On November 14, 1964 plaintiff Lillian Weiss was driving a 1960 Chrysler Impe*452rial Sedan along Route 123 in Connecticut when she sustained serious injuries upon hitting a tree. On June 22, 1967 she began this action against the defendants Chrysler Motors Corp. and Chrysler Corp. (hereinafter, “Chrysler”), sounding in products liability and negligence. Mrs. Weiss is a citizen of New York; the defendants Chrysler are citizens of Michigan. Jurisdiction is based on diversity of citizenship. A jury trial was held before Honorable Thomas P. Griesa, U.S. D.J., from February 20, 1973 to March 22, 1973. The jury returned a verdict for Chrysler, and the plaintiff appeals from that judgment. The ground of appeal relates primarily to the trial court’s rulings in connection with the admissibility of evidence by Chrysler and the exclusion of evidence offered by the plaintiff.
It was the plaintiff’s contention that while she was driving on Connecticut Route 123 on the day of the accident, she suddenly found that the steering wheel did not respond, and that shortly thereafter, the front wheels turned sharply to the right. She went off the road and into the tree at a distance of about 69 feet from the road. The plaintiff’s theory was that the accident was caused by a defect in the steering mechanism of the automobile. More specifically, she contended that the steering malfunction was the result of failure of a part of the vehicle steering assembly called a “Pit-man arm stud,” which was allegedly defective at the time the car was manufactured.
The evidence indicated that the Pit-man arm stud had indeed been fractured. Weiss contended that the fracture had occurred before the accident and was enough to cause failure of the steering device. Chrysler contended that the fracture was the result of impact during the accident itself. The plaintiff’s theory was that the fracture of the Pitman arm stud was a two-stage fracture, the first of which had been caused by fatigue due to improper design and construction. Plaintiff contended that the fatigue fracture progressed to a second, sudden fracture at the time of the accident. Chrysler sought to explain the accident upon the simple ground that the plaintiff went off the road because she was speeding and lost control of the car through her own negligence, not because of steering failure.
During pre-trial discovery there was a concession by a Chrysler metallurgist, Donald Gregory, that the Pitman arm stud showed two successive fractures. At trial Chrysler sought to refute appellant’s theory that the first stage fracture was due to fatigue, by contending that both fractures occurred after the car went off the road — the first stage caused by impact with a one inch stump before the car hit a ditch that was four feet wide and two feet deep, and the second stage occurring when the car hit a tree. There was a dent on the rim of the front right wheel. It became Chrysler’s theory that it was the load force of the stump impact on the wheel rim that had been sufficient to cause the first fracture rather than fatigue.1
The plaintiff thus relied on fatigue as causing the first stage of fracture ultimately sufficient to cause the steering failure before the car left the road. The defendant, on the other hand, relied upon impact with the stump as causing the first stage of the fracture,2 thus es*453tablishing that the car left the road before any fracture whatever. The defendant could rely, therefore, on speeding as the sole cause for the car’s leaving the road. The dissent somehow comes upon a third theory which it concedes “Chrysler did not stress,” namely, that the first fracture happened before the car left the road and the second after-wards.3 See the “threshold question” in the court’s charge in footnote 8.
Speeding, of course, would not be the cause of the accident if there were steering failure through a defective part. In this sense much of the dissent’s discussion of speeding elides the vital issue for the jury to decide. Speeding does not rule out a product defect, as Judge Grie-sa charged. To rule out expert testimony on product defect is hardly “harmless error,” as the dissent assumes, since we cannot say that the jury did not decide the whole case on lack of proof of product defect, never reaching appellant’s contributory negligence.
I
Appellant claims surprise with respect to the appellee’s expert testimony that impact with the stump caused the first fracture of the Pitman arm stud. In support of its argument of surprise, appellant contends that she had a right to receive answers to interrogatories submitted to Chrysler in the course of discovery, which would have given her notice that Chrysler intended to bring in expert testimony that impact with the stump caused the first stage fracture. She argues that the expert testimony of Sylvester Mazur, an employee of TRW (the manufacturer of the. stud), in support of that theory, should have been stricken from the defendant’s case because the defendants had failed to disclose the theory in their answers to the plaintiff’s interrogatories.
