The opinion
Taft, J.,
concurring. In cases of this kind, the plaintiff must establish by the preponderance of the evidence that his employment was (1) a proximate cause of an accidental iiijury and (2) that such injury was a proximate cause of the disability for which he seeks compensation (McNees v. Cincinnati Street Ry. Co., 152 Ohio St., 269 , 89 N. E. [2d], 138); and, since the testimony of lay witnesses is without probative value to establish the probability of a proximate causal relationship between such injury and such disability, medical testimony is essential for the purpose. Stacey v. Carnegie-Illinois Steel Corp., 156 Ohio St., 205 , 101 N. E. (2d), 897.
Such medical evidence, to be admissible, must tend to prove that proximate causal relationship. If it does not it must be rejected. Brandt v. Mansfield Rapid Transit, Inc., 153 Ohio St., 429 , 92 N. E. (2d), 1. Otherwise, even if there is other sufficient evidence tending to prove such proximate causal relationship, a *581 jury would be invited to give a weight to such inadmissible medical evidence which it does not have.
This does not mean that a medical witness must, in cases of this kind, testify that the accidental injury was the proximate cause of the disability. Hallworth v. Republic Steel Corp., 153 Ohio St., 349 , 91 N. E. (2d), 690. “Proximate cause” is a legal term. It is even arguable that the use of such a term in a question to an expert witness would be improper because it would make the question call for an answer which would be a legal conclusion and would also represent an invasion of the province of the jury.
In my opinion, the testimony of the first medical witness in case No. 33990, that the accidental injury to the plaintiff in that case was “a cause” of the heart condition, is too indefinite to be of any help to the jury in determining that the injury was the proximate cause of the heart condition; and it was therefore properly excluded.
A plaintiff in this kind of a case has the opportunity not only to select expert witnesses to testify on his behalf but also to frame hypothetical questions calling for answers which will definitely support his claim of proximate causation. Thus, there was nothing to prevent this plaintiff from asking a qualified expert witness a proper hypothetical question calling for his opinion as to whether plaintiff’s accidental injury was the “natural and probable cause” of the heart condition or whether the heart condition was “the natural and probable consequence” of the accidental injury.
If there was any merit to plaintiff’s claim, a qualified expert that he selected to support his claim should have been able to testify at least that plaintiff’s accidental injury was “the cause” (Hall v. Nagel, 139 Ohio St., 265 , 39 N. E. [2d], 612) or “the probable cause” (Halhoorth v. Republic Steel Corp., supra) of his heart condition or “an important contributing *582 cause.” “A cause” may just as probably be legally “a remote cause” as it may be a “proximate cause.” On the other hand, either “the cause” or “the probable cause” or “an important contributing cause” will more probably be legally “a proximate cause” than a “remote cause.”
As stated in paragraph four of the syllabus in Buretas v. Industrial Commission, ante, 549, “to permit a jury to make a choice between two irreconcilable inferences raised by the facts in evidence * * * is to substitute speculation and conjecture for proof.” Likewise, to permit a jury to make a choice between two equally reasonable inferences as to the meaning of expert opinion testimony offered to prove the existence of an ultimate fact in the case by the party having the burden of proving such fact, one of which inferences would tend to indicate existence of that ultimate fact and the other of which would not, would be to allow speculation and conjecture to supply the meaning which such testimony should have to make it relevant and which it does not have without such speculation and conjecture.
In case No. 34018 the medical witness was asked for “an opinion based upon a reasonable degree of medical certainty, as to whether or not there” was “any causal connection between” the accidental injury and the claimed disability. Obviously “any causal connection” could just as probably be legally a “remote” as it could be a “proximate” causal connection. Regardless of how probable or certain it was that there was merely “any causal connection,” there would be nothing to indicate any probability of a “proximate” rather than a “remote” causal relationship. In my opinion, this hypothetical question was objectionable.
However, the answer to this question was “that there was a direct causal relationship between” the accident and the claimed disability. If the ques *583 tion had asked for an opinion as to whether there was such a direct causal relationship, it would, in my opinion, have been a proper question and the answer given would have been admissible.
Such a question should not be considered as using the term “direct” as necessarily meaning “direct” in the legal sense as equivalent to “proximate.” Cross-examination of the witness might disclose that, in using the words “direct causal relationship,” he was describing what would be legally a remote instead of a proximate causal relationship, in which event it would appear that his testimony had no more significance than the answer to the hypothetical question in case No. 33990 and would be subject to a motion to strike. However, in the absence of something to indicate that the words “a direct causal relationship” were not intended to be given their ordinary meaning, those words will more probably indicate what is legally a “proximate” than what is legally a “remote” causal relationship.
If this were an ordinary civil aetion in which the trial judge had sustained an objection to the hypothetical question, his ruling on that objection would be affirmed on appeal notwithstanding the proffered answer. However, the record in a workmen’s compensation case like this is made before a rehearing referee. In this case, the referee overruled the objection to the hypothetical question. Thus, if the sworn answer of the witness is not allowed to cure the defect in the form of the question, counsel for plaintiff will be deprived of a right, which he would have had in an ordinary civil aetion, to remedy that defect by rephrasing the question, although the record clearly discloses that he could have remedied the defect in the form of the question so as to make the sworn answer of the witness admissible. Because of this and since there is nothing to indicate that plaintiff would secure *584 any advantage by having this hypothetical question in its present form instead of in the proper form which the sworn answer of the witness indicates it could have had, that answer should be considered as curing the defect in the form of the question.
