JOHN R. BROWN, Circuit Judge:
Prelude
It was a dark and stormy night. A patchy, low-lying fog covered the murky waters of the river and obscured the banks. Ships, passing in the night, were but phantoms, vague outlines disappearing into the mist. Ships’ whistles, echoing across the dark expanse, seemed like mournful cries from another world. Then suddenly, looming out of the darkness, another ship appeared. The distance was too small; time too short; before anyone could do more than cry out, the unthinkable occurred. The ships collided. The tug, helpless, drifted downriver. Floundering like some giant behemoth wounded in battle, the tanker came to ground and impaled itself on some [1047]*1047voracious underwater obstruction. And still the whistles, echoing, seemed like cries from another world.
This apparent screenplay for a Grade B film, in fact, came to Southwest Pass of the Mississippi River, about 3 miles below Head of Passes, in Louisiana, on February 1,1973. The cast of characters included the SS HESS REFINER (REFINER), a jumboized T — 2 tanker approximately 605 feet in length and 75 feet in beam and drawing, that fateful night, 32 feet, 1 inch forward and 34 feet, 7 inches aft; the SOCRATES, a 125-foot seagoing tug drawing 14 feet, equipped with twin pilot-house-controlled engines boasting 3200 h. p.; and its tow, the barge ALLIED CHEMICAL NO. 44 (AC-44), which measured some 340 feet in length, 68 feet in beam, drawing approximately 17 feet forward and 18 feet aft. REFINER is owned by Hess Tankship Company and was bareboat chartered to Amerada Hess Corporation (collectively Hess). SOCRATES is owned by Allied Towing Corporation (Allied Towing). AC-44 belongs to Allied Chemical Corporation (Allied Chemical), which, despite the apparent filial connection, bears no relation to Allied Towing.
The drama opens early on the morning of January 31, 1973.1 SOCRATES, with AC-44 loaded with liquid fertilizer, departed Allied Chemical’s dock at Geismer, Louisiana. It was made up to the stern of the barge, pushing, the flotilla being 455 feet in length. Bound for the Gulf of Mexico, the tug had a full crew and proper equipment. The clear weather and good visibility promised an auspicious voyage.
At some point between Pilottown and Beacon # 26 (located somewhat more than 3 miles below the Head of Passes on the east, or right ascending bank), Captain Robert Scott of SOCRATES spoke to the pilot of the upbound A VAX. A VAX informed him that there were high winds and 10-12 foot seas on the bar beyond the Pass. Preferring to avoid the rough weather, Captain Scott, at 2045 hours, E.S.T. (1945 C.S.T.),2 rounded up his tow and pushed into the west bank of the Pass.
They selected a position approximately one quarter nautical mile below a dike on the west (left ascending) bank, 3.05 statute miles below Head of Passes (dike 3.05) and abeam of Beacon # 26 on the east (right ascending) bank. The tow, having turned around so that it now headed north, was angled in toward the west (left ascending) bank. The vessel’s heading as shown on the gyro was 355-359°. Captain Scott believed that the forward port corner of AC-44 was aground.3 The tug, with its engines at half ahead with 20-30° left rudder, held the barge in to the bank. The flotilla was lighted in accordance with applicable rules governing a tug and tow underway.4
At this spot, the Mississippi is some 2,000 feet wide. The U.S. Corps of Engineers maintains a dredged channel 40 feet deep about mid-way between the banks. The deep channel is approximately 700 feet wide but is unmarked by buoys.
Captain Scott notified the pilot of the upbound CHRISTRIP5 of his position and requested him to alert other vessels.
Captain Scott and AB seaman Vernie Cossette stood a six-hour watch ending at midnight E.S.T. They verified that the tug and barge were maintaining their position [1048]*1048by making a radar check on the dike 6 and by comparing their position to two wooden stakes which stood on or near shore, off the port bow. Both men testified that the position relative to the stakes remained constant throughout the watch, as did the gyro heading of 359° and the ship’s relation to Beacon # 26 across the river.7 From midnight until 0600 E.S.T., relief captain-mate Blake and AB Goodwin maintained the watch. They, too, testified that no change in position took place.
By the end of the second watch (0600 hours E.S.T.), the weather had deterioriated and patchy fog had arisen. Captain Scott and AB Cossette returned to the pilot house. The fog grew so thick that the stakes, dike and beacon disappeared.
SOCRATES sounded no fog signals, either by whistle or bell, prior to the collision. The pilot house windows were closed, and Scott and Cossette remained inside. They periodically checked the radar for nearby traffic but made no radio calls.
Meanwhile, back at the tank farm, REFINER completed loading a cargo of crude oil at Ostrica, Louisiana, and started its downbound voyage in the Mississippi toward the Gulf. She too was fully manned and equipped. At the conn was James Moore, a pilot of the Associated Federal Coast Pilots of Louisiana.
As the vessel passed Venice, Louisiana it encountered light fog. The further downriver it sailed, the worse the fog became. At 0714 C.S.T., low fog was noted. At 0717, the second mate recorded “dense low fog”.
Despite the fog, the tanker proceeded beyond Pilottown, through the Head of Passes, and into Southwest Pass at 0726 (Bell Book Time). On the bridge, together with the pilot and mate Allen, were the ship’s master and the helmsman. A lookout was posted on the bow.
Pilot Moore testified that, as REFINER entered the Pass, fog obscured both banks. With engines at half ahead, the ship had a speed — in light of river current — of 12-14 m. p. h. over the ground.8 He made several calls on Channel 13 to ascertain traffic conditions in the Pass, but received no response.
As a part of his navigational duties, Moore also periodically checked the radar screen. He used the ship’s radar on a two-mile range to check for traffic ahead, switching occasionally to a one-mile range for navigational purposes. None of the bridge personnel were assigned to monitor the radar screen constantly or to make a radar plot.
