On the morning of June 16, 1958, the decedent, Rawson, and one Ross, fellow longshoreman, were on the weather deck of the S. S. Kenmar, the Calmar Company ship which was berthed at a dock in Seattle with cargo to be there unloaded. The night before, the crew of the vessel had uncradled the ship’s booms from their fixed positions in which they rode at sea. The port boom for the No. 3 hatch had been upright against the mast. The crew had lowered the boom perhaps to an angle of about 65° with the deck, or one-third of the way down.

The purpose of lowering the boom is to get it into a working position for use during the loading or unloading of cargo from the hatch. It is a normal occurrence, even though the crew has set the booms, that longshoremen find some adjustment of position is necessary. This, Rawson and Ross, who was working with him, were doing. Ross appears to have been acting under the direction of the former at the material time.

To adjust the boom it was necessary to use a topping lift winch, part of a ship’s normal equipment adjacent to a hatch. (This particular winch was perhaps rigged so that it could have been used in the unloading of cargo, but in practice it was only used on the Kenmar for raising and lowering or adjusting a boom.)

A drum around which a cable or “wire” was wound or spooled, a clutch, motive power (which was steam), various levers and safety devices were essential parts of the winch apparatus.

One adjustment on the position of the boom had already been made at about 8:00 a. m. by Rawson. But Rawson decided a further shift of the position was advisable. Ross joined him and was at a point near two controls — the steam handle and the handle (in a quadrant) controlling a dog pawl, a common safety device. Ross was several feet away at the clutch lever.

As Ross and Rawson were shifting the various control levers, something slipped. The drum ran out of control, the cable on the drum “free spooled,” the cable came loose from the drum and a loop from the wild cable ensnared Rawson’s head, crushed it. Death was apparently instantaneous.

The widow of Rawson (who did as a longshoreman have workmen’s compensation) close to sue the steamship company. She asserted both unseaworthiness and negligence of the ship. Either established as a cause would permit recovery. But in this judge-tried case, it was found that there was no negligence or unseaworthiness on the part of the ship. Further, it was found that the death was caused solely by the negligence of Rawson and Ross working under the former’s direction. The widow appeals in behalf of herself and her children.

Thus, we must consider the cause and ascertain if, in our opinion, the decree below is or is not clearly erroneous. For such consideration, further facts must be related.

Clutches on winches with pins to lock the clutch out of engagement or in engagement, it may be assumed, are old devices. Likewise, it is indicated that a safety dog pawl with pins on its control handle to lock it in engaged or disengaged position is an old device. Apparently the pins on both have always been inserted manually.

All agree that to prevent free spooling when the boom is out of the cradle, the clutch and the safety dog pawl, both subject to hand control, should not be disengaged at the same time. One should be disengaged when the other is engaged and vice versa. The operation of shifting the levers (and normally locking in position after the shift) can be done by one man in successive steps, or by two men in quick succession.

After the accident, pins were found out of the quadrant holding the locking dog lever and out of the locking device on the clutch handle with which the Ken-mar’s winch was equipped. The ship says that fact indicates recklessness on the part of Ross and Rawson — negligence. There was some testimony indicating that it was necessary to have both pins out at the same time, but it would appear that such evidence was pretty well discredited, even by Mrs. Rawson’s own experts. Had the alternate put and take of the pins been used, it would seem fairly certain there would have been no accident.

But the case really only begins with consideration of the clutch locking and the safety dog locking. Here Calmar had provided another safety device — a clutch guard or clutch pawl. We believe it fair to describe it cryptically as automatic or semiautomatic. Sometimes in the testimony it is called the “interlocking safety bar apparatus.” (At least one witness thought this description inappropriate.) At any rate, at the vital point of contact there was a finger-like piece of metal which rode against the clutch collar in sort of a ratchet arrangement. The clutch guard apparatus was intended to prevent or assist in preventing double disengagements of clutch and. safety dog pawls. Most of the witnesses agreed that Cal-mar's special device was an unusually good one, but one expert questioned its basic design.

Naturally, after the accident there was a detailed investigation of the topping lift winch and all of its component parts. A th inch clearance was found on the finger-like apparatus against the dogs of the collar of the clutch. This was not good. It would seem the clearance should not have been over one-eighth of an inch. How did the %eth inch get there? Cal-mar says it had to get there by misuse of the equipment during the very operation in which it failed. (The trial judge declared he was satisfied at the time of the slippage Rawson and Ross were engaged in a “flying switch,” comparable to the outlawed practice by the same name in railroading.) Calmar suggests that a simultaneous twisting of the shaft on which the metal finger rode could have produced the Tiath inch play. The shaft on testing proved to be .0015 inch off true. Appellant’s experts say that the shaft must have left the foundry that way — that such a variance, the approximate thickness of a piece of thin paper, is normal. Calmar experts said the variance (they would call it a bend) could have been produced by a twisting. Such a twisting, they say, would require the resetting of the mount of the pedestal out of which the locking bar or finger emerged (as was done after the accident) if the old shaft were continued in use. Appellant counters with an expert that says, if the shaft twisted, the crust of rust on the shaft would have flaked, and that there was no evidence of flaking.

