This is an appeal from a judgment on a jury verdict in favor of the defendant American Motors Corporation. The plaintiff-appellant, Bair, was injured on October 26, 1966, when, in an automobile accident, she was ejected from an American Motors Rambler which she had purchased in August, 1966, from an American Motors' dealer. She sued American Motors on both negligence and strict liability theories, alleging that the defective design of the door latch on the Rambler failed to prevent the door from opening when the car was subjected to upward and longitudinal stresses in the accident, thus permitting her ejection and increasing her injuries.1 Pennsylvania law governs liability.

The case was submitted to the jury on three interrogatories:

"1. Was plaintiff guilty of negligence which contributed to the happening of the collision between her Rambler and the Mustang?

2. Was defendant guilty of negligence in the design of the door latch on the 1966 Rambler?

3. Did defendant sell a product in a defective condition unreasonably dangerous to the user when it sold the 1966 Rambler equipped with the door latch described in this case?"

To each interrogatory the jury answered no.

Bair contends on appeal that the negative answer to the second and third interrogatories are the product of trial errors, and that a new trial should be granted. The errors alleged consist (1) of a ruling as to the admissibility of the results of statistical studies conducted by Cornell Aeronautical Laboratory, Inc., Automotive Crash Injury Research, offered during the testimony of the plaintiff's expert witness, Walter V. H. Pruyn; and (2) of allegedly disparaging treatment by the trial court of that witness.

Mr. Pruyn, an expert who gave opinion evidence, was the only liability witness for Bair, and American Motors did not put in any liability evidence. The district court held that plaintiff's liability case was sufficient to go to the jury. The jury, apparently, did not accept Pruyn's opinion. The excluded evidence was not cumulative of any other evidence in the case. Thus, if it was improperly excluded, and if it tended to support Pruyn's opinion, the error cannot be regarded as harmless.

The event was an intersection accident in which Bair's vehicle was struck on the left front fender by a vehicle travelling at 50 to 60 m. p. h. Pruyn's theory of the accident was that after the initial impact there was a second impact, as the two cars swerved from a right angle to a parallel alignment, in which the right rear of the striking vehicle struck the Rambler, and that following this impact the left front (driver's) door of the Rambler opened. As the Rambler continued to rotate Bair was ejected. Pruyn claimed that had the Rambler been equipped with door latches then used by other automobile manufacturers to prevent opening when the vehicle was subjected to upward and longitudinal stress such as occurred in the accident, Bair would not have been ejected.

The door latch of the Rambler was designed so that the bottom portion of the latch's ratchet wheel teeth, mounted on the door, was, when the door was closed, enclosed by a steel retaining plate affixed to the left door striker pillar. The upper portion and sides of the ratchet wheel were not so enclosed. Thus, when the door was subjected to upward and longitudinal stress which lifted the teeth of the ratchet wheel above and out from behind the steel retaining plate the door was free to open. To prevent this occurrence some manufacturers had as early as the 1961-62 model year equipped their doors with a T-head type bolt assembly in which, when the door was closed, the bolt was totally enclosed at the top, bottom and sides. Bair's claims, advanced through Pruyn's opinion testimony, was that American Motors' failure to adopt such a design by 1966 was negligent, and that the absence of such a latch mechanism made its product inherently defective.

The district court ruled that Pruyn was qualified to express such opinions; appellant claims unenthusiastically. During his direct testimony Bair sought to prove through him the findings of the three statistical surveys by Automotive Crash Injury Research of Cornell University as a basis for his opinion that the absence of the T-head type bolt assembly was a negligent design and produced an inherently unsafe vehicle. The surveys were objected to as hearsay, and the objection was sustained on the specific ground that since they were not made by Mr. Pruyn he could not testify to them. (N.T. 89).

Each of the surveys was made by Automotive Crash Injury Research of Cornell Aeronautical Laboratory, Inc., Cornell University. That organization is engaged in the study of injury causes among occupants in automobile accidents. Its research has been supported by grants from the National Institute of Health and the Division of Accident Prevention of the United States Public Health Service, the Automobile Manufacturers Association, Inc., and the United States Department of Transportation, National Highway Safety Bureau. The research methods pursued by that organization, as described in the exhibits, is, with the cooperation of police and medical authorities around the country, the analysis after the event of large numbers of actual injury-producing accidents to determine the safety performance of various components having injury causative or preventive potential. The studies here in issue are:

1. "An Evaluation of Door Lock Effectiveness: Pre-1956 vs. Post-1955 Automobiles", published in July, 1961. (Exhibit P-12)2

2. "The Safety Performance of 1962-63 Automobile Door Latches and Comparison with Earlier Latch Designs", published in November, 1964.

