Full opinion text
OPINION GRIESA, District Judge. This action rises from the grounding of the Liberian S. T. Argo Merchant December 15, 1976 approximately 25 miles southeast of Nantucket Island. The Argo Merchant was a tanker carrying a cargo of 27,566 long tons of fuel oil on a voyage from Puerto La Cruz, Venezuela to Salem, Massachusetts. After lingering in the stranded position for some days, the ship broke in two and sank. The entire cargo of oil spilled into the ocean and was a total loss. The shipowner is Thebes Shipping Inc., a Liberian corporation. The cargo of oil was consigned to Northeast Petroleum Corporation. The cargo insurer, Continental Insurance Company, having paid Northeast approximately $2 million for the cargo loss, is subrogated to the rights of Northeast. Procedural History On December 20, 1976 Thebes Shipping Inc. filed the present action — a petition for exoneration from or limitation of liability. On the same date an order was entered restraining all other actions against Thebes arising out of the grounding, and directing that claims be filed in the limitation proceeding. The following claims have been filed: (1) Continental Insurance Company— claim for loss of oil; (2) Northeast Petroleum Corporation— contingent claim for indemnity in the event of recovery against Northeast for environmental damage; (3) Jay Lanzillo et al. — claim on behalf of commercial fishermen for possible damage to fishing beds and fishing equipment; (4) Nantucket Conservation Foundation, Inc., et al. — claim for possible fouling of beaches and other types of environmental damage; (5) United States of America — claim for expenses of activities undertaken to protect against oil pollution; (6) Commonwealth of Massachusetts— claim for expenses of cleaning pollution and protecting against pollution; (7) State of Rhode Island — claim for expenses of protecting against possible pollution. Several actions filed in this court and in the United States District Court for the District of Massachusetts are under the restraint of the December 20, 1976 order. The claims of the United States of America, the Commonwealth of Massachusetts, and the State of Rhode Island have been settled for the total sum of $1.1 million— $20,000 being paid to the Commonwealth of Massachusetts, $5,000 to the State of Rhode Island, and the remainder to the United States of America. The major issue involved in Thebes’ petition for exoneration or limitation is whether Thebes will be liable to Continental for the value of the lost oil. With regard to the question of environmental damage, it appears that the favorable weather conditions resulted in the oil being washed out into the open sea. However, the environmental claims of Lanzillo et al. and Nantucket Conservation Foundation et al., and the contingent claim of Northeast, have neither been settled nor formally abandoned. At least in theory, the question of whether Thebes has a right to exoneration or limitation of .its liability bears upon these latter claims. The petition of Thebes for exoneration or limitation was tried to the court without a jury. This opinion constitutes the court’s findings of fact and conclusions of law. Statutory Provisions and Issues The basic substantive law governing the issues between Thebes and Continental is the Carriage of Goods by Sea Act (COGSA), 46 U.S.C. §§ 1300 et seq. Section 4(2)(a) of COGSA, 46 U.S.C. § 1304(2)(a), provides that neither the carrier nor the ship shall be responsible for loss or damage to cargo resulting from: “(a) Act, neglect, or default of the master, mariner, pilot, or the servants of the carrier in the navigation or in the management of the ship; . . . ” The shipowner has the burden of proving this “error of navigation” exemption from liability. Director General of India Supply Mission v. S. S. Maru, 459 F.2d 1370, 1372 (2d Cir. 1972), cert. denied, 409 U.S. 1115, 93 S.Ct. 898, 34 L.Ed.2d 699 (1973). If a shipowner proves error of navigation as a cause of loss, a cargo owner may defeat the owner’s claim of exemption under COGSA by showing that an unseaworthy condition of the vessel was a concurring cause. In the Matter of the Complaint of Grace Line Inc., 517 F.2d 404, 407 (2d Cir. 1975); Director General of India Supply Mission v. S. S. Maru, supra at 1372; J. Gerber & Company v. S. S. Sabine Howaldt, 437 F.2d 580 (2d Cir. 1971). If unseaworthiness is shown as a cause, the shipowner, in order to escape liability, must sustain the burden of proving that it exercised due diligence at the beginning of the voyage in respect to the condition found by the court to be unseaworthy. Section 3(l)(a) of COG-SA, 46 U.S.C. § 1303(l)(a), provides: “(1) The carrier shall be bound, before and at the beginning of the voyage, to exercise due diligence to— (a) Make the ship seaworthy; . .” Section 4(1), 46 U.S.C. § 1304(1), provides: “(1) Neither the carrier nor the ship shall be liable for loss or damage arising or resulting from unseaworthiness unless caused by want of due diligence on the part of the carrier to make the ship seaworthy, and to secure that the ship is properly manned, equipped, and supplied, . Whenever loss or damage has resulted from unseaworthiness, the burden of proving the exercise of due diligence shall be on the carrier or other persons claiming exemption under this section.” Thebes contends that it is entitled to exoneration under Section 4(2)(a) of COGSA because it has proven that the grounding of the Argo Merchant resulted from errors in navigation by the ship’s officers — i. e., steering improper courses and failing to make adequate use of available navigation techniques to ascertain that the vessel was in danger of grounding. Continental contends that defects in the vessel’s navigation equipment were concurring contributing causes of the casualty— specifically, that the vessel’s gyrocompass and radio direction finder were malfunctioning, and that certain charts used prior to the grounding were outdated. Also, Continental argues that the ship’s compliment of navigation gear was inadequate because it lacked a radio navigation device known as LORAN. As to any condition found to have been unseaworthy and a cause of the grounding, Thebes urges that there was an exercise of due diligence, and Continental argues to the contrary. Evidence and Findings The Argo Merchant left Puerto La Cruz, Venezuela the morning of December 5,1976 bound for Salem, Massachusetts. The officers of the ship having to do with navigation were the master, Georgios Papadopoulos; the chief officer, Georgios Ypsilantis; and two second officers, Anastasios Nisiotis and Georgios Dedrinos. On the voyage from Puerto La Cruz the chief officer and the two second officers had the following watches on the bridge.; Dedrinos Midnight — 4:00 A.M. and noon to 4:00 P.M. 4:00 A.M.