REFINER came down off the ranges on the east (right ascending) side of the river on a course of about 200°. She then angled to the other side, where the channel was deeper, on a course of 207°. The pilot noticed a small “blip” on the radar screen, which later proved to be SOCRATES.9 Although he felt it was perpendicular to his bow, he believed he could pass it to port. The engines remained at half ahead. Visibility remained limited.
Back on SOCRATES, AB Cossette too noticed a radar “blip”, REFINER, approximately V2-3/4 nautical miles upriver. He pointed it out to Captain Scott, who checked the screen and, feeling no concern, sat back down on the couch. He believed the other ship was in the channel, from which his tow was safely removed.
[1049]*1049
Deus Ex Machina
Suddenly, through the fog, like an apparition, the superstructure of SOCRATES appeared to the crew on board REFINER. It was about 5° off the port bow and not more than 3/i mile ahead. Simultaneously, the barge, which the low-lying fog had cloaked, appeared, directly in REFINER’S path. By then, it was too late.
Moments before impact, Captain Scott and AB Cossette saw the bow of REFINER looming in front of them. Scott ordered full astern, but the tow had not even begun to move when the collision occurred. As the District Court found, such action “would have been utterly fruitless” in any event. 526 F.Supp. 1333, 1339, -A.M.C. -,-(E.D.La.1979).
On seeing SOCRATES, Pilot Moore ordered full astern and hard right rudder. When he perceived the barge in his path, he rescinded his order. Hoping to strike the barge with the collision bulkheads on the bow, thereby minimizing the danger of fire, he ordered the rudder hard left. After checking the swing of the bow to the right, the rudder was put amidships. Seconds later, REFINER struck the starboard quarter of AC — 44. The cables linking the tug and barge parted; the ships began to drift downriver.
Following impact at 0740 C.S.T., the engines of REFINER remained full astern for three minutes.10 The vessel, with reduced headway, continued moving downriver but flanking toward the west (left ascending) bank, her bow pointing out into the channel and her stern toward the bank. As she drifted an uncharted and perilous course, REFINER struck an underwater projectile, a steel beam which gouged a thirty-foot, nine-inch gash in the side of the hull in the vicinity of the engine room. Act I ends at 0751, when REFINER went aground and the engines were stopped.
I.
FAULT: SEPARATE BUT EQUAL
Act II of the drama opens far from the waters of Southwest Pass, in the courtroom of the U.S. District Court, sitting in admiralty. There the parties, with their assembled documents, charts, maps, depositions, witnesses, and legal theories, under the “bakery theory of half a loaf is better than none,” Board of Commissioners of the Port of New Orleans v. M/V FARMSUM, 574 F.2d 289, 291 (5th Cir. 1978), sought collectively to pass the buck — in this case, some two million of them — and allocate fault among the actors.
Six separate actions, including a limitation of liability proceeding, involving both the vessels and their owners, charterers and underwriters, were consolidated for trial, with only the question of liability before the Court. In a lengthy and detailed Memorandum Opinion dated June 18, 1979, the District Court found REFINER and SOCRATES equally at fault and assessed liability accordingly. The damages occasioned by the contact with the submerged metal object, the Court found, were a “proximate and direct result of the initial collision [that] should be considered as part of the total measure of damages to be apportioned between the parties at fault.” 526 F.Supp. at 1335,-A.M.C. at-.
Specifically, the Court found that SOCRATES was a vessel underway for navigational purposes throughout its sojourn on the west bank and thus had a duty, a duty that it breached, to sound fog signals, post a lookout, maintain a radio watch in conformity with the Vessel Bridge-to-Bridge Radiotelephone Act, 33 U.S.C. § 1201 et seq., and transmit information relative to its position. The Court also found, relying on AB Cossette’s testimony, that the flotilla moved sometime during the night and stuck out beyond dike 3.05 into the channel, where it obstructed navigation.
[1050]*1050REFINER, the Court held, proceeded through the Pass at excessive speed, in violation of 33 U.S.C. § 192,11 and failed to post a proper lookout. The Court did find that REFINER was navigating a proper course in lawful waters and was not, as Allied Towing asserted, off course and aiming toward the west bank.
Because both vessels were guilty of statutory fault, the Court held that it would apportion damages in conformity with United States v. Reliable Transfer Co., Inc., 421 U.S. 397, 95 S.Ct. 1708, 44 L.Ed.2d 251 (1975). The sum at issue amounts to in excess of $2,000,000.
As the Court found the damage to REFINER was proximately caused by the collision, it declined to consider Hess’ claims against the other parties summoned as defendants, Great Fortune Navigation, Ltd., Savare DeFelice d/b/a DeFelice Towing Company, and the United States. Hess claimed that the underwater obstruction which holed REFINER was the rub rail of the M/V T-TRUC # 1, owned by DeFelice, which sank in the area following a collision with M/V GREAT FAITH, which belonged to Great Fortune, on May 10, 1972. It sought recovery against DeFelice and Great Fortune for negligence in sinking the boat and for failure to remove the wreck, and against all three parties for their failure to locate and mark it.
With no apology for its negligence, SOCRATES has appealed the District Court’s decision on fault and liability. Hess has appealed the Court’s decision that it may not explore the facts of the 1972 collision and that DeFelice, Great Fortune and the United States are not liable. Finding no clear error in the District Court’s careful and well-reasoned opinion, we affirm.
S.S. Scope of Review
Before embarking on our study of the District Court’s findings of fact and law, we must determine the proper scope of review. In McAllister v. United States, 348 U.S. 19, 20, 75 S.Ct. 6, 8, 99 L.Ed.2d 20, 24 (1954), the Supreme Court held that “in reviewing a judgment of a trial court, sitting without a jury in admiralty, the Court of Appeals may not set aside the judgment below unless it is clearly erroneous.” See Chaney v. City of Galveston, 368 F.2d 774, 776 (5th Cir. 1966)12; accord, Termar Navigation Co., Inc. v. SS POLANICA, 529 F.2d 1166 (5th Cir. 1976), cert. denied, 429 U.S. 862, 97 S.Ct. 165, 50 L.Ed. 140 (1976).
