This appeal brings before us a second time the great Cleveland gas tank disaster of October 20, 1944. The facts are fully set out in our earlier opinion, 3 Cir., 1948, 166 F.2d 908 and need not be repeated. In the meantime, a companion case has come to the Supreme Court of Pennsylvania. The view of the Pennsylvania law expressed in our earlier decision was confirmed by the state court. Foley v. Pittsburgh-Des Moines Co., 1949, 363 Pa. 1, 68 A.2d 517. The legal rights and duties of the parties have thus been settled. The present appeal concerns alleged errors in the second trial, which terminated with a verdict and judgment for the defendants. We proceed to consider the grounds for reversal upon which the plaintiff relies.

I. The Shape of the Tank as an Element of Negligence.

The plaintiff contends that the trial judge erred in refusing to submit to the jury the question of defendants' negligence in designing and building the tank in the shape of a cylinder rather than in the shape of a sphere.

At the time East Ohio, the public utility distributing gas in Cleveland, ordered the tank here involved from the defendants, it had standing on its premises three tanks of spherical design, each of which had a capacity half that of the proposed new tank. The spherical tanks had also been designed and built by the defendants. At East Ohio's request for additional unit capacity, the defendants prepared designs for a spherical as well as a cylindrical tank, and wrote to East Ohio that "As to the relative merits of the two designs we are confident that the toro-segmental bottom tank will be just as satisfactory as the spheres have been and we feel that they should be preferred because of the saving in cost." Despite this statement, the defendants make here the unmeritorious argument that they are not responsible for the design because once having entered into a contract with East Ohio for the construction of a cylindrical tank, they could not have built one in the shape of a sphere. The argument ignores the fact that the terms of the contract followed recommendations made by the defendants; if those recommendations were negligently made, then the defendants are responsible.

The trial judge refused to submit to the jury the question of negligence as to the form of the tank because he thought that the experience acquired in the construction and apparently successful operation of three spherical tanks of half the capacity of the cylindrical container was not sufficient to establish a custom or standard of usage in the industry. He said, "As a matter of law I am charging you you are not concerned or you should not consider whether the spherical tank was safer than the cylindrical tank; * * *"

This was error. We said in our prior opinion that the evidence required submitting to the jury the question of "The Use of a dangerous Cylindrical form * * * instead of the relatively safer Spherical form." 166 F.2d at 916-917. The evidence presented at the first trial on this point was almost identical with that submitted at the second. If that evidence was sufficient in the first trial to require submission to the jury, it presented an issue for jury determination in the second trial as well. It makes no difference that objections by the defendants to the admission of this evidence may or may not have been made upon different theories at the two trials. What does matter is that testimony bearing upon the allegedly defective form was in evidence and that we had already held the question raised by it to be one for the jury.

There seems to be some confusion of thought arising from the fact that no usage reaching the dignity of custom had been established in favor of the spherical shape. No one contends that there was a standard established by custom and usage in the industry; it is clear from undisputed statements of fact that this type of tank-building was new. Precedents on what is or what is not practice of the industry to show observance of due care have nothing to do with our situation in this case and are therefore irrelevant.

The problem was to determine whether the defendants had been negligent in designing, supplying and recommending a tank of cylindrical design. The evidence offered by the plaintiff tended to prove that reasonable persons of the skill which the defendants held themselves out to possess would not have constructed a tank of cylindrical form in this situation. The rest was for the jury.

It is significant here, too, that the Supreme Court of Pennsylvania, in the case mentioned above, involving this very disaster, held that under Ohio law the question of negligence involved in the supplying of the cylindrical tank was for the consideration of the jury. 363 Pa. at 11-13, 68 A. 2d at 522-523.

II. Questions Concerning Testimony of James O. Jackson.

The plaintiff called James O. Jackson as a hostile witness. Jackson admitted that he was an officer of the defendant corporation, and under Rule 43(b) of the Federal Rules of Civil Procedure, 28 U.S.C.A., it is quite clear that so to call him was proper.1 Five serious errors occurred in connection with the examination of this witness. (a) The trial judge restricted examination by the plaintiff's counsel to Jackson's activities relating to the corporation. This ruling was erroneous. Jackson said that fabrication was done at the corporation's plant by the corporation, but that outside assembly and erection were done by corporation employees on behalf of the partnership. Jackson himself admitted that for many years it had been the uniform practice for him to sign letters written on the partnership letterhead as manager of the partnership's engineering and purchasing departments. Dozens of such letters were admitted in evidence and Jackson admitted that this practice was known to and unobjected to by the partners. On this record, we think that Jackson was a "managing agent" of the partnership within the meaning of Rule 43(b) and that examination of him as an adverse witness should not have been restricted only to his activities on behalf of the corporation.

