CHAMBERS, Circuit Judge.

On February 5, 1955, Titus, a longshoreman, was injured on the deck of the S. S. Santorini as it lay at the dock in Coos Bay, Oregon, taking on a load of lumber. At the moment he was acting as hatch tender. As the load came, up in the slings attached to the hook, both a preventer wire and a rope guy broke. Each was fastened to the upper end of the starboard boom and to the vessel’s rail to give stability to the boom. As a result the winch tender naturally lost most of his control over the boom and the load in the slings. Titus got away from the load, but in doing so he slipped and suffered a bad break at his right ankle. He was an employee of the Independent Stevedoring Company which was connected with the ship and its owners only by its contract of employment. Passing over his compensation rights existing by virtue of his employment, he brought this suit in admiralty alleging negligence and unseaworthiness.

At the conclusion of the trial, the trial judge indicated his intention to rule in favor of Titus as libellant. After the case was briefed by counsel the court, reversed itself and entered findings of fact for the defendant. 1

*353 It is now the law that a longshoreman does not have to be at sea to take advantage of unseaworthiness (if that caused his injury) and he doesn’t have to be a seaman to obtain the benefits of the rule. See Seas Shipping Co. v. Sieracki, 328 U.S. 85, 66 S.Ct. 872, 90 L.Ed. 1099.

McAllister v. United States, 348 U.S. 19, 75 S.Ct. 6, 99 L.Ed. 20, laid to rest the right of appellate courts to try admiralty cases de novo, a prerogative which had been little used anyway in recent years.

But still this court must examine the facts to determine if the ruling was clearly erroneous. There is no suggestion in the proof that the flight of Titus when the wire and rope broke was ill-advised or that he acted unreasonably in the emergency. Possibly there was evidence upon which the trial court could have found unseaworthiness and therefore for Titus. But the trial court did not do so. Perhaps it might have done so if it had disbelieved some of the defense witnesses.

This court sitting on appeal a long distance from Coos Bay and without witnesses cannot announce what the cause of the breaking was. But from the evidence it believes it can put the possible causes in categories.

First, the wire and rope could have broken because they were inherently too small to do the job. They could have broken because they previously had been damaged or w-ere worn out. They could have broken because of some latent defect in the manufacture. And they could have broken because of some negligence in the method of their installation. But there was evidence which the court accepted (and made findings thereof) that the wire and rope were newly installed the day before, were in good condition and had great margins of safety in their tensile strength; that there was no latent or patent defect upon subsequent examination of the wire at its point of breaking; and that the method of installation, if not the best approved method, did not contribute to the breaking.

Second, there was the possibility that the rope and wire broke because of some defect in the winch equipment. Appellant’s witnesses were positive there was nothing wrong in the design or operating condition of the winches.

Third, there was the possibility that the winch was improperly operated or the loading was improperly done by a fellow longshoreman at the instant when the breaking occurred. In this field, the outlawed practice of winch operators “tightlining” their cables and producing an unusual strain on the wires and lines is the most typical. But there are others. There was no direct evidence that the loading was conducted in any way but in a skillful manner. But if the causation of the break was in this third category, negligence at the moment of a fellow longshoreman, the ship would not be liable. 2

Fourth, only to complete the group of possible causes, it may be mentioned that some mysterious outside force may have clipped wire and rope. All would agree this did not happen.

In this case appellant, as did appellee, devoted much attention to the wire, but appellant devoted little attention to the rope. 3 The rope, said the experts, was more than adequate for the normal stress under normal operation with the load here involved. 4 The guy rope and the preventer wire apparently supplemented each other and did not perform independent functions. Whether the boom and load would have gone out of control if only the wire or the rope had broken is not clear. But this simultaneous breaking was an arresting circumstance, one that could not help but cause a trial judge to wonder. Certainly this coinci *354 dence would cause a court to question whether there were latent defects in the manufacture of both the rope and the wire.

There is one possibility overlapping perhaps between our second and third groups of causes which the parties do not sharply point up. There was testimony (and possibly the knowledge thereof is in the general domain not requiring experts) that excessive stress may be applied to wire (here stretching) which is released before the wire breaks, yet leaves the wire so weakened that a later less powerful pull when applied will cause the wire to break at a point far below the wire’s rated tensile strength. One may assume the same rule of physics is applicable to rope.

Suppose the wire and rope were weakened by negligent or willful misconduct of the longshoremen’s crew, either yesterday or an hour ago. Wouldn’t that defect have to be charged to the ship as unseaworthiness ? 5 The court had .ultimately satisfied itself that appellant’s theories of causation were untenable. Further, it came up with an abiding doubt as to what caused breaking. So the question just posed does not arise and will have to be answered in this circuit in another case.

This court is not convinced that the instantaneous acts of a fellow longshoreman rendering the equipment unsea-worthy and injury to the longshoremen are chargeable to ship as unseaworthiness. 6 If this be characterized as slicing the law too fine, one may counter with an illustration. Suppose a longshoreman throws a match in a bucket of gasoline on the deck. The result is an explosion which simultaneously and immediately renders the ship unseaworthy and injures the second longshoreman. Would the ship be held on its implied in law warranty of seaworthiness for the injury to the second longshoreman?

Practically, the trial court having eliminated other causes, 7 it was left with one likely cause (uncertain in time of origin) of the breaking of the preventer wire and the rope guy: overstressing in the loading operation at Coos Bay either at the time the rope broke, or a pre-stressed, preweakening earlier in the day which later caused the breaking under pressure which was lower than would have been required to accomplish the breaks if the rated strengths of the wire and rope had been maintained.