The statement in paragraph three of the syllabus represents the expression of an opinion on a question of law which was neither briefed nor argued by the parties and which it is not necessary for this court to consider in deciding these cases.
Although, in case No. 34018, testimony of an expert medical witness as to his reasons for his answer to a hypothetical question was excluded, the decision of this court reversing the- Court of Appeals and the Common Pleas Court can be fully supported by our finding of error in sustaining the objection to the hypothetical question and its answer. If the proffered answer to that question had been admitted, there would have been sufficient evidence to justify submission of the cause to the jury, even in the absence of any further testimony of the witness as to his reasons for that answer.
Nelson v. Industrial Commission, 150 Ohio St., 1 , 80 N. E. (2d), 430, is the only authority cited in the opinion in the instant cases in support of the statement made in the third paragraph of the syllabus. Although the second paragraph of the syllabus in the Nelson case does state that “where a medical expert witness has answered a hypothetical question based upon testimony properly admitted, it is error to exclude the testimony of such witness as to his reasons-for his answer and opinion,” that same paragraph indicates that this quoted statement therefrom represented the expression of an opinion on a question of law which was not used-as the basis for the decision rendered by this court- in that case. The holding of the court was that, even if such testimony was con *585 sidered, there was insufficient evidence to justify submission of that case to the jury. The opinion in that case gives no reasons for this statement quoted above from paragraph two of that syllabus and does not even discuss the questions of law involved therein.
There is no doubt that, where a witness- -has answered a hypothetical question on direct examination, the opposing party on cross-examination may ask him to give the reasons for the opinion which he gave in answer to the hypothetical question. See Bellefontaine & Indiana Rd. Co. v. Bailey, 11 Ohio St., 333 , 336, 337. The primary purpose of cross-examination is to test the accuracy, truthfulness, soundness and thereby the credibility (i. e., its worthiness of belief) of the testimony given by the witness on direct examination. Martin v. Eldon, 32 Ohio St., 282 , 287; Legg v. Drake, 1 Ohio St., 286 , 291, 292. However, it was held by this court at a time when our statute' authorized a party to call the adverse party as a witness “in the same manner and subject to the same rules of examination, as other witnesses are compelled to -testify,” but not as now “as if under cross-examination” (see Section 11497, General Code), that such adverse party witness could not be cross-examined on behalf of himself with a view to his own impeachment “for the reason that he could not allege his own want of credibility;” See Legg v. Drake, supra, 290. Likewise, a party who calls a witness is not permitted to impeach that witness or even to cross-examine him in the absence of hostility of the witness (Hurley v. State, 46 Ohio St., 320 , 21 N. E., 645 , 4 L. R. A., 161) because, by calling the witness, he represents that his testimony is credible.
Ordinarily, cross-examination of a witness is an attack on the credibility of a witness.' Where conducted by a- party adverse to the party offering the witness, that is its usual purpose, although--sometimes the tes *586 timony has been so favorable to the adverse party that the purpose of his cross-examination may be to bolster the testimony of the witness.
To permit a party to ask of his own expert witness on direct examination questions, designed only to elicit answers giving his reasons for an opinion stated in answering a hypothetical question, appears to me to permit that party to cross-examine his own witness for the purpose of bolstering his credibility before that credibility has even been attacked. Such a bolstering of the credibility of one’s own witness would appear to be what might be referred to as “impeachment in reverse” of one’s own witness. It would also provide a very effective opportunity for a party to use his expert witness as a means of making an argument to the jury through the testimony of that witness as to the soundness of the conclusion of the witness before the credibility of that conclusion had even been attacked. The necessity, which justifies permitting an expert witness to state his conclusion in answer to a hypothetical question, would hardly seem to justify the giving of such an opportunity to a litigant before the credibility of that conclusion has even been attacked.
I do not mean to suggest that, on direct examination, an expert witness cannot be asked to give any explanation with respect to his answer. For example, in case No. 34018, the medical expert had examined the plaintiff about whom he was testifying. If, in the hypothetical question, he were asked to base his opinion on certain assumed facts and on his findings, it would be necessary to ask him to state to the jury what the findings were that he took into consideration in formulating his opinion. Otherwise, the jury would have no means of knowing the facts upon which his opinion was based. It is, of course, necessary that the record show the facts upon which the opinion is based *587 because, unless the jury finds that all those facts are established by the preponderance of the evidence, it should give no weight to the expert’s opinion as stated in his answer to the hypothetical question. See Haas v. Kundtz, 94 Ohio St., 238 , 113 N. E., 826 ; Williams v. Brown, Exr., 28 Ohio St., 547 .
Also, there may be instances where, after questions have been asked on cross-examination as to reasons for the opinion of the expert witness given in his answer to the hypothetical question, it may clearly be proper on redirect examination to inquire further with respect to such reasons.
It may be that the reasons for my doubt as to the soundness of the broad statement in paragraph three of the syllabus are not sound or would be outweighted by other reasons which are not now apparent to me, but I believe that this court should refrain from again making such a statement in a syllabus before the legal questions involved therein are presented to us for consideration.