Appellant’s second contention relates to an exclusionary ruling made as the plaintiff attempted to dispute Mazur’s theory on its rebuttal case. The plaintiff offered the expert testimony of Professor Dennis Rader of Yale University to rebut Mazur’s testimony that the dent on the wheel rim had resulted from sufficient force to have caused the first fracture of the Pitman arm stud. Though Judge Griesa permitted Dr. Rad-er to testify about other subjects, he refused to allow him to testify to an experiment which he performed after Mazur had testified, which was designed to show that the amount of force on impact required to dent the rim of Weiss’ front right wheel was of an order of magnitude of one-fourth of the force that would have been required to cause a first fracture of the stud if it were sound.4
The trial judge’s main ground for refusal to permit Rader’s rebuttal testimony was that the subject matter had been broached on plaintiff’s direct case, that Rader’s opinion testimony should have been introduced at that time, and that it was improper rebuttal. He noted that during the cross -examination of one of the plaintiff’s witnesses, Alfred L. Mose*454ley, Moseley purportedly admitted that the sudden motion of the right front wheel in hitting the stump exerted a pull on the steering link which supported the tie rod fitting for the left front wheel and that the forces generated — sufficient to separate the tie rod stud — would also be sufficient to cause a “failure” of the Pitman arm stud.
In this respect appellant takes sharp issue with the meaning of that testimony. She contends essentially that when Moseley spoke of a “failure” of the stud, he was speaking generally and was including not only the possibility of fracture, but also separation of the stud from its housing — a kind of “failure” which would require far less force than that allegedly necessary to precipitate an actual fracture of the stud in Weiss’ car. Be that as it may, appellant contends that the eliciting of. matter helpful to the defendant on cross -examination of plaintiff’s witness, as distinct from its being proffered on direct examination, is no ground for the exclusion of rebuttal testimony to the contrary by the proponent.
Chrysler contends 1) that the appellant was not “surprised” by the Mazur testimony concerning the impact with the stump, because it had been the Chrysler theory and the subject of expert testimony presented in an earlier action in the New York State court arising out of the same accident with which plaintiff’s counsel was familiar and 2) that the interrogatory, the answer to which failed to disclose Professor Mazur’s proposed testimony, was not directed to that type of testimony and was therefore not called for in spite of the district Judge’s order to keep the interrogátories up-to-date as expert testimony developed. Chrysler also contends that, since the theory in question was elicited in the first instance from Moseley, the plaintiff’s witness, plaintiff should have structured her case so as to bring Professor Rader’s expert testimony into its direct case rather than waiting for rebuttal. Chrysler further urges that the plaintiff made no proper offer of proof at the time the rebuttal testimony was excluded. It urges that, in any event, the proposed testimony of Professor Rader would not have been in direct contradiction of Mazur’s testimony, that his methods of measurement were different from those of Mazur and hence not proper rebuttal, and that its admission would have unnecessarily prolonged a long trial by inviting sur-rebuttal. Finally, Chrysler makes the contention that its case was so strong that, even had Rader been permitted to testify as expected, a verdict for the plaintiff could not have been sustained on a motion to set aside the verdict.
II
We have reviewed the matter with particular care in view of our reluctance to upset a jury verdict, to trench upon the trial court’s discretion, and to protract a litigation concerning an accident that happened more than a decade ago. Nevertheless, we feel constrained to hold that the plaintiff was improperly and prejudicially foreclosed from proving her case on the crucial issue of her claim against Chrysler by the rulings adverted to above.
1. The failure of discovery.
The initial question is whether the plaintiff received fair notice of the theory which Chrysler ultimately pressed to explain the two-stage fracture of the Pitman arm stud. We conclude that Chrysler’s failure properly to respond to plaintiff’s interrogatories denied plaintiff such fair notice.