II.
THE DISTRICT COURT’S FINDINGS
Equal Fault
The District Court found that SOCRATES and REFINER together caused the accident by proceeding “in weather conditions which dictated utmost caution and prudence, with courses of action born of their own inattention to or misapprehension of the situation as it was.” 526 F.Supp. at 1339, — A.M.C. at -. REFINER proceeded at excessive speed in a dense fog toward an unplotted radar target lying across its bow which did not answer its radio calls. SOCRATES, lying in the edge of a narrow and much travelled waterway, in dense fog, neglected adequately to verify or to advise other ships of its position. Pilot Moore believed he could pass the object before him without incident. Captain Scott apparently believed in good faith he was out of the channel. Both were wrong, the Court concluded, so both vessels must share the blame for the result.
[1051]*10511. Status of the Vessels
The Inland Rules of the Road, 33 U.S.C. § 154 et seq.,13 which apply to this collision, supply the following definition: “A vessel is ‘underway’ within the meaning of these rules when she is not at anchor, or made fast to the shore, or aground.” REFINER was a vessel underway. The status of SOCRATES is more difficult to ascertain. Allied Towing asserts that the tug was aground and therefore not required to sound fog signals, post a lookout, or maintain a listening watch. Such a status, moreover, would entitle it to the presumption of fault which arises whenever a moving vessel strikes one moored or anchored.
Allied Towing concedes that SOCRATES was neither at anchor nor made fast to the shore. The sole category that could apply, then, is that of a vessel aground.
SOCRATES seeks refuge in the analogy of the Western River Rules,14 applicable above the Huey P. Long Bridge. They distinguish between a vessel “underway and towing” and one “holding her position near or against the bank by using her engines”. 33 U.S.C. § 331. As this collision took place below the Huey P. Long Bridge, we need not dally with Allied Towing’s argument.
Although the evidence indicates that the bow of AC-44 may have been in the mud against the bank, and a barge and tug are indeed for many purposes regarded as one vessel, SOCRATES maintained this condition only by use of its engines pushing against the current to keep the barge from sliding. From the time that the tug rounded up until the collision, the engines were not shut down and no anchor was set out.
The District Court found that with SOCRATES running its engines to stay in place, the tow was not aground. This view comports with common sense. “Aground” implies that the vessel is stuck and cannot freely move, at least not until the tide changes or she is pulled free. Here, AC — 44 was kept “aground” only because SOCRATES, running its engines all night, kept it pushed into the bank.
The RUTH, 186 Fed. 87 (9th Cir. 1911), involved an accident between two steamers on the Willamette River in Oregon. OREGONA, in the lead, attempted to proceed over the Clackamas' Rapids. Even using her full power, she was unable to hold her position against the current, and she began to drift downstream. RUTH, meanwhile, had pushed her nose into the west bank and was holding herself there by the movement of her wheel. As OREGONA continued to drift, her captain and RUTH’s captain, standing on their respective decks, desultorily engaged in “some conversation” about moving downstream. RUTH’s captain declined. The wheel of RUTH picked up a line of OREGONA and wound it up rapidly, which caught and injured one of OREGONA’s crew. In the crew member’s suit for damages, the Court held RUTH liable. As a vessel underway, it had a duty to drop downstream at OREGONA’s approach. The Court rejected RUTH’s claim that she was a vessel at rest:
The fact that her nose was rammed against the shore while her wheel was turning at the rate of 30 revolutions per minute brought her no more at rest or anchor than she would have been had her ■ stem been pointed against the current and she had been held in position by the revolutions of her wheel.
186 Fed. at 90; Savoie v. Apache Towing Co., 282 F.Supp. 876 (E.D.La.1968). Citing The RUTH, a leading admiralty author has affirmed, “A vessel lying in a river with her stem against the bank and holding herself against the current by using her engines is underway.” J. Griffin, The American Law of Collision 19 (1949) (emphasis added).
The District Court correctly characterized SOCRATES as a vessel underway.
[1052]*10522. Statutory Violations Fog Signals
A vessel underway must sound whistle signals in fog. 33 U.S.C. § 191(2)(e)15; Barrois Brothers, Inc. v. Lake Tankers Corporation, 188 F.Supp. 300 (E.D.La.1960); Griffin on Collision, §§ 134, 139. The evidence establishes that fog blanketed the Southwest Pass that night, and that SOCRATES sounded no such signals. Allied Towing, citing Valley Towing v. AMERICAN WHEAT, 1978 A.M.C. 2175 (E.D.La. 1977), asserts that sounding fog signals “under circumstances in which none are prescribed” creates confusion. This argument overlooks what we have just concluded, that in these circumstances, fog signals were necessary. Allied Towing’s reliance on AMERICAN WHEAT is equally misplaced. There, three boats were involved. AMERICAN WHEAT and MAMA LERE collided. An explosion resulted which knocked MAMA LERE and the remnants of her burning tow against M/V ROY P. SCHOUEST, which had headed up to the west bank. While the Court did term ROY P. SCHOUEST’s failure to sound fog signals “reasonable, as it would have only added to the confusion,” 1978 A.M.C. at 2186, it continued: “Moreover, it can only be construed as a technical violation insufficient to invoke liability as there was no causal connection between such an omission and the collision.” Id. (emphasis added). ROY P. SCHOUEST was an innocent bystander, whose failure to sound fog signals, reasonable or otherwise, could not possibly have altered the events which enveloped it. Not so here: the District Court specifically found that if SOCRATES had sounded fog signals, it might have provided REFINER with sufficient warning to avoid the accident.
