This entirely factual appeal is by a shipowner from a judgment denying a claim for indemnity against a ship repair yard, following a $325,000 settlement paid by the shipowner to a seaman who had sued under the Jones Act, 46 U.S.C. § 688. The seaman, Leonard Imanuel, was injured when cables on an "engineer's platform hoist" broke. The "hoist" was in lay terms a small freight elevator or lift running from the main deck to the engine room with stops at the tween and lower decks. The accident occurred on December 10, 1971, when the elevator platform fell about 60 feet from the top to the bottom of the shaft. The elevator was on the S.S. Nancy Lykes, owned by Lykes Bros. Steamship Co., Inc. (Lykes), the appellant. The repair yard was operated by Todd Shipyards Corporation, impleaded below as a third party defendant and appellee here. The ground for liability asserted was that the cables had been damaged by Todd in the course of its performing ship repairs so as to breach its warranty of workmanlike service. The United States District Court for the Southern District of New York, Edmund L. Palmieri, Judge, sitting without a jury, held that Lykes had failed to sustain its burden of proof that Todd had caused damage to the cables of the elevator, and that Todd was not negligent and did not breach its warranty of workmanlike service. We affirm the judgment.

The Nancy Lykes went into drydock in June of 1971 and was extensively altered and repaired by Todd over a period of months ending in December of that year. In the course of making those repairs Todd, using an overhead crane, moved certain machinery through the engine room elevator shaft to the engine room. The heaviest object lowered through the shaft by Todd was a manifold weighing between 300 and 380 pounds. The district court found that there was no evidence from which it could be inferred that the manifold or any other objects moved through the shaft by Todd's crane dented the wall of the shaft, touched the cables, or interfered in any way with the platform lift equipment. Lykes' claim, that the only way the accident could have happened from the standpoint of the physical facts was by virtue of damage to the cables from being struck by an external object, will be examined in more depth subsequently. It is not denied in any event that use of the elevator shaft to do certain work in the engine room was necessary, since the lift platform itself had been deactivated. It remained at the bottom of the elevator shaft throughout the repair work and was not itself repaired or maintained by Todd.

Leonard Imanuel, the plaintiff seaman below, rejoined the Nancy Lykes with a second electrician rating in early December, 1971, to help prepare it for sea. It was his job, working with his immediate supervisor, Electrician Herman Haag, to put the engineer's platform hoist back into service. On the day of the accident the two men were, according to Haag, checking the automatic "limit switches" located in the shaft just below the several deck levels; these switches operated to cause the lift to stop at the proper level. Imanuel's initial memory was that he was on the lift platform to check interlock switches that required the platform to be at or near the top of the shaft, but he also testified that he stopped the platform about two feet below the main deck level to enable Haag to check the topmost limit switches. In any case, to make checks of the limit switches, the lift, absent use of a boatswain's chair, had to be stopped between levels, and Imanuel and Haag accomplished this by jamming the relay in the control room closed with a screwdriver, thereby bypassing the limit switches. Imanuel recalled that he was riding the lift and Haag jamming the relays in the control room just below the top deck, while Haag recollected the opposite. In either case the man in the control room left it to join his partner on the lift platform, incidentally activating the electrical brake on the cable drum so as to render the drum (and hence the cables and the lift) essentially immobile. When the platform itself was near the uppermost limits of the shaft at the main deck, Haag stepped off the lift platform and onto the main deck to get a meter. Imanuel, on the platform, heard a loud grinding noise and the lift began what appeared to be a free fall down the shaft with Imanuel aboard. He was seriously injured.

The two cables were found severed "as though with a knife" at about the same distance from the anchored ends, though the expert surveyors differed as to the extent of that distance (11 to 16 feet). The fall of the platform had not been prevented by the safety device under the platform, which operated mechanically to prevent just such an event the so-called "rope safety" device because Lykes had not maintained it properly; indeed, neither Haag nor Imanuel knew that there was any such safety device. Since the cables had parted, the brake on the drum was also ineffective to prevent the accident.

Haag's and Imanuel's testimony that the lift platform was about two feet below the top of the shaft to check the upper limit switch just before the fall was rejected by the district court quite properly: because the limit switches were two feet or so below the top of the shaft, and the platform would have had to have been about two feet below that to check them, Haag would have had to have stepped up four feet or so in order to get off the platform as he did, a rather remarkable acrobatic feat for a non-bionic 51-year-old man. The court found instead, and we do not think the finding can be faulted, that the platform was at or near its uppermost limits at the main or weather deck, as Imanuel himself originally testified.

In any event, Imanuel heard a loud grinding noise just before the platform dropped. Lykes points out that the cable parted "within a minute or two" after Imanuel and Haag had been on the platform together for the first time; the court below found, however, and we cannot say the finding was clearly erroneous, that a minimum breaking strength of 5,200 pounds per cable and a minimum lift strength of 20,800 pounds meant that the lift even with two crewmen on it did not approach a deadweight sufficient to part the cable even if 25% of the cable were damaged, as Lykes' experts appeared to contend.