Two sets of interrogatories were submitted to Chrysler which were broad enough to require Chrysler to divulge a theory like that which Mazur ultimately disclosed at trial, and the responses given fell far short of giving appellant any indication of the theory involved in Ma-zur’s expert testimony.5 Two motions *456were made to strike Chrysler’s answers for failure to respond to these interrogatories and, although both motions were denied, the transcript of a hearing on the second of these motions on February 7, 1973, makes it entirely clear that disclosure of such a theory should have been made. The only theory apparent at that time, on which Chrysler might depend, was that the first stage fracture had occurred in a prior collision. Plaintiff fairly assumed that Chrysler could not deny that the stud failure was, indeed, a two -stage fracture because of an admission to that effect in the deposition of Chrysler metallurgist Gregory. However, Chrysler’s responses to Interrogatories 3(b), 3(c), 45, and 46 gave no indication that Chrysler’s experts had made any tests to substantiate any theory on which it might rely at trial. Nor, in response to Interrogatories 50-52, was Mazur’s theory, or an experiment to support it, even hinted at. It was the lack of such response and the fear that some undisclosed theory would be sprung at trial which prompted appellant’s motion to strike. While the particular thrust of appellant’s argument at the hearing on the motion to strike was over possible use of an earlier accident to explain the first stage fracture, there is no doubt that the more general concern of both appellant’s counsel and the trial judge was that Chrysler had not disclosed any theory to explain the first stage fracture:
[Appellant’s counsel]:
“ . . . I can’t believe that they are going to come in with no information and no facts on prior fractures. It is clear that their experts have analyzed it in this regard. They know there were prior fractures; there is no question about it. Perhaps a question to be resolved is whether there was prior impact or prior fatigue, but this is central to the case. Now we know that Chrysler knows that this piece was fractured before the accident.
“The Court: What you are doing is forcing them to analyze the case and to analyze the existence of prior fractures. Maybe what should be done is, if they won’t disclose this, they should be precluded from putting in any evidence on the trial as to fractures.” (App. pp. 63-64).
We think it clear that the face of Interrogatory 46 coupled with the trial judge’s general warning that Chrysler might be “precluded from putting in any evidence on the trial as to fractures” clearly required Chrysler to divulge Ma-zur’s theory as a response to Interrogatory No. 46 as well as No. 52. Chrysler’s argument that its failure so to respond can now be justified after the fact because “plaintiff never complained of defendant’s answer to interrogatory # 46” (Brief p. 54) is disingenuous. Even if plaintiff had reason to believe that such testimony would be presented because of similar arguments in previous state court litigation, which is at best sharply disputed, that would not justify the failure of Chrysler to respond truthfully and completely to a properly framed interrogatory. “It is no objection to interrogatories . . . that the information sought is within the knowledge of the interrogating party.” Bowles v. Safeway Stores, 4 F.R.D. 469, 470 (W.D.Mo. 1945); Grand Opera Co. v. Twentieth Century-Fox Film Corp., 21 F.R.D. 39, 40 (E.D.Ill.1957). “A distinction should not be drawn between facts with or without the knowledge of the examining party.” 4 Moore’s Federal Practice ¶ 26.59 at 26—219 (2d Ed. 1974).
Consequently, to characterize the issue here as being determined by a lack of “surprise” distorts the essential problem. The policy which prompted amendment *457to Rule 26(b)(4) of the Federal Rules of Civil Procedure to allow more liberal dis-' covery of potential expert testimony was not merely for convenience of the court and the parties, but was intended to make the task of the trier of fact more manageable by means of an orderly presentation of complex issues of fact. “Mutual knowledge of all the relevant facts gathered by both parties is essential to proper litigation.” Hickman v. Taylor, 329 U.S. 495, 507, 67 S.Ct. 385, 392, 91 L.Ed. 451 (1947). See generally F.R.Civ.P. 26(b)(4), Advisory Committee Notes to 1970 Amendments.
In this case the expert testimony was crucial. As Judge Bartels has well put it: “Realistically speaking, the resolution of the entire case depends upon [medical] and expert testimony and opinion. The necessities of such a case transcend the usual limitations which may otherwise be imposed upon discovery proceedings.” United States v. Nysco Laboratories, Inc., 26 F.R.D. 159, 162 (E.D.N.Y.1960).
We think that this policy was defeated in this litigation by reason of Chrysler’s failure to disclose the theory of its potential expert testimony to explain the first of the two-stage fractures of the Pitman arm stud.
