(dissenting).

The reason I dissent is that I feel I must accept what the trial judge actually said about res ipsa loquitur, and must assume that he meant what he said. That is in Conclusion IV as follows: “That the doctrine of res ipsa loquitur is not applicable under th'e evidence presented in this case.” I perceive no difference between myself and the majority in the view that such statement is incorrect; it is not the law under the facts of this case. The consequence here should be the same as that applied by this court in Wilson v. United States, 250 F.2d 312 , where we held that where the trier of facts has applied improper standards of law to the facts of the case, there must be reversal. As there indicated this court had previously said the same thing in Mar Gong v. Brownell, 9 Cir., 209 F.2d 448 , and in Takehara v. Dulles, 9 Cir., 205 F.2d 560 .

No one supposes that the trial judge here was obliged to find for the plaintiff or to draw any inference of negligence because of the rule of res ipsa loquitur. That is not the point. In saying “the doctrine or res ipsa loquitur is not applicable” the court has said, as plainly as language can say it, that upon this record it was obliged to find for the defendant. What the trial judge has said is: “I cannot infer negligence here.” Plainly enough if this is a proper case for application of the doctrine of res ipsa loquitur, then the trial court was proceeding upon a misconception of what it could decide; and for the reasons stated in Wilson v. United States, supra, such an application of improper standards of law to the facts by the trier of facts, compels reversal.

What exposes a reversible error here is the manner in which the findings are framed. The majority opinion succinctly describes these findings in saying: “ * * * the findings do say there was no evidence of ‘this and that’.” In essence the findings say: No evidence was presented as to where the coil of wire had come from or that any such wire had been placed or seen in the yards, or that there was a failure to inspect, or to keep the area clear of wire, or that defendant negligently left it there, etc. In other words, the findings are limited to a statement that there was no evidence of any of these specific acts of negligence. But as stated in Sweeney v. Erving, 228 U.S. 233, 240 , 33 S.Ct. 416, 418 , 57 L.Ed. 815 , the meaning of res ipsa loquitur, where it applies, in respect to proof of negligence, is that though “direct evidence of it may be lacking,” yet the facts of the occurrence “furnish circumstantial evidence of negligence.” If the court was wrong in *803 its statement that the doctrine of res ipsa loquitur is not applicable here, then the court was erroneously saying: “there is no circumstantial evidence of negligence upon which a finding for plaintiff could be based.”

Suppose an eye-witness should testify that he observed a specific act of negligence but the court should erroneously hold that the witness was disqualified from testifying and strike his testimony from the record and then found for defendant for want of evidence. In my view there is no difference in the two cases. In the case of the eye-witness the judge would not be obliged to believe the testimony and in the present case the judge was not obliged to draw the inference of negligence from the circumstances of the case. But there is just as much reason for reversal in this case as in the supposititious one, for the trier of facts has proceeded under a misapprehension as to what decision is open to him.

Now, if we were to come to the question which I think we cannot properly avoid, namely, whether this was a proper case for the application of the doctrine of res ipsa loquitur, we must note that that presents a federal question. Jesionowski v. Boston & M. R. Co., 329 U.S. 452, 457 , 67 S.Ct. 401 , 91 L.Ed. 416 . As indicated in that case and in Johnson v. United States, 333 U.S. 46, 49 , 68 S.Ct. 391 , 92 L.Ed. 468 , the Supreme Court has rejected interpretations of the doctrine which unduly narrow it. In Bedal v. Hallack & Howard Lumber Co., 9 Cir., 226 F.2d 526, 538 (footnote 18), we took occasion to note that the Supreme Court “has given the rule what may be an exceptionally broad and liberal application.” The same conclusion is drawn by Prof. McCormick, McCormick on Evidence, 1954, p. 644, (footnote reference to these two cases).