Although neither we nor the District Court can ever know exactly what happened that morning, the rule of The PENNSYLVANIA, 86 U.S. (19 Wall.) 125, 22 L.Ed. 148 (1873) requires SOCRATES to prove that its failure to sound fog signals could not possibly have contributed to the accident. That time-honored rule places a heavy burden of proof upon the party attempting to refute fault.
The liability for damages is upon the ship or ships whose fault caused the injury. But when, as in this case, a ship at the time of a collision is in actual violation of a statutory rule intended to prevent collisions, it is no more than a reasonable presumption that the fault, if not the sole cause, was at least a contributory cause of the disaster. In such a case, the burden rests upon the ship of showing not merely that her fault might not have been one of the causes, or that it probably was not, but that it could not have been.
86 U.S. at 136. This rule still floats, in the wake of U.S. v. Reliable Transfer, supra, which only overruled The PENNSYLVANIA on the point of allocating comparative fault. See Atlantic Mutual Insurance Co. v. ABC Insurance Co., 645 F.2d 528, 531 n.9 (5th Cir. 1981).
As the District Court found, if SOCRATES had sounded fog signals, REFINER would have had the duty to stop and ascertain its position. 33 U.S.C. § 19216; see Hess Shipping Corporation v. SS CHARLES LYRES, 417 F.2d 346 (5th Cir. 1969), cert. denied sub nom., Lykes Bros. S.S. Co. v. Hess Shipping Corp., 400 U.S. 853, 91 S.Ct. 56, 27 L.Ed.2d 91 (1970); Getty Oil (Eastern Operations) v. SS PONCE DE LEON, 409 F.Supp. 909 (S.D.N.Y.1976), aff’d., 555 F.2d 328 (2nd Cir. 1977). Such signals, unlike a ship’s lights, will better penetrate a fog and make known the presence of some vessel. And as REFINER had posted a bow lookout, it is reasonably conceivable that he would have heard any signals SOCRATES sounded. Proper signals would have alerted REFINER to the presence not just of the tug but of the low-lying barge as well,17 which could have given it [1053]*1053time to adjust its course and avoid the collision. While we can never know precisely “what if”, we hold that with her clear statutory fault SOCRATES failed to carry her burden under the rule of The PENNSYLVANIA, and affirm the trial court’s holding.
Lookout
33 U.S.C. § 22118 requires a vessel to keep a proper lookout. SOCRATES had posted no lookout at any place on the tow, which extended 320 feet ahead. Although two men stood watch in the pilot house, we believe that Captain Scott, sitting on a couch, was a less than ideal lookout.
The duty of a tug with tow to use due care “also includes the duty to keep a proper lookout,” Cenac Towing Company v. Keystone Shipping Company, 404 F.2d 698 (5th Cir. 1968); Griffin on Collision, §§ 102, 104, 113; cf., China Union Lines Ltd. v. A. O. Andersen & Co., 364 F.2d 769 (5th Cir. 1966), cert. denied, 386 U.S. 933, 87 S.Ct. 955, 17 L.Ed.2d 805 (1967) (failure to post lookout could not have made any difference). Having approved the trial court’s finding that SOCRATES was a vessel underway, we also approve its finding that it is likely that a lookout could have heard REFINER’S fog signals in time to take action. It is impossible to conclude that failure to post a lookout could not have contributed to the collision, The PENNSYLVANIA, supra.
Radar
Both vessels were equipped with operational radar on the night of the accident. Although Captain Scott sporadically checked the radar screen, he made no attempt to analyze the data (blips, etc.) it provided or to plot REFINER’S position. The District Court found this watch neither proper nor sufficient. Such neglect was clearly sufficient to ground fault.
The Second Circuit, following a lengthy discussion of the pros and cons of radar, observed:
It is likely that there was time for plotting, and this would have revealed the exact speed and course of the BURGAN. We cannot be sure of this. Although her speed was diminishing, the estimate of 5 knots could not have been much out of the way; and it would seem to be a good general rule that if a radar pip shows another vessel nearby and forward on your starboard bow, you should stop and make the time for a plot.
Afran Transport Co. v. The BERGECHIEF, 274 F.2d 469, 475 (2nd Cir. 1960); SS PONCE DE LEON, supra; see also Polarus Steamship Co. v. T/S SANDEFJORD, 236 F.2d 270 (2nd Cir. 1956), cert. denied, Rederi v. Polarus S.S. Co., 352 U.S. 982, 77 S.Ct. 383, 1 L.Ed.2d 365 (1957).
The rule of Afran Transport necessitates only a modicum of effort for great potential benefits. By plotting the movement of radar blips representing REFINER, SOCRATES could have detected the possibility of a collision sufficiently in advance to take appropriate (including warning) action. At a minimum, it could have radioed REFINER to advise her of SOCRATES’ presence. We approve the District Court’s holding that both vessels should have assigned a crew member exclusively to watch the radar.
Radio
Having found that SOCRATES was-a vessel underway, the District Court assessed fault for its disobedience in failing to maintain a proper watch as required by the Vessel Bridge-to-Bridge Radiotelephone Act, 33 U.S.C. § 1201 et seq. Section 1204 requires a “listening watch” for a towing vessel while navigating. Nowhere does the statute define “navigating”. Allied Towing makes a good argument, based on the preamble and other language of the Act, that it applies only to approaching vessels. Since SOCRATES was holding her position, she does not come within the ambit of the rule.
[1054]*1054We disagree. Section 1203 also applies to dredges and floating plants “engaged in or near a channel or fairway in operations likely to restrict or affect navigation of other vessels.” That language testifies to Congress’ intent to cover all vessels whose operations might affect or impede navigation in navigable waters. Our impression is borne out by the Legislative History, [1971] U.S. Code Congressional & Administrative News 1237, which declares: “The purpose of the bill is to reduce vessel collisions and other mishaps [for] the most modern and efficient radio equipment in the world is of little use if no one is listening.” Id. at 1237-1239. We hold that a vessel which is “underway” for the purposes of the Inland Rules must obey the Vessel Bridge-to-Bridge Radiotelephone Act. Hogge v. SS YORKMAR, 434 F.Supp. 715, 735, 1977 A.M.C. 805, 830 (D.Md.1977).