Similar considerations lead us to reject Chrysler’s arguments that its duty to respond to the interrogatories concerning the expert testimony ended when the trial began because the tests on which it was based were not completed until after the trial had begun. Even if appellant is wrong in contending, to the contrary, that the crucial tests were actually completed before trial, Chrysler’s obligation to supplement its response to interrogatories on the “subject matter” and “substance” of expert testimony was a continuing obligation. Rule 26(e)(1)(B). See Diversified Products Corp. v. Sports Center Co., 42 F.R.D. 3, 5 (D.Md.1967). Nothing in either Rule 26(e) or the trial judge’s direction to the parties indicates that that obligation contained any limitation of time. Particularly in a case like this, where trial was reached nearly nine years after the accident, there is no justification for discovery to be incomplete prior to trial because data for expert testimony has not been compiled, and we refuse to give an interpretation to Rule 26(e) that would provide such a tactical justification.
The dissent, we respectfully suggest, misses the point. We are not interfering with the judge’s discretion by reversing him on this error. The point of this discussion of failure to answer the interrogatories properly is simply to emphasize that appellant was under no warning to put in her evidence that the stump impact could not have caused the first fracture in her direct case, because the appellee had hidden its theory — to be sprung only after plaintiff had rested.
Having concluded that appellant was unjustifiably denied discovery of Chrysler’s fracture theory, we turn to the rulings of the trial court which made that fact critical to appellant’s right to a fair presentation of its ease to the jury.
2. Exclusion of Appellant’s Rebuttal Evidence
In the defendant's case the expert Mazur gave his opinion of the cause of the two-stage fracture based on a test which had not been disclosed in the answers to the interrogatory. Based upon that test, Mazur gave it as his opinion that the force of impact on the rim of the right wheel from hitting the tree stump had been sufficient to fracture the Pitman arm stud, thus accounting for the first fracture. On rebuttal, plaintiff offered to prove by its expert, Rader, that by actual experiment, the force required to cause a dent in the rim was shown to be far less than that necessary to fracture that stud. This proffered testimony was excluded.
While a trial judge has discretion to exclude rebuttal evidence which would have been admissible if offered as evidence in chief, see French v. Hall, 119 U.S. 152, 7 S.Ct. 170, 30 L.Ed. 375 (1886); Casey v. Seas Shipping Co., 178 F.2d 360, 362 (2 Cir. 1949), such discretion should *458be tempered greatly where the probative value of proffered evidence is potentially high and where such evidence, though admissible on the case in chief, was unnecessary for the plaintiff to establish in its prima facie case. Throckmorton v. Holt, 180 U.S. 552, 563-65, 21 S.Ct. 474, 45 L.Ed. 663 (1901); French v. Hall, supra; Zurich v. Wehr, 163 F.2d 791, 794 (3 Cir. 1947). See National Securities Corp. v. Heinbokel, 154 F.2d 266 (3 Cir. 1946). We believe that the proffered testimony of Rader, even if it might have been part of plaintiff’s case in chief, was not merely cumulative and should have been admitted' in the exercise of sound judicial discretion. Smith v. Dravo Corp., 203 F.2d 369, 377 (7 Cir. 1953); Friend v. C.I.R., 102 F.2d 153, 155 (7 Cir. 1939).
But we also believe that Rader’s testimony was not necessarily part of plaintiff’s case in chief. Plaintiff presented a prima facie case that the steering mechanism was defective. Her own testimony that the steering failed, the finding of a twice-fractured Pitman arm stud after the accident, and the testimony of her experts 6 that the first fatigue fracture could have been followed by a final complete fracture before the car left the road satisfied her burden to go forward. She had made her prima facie case.7 In *459the- words of Wigmore “[F]or matters properly not evidential until the rebuttal, the proponent has a right to put them in at that time, and they are not subject to the discretionary exclusion of the trial court . . . matters of true rebuttal could not have been put in before, and to exclude them now would be to deny them their sole opportunity for admission.” Wigmore, Evidence § 1873 at p. 517 (3d ed. 1940) (emphasis in original).