To understand appellant's contentions a more precise description of the elevator itself is required. The shaft was four feet six inches by three feet six inches, the platform four feet by two feet nine inches. The control room, which contained the motor, electrical brake and other electrical equipment in a unit with the drum for the cable, was on the level just below the weather deck. The two cables, 5/16 inch in diameter, were attached to the drum and wound in grooves around it. These cables then extended upward from the lower side of the drum to a stationary wall sheave (or pulley), also double grooved, located just below the top of the shaft. The cables then went back down under the lift platform, passing under double grooved sheaves underneath and at both sides of the narrow part of the platform to the other side of the shaft. They then ran back up the opposite side of the shaft and were anchored near its top or "bitter end." The platform rode on guide rails located on the same opposite sides of the narrow part of the shaft as the wall sheave and cable anchors above referred to. At the top of each of the guide rails there was a hole for the insertion of a bolt, the insertion of which formed a "T" and stopped the platform in the event that the limit switches were overrun at the uppermost point of movement. Neither of the bolts, the trial court found, was in place at the time of the accident, and Lykes' post-trial suggestion that they were removed only after the accident in order to permit removal of the platform for purposes of photography is not sustained by the record.

Lykes argues that the facts compel a finding of Todd's liability. These facts, it claims, are that the cables received a horizontal blow when they were in a vertical position, which caused, according to Lykes' experts, "nicking and flattening" of the cables. The blows are said to have occurred when the 48 by 24 by 16 inch, 300- to 380-pound manifolds were moved up and down the shaft by Todd, a possibility that might strike the casual reader as more likely than it did Judge Palmieri. Lykes points out that the manifolds were supported by more than 100 feet of cable from the end of the boom of a gantry crane and could easily have swung and caused the damage, especially as no tag lines were attached to the manifolds to steady them as allegedly required by safety regulations. Lykes' photographs and evidence indicate a couple of rather fresh horizontal three-inch dents in the sides of the shaft near the top which "could have been caused by a swinging manifold." Lykes claims that "there was no factual evidence submitted of any other possible cause for the damage which resulted in the parting of the cables,"1 so that its theory is the only tenable one, however circumstantial its factual premises.

Lykes' surveyor, Brierly, who examined the shaft and cables the day after the accident, testified that the breaks were straight across as though the cable had been cut and that on external visual examination the cable otherwise appeared in satisfactory condition. He said the lays of the cable on the drum were all even and had not jumped or crisscrossed, that the cables were well greased and not rusty, and that the cut, 11 feet 8 inches from the "bitter end," was clean except for one 12- to 15-inch strand. The London Salvage surveyor, Thompson, stated in his report that he could find no mechanical fault that could have resulted in failure of the cable; it was he who suggested the possibility that in the course of the repairs materials had been removed through the elevator shaft causing the damage. He too said that the cables were lying within the grooves of the drum and, except for the breaks, were in good condition without readily visible broken wires or frays. He testified that both cables were broken at a point 16 feet 7 inches from the bitter end, and that the breaks were not caused by metal fatigue. After the cables had been removed on December 15, 1971, cut up and delivered to a testing laboratory, a metallurgical expert, Kiste, examined them. He testified as to the nicking and flattening of the rather slender wire ropes at the fracture site of the outer layer wires; he also observed abrasion and flattened areas. He testified that the angle of impact was approximately horizontal to the vertical axis of the wire ropes, and that the damage that he found could have been caused by impact from a metal-like object such as the manifold.

Appellant argues that there is no evidence of any damage to the cables at the time of the accident. "There is nothing that could have delivered to the cables a horizontal blow." Therefore, appellant reasons, the blow must "have taken place" as Todd was moving objects, particularly the manifold, through the elevator shaft.

Judge Palmieri in his findings advanced three possible theories to explain how the accident could have occurred. One, suggested by Todd, was that by bypassing the safety features, including door interlocks, limit switches and a "slack cable" switch, Imanuel and Haag allowed the lift to function without control and thereby created a risk of reverse spooling on the drum and of contact between cable and cutting edges such as the sheaves, the guard over the wall sheave, or the machined opening in the bulkhead wall where the drum was located. But appellant effectively rebuts this by pointing out that there was no evidence of any reverse winding of the cables on the drum. Quite to the contrary, when surveyors Brierly and Thompson examined the cables they were on the drum in the grooves, which would not have been true had there been a reverse winding. Indeed, photographs in evidence clearly show that several turns of the cables on the drum had never been off it, and in order to begin a reverse winding all the cable would have had to come off.