(b) Jackson's testimony when called by the plaintiff related to certain patents which he had applied for and which he had assigned to the defendant corporation. In these patents were certain statements about the minimum requirements for the steel which was to be used in gas storage tanks, which the plaintiff correctly contended constituted admissions. Despite the express provision of Rule 43(b) that witnesses in such situations may be cross-examined only upon the subject matter of the examination in chief,2 the trial judge permitted cross-examination by defendants' counsel to range far afield. The transcript indicates that a Pennsylvania decision was cited to the trial judge as a basis for the ruling. It is clear, however, from the authorities cited that this is not a state law matter at all, but one to be governed by the federal rules and decisions.

It may well be that this mistake, by itself, would not be so prejudicial as to require a new trial, though the contrary has been declared. Jackson could and undoubtedly would have given the same testimony if and when called by the defendants for direct examination, so the testimony given would have been before the jury anyhow. But since there must be a reversal on other grounds, we call attention to Rule 43(b) for the trial judge's future guidance.

(c) We find, moreover, still a third error in the trial court's handling of the Jackson testimony and this one was highly prejudicial. The jury was instructed, with regard to it, that "to the extent that testimony as produced from a person called for cross examination is not rebutted by either direct proof or circumstances it is conclusively taken to be true." A fair inference from the entire charge was that in the absence of rebuttal the plaintiff was bound by everything Jackson said.

We do not have here a case in which a party calls a supposedly favorable witness who gives unfavorable testimony. It is frequently said in such situations that the party calling the witness is bound by his testimony. That rule has been assailed by Wigmore as a "primitive notion" which "no longer finds defenders."3 But we need not in this case either affirm or repudiate that rule. Here, Jackson was called in the first place as an adverse witness under Rule 43(b), which expressly provides that such an adverse witness may be contradicted and impeached. Rule 43(b), we think, is utterly inconsistent with any notion about being bound by his testimony. It seems to us that any statement to the effect that a party is bound by the testimony of a witness whom he is free to contradict and impeach is inherently anomalous.

(d) The plaintiff complains that the witness Jackson was permitted to state that certain things done by East Ohio, the operating company, increased the hazard in the storage of the gas. There is no merit in this objection. The witness was an expert. The claim that acts of a third person increased the danger was part of the problem before the jury upon the question whether the defendants were negligent and whether their negligence caused the harm.

(e) Exclusion of the Patents. The plaintiff offered in evidence six patents or patent applications. Two of these were admitted generally. The remainder were excluded except for the purpose of impeaching James O. Jackson. It is contended by the plaintiff that the patents were substantive evidence and should have been admitted as such. The patents represented designs for gas storage tanks the same or comparable to the one which was the origin of this disaster. All of them were the work of Jackson and all had been assigned by Jackson to the defendant corporation. We think that exclusion of these patents was error. The rule is well settled which permits the receipt of admissions made by an agent of a party when the agent's powers are broad enough to constitute him the general representative of the principal with broad managerial responsibilities. McGrath v. Pennsylvania Sugar Co., 1925, 282 Pa. 265, 127 A. 780; York Mfg. Co. v. Chelten Ice-Mfg. Co., 1924, 278 Pa. 351, 123 A. 327; 4 Wigmore on Evidence § 1078 (3d ed. 1940). Since Jackson was manager of the defendants' purchasing and engineering department and since he designed the tank for the defendants, concluded negotiations on their behalf with East Ohio and finally supervised assembly of the tank, we have no doubt that his statements in the patents should have been received as admissions against all the defendants. Moreover, it was Jackson himself who applied for the patents in question and then assigned them to the corporate defendant. He was therefore a "privy in title" whose admissions could be received against his successors in interest. See 4 Wigmore on Evidence § 1080 (3d ed. 1940).

Here again we are not at all sure that this limitation imposed upon the reception of part of these documents would alone constitute reversible error. The jury had the benefit of most of the testimony in those which were admitted; the additional material would have been in part cumulative. But since the case must go back for a new trial the point is made clear for the benefit of the parties and trial judge.

(f) Writings of James O. Jackson. James O. Jackson had written several technical studies of the East Ohio storage tanks, and these the plaintiff offered as admissions. For the reasons given in the discussion of the patents, supra, we think that the trial judge erred in excluding them.