That does not mean, of course, that in every case where evidence is improperly excluded on rebuttal there must be a new trial. On the contrary we must approach such questions with a rational liberality. The problem here is that the issue was crucial.8 The jury was entitled to hear the evidence, and the plaintiff had no cause or duty to go forward to negative in its case in chief the defense opinion of Mazur negating as the cause of the accident the defect in the Pitman arm stud. “The plaintiff was not required to offer evidence which positively excluded every other possible cause of the accident.” Rosenberg v. Schwartz, 260 N.Y. 162, 166, 183 N.E. 282, 283 (1932). As we have noted, contrary to what the dissent suggests, Chrysler’s proof of stump impact was not directed at establishing a “third alternative — that one stage of the fracture occurred prior to the car’s leaving the roadway and the second afterwards.”
The argument that the defendant’s theory was elicited on the cross-examination of plaintiff’s expert, Moseley, and that the plaintiff should, therefore, have tendered his expert testimony on the subject in her direct case is not correct.
It is true that Moseley was asked a question on cross-examination about a possible connection between the dent in the rim and the fracture of the stud. That did not make his testimony in response a part of plaintiff’s direct case. Moreover, Moseley’s response to that question was equivocal (his attempted explanation was cut off by Chrysler’s counsel). Its significance to the case was not apparent at the time because' the theory which Chrysler would ultimately develop in Mazur’s testimony was still unknown to the plaintiff. We choose to follow the Wigmore rule that the rebuttal evidence should have been admitted, even though appellant was permitted, on redirect, to question Moseley on his response in the cross-examination.
“[W]here the proponent has found it necessary or desirable, by reason of the opponent’s cross examination, partly to anticipate his case in rebuttal by going to it during his case in chief,— for example, on a re-direct examination; here he may take up the same subject again during the rebuttal.” VI J. Wigmore, supra, § 1873, at 517 (3d ed. 1940) (emphasis in original).
Since we have found that the exclusion of Rader’s testimony was critical because of its relevance to the central issue of the case, its exclusion is manifest error. The testimony excluded in this bat-*460tie of experts might have changed the verdict.9
Reversed and remanded for a new trial.
. Chrysler abandoned another theory during pre-trial discovery that the first stage fracture may have occurred three months before the accident during a minor collision between plaintiff-appellant’s car and a car owned by Frederick Elfers. A separate cause of action against Elfers was discontinued before submission to the jury.
. The issue on appeal is summarized by Chrysler’s brief as follows: “Plaintiff contended at trial that the steering failure was caused by the failure of a part of the vehicle known as the Pitman Arm Stud, asserting that the part failed in two stages. The first stage was occasioned by a product defect, a fatigue fracture, which ultimately progressed under ordinary driving stress to a second, sudden and complete fracture at the time of the accident. Defendants contended however that the two-stage fracture of this part was occasioned by a series of impacts which followed plaintiff’s loss of control due to her extremely high rate *453of speed on a Connecticut country road whose topographical configuration would cause a vehicle to go out of control at such speed.” (Emphasis added).
. Judge Griesa charged the jury that Chrysler contended that the first stage fracture could have occurred as a result of the earlier collision with the Elfers car. Contrary to an inference that might be drawn from note 3 of the dissenting opinion, it was Chrysler’s counsel who objected to that charge on the ground that it was not their theory of the case. He stated: “Chrysler Corporation does not contend that any condition arose to the Pitman stud from the Elfers accident.” (p. 38). Finally, after discussion, however, Judge Griesa did not withdraw this portion of the charge even though Chrysler was disavowing the contention, probably because he felt any further instruction would only confuse the jury. In any event, the theory on which Chrysler relied most heavily was that both cracks had occurred after the car left the road. This is enough to make the exclusion of testimony prejudicial error even if there was some vague hinting of another theory.
. The dissenting opinion notes that Dr. Rader was permitted to testify on other matters. We are at a loss to understand how this cures the error in not letting him testify to the crucial issue. No one contends that Judge Griesa was prejudiced against appellant.
. These interrogatories and Chrysler’s entire response are as follows:
“45. Following the alleged failure of the steering and the stud on the occasion in ques*455tion on November 14, 1964, were any tests, examinations or inspections made by defendant, or in defendant’s behalf, in an attempt to determine the cause of the failure.