-8:00 A.M. and 4:00 P.M.-8:00 P.M. Ypsilantis 8:00 A.M.-noon and 8:00 P.M.-midnight Nisiotis From Puerto La Cruz the vessel sailed northwest about 500 miles to Mona Passage, between Puerto Rico and the Dominican Republic. The vessel sailed through Mona Passage on December 7. From Mona Passage the vessel continued northwest to Cape Hatteras, North Carolina, a distance of approximately 1100 miles. The vessel arrived at a point off Cape Hatteras shortly before midnight December 12, at which time it turned northeast and set course for Nantucket Lightship. The intention was to pass Nantucket Lightship to the east, and sail around Cape Cod into Salem. The distance from Cape Hatteras to Nantucket Lightship is about 435 miles. The navigating officers expected to sight Nantucket Lightship at about 3:00 A.M. December 15. They did not do so, kept sailing, and ran aground at 6:00 A.M. December 15, at 41° 02' N latitude, 69° 28' W longitude, in Fishing Rip, Nantucket Shoals, 32 miles north of Nantucket Lightship. The vessel actually sailed past Nantucket Lightship about 18 miles west thereof. The vessel’s intended course would have taken her about three miles east of the lightship. The vessel was thus off course by some 21 miles. As would be expected, the vessel was equipped with both a gyrocompass and magnetic compasses. The master gyrocompass was located in a room beneath the bridge. There were two gyrocompass repeaters on the bridge — one located to the right of the helmsman and one on the forward bulkhead of the bridge. The gyro repeater to the right of the helmsman was attached to the automatic steering mechanism. There was also a magnetic compass located to the left of the helmsman — referred to as the “steering compass.” Another magnetic compass, referred to as the “standard compass,” was located on the flying bridge. In the chart room located behind the bridge there was a radio direction finder. There were also two radars and a fathometer in the bridge area. The evidence indicates that, at various times in the open seas under favorable conditions, the automatic steering mechanism would be used. However, the indications are that the automatic steering mechanism was not used after Cape Hatteras, and that there was manual steering between Cape Hatteras and the grounding. Although most of the Argo Merchant’s navigation materials and equipment went down with the ship, the two charts used on the vessel from Cape Hatteras to the grounding were rescued. These two charts are C. & G.S. No. 1000 (Cape Sable to Cape Hatteras), and C. & G.S. No. 1107 (Georges Bank and Nantucket Shoals). The rough deck log was rescued, as were the “deviation cards” for the standard and steering magnetic compasses. The officers and crew of the Argo Merchant were rescued. Depositions of the captain and others were taken commencing a few days after the grounding. Also, testimony was taken in the spring of 1977 before a Marine Board of Investigation appointed by the Republic of Liberia. Both the depositions and portions of the Liberian Board of Investigation testimony were introduced at the trial of this action. Officers’ Testimony Regarding Navigation from Cape Hatteras The summary in this section of the opinion is based mainly upon the testimony of the captain, the chief officer, and the two second officers, relating to the navigation of the Argo Merchant from the time of the turn at Cape Hatteras until the grounding. Although there are questions about the reliability of certain important features of this testimony, the basic story told by the navigating officers should be understood. The turn at Cape Hatteras was made at 11:00 P.M. December 12. The chart shows that the turn occurred ten- miles off Diamond Shoal Light. A course of 040° was laid out on chart C. & G.S. No. 1000 (Cape Sable to Cape Hatteras), which would take the vessel slightly to the east of Nantucket Lightship. When testifying about the vessel’s courses, the navigating officers referred primarily to' what they said were courses on the gyrocompass. The magnetic compass served mainly as an auxiliary, to be used for steering if the gyrocompass failed. The magnetic compass was used for steering after trouble occurred with the gyrocompass the evening of December 14. According to the navigating officers, until this problem arose, the helmsmen steered by the gyrocompass repeater immediately to the right of the wheel. The ship’s courses were entered in the deck log every four hours — midnight, 4:00 A.M., etc. Course changes were supposed to be entered when made, although, during the period from Cape Hatteras to the grounding, the three course changes which occurred were not promptly logged, and were not entered until the next regular four-hour recording time. The logbooks of the Argo Merchant contained printed forms in which the courses were recorded. There were three columns —“Standard,” “Gyro” and “Steering.” On the final voyage of the Argo Merchant no entries were made in the “Gyro” column, and only the “Standard” and “Steering” columns were filled out. The heading “Standard” was usually crossed out by hand and the word “True” printed in. The testimony of the navigating officers is that this column referred to gyrocompass courses, and they were put in the “True” column because the gyrocompass had been corrected so that the gyro course was the same as the true course. The course entries labeled “Steering” referred to the magnetic steering compass. Dedrinos had the deck watch from midnight December 12 to 4:00 A.M. December 13. At some point during this watch he called the captain, said that the weather was pushing the vessel