III. Bureau of Mines Report as Competent Evidence.

The plaintiff offered in evidence a report of the Cleveland disaster by the Bureau of Mines of the United States Department of the Interior. The report, signed by five Bureau experts who made their own investigation, concluded that the tank disintegrated rather than exploded and that "the fractures were characteristic of failure due to low-temperature embrittlement." Both conclusions supported important elements of the plaintiff's case. The offer of the report was based upon the Uniform Business Records as Evidence Act, as adopted by the General Assembly of Pennsylvania, 28 P.S. §§ 91a-91d, and upon the similar federal statute, 28 U.S.C.A. § 1732.4 If the report was admissible under either statute, its exclusion was error. Fed.R.Civ. P. 43(a).

The Bureau of Mines is charged by Congress with the duty of investigating the "preparation, treatment, and utilization of mineral substances with a view to improving health conditions, and increasing safety, efficiency, economic development, and conserving natural resources * * * in the * * * mineral industries * * *" 30 U.S.C.A. § 3. Moreover, the statute provides that the Bureau "shall prepare and publish * * * reports of inquiries and investigations, with appropriate recommendations of the bureau, concerning the nature, causes, and prevention of accidents, and the improvement of conditions, methods, and equipment, with special reference to health, safety, and prevention of waste in the * * * mineral industries. * * *" 30 U.S.C.A. § 5.

The report which the plaintiff offered had been prepared in obedience to the above statutory provisions. The making of such reports is, by command of Congress, part of the "business" of the Bureau of Mines.5 The report was a record of an "occurrence" or "event" and it was required by act of Congress. The document therefore comes squarely within the language of the federal statute and we think that the District Court erred in excluding it.

The report is no less admissible because it contains conclusions of experts which are based upon hearsay evidence as well as upon observation. These circumstances, by virtue of express statutory provision, go to weight rather than to admissibility.6 Moreover, this Court has several times held that hospital records are admissible under the statute, and certainly medical diagnosis is no less a matter of opinion based upon observation and perhaps hearsay than this report of the Bureau's investigation. Bartkoski v. Pittsburgh & Lake Erie R. Co., 3 Cir., 1949, 172 F.2d 1007; Norwood v. Great American Indemnity Co., 3 Cir., 1944, 146 F.2d 797; Pollack v. Metropolitan Life Insurance Co., 3 Cir., 1943, 138 F.2d 123.

IV. The Allegedly Defective Steel.

There is evidentiary support for the conclusion that before the erection of the tank was completed, it was discovered that one component batch or "heat" of steel showed a "Charpy impact" strength of two foot pounds. In his patents, however, James O. Jackson had said that the minimum strength of steel for such tanks should be of the order of ten foot pounds. The plaintiff sought to establish that not only was the alleged defect known to the defendants but that at one time they considered tearing down the tank and fabricating a new one. The plaintiff did prove that James O. Jackson directed Stuchell and Garner, two employees of the defendants, to consult the Carnegie-Illinois Steel Corp., which had manufactured the steel, and find out what caused the low test strength. Then the plaintiff called a witness from the Carnegie-Illinois complaint department. This witness indicated to the District Court at side bar that he would testify to the effect that at a conference with Stuchell and Garner one of them had said that if later tests confirmed the low impact value of the questioned heat of steel, the tank would have to be torn down. The trial judge then refused to permit plaintiff's counsel to elicit this information, on the ground that Stuchell and Garner were without authority to make admissions for the defendants. We agree. Stuchell and Garner were apparently research engineers. They were nominally on the payroll of the Mellon Institute under "fellowships" paid for by the defendants. The evidence establishes that they in fact took orders from James O. Jackson and that they were reimbursed for expenses by the defendants rather than by the Mellon Institute. But it falls far short of proving that Garner and Stuchell held such managerial responsibilities that what they said could be received as admissions against their employers. See the decision cited earlier in this opinion.

The plaintiff then attempted to prove that the defendants had considered tearing down the tank by eliciting an admission to that effect from James O. Jackson on cross-examination. The trial judge, however, refused to permit questions along that line unless the plaintiff was prepared to contradict with evidence a possible negative response by Jackson. Plaintiff's counsel frankly conceded that although he expected an affirmative reply from Jackson, he would not be able to disprove a negative answer. We think that this limitation of cross-examination was within the area of discretion which is always reposed in a trial judge. The precise wording of a question along the forbidden line of inquiry would of itself, regardless of Jackson's answer, carry an implication of callous disregard for safety on defendants' part. The question would seem to presuppose knowledge by the plaintiff that defendants at one time thought of tearing down the tank. The trial judge might well have concluded that such an implication would be so prejudicial that it should not be permitted without proof.

We have examined the remaining allegations of error and find them to be without merit. But for the reasons given above, the judgment of the District Court will be reversed and the cause remanded for proceedings consistent with this opinion.