[II] In addition to its failure to maintain a listening watch, SOCRATES neglected its affirmative duty to transmit information on its position, information vital for safe navigation, as required by 33 C.F.R. § 26.04(b).19 The record shows that SOCRATES transmitted its position to CHRIS-TRIP shortly after heading in to the west bank, but made no further calls, despite the deterioration of the weather conditions and visibility.
In SS PONCE DE LEON, supra, the Court held that a vessel’s failure to issue “periodic security calls after anchoring” to advise subsequently arriving vessels of her situation violated the General Prudential Rule, 33 U.S.C. § 221.20 The court reasoned:
Such additional security calls would have given advance warning of the [vessel’s] status to oncoming vessels. Considering the weather and traffic conditions, such additional security calls would not have placed an undue burden on the WILMINGTON GETTY considering its choice of a hazardous place of anchorage.
409 F.Supp. at 919. We believe the District Court operated well within its discretion when it found that SOCRATES should have updated its warning to other vessels in the vicinity of its location.
Where Was SOCRATES?
The District Court concluded that although SOCRATES headed into the west (left ascending) bank of the Mississippi, a location in which it posed no danger to traffic in the deepwater channel, it moved during the night. “[A]t some point and in some manner, the entire flotilla or part of it moved so as to extend at least the stern of the barge and the tug beyond the end of dike 3.05, whereupon it became an obstacle in the path of the HESS REFINER. 526 F.Supp. at 1341, -A.M.C. at-.”
Support for the Court’s finding comes from the testimony of Blake and Cossette, crew members of SOCRATES on the second (midnight to 0600 E.S.T.) and third (from 0600 E.S.T.) watches, respectively. Blake said in his deposition that he could see Beacon # 26 across the river “straight out the side window” of the wheel house and “about directly abeam on the starboard side.” Cossette, on the other hand, said that when he returned to what would have been the third watch, the beacon was visible through the back window if he leaned forward in his seat at the wheel.21 Allied [1055]*1055Towing suggests that perhaps Cossette saw some other light. Even were that so, it would hardly generate confidence in SOCRATES’ crew’s ability to figure out, let alone maintain, their position. The judge could conclude it far more likely that the boat had slowly drifted during the night, so that its position relative to the light had changed.
Captain Scott conceded that a ship passing SOCRATES could create a suction that, despite the continuous use of the engines, could pull the flotilla away from the bank. He did not remember how many vessels passed SOCRATES during the night but admitted that there were several. Scott, whose testimony was largely discredited when he admitted that, contrary to his testimony at the Coast Guard hearing shortly after the collision, he could not see the radar from his spot on the couch, insisted that AC-44 was aground because of a six-inch list he detected. Counsel for Hess showed that the list amounted to a lk° variation — which hardly would be discernible, especially when the fog hid the bow of the barge.
The parties offered in evidence numerous hydrological charts, which map out the depth of the river according to contour lines, and testimony interpreting those charts. Hess asserts that AC-44, which drew approximately 18 feet, would have stuck in the mud along the river bottom well away from the bank. In such a position, the flotilla would have extended out well beyond dike 3.05. Although the District Court did not rely on this data, we believe it further supports the Court’s findings.
The District Court had the opportunity to hear all the witnesses and to study, aided by the parties, the mass of exhibits. Its conclusion that SOCRATES had moved, although not the only possible answer, is a compelling one which is well above the Plimsoll line of F.R.Civ.P. 52(a), Drachenberg v. Canal Barge Co., 571 F.2d 912, 918 (5th Cir. 1978).
SOCRATES bitterly contests this finding, pointing to the testimony of six bar pilots, neutral witnesses who passed SOCRATES prior to the collision, each of whom testified that it posed little or no obstacle to navigation. Pilot Michell, a witness for Hess, twice passed SOCRATES. He testified that it extended “slightly out” beyond the end of dike 3.05. Pilot Booksh said the stern of the tug was inside the line of the dikes. Pilots Malasovich, Myers and Mott testified to the same effect.
Pilot Miller passed SOCRATES only 15 minutes before the accident. He testified that he got a very good look at the flotilla and that the barge had a decided starboard list, which would indicate it was aground. The Court gave no credence to Miller’s testimony, however, because his story was at odds with the facts as otherwise shown in the record. Both Captain Scott and Cossette testified that visibility was extremely limited for 30-45 minutes prior to the collision. Another witness, Robert Owen, a wire line operator who had observed REFINER while working along a canal adjacent to the channel, said there was a dense fog bank about 5 feet in height hanging over the river. Under these circumstances, the Court refused to accept Miller’s version of SOCRATES’ location.
To believe Allied Towing’s story, not only would the Court have to accept Miller’s testimony in the face of contradictory evidence as to visibility, but it would also have to accept Allied Towing’s projection of REFINER’S course. Allied Towing theorizes that REFINER had come off the ranges, turned to a course of 207° but failed to straighten up and, instead, kept on turning to a course of 217° or 227°, which would take it in towards the west bank at a rather acute angle.22 Only such an acute angle [1056]*1056could explain how REFINER could have struck SOCRATES if she were, in fact, safely behind the dike.
This theory, like T-TRUC # 1, promptly sank, the District Court finding it impossible — certainly a major flaw. If REFINER had approached SOCRATES on a course of 217° or 227°, then it would inevitably have continued ahead and run into the west bank, there going aground. “Following collision on such a course, the HESS REFINER, considering its weight, size and speed, would have plowed on and into the bank, and no maneuver or attempted maneuver could have prevented it.” 526 F.Supp. at 1342, - A.M.C. at -. Allied Towing suggests that the collision altered REFINER’S course. How REFINER could have bounced off AC-44 and changed course without flattening SOCRATES in the process, we are at a loss to explain.