3. "Comparison of Door Opening Frequency in 1967-1968 cars with Earlier Model U.S. Cars", published as a final report in May, 1969.

These reports tend to support Pruyn's opinion that the T-head type bolt assembly was significantly more likely to prevent ejection. The first report referred to the pioneering 1954 Automotive Crash Injury Research report which established that, contrary to a widely held pre-1954 general belief, passenger car occupants ejected during an accident had a lower chance of avoiding serious injury or death than those not ejected. The 1961 report studied 14,135 automobiles from injury-producing accidents involving 31,855 occupants. It compared ejection rates for pre- and post-1956 automobiles, since in that model year American manufacturers for the first time introduced a modified safety door latch designed to reduce door opening on impact. It found a significant lowering of the incidence of ejectment. The 1964 report studied 24,342 cars from injury-producing accidents. It compared ejection rates for pre- and post-1963 automobiles, since in the 1962 and 1963 model years some manufacturers made further improvements in the door latch mechanism. It found a further significant lowering of the incidence of ejectment. Essentially, according to Pruyn, the 1966 Rambler had the type latch adopted by all manufacturers in 1956, while he contended it should have had the type latch adopted by other manufacturers in 1962-1963. Had Pruyn been permitted to support his opinion by reference to these studies, his contention that American Motors had neglected to incorporate in its 1966 door latch design the current state of the art in safe design would undoubtedly have been reenforced.

Bair refers to that part of Rule 43(a), Fed.R.Civ.P. which provides:

"All evidence shall be admitted which is admissible under the statutes of the United States, or under the rules of evidence heretofore applied in the courts of the United States on the hearing of suits in equity, or under the rules of evidence applied in the courts of general jurisdiction of the state in which the United States court is held."

She urges that the excluded evidence, though hearsay, was admissible on all three grounds.

She contends, first, that the reports in issue were admissible under 28 U.S.C. Sec. 1732, as business records of Cornell Aeronautical Laboratory, Inc. made in the regular course of its business. No objection was made to the authenticity of the reports, and their regular method of compilation is set forth on their face. Possibly a case can be made out for their admissibility under 28 U.S.C. Sec. 1732. That statute as a ground for admissibility was not urged upon the district court. The defendant had no opportunity to develop possible grounds for objection to admissibility as business records. A ruling that the studies were admissible under 28 U.S.C. Sec. 1732 would make them admissible for all purposes. We are not prepared to go so far on this record.

Bair contends, next, that the contents of the reports were admissible under "the rules of evidence heretofore applied in the courts of the United States . . . [in] suits in equity." She refers to a rule that there may be introduced in evidence in the direct examination of an expert statements contained in so-called learned treatises relied upon by him in the formation of his opinion.3 This exception to the hearsay rule has been urged by commentators. VI J. Wigmore, Evidence, Secs. 1690-92 (3d ed. 1940); C. McCormick, Handbook of the Law of Evidence, Sec. 321 (2d ed. 1972). There is some federal authority which strongly supports the commentators' position. In Western Assurance Co. v. J. H. Mohlman Co., 83 F. 811, 821 (2d Cir.), cert. denied, 168 U.S. 710, 18 S. Ct. 949, 42 L. Ed. 1213 (1897), the Second Circuit recognized that an expert could read from excerpts from reports prepared by the United States Department of Agriculture under the direction of the Division of Forestry, containing tables which comprised the results of 2,000 tests by the Government on the crushing strength of different kinds of timber, and from two other standard reference books listing such crushing strengths. The books and report were identified as recognized authorities among builders and engineers. The Second Circuit relied as authority for admissibility upon Vicksburg & Meridian R. R. Co. v. Putnam, 118 U.S. 545, 7 S. Ct. 1, 30 L. Ed. 257 (1886), in which the Supreme Court recognized that standard life and annuity tables showing at any age the probable duration of life, are competent evidence; ". . . and yet these tables show merely the deductions from records of past transactions, when neither the record of the transactions nor the individual who has worked out the deductions is called to testify to the accuracy of his work, or to the conditions under which it was performed." Western Assurance Co., supra, at 820-821. The Second Circuit opinion observes that the learned treatises most commonly dealt with in the cases are medical works, but that there are many other such works, including, besides annuity tables, almanacs, astronomical calculations, tables of longarithms, interest tables, weather reports, and tables of the rise and fall of the tide, all of which have been admitted in evidence. See also Woelfle v. Connecticut Mutual Life Insurance Co., 103 F.2d 417 (8th Cir. 1939).