“46. If so, for each study, state:
(a) its description
(b) the time and date it was performed
(c) the place it was performed
(d) the name, address, telephone number and job title, or capacity of the person who performed it
(e) the nature and substance of all information which was made available to the person conducting the study with respect to the item and the circumstances of its failure
(f) the nature and extent of the experience in such work of the person who conducted the study and interpreted the results
(g) the name, address, telephone number and job title or capacity of the person who authorized its performance
(h) the description of any equipment used (stereo-binocular microscope, etching chemicals, etc.)
(i) the description of the procedure used
(j) its results and the conclusions arising therefrom.”
(Interrogatories submitted April 24, 1970).
Further Interrogatories submitted November 24, 1970 included:
“50. Identify each person whom the defendants Chrysler expect to call as an expert witness at trial, setting forth: . . . [description of expert witness]
“51. As to each such person named in response to Interrogatory No. 50 state the subject matter on which that expert [is] expected to testify.
“52. As to each such person named in response to Interrogatory No. 50, state the substance of the facts and opinions as to which each such expert is expected to testify and set forth a summary of the grounds for each opinion.”
Response submitted May 12,' 1971:
“46. In September, 1968 an electron microscopic examination of a portion of the fracture force of the Pitman arm stud was conducted by Mr. Thomas B. Turnbull, an Electron Microscopist employed by Ernest P. Fullam Inc., in Schenactady, New York, which examination was authorized by John H. Dudley an attorney, employed by Chrysler Corporation, P.O. Box 1919, Detroit, Michigan. The address and telephone number of Mr. Turnbull is now known. Upon information and belief he is an expert of many years in the field of electron microscopy. The information available to him consisted of a portion of the fracture face of the Pitman arm stud alleged to have come from the vehicle which is the subject matter of this litigation, and upon information and belief, also one photograph, these items having previously been obtained from Herman J. McCarthy, Esq., attorney for Lillian Weiss. Upon information and belief, Mr. Turnbull utilized an electron microscope and etching chemicals, and arrived at his conclusion of an impact load induced failure to the Pitman arm stud resulting from the happening of the accident.
“Also, in September, 1968, Donald Gregory, a metallurgical engineer with many years of experience in failure analysis, employed .by Chrysler Corporation performed a visual and microscopic examination by means of microscope under the authorization of John H. Dudley of the same portion of the fracture face of the Pitman arm stud, and arrived at a conclusion similar to that of Thomas B. Turnbull.”
In response to No. 50, there was included “Sylvester Mazur, Products Engineer, employed by T.R.W., Inc., who was previously deposed by the plaintiff.”
In response to Nos. 51 and 52, the response regarding Mazur was:
“Sylvester Mazur will testify, it is expected, regarding the manufacturing techniques, procedures and inspections by his employer, Thompson Ramow Wolldridge, in its capacity as manufacturer of the Pitman arm stud and related components of the. suspension. More specific information as to Mr. Mazur’s anticipated testimony is not known at the present time, although it is noted that he will also testify regarding matters covered in the deposition of Sylvester Mazur, conducted by the plaintiff in the within action on February 17, 1970.”
Supplemental Interrogatories submitted October 11, 1972:
“3.a. State all the causes which contributed to causing the accident of November 14, 1964;
b. Do defendants Chrysler contend that the Chrysler automobile here involved sustained any damage in the accident of August 3, 1964 that in any way contributed to the happening of the accident of November 14, 1964, and if so, how?
c. Set forth in detail what portions of the Chrysler vehicle here involved were damaged on August 3, 1964, what the nature of that damage was, how such damage affected the operation of the vehicle, when, how and to whom such damage or malfunction first became to [sic] any of the defendants Chrysler, the precise location of such damage, the manner, if any, in which such damage contributed to the happening of the November 14, 1964 accident, and set forth all the facts which demonstrate that the damage sustained by the Chrysler automobile on August 3, 1964 contributed to causing the accident of November 14, 1964 and identify all documents and the present location and custody of all documents believed by defendants Chrysler to be relevant to such contention.”
Response submitted November 8, 1972:
“3. The defendants respond as follows:
(a) Objection
*456(b) The defendants at the present time make no contention, one way or another, with regard to the prior accident since it is the plaintiff who claims a casual connection in this regard.