to the right, and recommended that there be a change of course to the left in order to compensate. The captain approved a 4° change from 040° to 036° on the gyro. The course entries in the log commencing at 4:00 A.M. December 13 and continuing through midnight of that day were 036° True, 050° Steering. The most precise method for establishing a ship’s position by celestial navigation is to take star sights at dawn and dusk of each day. However, because of overcast weather, no such star sights could be obtained on the Argo Merchant after Cape Hatteras, until one-half hour before the grounding when a star fix was obtained which will be discussed later in this opinion. Although no fix was obtained by star sights on the morning of December 13, a position was obtained at noon that day by sun lines. This position was nearly ten miles to the right, or east, of the intended course line drawn from Cape Hatteras to Nantucket Lightship. This means that, while the vessel had supposedly been steering 040° from 11:00 P.M. December 12 until some time during the midnight-4:00 A.M. December 13 watch, when the course was changed to 036°, the vessel had actually made good a course of 044° during this time. The vessel did not change course at noon December 13. The captain testified that the reason for no course change was that Nantucket Lightship bore 036° from the noon December 13 position, and the vessel was already on this course. At about 4:00 P.M. December 13 the vessel started experiencing a strong gale from the north northwest. The captain gave orders to reduce speed. The gale started to subside after midnight December 13 and was over at least by 8:00 A.M. December 14. Commencing at 4:00 A.M. December 14 there was a change in the course entries in the log which is the subject of some confusion — a change from 036° True, 050° Steering, to 036° True, 047° Steering. Thus there was a change in the steering, or magnetic, compass entries from 050° to 047°, but the entries for the true, or supposedly gyrocompass, entries remained the same— 036°. When the captain first testified about this, he stated that there was in fact a 3° course change to the left, as reflected by the change in the magnetic compass entries from 050° to 047°. The captain testified that this course change was made to compensate for weather and current from the port side of the ship. He testified that his information regarding the currents came from a chart entitled “Pilot Chart of the North Atlantic Ocean.” These pilot charts are published by the United States Defense Mapping Agency, and are issued monthly. The captain used the pilot chart for November 1976. He did not have the December 1976 chart. There is a question, which will be discussed later, as to whether the captain was misled in any material degree by having available only the November chart. At a later point in his testimony, the captain stated that there was no course change at 4:00 A.M. December 14, judging from the fact that there was no change in the gyrocompass entries. He could not explain why the magnetic compass entries went from 050° to 047°. Still later, the captain testified that he could not remember whether there was or was not a course change at this time. Dedrinos had the deck watch from midnight December 13 to 4:00 A.M. December 14. He testified that he could not remember any order for a course change during his watch and that during the entire time the helmsmen were being directed to steer 036° gyro. He testified, however, that there was a blackboard on the bridge on which the compass courses were written, and that during his watch the figures on the blackboard were changed with respect to the magnetic course from 050° to 047°, while the figures relating to the gyro course remained 036°. He cannot recall any discussions about this change. The only reason he can recall for the change in the blackboard notation for magnetic was that the heading on the magnetic compass changed. Ypsilantis assumed the deck watch at 4:00 A.M. He testified at one time that he did not know whether there was a course change before 4:00 A.M. December 14, and later testified positively that there was no change. Nisiotis, who came on watch at 8:00 A.M., testified that there was no course change, since the gyrocompass course entries remained the same in the log and the vessel was steering by gyro. Positions were obtained by sun lines at noon and 2:00 P.M. December 14. The noon position was about 2V2 miles to the left, or west, of the intended 040° course line which had been drawn from Cape Hatteras to Nantucket Lightship. The 2:00 P.M. position was about 3!/2 miles to the left of that course line. This meant that, from noon December 13 to noon and 2:00 P.M. December 14, the vessel had made good a course of 035°. Nisiotis, who was on watch from 8:00 A.M. to noon and calculated the noon fix, informed the captain, who was on the bridge at noon, that the vessel was to the left of the course. However, he did not calculate the course made good from noon of the previous day and he did not discuss a possible problem or course change with the captain. Dedrinos came on watch at noon, and' calculated the 2:00 P.M. fix. He testified at first that he did not give any thought to the course made good since noon the previous day. Later he testified to the contrary, and stated that he did discuss with the captain the course made good, but did not discuss a possible course change. The captain’s testimony is that he did not direct or consider any possible course change at midday December 14. Thus, according to the testimony of these officers, the vessel continued to steer 036° gyro and 047° magnetic after the noon and 2:00 P.M. fixes December 14. If the vessel continued to make good 035°, as it had from noon December 13, the vessel would pass to the west of Nantucket Lightship headed into shoal waters. This would also be true if the vessel made good 036° from the noon and 2:00 P.M. December 14 positions. According to the navigating officers, trouble with the gyrocompass commenced about 6:00 P.M. December 14. The gyro repeater to the right of the helmsman started fluctuating rapidly over an arc of about 12-14°. From this time until the grounding the