Appellant has a strong argument when he points to the sentence in the court’s opinion 8 as follows: “There was some suggestion that the break could have been caused by tight lining- — one winch pulling against the other until the force exerted was greater than the strength of the wire causing it to give way. It should be noted that this was only a suggestion; there is no evidence in the record that such was the case.” One should note that this comment does not eliminate other mishandling of the equipment. Further, the opinion as a whole shows that the court (devoting almost its sole attention to the wire to the exclusion of the rope, as did counsel) was expressing an abiding belief that appellant had not proved what caused the breaking other than that it had been shown that “the force exerted on it was greater than its breaking strength.” The court was impressed with the ship’s evidence which negatived most of the possible causes of breakage.

But the opinion cannot help the libel-lant because when rather full findings were made seven weeks later their function was to supersede the opinion, thence *355 forward the opinion having no more standing than random observations made by a trial judge in the course of a trial. It may be that the Supreme Court decisions will reach the point where the shipowner is liable for unseaworthiness based upon the very act of a fellow longshoreman which causes the injury, but it does not appear that the Supreme Court has gone that far. It would seem that unseaworthiness must be a preexisting condition. 9 Perhaps, the preexistence will be held to be ever so short. One cannot say. But this question must await a different case. The court did not find facts requiring the problem to be answered.

Permeating the trial court’s decision and findings is the final conviction that on the morning of February 5 the ship supplied the stevedoring company with good and adequate, as of then, equipment. But it was unable to find on what day or what time of day the overstraining occurred which eventually produced the break. This court cannot solve the mystery, but certainly the simultaneous breaking of the wire and the rope, each rated strong enough to do the whole job, at the time of the injury, does give cause for pause, does make one doubt that there was a prestressing of both wire and rope, almost but not quite to the breaking point of each, at some prior time. 10

As above indicated, it is possible that if the court had decided from all the

circumstances here that the rope and wire must have been preweakened, its decision would have had a rational basis. But in this court’s view it had also the choice of concluding that the handling of the last load caused the break or the choice of being honestly unconvinced as to when the rope and wire were first weakened — thus a failure of proof.

This court is fully mindful of its decision in Petterson v. Alaska S. S. Co., 205 F.2d 478, 479. There a snatch block was being used in the gear lifting the cargo. The block broke with resulting injury to Petterson. The trial court found that the block was used in a “customary and usual manner” and that it “was of a type ordinarily and customarily used and proper for the use to which it was being put upon the occasion in question.” The whole opinion proceeds on the basis of the initial statement, “while being put to a proper use, * * * the block broke.” And, it is significant that the case says, “There was no proof as to the condition of the block prior to its use other than what may be implied from the accident.” 11 Thus, it was permissible to infer that the block was defective when the day started. Here the fact situation was different and, in the end, Titus came out without the helpful fact findings which Petterson had.

But in the light of McAllister v. United States, supra, Petterson must be hereafter approached with a caveat. The *356 author of Petterson at page 479 says: “But an admiralty appeal is a trial de novo, * * *, and this court may make its own inferences from the facts as found where it does not upset the findings based upon the credibility of witnesses.” So, in Petterson starting with some few findings favorable to appellant, it may have been adding its own expertise, which by McAllister is no longer permitted.

The trial court on the facts here had the privilege of being honestly puzzled as to causation. Without a conviction that the accident happened because of a definite condition for which the appellee was to blame, it had no duty to find for the injured stevedore.

The decree is affirmed.

Notes

. The findings of fact, conclusions of law, and decree were entered March 1, 1957.

. Freitas v. Pacific Atlantic Steamship Company, 9 Cir., 218 F.2d 562.

. The rope guy is nowhere mentioned in the pretrial order.

. This was contradicted by a fellow stevedore.

. See Grillea v. United States, 2 Cir., 232 F.2d 919.

. This concept is very clearly expressed by Circuit Judge Learned Hand at page 922 of the Grillea decision.

. One should remember that circumstances ■which one feels one must accept may upset in the end direct testimony which, at the time of its reception, appeared to be wholly reliable.

. The opinion was filed January 8, 1957, D.C., 152 F.Supp. 63-65.

. The concept that the negligence producing unseawortliiness must precede in point of time (at least a second arc of the apparatus) to establish a pre-existing condition seems to be the underlying basis for a rationalization of Manhat v. United States, 2 Cir., 220 F.2d 143, although neither the majority or minority opinions address themselves sharply to the point of whether enough time had elapsed to result in unseaworthiness as was done by Judge Hand in Grillea.

. One point on which the evidence is bare is this: As to the wire and ropes there was testimony as to the tensile strengths of each. These ratings wore based on a direct pull. It would seem that in the lifting of the ton or ton and a half load of lumber there would have been some factor of leverage increasing tlio pull beyond that of a vertical direct pull. How much such leverage might increase the strain, no one has told us. The responsibility for that point was on him who had the burden of proof.

. In Potterson’s case the ownership of the block was never solved, only its use. When it broke, it fell into the water and was not recovered. Also, in Petterson, from libellant's standpoint, the amount of stress on the gear in the normal operation of lifting the load of lumber up over the deck load before lowering through the hatch to the hold was better developed. There is naturally more stress on the gear when the load has to be lifted up and over a deck load.