Plaintiffs appeal of right from a directed verdict. It was entered at the conclusion of all the proofs offered. The trial judge allowed the case to go to the jury as to co-defendant Eaton County. A verdict of no cause of action was returned. No appeal was taken. The case stands as between plaintiffs and General Motors Corporation.
*211In a sense this is jurisprudential old hat. decisions based upon an implied warranty of fitness of a product for the purpose for which it was manufactured fairly abound.
We could take the narrow approach and add one more opinion to the skyscraper-high pile that merely says that in our view there was or there was not enough evidence to make jury-submissible the question of the breach pleaded and argued, which in turn would permit the jury to make reasonable inferences from established facts arid decide the questions of breach, proximate causality and damages, if any.
The counsel in this case have earned the right to more. The comprehensiveness and excellence of their briefs, the forceful and analytically disciplined oral argument, and the precise unequivocal answers to very precise, extensive bench questioning impel us to try, at least, to hone and refine the issue by re-reference to key cases in the products liability field. This to the end that in the event of review of our decision by the Supreme Court, we will at least have pinpointed the question which we believe controls.
First, we will record certain material facts which are uncontroverted.
At some point in time someone bought a new, unused 1967 Chevrolet. That someone sold it to someone else. That second someone sold it to plaintiff administratrix’s decedent.
In May of 1971 decedent was operating the vehicle on a secondary road in Eaton County.
Now the facts become controverted. There was testimony that the road in question was pockmarked with chuck holes of various depths and dimensions. There is other testimony that it was not any worse or any better than most similar *212roads. One witness testified he observed evidence of recent grading.
The speed of the vehicle at the time of the accident was placed at between 50 and 55 miles per hour. This fact is not seriously challenged.
For some reason or another, while so being driven, the right front ball-joint assembly broke. The car left the highway, crashed into a telephone pole, killing the driver and injuring his minor daughter.
Whether the assembly broke on the highway causing the car to leave the road, or broke after the car left the highway and hit the pole is vigorously contested.
Now comes, however, the argument-stressed position of the defendant. Whether it broke on the highway before the crash or after it left the highway and hit the pole is not the real question. For, says defendant, in either event plaintiffs failed to introduce any evidence of defective design, defective or unsuitable material, or any other specific act or omission in the manufacture of the vehicle which existed at the time it left its control and was placed upon the open market.
Two experts testified after due qualification; one a physicist with experience in accident reconstruction, the other a metallurgist. Each deserves commendation for his candor and admission of testimonial limitations.
Plaintiffs’ expert testified in substance that in his opinion the assembly broke while the vehicle was on Nixon Road, and was caused by the rough and bumpy condition of the road. He conceded that he could express no opinion on "metal fatigue” ( i.e., defective material), but stated one did not have to be a metallurgist to state with confi*213dence that the assembly broke on the road and likely due to the road conditions.
Defendant’s expert testified he could find no evidence of "wear out” or metal fatigue. It was his opinion that the break did not occur while the vehicle was on the road. He, with like candor displayed by plaintiffs’ expert, testified he knew nothing about skid marks, tire marks or chuck holes.
So here we are with two experts expressing antithetical opinions, and a raft of lay and police witnesses testifying in sharp disagreement about road conditions, skid marks and the like.
Where is all this conflict to be resolved except in the jury room, ask plaintiffs.
Right where it was, answers defendant, on the bench, by the judge, because no one of the initial prerequisites of a jury-submissible case on defect of design, material, or any other act or omission was ever adduced by plaintiffs. Without that cornerstone, no jury could do anything but speculate, and this is impermissible under the law.
Hold on, say plaintiffs. Michigan’s implied warranty law is not that strangulating. While we continue to protest that we don’t have res ipsa loquitur by that name, we, in effect, apply it under the doctrine of permissible inference from established facts. In this case there are enough such facts to go to the jury on defect and causality.
We know not what to do but to go to the landmark cases, reexamine them, do our best to construe them and leave the rest to our Supreme Court if it chooses to act.