next morning the steering was by magnetic compass. However, no entry about the gyro problem was made in the log. Also, entries continued to be made in the “True” course column in the log. Thus the entries at 8:00 P.M. December 14 were the same as they had been the entire day— 036° True and 047° Steering. The captain testified that, after the trouble began with the helmsman’s gyro repeater, he and Ypsilantis checked the master gyro and found that it was also defective, although they could not ascertain what was the source of the problem. Ypsilantis, who was on watch from 4:00 P.M. to 8:00 P.M. December 14, testified that the only trouble was with the helmsman’s gyro repeater, and that he and the captain checked the master gyro and found no problem there. Nisiotis, who came on watch at 8:00 P.M. December 14, agrees with Ypsilantis that there was a problem with the helmsman’s gyro repeater but that the master gyro continued functioning properly. Indeed, Nisiotis testified that he checked the master gyro and that it read 036° while the magnetic compass on the bridge read 047°. There is a conflict about the condition of the other gyro repeater on the bridge — the one at the forward bulkhead. This repeater was the one which was supposed to be available for the officer of the deck to watch, while the helmsman was watching the repeater to the right of the wheel. The captain could not recall whether the problem which commenced the evening of December 14 affected any repeater other than the one near the helmsman. Nisiotis testified that he believes that the gyro repeater on the forward bulkhead continued to operate normally after the evening of December 14. However, Ypsilantis testified that the repeater on the forward bulkhead was not operating at all during the entire voyage. It is worth noting that there is no testimony by any officer about any use being made of the repeater on the forward bulkhead. There appears to be unanimous agreement that a third gyro repeater, on the flying bridge, was inoperative during the final voyage. Following the onset of the gyro problem, the helmsmen steered by the magnetic steering compass. For a short time this course remained 047°. Nisiotis came on watch at 8:00 P.M. He recalled being to the left of the course line at noon that day. He also was aware, both from observation and from looking at a chart, that the current was pushing the vessel to the left. He discussed this with the captain, who ordered a change of 3° to the right — i. e., from 047° to 050° magnetic. According to Nisiotis, this course change was made early in his watch, about 9:00 P.M. However, the first record in the log was at midnight, when the entry of 050° Steering was made. The entry in the True column remained 036°. The course, and the log entries, remained the same until the grounding at 6:00 A.M. the next morning, December 15. Dedrinos had the deck watch from midnight to 4:00 A.M., and Ypsilantis the watch from 4:00 A.M. until the time of the grounding. Dedrinos remained on the bridge with Ypsilantis during the latter’s watch. The captain was on the bridge commencing at 1:00 A.M. in order to monitor the expected sighting of Nantucket Lightship and the changes of course that would occur thereafter. The officers expected to sight Nantucket Lightship visually at approximately 3:00 A.M. and pass it by 4:00 A.M. Ypsilantis put down a midnight “dead reckoning” position on a large scale chart — C. & G.S. No. 1107 (Georges Bank and Nantucket Shoals). The midnight dead reckoning position was calculated by simply estimating the distance which the ship had traveled to that time, and marking a position on a line which corresponded to the original 040° course line from Cape Hatteras to Nantucket Lightship. No adjustment was made to take into account information such as the fact that between noon December 13 and noon December 14 the ship had made good a course of 035°, and that the ship was to the left of the 040° course line as of noon December 14. On the large scale chart used on the Argo Merchant in the approach to Nantucket Lightship, the visibility of the light is described as 14 miles. However, on a more recent version of the chart, which was apparently not present on the Argo Merchant, the visibility of Nantucket Lightship is shown as 23 miles. Nantucket Lightship had a radio beacon, which could be used by vessels to obtain a radio bearing on that lightship. The Argo Merchant had a radio direction finder (“RDF”) to pick up such a bearing. The Nantucket Lightship radio beacon had a listed range of 100 miles. At the time the captain came on the bridge at 1:00 A.M. Nantucket Lightship was about 25 miles away. This 25 mile distance is what we now know on the basis of working back from the grounding position. Of course, the navigating officers had a different perception of the location of the ship from what actually turned out to be the case. The officers’ estimates of position placed the ship about 16 miles southeast of where it actually was. However, it happens that, even under the assumptions of the officers regarding the ship’s position, the ship would have been only 28 miles from Nantucket Lightship at 1:00 A.M. It is apparent from the officers’ testimony that the RDF should have been in use commencing at least by 1:00 A.M. This was well in advance of the time when the officers expected to pass Nantucket Lightship. Even according to what we now know about the actual 1:00 A.M. position, the ship was in safe waters west southwest of Nantucket Lightship. Dedrinos testified that the captain used the RDF every few minutes commencing at 1:30 or 2:00 A.M., although Dedrinos testified that he himself never used the RDF that night. The captain testified at first that he used the RDF commencing at about 1:00 A.M., but later testified that he did not use the RDF before 4:00 A.M., and that if anyone did use the RDF before 4:00 A.M., it was Dedrinos. The Nantucket Lightship was not sighted visually or by radar by 3:00 A.M. as expected, or at any time thereafter. The captain, Ypsilantis and Dedrinos testified that commencing at 4:00 or 4:30 A.M. the RDF was used continuously to obtain a bearing on Nantucket Lightship. According to the testimony, the