Owen, the first-hand, wholly disinterested observer, related that REFINER followed a straight and narrow course about 200-250 feet off the wing dams. The District Judge noted that REFINER neither hit dike 3.05 nor came close. As the facts belied the possibility that REFINER was off course, the Court held, the only alternative, given that the accident did in fact occur, was that SOCRATES and its tow obstructed navigation in the channel.
REFINER’S Course
SOCRATES maintains that, even if it had somehow drifted into the deepwater channel, REFINER alone was at fault because it was navigating outside the channel where it belonged. To this objection, the Court replied that REFINER need not remain in the channel. “We are aware of no rule of law which required her to be navigated within the limits of the dredged channel.” Reading Company v. Pope & Talbot, Inc., 192 F.Supp. 663, 666 (E.D.Pa.), aff’d., 295 F.2d 40 (3rd Cir. 1961), citing American Dredging Co. v. Calmar S.S. Corp., 121 F.Supp. 255, 263 (E.D.Pa.1954), aff’d., 218 F.2d 823 (3rd Cir. 1955):
That the steamship CALMAR was navigating in the extreme easterly portion of the ship channel and that its course would and did take it slightly outside of the channel does not, in my opinion, constitute any fault on the part of the vessel. Certainly, it is no fault of which this libellant can complain. I know of no rule of law which limits the navigation of a vessel to a ship channel, as a land vessel such as an automobile would be limited to operation on a highway.
Several witnesses told the Court that it was customary practice to navigate outside the 40-foot channel.23
SOCRATES, hoping to educate us as to its innocence, refers to two cases to bolster its argument. Each is readily distinguishable. In Chesapeake Bay Bridge and Tunnel District v. Lauritzen, 404 F.2d 1001 (4th Cir. 1968), BELLA DAN strayed 1000 feet from a channel reserved for vessels its size. Paying no attention to charts, BELLA DAN struck a submerged light tower which had collapsed during a storm. The Fourth Circuit reversed the District Court’s finding of no liability.
We disagree with the District Court’s vindication of the motorship. Her pilot quite plainly was negligent. Carelessly conned, she ran afoul of the submerged light tower. With full knowledge of its existence in the waters into which she chose to stray, her path led from the [1057]*1057auxiliary channel into the threatened area.... Admittedly, his vessel was not obligated to keep within the auxiliary and was in no danger of stranding, but nonetheless he was ignoring the known portending menace.
404 F.2d at 1006. REFINER, by contrast, had no knowledge of an imminent danger: it could not see SOCRATES and with no fog signals from her, her presence was completely unknown.
In Warrior & Gulf Navigation Company v. The S.S. STEEL VOYAGER, 237 F.Supp. 200 (E.D.La.1964) (Ainsworth, D. J.), aff’d. sub. nom., Isthmian Lines, Inc. v. Warrior & Gulf Navigation Co., 359 F.2d 68 (5th Cir. 1966), STEEL VOYAGER collided with a tow along the west bank of the Mississippi. It is true, as Allied Towing points out, that the Court assessed sole liability to STEEL VOYAGER for navigating “too close to the west bank of the Mississippi River . . . when there was ample space in the river ... to maneuver.” 237 F.Supp. at 203. Yet the two was moored in a facility known as Ralph’s Fleet in New Orleans harbor, a well-known anchorage which STEEL VOYAGER should have taken care to avoid. That case is a far cry from the instant action, where even if REFINER was too close to the west bank (the District Court expressly found that it was not), SOCRATES’ blocking the channel, in an area not designated an anchorage, certainly constituted fault.
REFINER’S Speed
The District Court characterized REFINER’S speed as excessive and its radar lookout as improper. Hess, content with its 50% share of fault and damages, tells us that these findings are not clearly erroneous. After reviewing the lower court’s findings, we agree and so affirm.
III.
ALLOCATION OF FAULT
The District Court found the parties equally to blame and allocated fault accordingly in conformity with Reliable Transfer, supra. That flood-tide case, which at the stroke of a pen erased a century’s precedent, introduced into the admiralty the concept of comparative fault. Traditionally in collision cases, whenever both parties were guilty of contributing fault, no matter how slight, the Court would divide the damages equally. The Schooner CATHERINE v. Dickinson, 58 U.S. (17 How.) 170, 15 L.Ed. 233 (1855); Griffin on Collision § 245.
To escape the obvious unfairness of this rule, courts engrafted the concept of major/minor fault. Whenever one party’s fault far outweighed another’s, the courts in effect would excuse the less serious offender. See Griffin on Collision § 224. This exception gradually swallowed the rule, leaving no clear course for an admiralty court to follow.
Reliable Transfer changed all this. “It is no longer apparent, if it ever was,” the Court declared,
that this Solomonic division of damages serves to achieve even rough justice. An equal division of damages is a reasonably satisfactory result only where each vessel’s fault is approximately equal and each vessel thus assumes a share of the collision damages in proportion to its share of the blame, or where proportionate degrees of fault cannot be measured and determined on a rough basis.
421 U.S. at 405, 95 S.Ct. at 1713. A court now assesses damages in whatever ratio it deems appropriate. See Atlantic Mutual Insurance Co., supra; Nutt v. Loomis Hydraulic Testing Co., 552 F.2d 1126 (5th Cir. 1977).
Our review of the Court’s apportionment of damages is governed by the “clearly erroneous” standard. Termar Navigation, supra ; Florida East Coast Railway Co. v. Revilo Corp., 637 F.2d 1060, 1067 (5th Cir. 1981); Movible Offshore, Inc. v. M/V WILKEN A. FALGOUT, 471 F.2d 268, 272-273 (5th Cir. 1973); A/R SELJAN v. Pioneer Leasing Corp., 406 F.2d 768, 769 (5th Cir. 1969). The equal 50/50 allocation of fault finds ample support in the facts and the record, and we affirm.