The information in the Cornell Aeronautical Laboratory reports, though not statistical tabulations of experiments conducted by that organization, and hence in one sense distinguishable from at least one of the three texts referred to in Western Assurance Co., is closely analogous. Timber can be subjected to crushing stresses in laboratory conditions, while people may not be. The compilation of data from the unfortunately all-too-frequent actual injury-producing accidents is the best substitute available. That courts must, on hearsay grounds, be deprived of the use of the collected data on which other departments of government, industry, and the engineering profession obviously rely, makes no more sense now than in 1897, when that notion was rejected by the Second Circuit.

Neither Western Assurance Co. v. J. H. Mohlman Co., supra, nor Vicksburg & Meridian R. R. Co. v. Putnam, supra, are literally authority as to the rules of evidence applied in the United States courts in suits in equity. Rule 43(a). This would present no insurmountable objection to our reliance upon them, for it is now clear that we will assume, in the absence of an equity precedent, that the equity courts would have followed evidentiary rulings from the law side. Treharne v. Callahan, 426 F.2d 58, 63 (3d Cir. 1970); United States v. 60.14 Acres of Land, 362 F.2d 660, 666 (3d Cir. 1966). Cf. Grossman v. U. S. Slicing Machine Co., 365 F.2d 687, 689 (3d Cir. 1966).4 In this instance, however, we have what is usually not available, an equity precedent expressly approving the rule of Western Assurance Co. v. Mohlman Co., supra, for in G. & C. Merriam Co. v. Syndicate Publishing Co., 207 F. 515, 518 (2d Cir. 1913), appeal dismissed, 237 U.S. 618, 35 S. Ct. 708, 59 L. Ed. 1148 (1915), the Second Circuit approved the district court equity decree and opinion to that effect. Thus, under Rule 43(a) Bair's expert should have been permitted to refer to the Cornell Aeronautical Laboratory, Inc. reports.5

Since it is possible that a retrial of this case might not be completed until after July 1, 1973, the effective date6 of the new Rules of Evidence for United States Courts and Magistrates, 56 F.R.D. 183 (1972), it is appropriate to comment that under those rules the new federal rule would be the same. Rule 803 governs hearsay exceptions. Rule 803(18) provides:

"Learned treatises.-To the extent called to the attention of an expert witness upon cross-examination or relied upon by him in direct examination, statements contained in published treatises, periodicals, or pamphlets on a subject of history, medicine, or other science or art, established as a reliable authority by the testimony or admission of the witness or by other expert testimony or by judicial notice. If admitted, the statements may be read into evidence but may not be received as exhibits."

As originally proposed this rule would have permitted the treatises to go to the jury. See Rule 8-03(b)(18) Preliminary Draft of Proposed Rules of Evidence for the United States District Courts and Magistrates, Committee on Rules of Practice and Procedure of the Judicial Conference of the United States, March, 1969, 46 F.R.D. 161, 349. The last sentence was added with the explanation:

"The rule avoids the danger of misunderstanding and misapplication by limiting the use of treatises as substantive evidence to situations in which an expert is on the stand and available to explain and assist in the application of the treatise if desired. The limitation upon receiving the publication itself physically in evidence, contained in the last sentence, is designed to further this policy." Revised Draft of Proposed Rules of Evidence for the United States Courts and Magistrates, Committee on Rules of Practice and Procedure of the Judicial Conference of the United States, March 1971, at 120. [51 F.R.D. 315, at 434].

The permitted use is precisely that approved in Western Assurance Co. v. J. H. Mohlman Co., supra, and Woelfle v. Connecticut Mutual Life Insurance Co., supra.

Finally Bair contends that Pennsylvania law also recognizes a learned treatise exception to the hearsay rule. Clearly it recognizes such a general exception with respect to mortality tables. E. g., Rosche v. McCoy, 397 Pa. 615, 156 A.2d 307 (1959). Bair contends that in Commonwealth v. Thomas, 444 Pa. 436, 282 A.2d 693 (1971), the Pennsylvania Supreme Court adopted a rule as to what matters an expert could refer to at least as broad as that announced in Western Assurance Co. v. J. H. Mohlman Co., supra. We do not find the statement of the Pennsylvania Supreme Court in Thomas to be as clear an adoption of this position as Bair suggests. Closer in point, perhaps, is Grantham v. Goetz, 401 Pa. 349, 164 A.2d 225 (1960), in which the Pennsylvania Supreme Court upheld a trial court's ruling which permitted use of literature prepared to accompany a prescription drug only for the limited purpose of reading and comment by an expert witness.7 Neither case gives as clear guidance as the federal authority to which we have referred, and since until July 1, 1973, under Rule 43(a) we must follow the rule favoring admissibility, we need not here decide the full scope of the learned treatise exception in Pennsylvania.

Since the erroneous exclusion of evidence requires a new trial, there is no occasion to rule upon Bair's additional contention that the court unfairly disparaged her expert.

The judgment of the district court will be reversed, and the case remanded for a new trial.