The second theory, advanced by the court in its Finding No. 34, is that without the restraining bolts at the top of the rails and with the platform operated at its uppermost limits, the cables under the platform could have been disengaged from their sheave grooves, thereby coming into contact with the machined edge of the under-platform sheave, a danger which would be increased if that sheave were restricted in free movement or indeed seized. The court's suggestion is that because there was no lubrication or maintenance of the entire undercarriage over a long period of time, as evidenced by the lack of performance of the mechanical brake safety device, the cable broke at the time of the accident because of such a seizure. Appellant we think adequately refutes this by pointing out that the cable broke at a point at least 11 feet and perhaps 16 feet 7 inches from the anchored end. According to the dimensions of the platform, giving its position the one described in Finding No. 34 as "uppermost limits," a break under the circumstances suggested by the court would have been at a point seven or eight feet from the anchored end. Moreover, in all probability the cutting in such an event would not have been horizontal as the expert testimony indicated it was, but oblique. Certainly there was no evidence of any restriction in the free movement of the platform sheaves prior to the platform's fall to the bottom of the shaft.

In Finding No. 38, however, the court developed a third theory for the accident. The drum was located approximately ten feet below the top of the shaft and the cables played out from its lower side, as we have said. As the platform was raised near the top of the shaft, the platform sheaves would eventually approach the level of the wall sheave and the cable running from the drum to the wall sheave would then begin to assume a horizontal path on leaving the wall sheave rather than going straight back down as would occur when the platform was lower in the shaft. If the platform were raised to its uppermost limits, the cables might well work their way off the wall sheave at which point they would no longer support the platform. They would be held up on one side of the shaft by the anchor and on the other only by the fact that they were wrapped around the drum below the wall sheave. Consequently, Judge Palmieri reasoned, the cables and platform would fall and as the platform fell with the slack cables beneath it the cables would eventually become taut somewhere just below the drum, which was electrically braked and set in a fast position. In that position the cables would run from beneath the drum out and down over the machined edge of the bulkhead beneath the platform and back up to the anchor so that, the judge suggests, the falling weight of the platform with Imanuel on it could have snapped the cables against the bottom machined edge of the bulkhead with enough force to sever them cleanly and evenly.

This theory has additional support in the evidence in that the guard over the wall sheave appears bent at both ends in the photographs, indicating a throwing of the cables. Moreover, the break at a point 11-16 feet from the anchor would fit this theory rather closely, allowing a 10-foot length for the distance from the top of the shaft to the drum, 4 feet for the width of the platform and 21/2 feet additional as the point of drop at which time the cables snapped. The existence of numerous coils around the drum we count in the photograph (Lykes Exhibit 3) approximately 40 would tend to create a considerable weight that would act, along with the electrical drum brake, as a drag preventing the drum from unrolling while the platform was falling. The rather apparent sharply machined edge of the drum box on its lower side would also be adequate cause for the kind of clean break which the expert witnesses described.

The only answer to this theory made by Lykes is that if this had happened the cables would have been off the sheave before Imanuel left the control room. They point out that his testimony was that he had turned off the power in the control room, set the brake on the drum so that it could not move, then gone up to the weather deck, joined Haag on the platform for the first time and examined the limit switch for a bit of time before the fall occurred. Thus, Lykes reasons, if the cables had come off the sheave it would have occurred at the time the platform stopped and this could not have happened or Haag would have fallen with the platform also. But this suggestion misses the point. Assuming that when the lift was first stopped the cable was drawn as tightly as it could be, at the precise mechanical position immediately prior to throwing the sheave, and that then Imanuel stepped on the platform, his additional weight would tend to stretch the cable or at least keep it taut at that point. But then when Haag stepped off the platform, it might well have moved slightly upward, as one sees old-fashioned small elevators move when a passenger steps on or off, just enough to permit the platform to rise and throw the cable off the sheave. The grinding noise heard by Imanuel would have been the cable throwing the sheave and contacting the sheave guard. The fall would then have occurred as suggested by the judge.

At the very least, it cannot be said that there was insufficient evidence to sustain this theory of how the accident happened; indeed, to us it seems altogether the most probable way. In any event, it negates the argument that the accident had to happen as Lykes suggested, by virtue of impact of the manifold or other machinery being lifted into the elevator shaft by Todd, and make it impossible for us to say that Judge Palmieri's conclusion that there was insufficient evidence to support a finding of negligence or breach of warranty on Todd's part was "clearly erroneous," as we must in order to reverse. Fed.R.Civ.P. 52(a); Tupman Thurlow Co. v. S.S. Cap Castillo, 490 F.2d 302, 304 (2d Cir. 1974).

Judgment affirmed.

TIMBERS, Circuit Judge, concurring in the judgment and dissenting from the opinion:

I concur in the judgment of affirmance to be entered by the Clerk. Such affirmance is required on the ground that Lykes failed to sustain its burden of proving by a preponderance of the evidence its indemnity claim against Todd.

I regret that I am not able to concur in Judge Oakes' opinion. Aside from the speculation with respect to the occurrence of the accident, in the end the only accurate statement that can be made as in the cases of so many marine accidents is that no one knows how this one occurred. In short, Lykes failed to prove its case. I would say so and stop.