We begin with Piercefield v Remington Arms Co, Inc, 375 Mich 85; 133 NW2d 129 (1965).
In that case defense counsel thus bluntly summarized the proposition in his brief:
*214" 'If we start from what appellees considered the incorrect assumption that all injured persons are entitled to recover from a manufacturer who puts the product in the market, merely because of the injury, then this ends the matter and the court can, and should, say so, in one crystal-clear paragraph, unencumbered with sophistries or other unnecessary explanations.’ ” 375 Mich at 91-92.
Former Justice Black, speaking for the majority, just as bluntly replied:
"A more accurate statement of 'the proposition’ is that something more than injury must be shown in instances such as this plaintiff has pleaded, that is to say, he must allege and prove (a) the defect of manufacture upon which he relies, and (b) injury or damage caused by or resulting from such defect.” 375 Mich at 96.
Further in the opinion he added:
"Some quibbler may allege that this is liability without fault. It is not. As made clear above, a plaintiff relying upon the rule must prove a defect attributable to the manufacturer and causal connection between that defect and the injury or damage of which he complains. When able to do that, then and only then may he recover against the manufacturer of the defective product.” 375 Mich at 98-99.
This was the state of the law until Bronson v J L Hudson Co, 376 Mich 98; 135 NW2d 388 (1965), came along. A new test was announced clearly incompatible with Pierceñeld, supra. We quote:
"In the instant case, there was no competent direct proof of what specific irritant, if any, was present in the slip cloth. Upon favorable view, did the facts, along with legitimate inferences to be drawn therefrom, constitute a prima facie case sufficient in law to prevail *215against defendant’s motion for directed verdict? More specifically, was there evidence from which it might be legitimately inferred that an irritant was present in the cloth at time of purchase, which irritant was a proximate cause of injury?
"We are aware that in the typical products-liability case, plaintiff’s counsel is usually able to produce some direct proof of defect and causal relationship between that proof and injury. Here, the trial court found no competent direct evidence and, without detailing what was presented, we agree. However, we are moved to disagree with the trial court’s direction of verdict at the close of plaintiffs’ proofs. We cannot say that reasonable minds could not differ as to what might be legitimately inferred therefrom, so that we find it error to have eorieluded, as a matter of law, that plaintiff did not establish a prima facie case. Our opinion is that under the circumstances, upon favorable view, a legitimate inference from plaintiffs’ proofs was that an irritant was present in the cloth, which irritant caused Mrs. Berke’s injury.” 376 Mich at 101-102.
Either we have to resign ourselves to an ad hoc disposition, casé by case, or choose between what appear to be two mutually exclusive rules of law.
It was suggested on oral argument that we might find some "middle ground” allowing for age of the product and other variables. We doubt our authority to promulgate a new test. We are oath-obligated to work within Supreme Court authority as best we can. When two lines of authority exist we believe it is our duty to choose and explain our choice.
We elect to follow Pierceñeld because we think it is a rule fair to the purchaser and fair to the manufacturer. We think there must be more than accident and injury to constitute a jury-submissible question of breach of implied warranty of fitness for the purpose intended.
Search as we will, and we did, the record does *216not contain any testimony of defective design, defective material, improper assembly — nothing.
A car left the road. A man was regrettably killed and a child injured. A piece of machinery was found broken. No one qualified to say so testimonially claimed it was too weak, too long, too short, or too anything. Neither did anyone say it was not designed correctly, the metal too light or that it was not too anything. Under the circumstances we feel obligated to affirm the trial judge.
If the case goes on to further review we should pass on one additional issue to avoid a needless remand. There was an express warranty on the vehicle that had not yet expired (50,000 miles; the car had been driven some 47,000).
The warranty was limited to repair or replacement of the warranted part. This is not the kind of disclaimer that in any sense purports to limit the manufacturer’s liability under the implied warranty.
The order directing a verdict is affirmed. Costs to the appellee.
Allen, P. J., concurred.