captain, Ypsilantis and possibly Dedrinos (Dedrinos testified about the captain and Ypsilantis using the RDF but denied that he himself used it, whereas the captain testified that all three of these officers used the RDF) obtained frequent bearings on Nantucket Lightship, and those bearings were all on the bow — i. e., 000° relative. According to the testimony of the officers, this information from the RDF meant that the vessel was not off course, and had not passed Nantucket Lightship, but was on course, headed directly toward the lightship, the only problem being that the vessel had been traveling at a slower speed than estimated and was thus behind schedule. We now know that during the entire period from about 1:00 A.M. until the grounding, the relative bearing of Nantucket Lightship from the Argo Merchant was never on the bow, or 000° relative. At 1:00 A.M. the lightship bore about 040° relative. The Argo Merchant passed Nantucket Lightship at a distance of about 18 miles to the west of the lightship at about 3:00 A.M. and was off course by a total of 21 miles. Thus at about 3:00 A.M. Nantucket Lightship was on the starboard beam of the ship — i. e., 090° relative. At all times after 3:00 A.M. Nantucket Lightship bore aft of the starboard beam of the Argo Merchant. At 4:00 A.M. the lightship bore about 115° relative. At 5:00 A.M. the lightship bore about 140° relative. As dawn approached, it appeared that the weather would permit the taking of star sights and the establishment of a fix. At 5:30 A.M., although the light was not yet optimum, Ypsilantis took certain sights. He calculated a fix showing the vessel some 38V2 miles southwest of Nantucket Lightship. Although this fix placed the ship somewhat to the left, or west, of the intended course line, it was still in safe waters. The officers considered this to be a somewhat doubtful fix and hoped to take a more reliable one in another half hour. Shortly before the grounding, the officers noted that the fathometer showed a depth of 15-20 fathoms. This fathometer reading was totally inconsistent with the information they say was obtained from the RDF. It should have indicated to the officers that they had already passed Nantucket Lightship and were on the verge of running aground. However, Ypsilantis, who was on watch at the time, cannot recall checking the fathometer reading with the chart. The officers ignored the fathometer reading, and kept steaming ahead in hope of getting a new star fix in a few minutes. The vessel ran aground at 6:00 A.M. — 32 miles north of Nantucket Lightship. At the trial Ypsilantis was asked to recalculate the 5:30 A.M. fix based on his notes of the star sights. This showed that the fix was erroneously calculated — off by some 39 miles. The position which should have been laid down from his star sights would have shown that the .vessel had passed Nantucket Lightship and was substantially inshore of it. Between noon December 14 and the grounding, the course actually made good by the vessel was 033°. Findings Regarding Navigation and Equipment Despite contradictions and inconsistencies, it is apparent that the navigating officers of the Argo Merchant were telling the following basic story. A course line of 040° was drawn from Cape Hatteras to Nantucket Lightship. For a short time after the turn at Hatteras the vessel steered 040° True, and then turned 4° to the left to 036° True in order to compensate for weather and current running to the right. The vessel stayed on the course of 036° True for more than 40 hours (from sometime in the midnight — 4:00 A.M. watch December 13 until sometime in the 8:00 P.M. — midnight watch December 14). The captain believed that, because of weather and current moving generally to the right, steering a course of 036° True would cause the ship to make good the basic course of 040° laid out on the chart. Finally, between 8:00 P.M. and midnight December 14, the ship changed course 3° to the right to compensate for current now moving leftward. By this time the steering was by the magnetic compass because of the trouble occurring with the gyrocompass in the early evening of December 14, so that the course change was from 047° to 050° magnetic. Nantucket Lightship was not sighted at about 3:00 A.M. as expected. The RDF was being operated and the officers were misled by erroneous signals on the RDF into believing that the ship was on course, in safe waters, and that Nantucket Lightship was dead ahead. The officers ignored a fathometer reading showing shallow water. The chief officer took star sights at dawn of December 15, but improperly calculated the fix based on these sights. The vessel went aground. From this version of the facts one could readily conclude that there were errors of navigation — the adherence to a course of 036° even after the dangerous nature of this course should have been obvious; the failure to heed the warning of the fathometer reading; the improper calculation of the fix; the failure to stop the vessel or reverse course in view of the uncertainty of the ship’s position after the failure to sight Nantucket Lightship. A further conclusion would be that a defective RDF contributed to the casualty. With regard to the gyro-compass, the officers’ testimony indicates a malfunction beginning the evening of December 14, but there would be a serious question about whether it contributed to the grounding, since, as far as the officers’ testimony showed, the difficulty was obvious and the ship immediately started steering by the magnetic compass. The problem is that this story is inaccurate in material respects. The evidence as a whole, including expert testimony at the trial, leads to the conclusion that the navigational errors, the mismanagement of the ship, and the deficiencies in equipment, were even greater in magnitude than indicated by the version summarized above. Certain specific conclusions are: (1) The course entries in the log referred to by the officers as “True” or gyrocompass courses were neither true courses nor gyrocompass courses. The only reasonably reliable record of compass courses steered is the record of the magnetic compass courses. (2) The testimony, about steering 040° true for a short time