[1058]*1058
Recovery Over Against Allied
Allied Towing labels as error the District Court’s holding that Allied Chemical may recover its full damages from Hess. The reason for this counter-intuitive strategy is that, under long-standing admiralty practice, this loss goes into Hess’ damage claim which, on striking the balance, means that Hess will be entitled to recover 50% of such amount against Allied Towing. The CHATTAHOOCHEE, 173 U.S. 540, 19 S.Ct. 491, 43 L.Ed. 801 (1899); The NEW YORK, 175 U.S. 187 (1899); Griffin on Collision §§ 245, 246.
Allied Towing informs us that Alamo Chemical Transportation Co. v. M/V OVERSEAS VALDES, 469 F.Supp. 203 (E.D.La.1979), interpreting the Reliable Transfer mandate, supplants that traditional rule. We disagree. In OVERSEAS VALDES, a tanker collided with a tug, HARD WORK, and its tow. Firestone Tire & Rubber Co. sued for damages to its cargo upon the towed barge. Maritime, the owners of OVERSEAS VALDES, sought recovery over against Alamo if found liable; Alamo sought recovery over against Maritime in the same event. In its trial of liability, 398 F.Supp. 1094 (E.D.La.1975), the Court had assessed liability at 20% to OVERSEAS VALDES and 80% to HARD WORK.
In the subsequent trial, the Court, finding a conflict between Reliable Transfer and the rule of the CHATTAHOOCHEE, refused to allow Firestone to recover all its damages from Maritime, with recovery over against Alamo. The Court acknowledged two choices: full recovery or only 20%. Concluding that Reliable Transfer dictated the 20% solution, the Court said: “To grant Firestone a judgment for the full amount of its damages against the M/V OVERSEAS VALDES . . . would be in violent contradiction to Reliable Transfer.” 469 F.Supp. at 214. The District Judge based his decision upon two factors: what he termed the Supreme Court’s “disdain for the prior rule” and the presence of a valid Both-to-Blame clause which limited Alamo’s liability to Firestone.
In arriving at that decision, the OVERSEAS VALDES Court explicitly rejected prior contrary holdings. In Complaint of Flota Mercante Grancolombiana, S.A., 440 F.Supp. 704, 725 (S.D.N.Y.1977), the Court, taking due notice of Reliable Transfer, staffed:
This Court believes that the better rule would be to allow ... a full recovery against the non-carrying vessel, allow the latter to add the payments it makes to cargo to its own damages, and then to permit it to recover this total from the carrying vessel in proportion to the fault of the carrying vessel. This resolution would be more in accord with the principies enunciated in Reliable Transfer [which] was intended to change the rule of damages so that the burden would fall more equitably on those parties whose fault contributed to the collision.
We believe that Flota Mercante more nearly charts a Reliable course. The Supreme Court acted to abolish the doctrine of mutual fault-equal contribution, an anachronism which every other maritime nation had already rejected. The decision only alters the proportions in which dam-ages are fixed — it does not abolish the procedure by which an innocent cargo owner traditionally could recover his damages in full from the tortious non-carrying vessel ¡ no matter what the degree of fault between ' the tortious actors.24
The Supreme Court provides support for our view in Edmonds v. Compagnie Generale Transatlantique, 443 U.S. 256, 99 S.Ct. 2753, 61 L.Ed.2d 521, 1979 A.M.C. 1167 (1979). Edmonds, a longshoreman, was injured while unloading cargo from a vessel, the SS ATLANTIC COGNAC, owned by [1059]*1059CGT. The jury found that he had suffered $100,000 damages but was responsible for 10% of his injuries. The stevedore, Edmonds’ employer, was found 70% liable as a result of a co-employee’s negligence. Construing 1972 Amendments to the Longshoremen’s and Harbor Workers’ Compensation Act, 33 U.S.C. § 901 et seq., the Fourth Circuit, sitting en banc, concluded that CGT was liable only for the share of total damages equivalent to the ratio of its fault to the total fault.
The Supreme Court reversed. Citing Reliable Transfer, the Court observed “Reliable Transfer merely changed the apportionment from equal division to division on the basis of relative fault. But we did not upset the rule that the plaintiff may recover from one of the colliding vessels the damage concurrently caused by the negligence of both.” 443 U.S. at 272, 99 S.Ct. at 2762, 61 L.Ed.2d at 534, 1979 A.M.C. at 1180 (n. 30).
OVERSEAS VALDES, decided before Edmonds, involved a Both-to-Blame clause that recovery under the traditional rule would sidestep. The Court, by confining Firestone to its percentage recovery against Maritime, did an end-run around the problem. In so doing, it created a greater problem.
The Court’s decision in OVERSEAS VALDES leaves the innocent cargo, “the one party both aggrieved and free to do something about it,” O/Y FINLAYSON-FORSSA v. Pan Atlantic Steamship Co., 259 F.2d 11, 13 (5th Cir. 1958), cert. denied, 361 U.S. 882, 80 S.Ct. 153, 4 L.Ed.2d 119 (1959), in the lurch. Where a cargo suffers damages from a collision where both vessels were negligent, it is the only innocent party. Reliable Transfer certainly did not intend to shift the risk of loss away from the negligent parties. Yet OVERSEAS VALDES does precisely that.
We reject Allied Towing’s argument and the decision in OVERSEAS VALDES to the extent it is inconsistent with our holding.25 We therefore affirm the District Court’s grant of full recovery against Hess, which automatically goes into its damages for ultimate division.
IV.
THE LOCH NESS MONSTER OF SOUTHWEST PASS
Hess appeals the District Court’s dismissal of DeFelice, Great Fortune and the United States and its refusal to consider whether those parties should shoulder some part of the liability for the damage to REFINER’S hull when it encountered the mysterious underwater object. This appeal presents an interesting question of proximate causation.
