WILLIAM E. DOYLE, Circuit Judge.

This is a products liability case in which the appellant was injured while driving a 1969 Ford Ranch Wagon, one which had been driven 44,000 miles. The judgment which is appealed from is based upon a jury verdict in favor of the defendant Ford Motor Company. The cause was appealed previously. There had been a jury trial on the other occasion. Following the inability of the jury to agree, the court granted a mistrial. After that a motion to dismiss filed by Ford was granted. Judgment based upon the dismissal order was reversed by us in an opinion by Judge Hill reported in 502 F.2d 1292. The question then presented was whether the evidence was sufficient to justify submission. On that occasion the trial court had based its final decision on the fact that the evidence failed to show that the metal in the ball joint or in the socket was too soft, saying that there had been no test to determine the softness or the hardness in the metal; that it could have been made but wasn't.

The evidence at bar showed that on July 10, 1970, the appellant was driving 65-70 miles per hour on an interstate highway when he passed a car and pulled back into the right lane. When he sought to straighten up the car the left front end raised up and the car immediately pulled hard to the left. He was unable to control the vehicle; it veered to the right and hit a reflector pole and went off the shoulder, across an access road and proceeded down an embankment. Following the accident, the plaintiff complained that he had back and other injuries which had resulted from the accident. At the subsequent, trial his theory of recovery was that of breach of implied warranty of fitness arising from the left front wheel's control arm being defective and falling out of its socket causing the car to become uncontrollable. Ford's position was that there was no defect and that there had not been a failure prior to the vehicle's becoming uncontrollable.

On this present appeal appellant contends:

1. That the court erred in failing to acquaint appellant's counsel with all of the instructions which it intended to give.

2. That it was error for the court to strike the testimony of the witness Hopper, the metallurgist; that it was also error for the court to strike the testimony of the witness Schoelzel and that of Dr. Jerome B. McCoy based on the court's determination that the witness was not qualified.

I.

The instruction to which particular objection is raised pertains to the credibility of the witnesses. In instructing the jury that if they found that false statements had been made they were to disregard the testimony, the judge said:

You are instructed that if you find that the plaintiff or any other person have (sic) made false statements in this cause as to the cause of the accident, or false statements as to his actual losses he has suffered, you may disregard his testimony or you may give it such weight as you think it deserves.

It is not impermissible to state that when a party is a witness in the case he is subject to all of the tests of credibility applicable to other witnesses and to tell the jury further that if it finds that a false statement was made by any witness, including the parties, that the jury is at liberty to disregard all or part of the testimony given by the said witness.

The instruction in question is somewhat more specific and personal than is the general instruction which is ordinarily given. The pattern instructions pertaining to the credibility of witnesses in civil actions do not, in discussing credibility of witnesses, focus on a party as a witness. However, if the instruction is general and impersonal, there is no reason why it should not be a statement concerning how a jury should treat the testimony of a witness who has testified falsely. There is no reason, however, for zeroing in on the testimony of the plaintiff as the court did in the case at bar. If it were justified, the court ought to have explained that the jury was not bound by its opinion.

Considered alone, the action in question might not justify reversal. On retrial, however, the instruction should not be given in a way that would suggest to the jury that the judge considered the witness a liar. The judge might have thought that the plaintiff had lied, but if so there should have been at the very least prior warning that such a statement was to be made. It is clearly improper for the court after having notified the parties as to intended instructions to give one which had not been among the submissions, one which placed counsel in a dilemma as to whether he should or should not object or as to whether one action or the other would be the more prejudicial.

II.

The evidence showed that the witness McCoy was a physicist and a professor of physics at the University of Tulsa. He was shown to have a Bachelor's Degree from Williamson College, a Master's Degree in the same subject from the University of Missouri, and a Doctor's Degree from the University of Helsinki in Finland, also in physics. He explained that physics is a study of energy, motion, materials and basic natural phenomena. He also said that he was Chairman of the Division of Physical Sciences in the College of Engineering at Tulsa and had occupied that position for two and one-half years. He described his experience at White Sands Proving Ground testing guided missiles and carrying out failure analysis. He also explained that the science of physics develops basic laws which are applied in the field of engineering. The plaintiff sought to have him testify concerning the front suspension system in a Ford automobile and particularly the construction of a ball joint. At that point counsel for Ford objected because the witness was not an automotive engineer, an expert in automotive design. However, the plaintiff explained that he was offering the witness to explain the suspension system, particularly the control arm in the front of a Ford vehicle, and he referred to an exhibit. The court refused to allow him to testify on this subject or any other on the ground that he was not shown to be an automobile engineer. The witness McCoy had, however, testified at the prior trial and on the appeal this court considered the testimony of Professor McCoy in connection with the issue of sufficiency of the evidence. McCoy's testimony was cited as supporting the theory that there was a defect in the vehicle. Our opinion in that case said:

Appellant called Professor McCoy to establish the fact of a defect in the vehicle. In response to questions by appellant's counsel, McCoy testified, in his expert opinion, that a defect was present in the upper control arm ball joint since installation of the part. He gave reasons why he believed the upper control arm was out of the socket at the time impact damage was done to the lower control arm. McCoy compared the mechanism from appellant's vehicle with one pressed out by a hydraulic press. Appellee argues McCoy's testimony was incredible and should not be considered as evidence of the defect. It is true that McCoy did not run metal hardness tests or examine the grease to substantiate his abnormal wear theory and that he did not examine the vehicle's spindle. These matters, however, go to the weight which the trier of fact should accord the evidence and do not make the testimony incredible.

502 F.2d at 1294.

Thus it was recognized in our prior opinion that McCoy's testimony was valid as against the contention that it was incredible and not to be considered evidence of the defect.

We also pointed out that the trial court had denied the contention that McCoy was not qualified as an expert. On this we said:

The trial court allowed McCoy to testify to the ultimate conclusion that a defect existed. The court specifically overruled a directed verdict motion based on appellee's contention McCoy was not qualified to testify as an expert. Thus there was expert testimony of the defect which was available for jury consideration. This expert testimony, appellant's narrative of the accident, appellant's other evidence, and the reasonable inferences resulting therefrom made a submissible case for jury consideration. Thus, the motion granting a directed verdict was error.

502 F.2d at 1295.

In view of the position that was taken by the trial court and by this court, McCoy's expertise to testify concerning the presence of a defect was established. Accordingly, the trial court at the second trial was not in a position to reach a different conclusion and particularly to do so without any explanation. Recognizing that McCoy's testimony was ruled vital and that plaintiff was deprived of it in the present proceedings, we are forced to conclude that the court committed error which requires reversal and a new trial.

It is not necessary that a witness who has broad and extensive training in physics, a field which embraces the science of mechanics, must be specifically qualified as an automotive engineer before he is allowed to testify. See Bratt v. Western Air Lines, 155 F.2d 850 (10th Cir. 1946); Sitta v. American Steel Wire Division of U.S. Steel, 254 F.2d 12 (6th Cir. 1958). See also Roth v. Bird, 239 F.2d 257 (5th Cir. 1956).

We are aware that the trial court's discretion in receiving expert testimony is broad. However, the court in this instance abused its discretion in refusing to allow the witness to testify.

First, the court refused to allow the witness to even make observations with respect to the part which was alleged to have failed.

Secondly, the court cut the testimony off at its inception based on the proposition that the witness was not an automotive engineer. This action was taken notwithstanding that there had been no opportunity to prove his knowledge quite apart from automotive engineering, and we do not see any magic in having this particular skill.

Thirdly, the testimony of this witness had been received before and had been considered crucial to the plaintiff's case, being testimony that justified, according to our previous decision, submission of the case to the jury. Thus, it was in effect the law of the case. The trial court was not in a position to reverse this determination.

We have considered the contentions of appellant concerning the trial court's ruling striking the testimony of Edwin H. Schoelzel and that of Robert E. Hopper, metallurgist. We do not believe that the court erred in its rulings with respect to the testimony of the witness Schoelzel.

As to the witness Hopper, the court ruled him out because of the fact that he had not examined the particular connecting spring. We do not see this factor as an essential basis for his giving testimony. The incompetency might have appeared, however, after he had given his testimony. He was called as a metallurgist and there was no deficiency in his qualifications as such. At least, therefore, the court should have heard his evidence before excluding it. On retrial this can be handled by offer of proof of a statement as to what the witness proposed to say or by an in camera method.

The judgment of the district court is reversed and the cause is remanded for a new trial consistent with the views expressed above.