after Hatteras and then steering the course of 036° true for more than 40 hours, followed by a 4° change to the right the evening before the grounding, is wholly inaccurate. The actual courses steered by the ship ranged from 046° to 033° true and were even more dangerous than the ones testified to by the officers. (3) The helmsmen were steering by the magnetic compass the entire time after the turn at Hatteras. The steering gyro-compass repeater near the helm was either inoperative during the final voyage or had been adjusted in an improper manner which meant that it was not functioning as a compass. (4) The problem with the steering gyro repeater was in part responsible for the steering of the courses which led to the grounding. The probabilities are that at least some of the navigating officers relied on the flawed steering gyro repeater, failed to analyze the magnetic compass courses, and thus did not comprehend the actual courses the vessel was steering. (5) The problem with the steering gyro repeater antedated the final voyage and made the ship unseaworthy. The shipowner has not shown that it exercised due diligence to detect or correct the problem. (6) Although there are doubts about the testimony of the officers regarding the use of the RDF, the probabilities are that the officers did try to use the RDF to home in on Nantucket Lightship, and that this apparatus malfunctioned. The shipowner has not shown that it exercised due diligence in the maintenance of the RDF. (7) One concurring cause of the casualty, although not a major one, was the failure to have the December 1976 pilot chart on the ship. The November chart, used by the navigating officers, showed the current going offshore in the waters south of Nantucket Lightship, whereas the December chart showed the current going inshore. Before proceeding further it is necessary to discuss the concept of compass error. This does not refer to a defect or malfunction of a compass. Compass error refers to errors induced by normal causes, which can be easily compensated for when compass readings are made, and in some cases can be mechanically corrected. Both magnetic compasses and gyrocompasses are susceptible to certain kinds of errors. Compass errors are differences between the directions shown on the compass and the true directions on the earth’s surface. One type of error in the magnetic compass is “variation.” This refers to the fact that the magnetic poles of the earth are 'located in different places than the true poles. The variation error changes depending on the location of the compass on the earth’s surface. Another type of magnetic compass error is “deviation.” This refers to error induced by the ship’s metal. The deviation error changes depending on the heading of the ship. A gyrocompass may have certain types of errors, the details of which need not be discussed here. There are means of determining these errors so that they can be known and allowed for. In the case of a gyrocompass certain of these known errors can be mechanically corrected. All types of compass errors are referred to as being so many degrees east or west. The amount of variation error for a magnetic compass at any given location will be shown on the nautical chart for that area. Deviation error for a magnetic compass and any uncorrected gyrocompass error should be recorded on the ship. In making corrections from the compass reading to the true direction, westerly errors are subtracted from the compass reading, and easterly errors are added. The variation errors for the magnetic steering compass, following the turn at Hatteras, ranged from 8° W near Hatteras to 16° W in the location of the grounding. According to the ship’s deviation card for the magnetic steering compass, and according to the officers’ testimony, there was no deviation error for the headings which the ship was on after Hatteras. The question of gyrocompass error is the subject of some confusion. At several places in his testimony, the captain stated that the gyrocompass error was between 1° and 2° W. At one point the captain specified that the error was 1.5° W. The log contains a space for entering gyrocompass error whenever courses are recorded. There are only two notations for gyrocompass error during the final voyage: Dec. 8 7:00 A.M. 1.5° E Dec. 9 7:30 A.M. 1.5° E There are no entries in the log for gyrocompass error following the turn at Hatteras. The 1.5° amount of gyrocompass error recorded in the two log entries is consistent with the amount of error testified to by the captain. However, the direction of the error recorded in the log is east, whereas the captain testified that the error was west. No questions were asked at the trial about this discrepancy, and it remains unexplained. The captain testified that there was a book on the Argo Merchant, containing a record of the azimuth readings of the sun, by which compass errors were calculated. However, this book was not rescued from the ship. There is extensive testimony by the captain about correcting gyrocompass error. It was apparently possible to adjust the steering gyro repeater — the repeater by the helm — to eliminate the 1.5° east or west error. Such a correction would mean that the directions and courses shown on the steering gyro repeater would be true directions and courses. The precise meaning of the captain’s testimony about gyrocompass corrections is not entirely clear. One possible interpretation is that the 1.5° east or west error was corrected. Another possible interpretation is that an adjustment of an entirely different nature was made to the gyrocompass. The evidence about this other adjustment will be discussed later in this opinion. It should be noted that, according to the captain’s testimony, whatever correction or adjustment was made was carried out with the steering gyro repeater, rather than with the main gyrocompass. Table I below sets forth information which will be used in the subsequent discussion. Columns (1) and (2) show the “True” and magnetic course entries in the log from Cape Hatteras to the grounding. Column (3) shows the variation errors, derived from the navigation charts after estimating the positions of the ship at the various