(c) Upon information and belief the damages as sustained in the prior accident are reflected in the appropriate repair records.”
. (a) Gordon, the Yale metallurgist, testified as follows:
“I believe that the sudden steering failure was a consequence of the final failure of the Pitman arm pin, which had been weakened over a considerable period of time by the growth of one or two fatigue "cracks; that when the cross-section of the pin had been so reduced by the growth of these cracks that it could no longer sustain the forces applied to it, in the ordinary operation of the vehicle that a sudden failure of the remaining section of the pin took place, leaving the driver with no control over the steering of the front-wheels.” (App. p. 582)
s(s * * * * *
“[A] fatigue crack grows under the application, repeated application, of small stresses, stresses which are far below the stress required to produce failure in the sound material. No unusual events are required to cause the fatigue crack to continue to grow.
“Q. So that the final sudden failure could occur during ordinary use?
“A. Yes.” (App. p. 635)
(b) Moseley testified that his analysis of the scene of the accident led him to believe that steering failure occurred on the road, before the accident began to occur:
“Q. When the pitman arm pin — -just to be clear on it, Mr. Moseley, we will come back to it — but am I correct that it is your opinion within a reasonable degree of scientific certainty, that the pitman arm pin that the pin went into failure, that it fractured, it stopped connecting the steering wheel in the driver’s hands to the mechanism below, sometime before this accident began to occur?
“A. Yes.
“Q. The moment before the accident began to occur.
“A. Yes.
“Q. That is your opinion?
“A. That is correct.” (App. p. 334)
******
“A. My findings and my opinion is that the vehicle went out of control because the pin connecting the steering link to the pit-man arm failed while the vehicle was in motion on the road.” (App. p. 357).
(c) This is generally corroborated by the testimony of Rader. At App. pp. 1089-95, he describes the weakness of a fatigue-fractured Pitman stud, and its load-bearing abilities, on the basis of Gordon’s prior testimony that the fatigue fracture extended throught 44% of the stud’s diameter. Thereafter on cross, Rader describes the extra stress on a Pitman stud in driving on a road like Route 123:
“Q. We are going down a straight and level country road and we hit a few bumps; you say that that changes the direction of travel and, therefore, you correct it with the wheel; is that correct?
“A. That’s right.
“Q. That’s enough to make any significant increase in load on the Pitman stud?
“A. It could be very significant.” (App. p. 1160).
. Because of our holding that the preclusion of Rader’s rebuttal testimony requires a new trial, we need not reach the question whether the trial judge properly denied appellant’s application to strike Mazur’s testimony on this subject, which included as a ground that plaintiff had not been served with any updated information with respect to Interrogatory No. 46. We note that at a hearing on plaintiff’s motion to strike Chrysler’s answer for failure to respond to interrogatories, the trial judge had raised the possibility that Chrysler might be “precluded from putting in any evidence on the trial as to fractures.” (App. p. 64). On a new trial the problem will have become moot because of the disclosure at the first trial.
. The court charged on the issues under the breach of warranty claim in part as follows:
“The threshold question, the first and basic question you must decided is whether the plaintiff has proved that in the November 1964 accident the plaintiff’s car left the road because of a steering defect. To be more precise, did this Pitman arm pin fracture on the road causing the car to go out of control, or did the car go out of control because of speed and/or road conditions with the fracture of the Pitman arm occurring from one or more of the impacts received on the embankment?
“The major part of the evidence at the trial has been devoted to this question, and 1 won’t attempt to restate it.” (Emphasis added).
He charged further:
“If you believe Chrysler’s version, or if you find that Chrysler’s version about fracture upon impact after the car left the road is just as likely as plaintiff’s version about steering failure on the road, that is, if the scales are evenly balanced on this question, then plaintiff has not sustained her burden of proof on this threshold question.”
. On appellee’s contention that there was an insufficient offer of proof, the trial court said that it understood “your offer of proof.” The related argument that Rader’s proposed rebuttal was inadmissible because his method of measurement was unlike Mazur’s is specious. Rader’s experiment was of obvious relevance to the conclusion of Mazur that the force of the stump impact could have fractured the Pit-man stud.
In view of our conclusion, it is unnecessary to consider appellant’s other claim of error.