran’s test, would you have rejected the test result and moved to some other way to measure his breath alcohol?”
Officer Maxwell: “Yes, I would have.”
Thereupon defense counsel moved to strike the test results. The motion was denied. Earlier the defendant had 'moved in limine to exclude the breath test because of the simulator deviation.
Extensive discussion of the legal and scientific underpinnings of breathalyzers appears in:
Commonwealth
v.
Brooks,
. Implicit in the assumption of reliability is that the device is working properly. That is why breathalyzers come with simulator test ampules and courts have relied on the simulator test as establishing that the device is functioning accurately. See
Commonwealth
v.
Neal, 392
Mass. at 13;
Commonwealth
v.
Doyle,
Indeed, the Department of Public Safety promulgated a regulation effective July 1, 1987 (too late for this defendant), 3 which invalidates breath tests if the reading of the simulator solution exceeds the plus or minus .01 deviation. 4
*264
On the test performed on Cochran, it will be recalled, the simulator registered .13, a deviation of minus .02 and beyond the permissible tolerance. The officer who performed the test conceded that he would have rejected the test had he noticed the discrepancy. We are of opinion that the results of breath measuring tests ought not to be received in evidence unless the Commonwealth has adduced evidence through the simulator or similar control mechanism that the measuring device was working. The recent decision in
Commonwealth
v.
Yameen,
Our view is consistent with that adopted by a majority of jurisdictions which have considered the question. See
Moore
v.
State,
*265
step rather than .00);
State
v.
Fairleigh,
Although there was other potent evidence of Cochran’s intoxication, we cannot say that failure to strike the breathalyzer test results was harmless error. That test was quantitative and it had the seal of scientific approval. Compared to other evidence, the test was likely to be thought the most powerful. It was not merely cumulative. Compare
Commonwealth
v.
Lowe,
We are not persuaded by the argument that if the simulator reads on the low side of the standard .15, the defendant has an advantage because his breath readings will be comparably lower than the reality. See, e.g.,
State
v.
Shaping,
Having thus disposed of the case, it is not necessary to consider the defendant’s grievance that he was deprived of permissible cross-examination of Walton, the passerby witness. In the context of the over-all cross-examination, the restraints imposed by the judge did not constitute reversible error. We do think, however, should the case be tried again and should Walton again testify that the defendant’s breath smelled of gin or vodka, rather than beer, that the defense is entitled to test
*266
by cross-examination what basis Walton has for sniffing that distinction. Compare
Commonwealth
v.
Flynn,
Judgment reversed.
Verdict set aside.
Notes
Being under the influence of intoxicating liquor connotes a less disabled state than drunkenness, i.e., one may transgress the statute without being drunk.
See Commonwealth v. Griffin,
Guidelines, Infrared Breath Testing Program (Massachusetts Department of Public Safety). See also Mann & Murphy, Infrared Breath Testing Operator’s Manual (Massachusetts Department of Public Safety).
The trial occurred on October 20 (motions only), October 24, and October 27, 1986.
See 501 Code Mass. Regs. § 2.56 (1987), the text of which reads as follows: “(2) A test shall not be considered valid if the defendant fails to supply two adequate breath samples upon request (in which [case] it shall
*264
be deemed a refusal under