times. There is no column for deviation error, since the evidence is that there was no deviation in the magnetic steering compass on the headings steered after Cape Hatteras. Column (4), entitled “Actual True,” contains a list of courses calculated by taking the magnetic courses and correcting for the variation errors. In each case the variation error is subtracted, as is proper for a westerly error. For reasons to be explained, it is my finding that Column (4), rather than Column (1), represents the true courses actually steered by the Argo Merchant after the turn at Cape Hatteras. Table I (1) “True” Date & Time in Log (2) (3) (4) Actual Mag. Var. True Dec. 12 11:00 P.M. 040 054 8 W 046 Midnight 040 054 8 W 046 Dec. 13 4:00 A.M. 036 050 8 W 042 8:00 A.M. 036 050 9 W 041 Noon 036 050 10 W 040 4:00 P.M. 036 050 10 W 040 8:00 P.M. 036 050 11W 039 Midnight 036 050 11 W 039 Dec. 14 4:00 A.M. 036 047 12 W 035 8:00 A.M. 036 047 12 W 035 Noon 036 047 13 W 034 4:00 P.M. 036 047 13 W 034 8:00 P.M. 036 047 14 W 033 Midnight 036 050 15 W 035 Dec. 15 4:00 A.M. 036 050 15 W 035 6:00 A.M. 036 050 16 W 034 (1) Fallacy of “True” Course Entries in Log and Validity of Magnetic Entries It should be remembered that, during the final voyage of the Argo Merchant, the “Gyro” course column in the log was not filled out, the explanation being that the gyrocompass had been corrected to true so that the gyrocompass courses were entered in the column labelled “True.” Thus, if the evidence invalidates the “True” column as a record of true courses steered by the ship, it also destroys the theory testified to by the officers that the “True” column represents corrected gyrocompass courses. There is no evidence anywhere in the record that the “True” column in the log represents the uncorrected gyrocompass courses — /. e., 1.5° east or west of true. Thus, if the officers’ testimony about the “True” column representing corrected gyro courses is shown to be invalid, it follows that the “True” column is not a record of gyrocompass courses either corrected or uncorrected. An analysis of the “True” and magnetic course columns in the log shows that they cannot be reconciled in any proper navigational sense. If the “True” entries in the log are correct (as representing true or corrected gyro courses steered), then the magnetic compass entries are wrong. If the magnetic compass entries are correct, the “True” entries are wrong. It is a fact of mathematical certainty, recognized by all navigators, that the true course of a vessel is equal to the magnetic compass course adjusted for variation and deviation. Where there is no deviation, as in the present case on the headings steered after Cape Hatteras, and where there is westerly variation, the true course is the magnetic course minus the variation. If a vessel has both gyrocompass and magnetic compass, and is steering primarily by a properly corrected gyrocompass, the gyrocompass course will correspond to the true course, and the magnetic compass will differ from the gyro/true course by the amount of the variation (assuming no deviation). If the gyro/true course of the Argo Merchant was in fact 040° at 11:00 P.M. and midnight December 12, then the magnetic compass course would have had to be 048° (040° + 8°), not the 054° shown in the log. A quick analysis of each of the “True” entries in the log, as shown in Column (1) of Table I, will show that in no case would the magnetic compass course be what was shown in the log if it is assumed that the “True” entries in' the log were the gyro/true courses. This is shown by simply taking the sum of the “True” entry and the variation, and comparing it with the magnetic log entry. The gyrocompass is not, of course, affected by variation. If the vessel intends to maintain a steady true course, it can maintain a steady gyrocompass course, but the magnetic course will change as the variation changes. Referring to Table I — if the gyro/true course was in fact 036° from 4:00 A.M. through midnight December 13, then the magnetic courses would have been as shown in Table II below. Table II Gyro/True Var. Mag. 4:00 A.M. 036 8 W 044 8:00 A.M. 036 9 W 045 Noon 036 10 W 046 Midnight 036 11W 047 Thus if in fact the ship was maintaining a steady gyro/true course of 036°, it would have been impossible for the magnetic compass course ever to be 050°, as shown in the log, and it would have been further impossible for the magnetic course to be steady at 050° or on any course, since the variation was changing. If we change our assumption- — i. e., if it is assumed that the magnetic entries in the log were accurate, then the “True” entries could not be a correct statement of true courses steered or gyrocompass courses corrected to true. Column (4) of Table I shows courses which result from taking the magnetic entries and correcting them for variation. If the magnetic entries are correct, then it follows inevitably that the true courses actually steered by the vessel were the courses shown in Column (4). In no case do these courses correspond to the “True” entries in the log. What has been said indicates that one or the other of the course columns of the log— or perhaps both — must be rejected. I conclude that the magnetic course entries should be accepted as basically accurate, and that the “True” entries do not constitute a record of either true courses or gyrocompass courses. There is no evidence — or any claim by any party — of any malfunction in the magnetic steering compass. As to the magnetic course log entries, the testimony justifies no other conclusion than that they were accurate notations of the courses steered on the magnetic steering compass. On the other hand, there is ample reason to doubt the validity of the “True” column as representing true or corrected gyrocompass courses, as testified to by the officers. One circumstance is the departure from the form prescribed in the log and the failure to make entries in the “Gyro” column. Moreover, Thebes’ own navigation expert witness, Captain Richard Patterson, testified that, in his opinion, the course entries in the “True” column in the log did not reflect the true courses steered by the vessel, but were courses intended to be made good between