Proximate causation lies among the most mysterious shoals of our jurisprudence. Courts often are asked to determine that some prior negligent act, no matter how minor, caused some subsequent and totally unforeseen harm. In this case, Hess does not ask us to excuse its negligence. Rather, it claims that the negligence of Great Fortune and DeFelice in colliding, and of the United States in not marking the wreck, also contributed, as a substantial factor, to the collision. The District Court disagreed.
Judge Boyle found that the sinking of T-TRIJC # 1 was not a proximate cause of the damage to REFINER. As he concluded:
Undoubtedly the position of the HESS REFINER at the point of her contact with the metal object resulted directly from the occurrence of the collision. . . . In this case, the HESS REFINER came into contact with a submerged object as the result of a collision occasioned by her own negligence and that of the SOCRATES. No intervening negligent act occurred to interrupt the chain of causation begun by the collision between the vessel and the AC — 44. Even assuming that the submerged object’s presence was due to the negligence of some other party or parties, the sole cause of the presence of [1060]*1060the HESS REFINER in the place in which it sustained damage was the collision which caused the HESS REFINER to deviate from her pre-collision course.
526 F.Supp. at 1347-48, -A.M.C. at-. To use the well-worn vernacular of proximate causation, but for the negligent conn of both REFINER and SOCRATES, the tanker never would have been in the vicinity of the submerged obstruction. She would have remained in or near the deepwater, continued her traverse of the channel, missed the undisclosed steel beam, and most likely negotiated the Southwest Pass without injury. The collision was a proximate cause of REFINER’S damages.
The District Judge made no specific findings as to the conduct of DeFelice, Great Fortune and the United States, holding that the accident resulted solely from the tug’s and tanker’s failure to exercise the wisdom of Socrates. Here, we believe, he was technically mistaken. The concept of new and intervening cause would not displace fault, if any, on the part of DeFelice, Great Fortune and the United States toward REFINER. A subsequent negligent act does not excuse prior negligence except in most unusual circumstances. It would only mean that all actors, past and present, must divvy up the liability. See Restatement of Torts 2d, § 442B.26
Yet the Court’s finding necessarily assumes that Hess failed to meet its burden of proof as to a number of points crucial to its argument. The record does little to dispel uncertainty as to where the object was located27 or, indeed, what it was.28 Despite a virtual chorus line of first-hand and expert witnesses, Hess failed to prove that the piece of metal which impaled REFINER came from T-TRUC # 1. Savare DeFelice stated that it did not. Reginald E. Daughdrill, Jr., a chemist who treated the metal, opined that it had been in the water for a minimum of 7 years and more probably about 30 years. Testimony as to whether a “curl” in the metal had resulted from flame cutting or shearing away was indecisive at best, as was evidence about the thickness of the metal, its hardness, original form, etc. Hess’ failure to supply a definite, let alone definitive, answer regarding the involvement of T-TRUC # 1 leaves us, as it no doubt left the District Court, in the position of scientists attempting to document the existence of the Loch Ness Monster. After reviewing the record evidence, we believe the District Judge’s failure to find liability as to these third parties finds adequate support, and so we affirm.
The potential liability of DeFelice, Great Fortune and the United States also depends on whether or not they failed in their duty to raise or mark the wreck. The Rivers & Harbors Act of 1899, as amended, 33 U.S.C. § 401 et seq., mandates that the owner of a sunken craft has a duty to mark and remove it, see § 409, commonly called the Wreck Act. The goal of § 409 indeed is to prevent incidents such as this collision.
Wyandotte Co. v. United States, 389 U.S. 191, 88 S.Ct. 379, 19 L.Ed.2d 407 (1967), emphasizes the importance of this requirement. There, the owner of a sunken barge loaded with poisonous chemicals had to reimburse the United States for its removal after refusing to do so itself.
[1061]*1061Humble Oil & Refining Co. v. Tug CROCHET, 422 F.2d 602 (5th Cir. 1970), proclaims the duty of the owner of a sunken craft to remove it or pay the price — in that case, a share of the cost of a subsequent accident. A barge sank in the Mississippi River. Its owner, despite full knowledge of its location, failed to remove it, a statutory fault under the Wreck Act. The United States thereupon placed a lighted buoy on the spot, but, unknown to anyone, the light was not functioning on the night when CROCHET ran upon the wreck. Despite faulty navigation almost comical in its scope on CROCHET’S part, the District Court also found fault on the part of Car-gill, the owner of the sunken barge. Its clear violation of its duty to mark and promptly to remove the wreck, which the United States’ assumption of the job could not displace, grounded liability.
CROCHET is an important case which fills Hess’ sails with a full wind. Yet we find it unsupported by the facts in this case. The record shows that DeFeliee made a sincere, diligent effort to locate T-TRUC # 1. Starting the day after the collision, teams of marine surveyors and divers combed the Southwest Pass from a point about 3 miles below Head of Passes all the way to the sea buoy. Employing sophisticated equipment and techniques, they uncovered no trace of T-TRUC # 1. Shell Oil Company also had 3 boats out searching. One expert stated that he believed that T-TRUC # 1 had sunk beyond the sea buoy in the Gulf of Mexico.
Hess, not for lack of trying, could not show any deficiencies in DeFelice’s search. That fact finds confirmation in the Coast Guard’s inability, in its own search, to locate T — TRUC # 1. On these facts, we believe DeFeliee has overcome the burden of CROCHET. However strong the requirement that the owner of a sunken vessel must mark and remove it, a requirement that we have no wish to weaken, it simply cannot apply when the owner has made a full, good faith search for his craft and cannot find it.
Since Hess failed to prove that the metal object came from T-TRUC # 1 or that DeFeliee was negligent in not marking the wreck, we believe the District Court judgment must stand.
Affirmed.