Calmar buttresses its technical arguments with the fact that the device had been used once a day, on an average, for many months with no slippage. This proves, it says, that the apparatus was in good adjustment until the very operation that resulted in the injury; that the free spooling was the end result of what Ross and Rawson were doing at the very moment. However, there is one reason that this conclusion might not be true— a reason having nothing to do with torsion. Perhaps, prior to this time no one had ever operated the winch without keeping one or the other, the clutch or the safety dog pawl, hand-pinned properly.

Here there is just no substantial evidence that safety could not have been completely maintained by the proper use of the hand pins provided to insure that the safety dog pawl and the clutch would offset each other in engagement and disengagement. Thus, we have a case as to negligence, the two hand safety locking pins both dangling out on their chains, which ought to be affirmed per curiam because it is obvious that Ross and Rawson had the pins out.

But what of unseaworthiness? Assuming unseaworthiness and causation, the liability is absolute. See Seas Shipping Co. v. Sieracki, 325 U.S. 85 , 66 S.Ct. 872 , 90 L.Ed. 1099 ; Petterson v. Alaska Steamship Co., 9 Cir., 205 F.2d 478 , affirmed 347 U.S. 396 , 74 S.Ct. 601 , 98 L.Ed. 798 ; and Mitchell v. Trawler Racer, 362 U.S. 539 , 80 S.Ct. 926 , 4 L.Ed.2d 941 .

There is no doubt that the trial court was greatly impressed by the fault-free history of perfect performance without incident of the winch for many months previous to the morning of June 16,1958. It is obvious that it thought that the accident was caused by mishandling of equipment by Ross or Rawson at the very instant of the free spooling, which may or may not have resulted in the %«th inch clearance instead of the normal %th inch on the other winch of the Kenmar and other Calmar ships. This normal clearance was established and reestablished on the Kenmar after the accident by resetting the mount of the pedestal in which the shaft and the bar were connected.

We know that contract longshoremen as well as seamen create unseaworthy conditions which very soon thereafter can cause injury for which the ship can be held liable. Mitchell v. Trawler Racer, supra.

But we are not convinced that the cases have reached the position of embracing instantaneous unseaworthiness in a situation created by the injured longshoreman himself; that is to say, we do not believe that the very act of negligence of a longshoreman which results in an injury to him from contemporaneous damage to the gear can be labeled as unseaworthiness. (We are not here concerned with injury to a non-actor.) We would think unseaworthiness implies an antecedent condition, although it is quite apparent that the pre-existence of the condition may be very short. Mitchell v. Trawler, supra. For example, we would not think as one continuous act a longshoreman can negligently apply pressure with the winch to a seaworthy line and reduce it to unseaworthy strength, and as part of the same pull keep on applying the pressure until the line breaks, injuring the same longshoreman, and then validly assert unseaworthiness as the cause of his injury. Surely there must be some interval, two acts and not the same one. Cf. Titus v. The Santorini, 9 Cir., 258 F.2d 352 . The Kenmar’s position would appear to be far stronger than the Santorini’s.

We are fully cognizant of the testimony of appellant’s engineer Blumberg on a reopening of the case. He was an excellent witness. He was convinced the clearance of %eth inch had existed since the original assembling of the winch months before. It was his belief that even though the pins in and pins out procedure was not observed, the safety clutch guard with the %sth inch clearance would normally have held (and would always have held with y8th inch clearance). Said he, even with %6th, it would be rare that the safety guard didn’t hold on the clutch collar, but the “excessive clearance” created the possibility — which happened. And he assured that physical conditions indicated there was no other factor. We think Blumberg’s testimony could have per mitted the trial court to find for Rawson, but we do not think the trial court was required to accept his theory. Thus, we think the trial court’s conclusion on unseaworthiness was not clearly erroneous.

We think we should note herein that counsel fbr appellant relies too heavily on appeal on the oral comments of the trial court in announcing its decision. Such comment is superseded by the findings of fact. The trial judge is not to be lashed to the mast on his off-hand remarks in announcing decision prior to the presumably more carefully considered deliberate findings of fact. Davis v. Rhay, W.D.Wash., 156 F.Supp. 114 , affirmed 9 Cir., 256 F.2d 617 ; Plastino v. Mills, 9 Cir., 236 F.2d 32 ; Brooks Bros. v. Brooks Clothing of California, D.C.S.D.Cal., 5 F.R.D. 14 .

Further, we would comment on the tedious problem of reviewing the record in this case. Much of the testimony is the identification of pieces or parts of the apparatus as found in blueprints, photographs and a very excellent model prepared by appellant. But dozens and dozens of times the witness points out something with his finger or a pointer. Only on four or five occasions is the place marked on the object where the witness pointed. Over and over the cold record says, “indicating,” or “here,” but we are left to guess where “here” was and what was indicated under “indicating.” The situation is not hopeless. After one goes over and over the record with pictures and blueprints at hand, one can finally come up with a firm conviction as to what “here” and “indicating” meant. But this indifference to the appellate court during the trial of the case has required repeated review to grasp the facts. This would not have been required if the testimony had been “tied” to the exhibits. Such a condition of the record largely explains what otherwise might be considered an inordinate delay in deciding the case.

The decree is affirmed.