It seems to me that if we were to apply the rules of Jesionowski and Johnson to this case we would say that in the ordinary course of things such a wire as caused this accident to Tittman would not have been where it was if the defendant used proper care to inspect the tracks and yards as a part of its duty to maintain them. That it had such duty, and that it had possession of the yards as a part of its line and system the court has found in Finding IV quoted in the opinion. The employer defendant owed the duty to use care to provide plaintiff a reasonably safe place to work. That involved the duty of inspecting its premises. It is true that witnesses for the defendant testified that there was inspection of the premises and that nothing amiss was discovered, but the court was not obliged to accept that testimony as conclusive. 1 As stated by Judge McAl-lister in his dissenting opinion in Baltimore & O. R. Co. v. O’Neill, 6 Cir., 211 F.2d 190 , 197: “It is not enough that the evidence of the defendant would, if true, be sufficient to rebut the presumption, because it is for the jury to pass upon the credibility of the witnesses and the truth of the testimony.” That dissenting opinion is peculiarly significant since Judge McAllister there expressed the view that where a new bolt, never previously used, which the employee took from the railroad’s tool room, broke and caused injuries to the employee, the rule of res ipsa loquitur warranted a finding of negligence on the part of the employer railroad. The contrary view of the majority of the court was summarily reversed in O’Neill v. Baltimore & Ohio Railroad Co., 348 U.S. 956 , 75 S.Ct. 447 , 99 L.Ed. 747 , which suggests that the Court approved Judge McAllister’s views.

In Webb v. Illinois Central Railroad Co., 7 Cir., 228 F.2d 257 , the plaintiff, a *804 brakeman, working along a switch yard between the tracks fell when he tripped on a clinker about the size of his fist. The court held that there was no sub-missible case made and that the fault or negligence could not be inferred from the mere existence of the clinker and the happening of the accident. Certiorari was granted and the decision of the court of appeals reversed in Webb v. Illinois Central R. Co., 352 U.S. 512 , 77 S.Ct. 451 , 1 L.Ed.2d 503 . As the decision there shows, there was a substantially greater amount of evidence as to the origin of the clinker than there is as to the source of the wire in the case now before us; but what seems to me to be peculiarly significant is the statement made by the court, 352 U.S. on page 513, 77 S.Ct. on page 453 of the opinion as follows: “The Court of Appeals viewed the evidence as insufficient to raise a jury question because the petitioner did not adduce proofs showing what standard procedures were followed to prevent large clinkers from being used in road ballast and in inspecting roadbeds for hazards to firm footing. We do not think that the petitioner’s evidence was lacking in such proofs even if we assume, and we question, that he had that burden.” (Emphasis mine.) It seems to me important that the Supreme Court stated that it questioned that Webb had any burden to adduce proofs showing what standard procedures were followed in inspecting the roadbeds for hazards to firm footing.

To my mind this sounds very much like a statement of the Court that the proof of the presence of the clinker and of the plaintiff’s fall thereon, satisfied plaintiff’s burden of going forward with the evidence. Perhaps it is dictum, but it comes from a high source.

What gives me the greatest assurance that I am right is the extraordinary length to which the majority have gone in correcting this defective conclusion. They amend the statement, which they do not purport to defend as written, by saying of the trial judge “he meant to say”, and then putting the matter as they would amend the conclusion. As I understand the opinion, it suggests that the judge could not possibly have meant what he said about the doctrine of res ipsa loquitur, so they add embellishments such as “It doesn’t help me”, “I find ‘res ipsa loquitur’ of no use here”, and so forth. I would not doubt at all that the findings were drawn by prevailing counsel and then presented to and signed by the judge. But I do not know any such thing and I must take it that the language used in the findings was that of the judge. Similarly it is my view, in contrast with that of my associates, that I must assume that the judge meant what he said, and that if the findings are defective, the case must be returned for him to correct them. Correcting them is not our function.

I think the case should be remanded with instructions to make further findings. I realize of course that on such a remand the court could properly find against the plaintiff. In my view that is beside the point, for the teaching of Wilson v. United States, supra, is that a litigant is entitled to have the facts found by a trier of the facts who is not applying improper standards of law.

. The court noted that this testimony was “undisputed,” By this it must have meant that there was no direct evidence to the contrary. Once it appears that the rule of res ipsa loquitur may be applied here, and that the facts therefore permit an inference of negligence (including a failure to inspect), the trial court is confronted with the question whether the inspector really inspected as he says he did, or whether he just said so.