various points. Patterson’s testimony was based upon an analysis of the log entries and courses of the Argo Merchant during its final voyage leading to the grounding, and also voyages in July and September 1976. A course steered is different from a course intended to be made good. The captain of a vessel may intend to make good a course of 010° between point A and point B. However, the captain may estimate that weather and current will push the vessel to the right. Thus, in order to make good the course of 010°, the vessel must compensate — i. e., must steer a course something to the left of the intended course. If we assume that the captain estimates that the effect of the weather and current will be 4° to the right, the captain will compensate by steering 4° to the left of 010° — or will steer a course of 006°. There is some question about whether Patterson’s testimony is precisely correct about the “True” courses in the log being courses intended to be made good between geographical points. However, Patterson’s testimony lends support to the conclusion that the “True” courses in the log are not a record of the true courses steered by the ship. It is obviously true that a proper log must have a record of courses steered. If the vessel is using both a gyrocompass and a magnetic compass, the courses on both compasses must be recorded. I find that during the final voyage of the Argo Merchant, there was no record of gyrocompass courses steered. The only record of compass courses was for the magnetic compass. As for true courses steered, whatever it is that the “True” column in the log reflects, it did not show such courses. These can only be derived by taking the magnetic courses and adjusting for variation, leading to the “Actual True” courses shown in Column (4) of Table I. (2) Effect of Courses Steered It follows from what has been said that the Argo Merchant was steering the magnetic compass courses shown in the log and in Column (3) of Table I. Regarding the contradictory testimony as to whether there was a course change of 3° to the left between midnight December 13 and 4:00 A.M. December 14, I find that there was such a change — from 050° to 047° magnetic. As already indicated, I find that the true courses actually steered were those shown in Column (4) of Table I. Of course, there was no record on the Argo Merchant of these true courses. However, a competent navigator could have readily calculated these true courses at any point. The testimony of the officers about steering 040° true for a few hours after Cape Hatteras and then 036° for a period of over 40 hours until a course change of 4° to the right the evening before the grounding — all this is inaccurate. The actual courses steered ranged from 046° to 033° true. If we examine the actual courses beginning at the time of the noon fix of December 14, they are all to the left, or inshore, of the 036° listed as the “True” course in the log. The actual course steered at noon whs 034°, and went as far left as 033° by 8:00 P.M. Even after the 3° course change to the right after 8:00 P.M., the actual courses steered were 035° and 034°. These courses were even more dangerous than the 036° listed in the log and testified to by the officers. The courses actually steered provide an obvious explanation as to why the vessel made good a course of 033° after noon December 14 and was off its intended course by 21 miles. (3) Use of Magnetic and Gyro Compasses If there was no record of gyrocompass courses steered, then the question arises as to whether the Argo Merchant’s gyrocompass was inoperative during this final voyage. The further question is whether the officers’ testimony should be entirely rejected as to steering primarily by gyrocompass until the alleged malfunction the evening of December 14. It should be noted at this point that the two helmsmen, Roach and Rivera, who were on duty during the 4:00 to 8:00 watches (and who were thus on duty during the final watch before the grounding) testified that, whenever the ship was being steered by hand, in contrast to steering by the automatic steering device, they steered by the magnetic compass. They testified that the same was true for all helmsmen on all watches. Both of these helmsmen testified that there was some use of the automatic steering device during the early part of the voyage after leaving Puerto La Cruz. Roach testified, however, that there was no automatic steering after the vessel was in the area of Florida. Rivera testified that automatic steering was suspended after Cape Hatteras. I conclude that the steering was by hand during the final voyage, at least commencing with the turn at Cape Hatteras, and that during this time the helmsmen steered by the magnetic steering compass. The gyrocompass on the Argo Merchant was manufactured by S. G. Brown Ltd. of It was a Brown Type Watford, England. A, manufactured in about 1957. It was installed on the Argo Merchant in 1971 as a factory overhauled model. Brown had stopped making the Type A in 1962. Spare parts were not manufactured after 1972, and in 1975 Brown ceased stocking spare parts. In August 1977, Brown wrote a shipping company other than Thebes stating that the Brown Type A gyrocompass was “totally obsolete.” There were frequent problems with the Argo Merchant gyrocompass going back at least into early 1975. The evidence at trial shows repairs on the following dates: April 23,1975 July 7,1976 December 18,1975 July 81,1976 January 3,1976 August 6,1976 June 28,1976 August 26-27,1976. These repairs were apparently successful in at least temporarily remedying the various malfunctions. However, at the completion of the work done on August 26-27, 1976, the technician made the following note: “IT IS IMPERATIVE THAT THIS GYRO COMPASS BE REPLACED. PARTS ARE NO LONGER READILY AVAILABLE FOR THIS COMPASS AND WE CONSIDER COMPASS TO BE BEYOND ITS NORMAL SERVICE LIFE. RELIABILITY AS IS NOW, VERY POOR.” This notation came to the personal attention of Thebes’ Marine Superintendent, J. Skarvelis, and also to the attention of Port Captain Vasilakis, who was traveling on the Argo Merchant from November 26, 1976 through the time of the grounding, ostensibly to exercise some kind of surveillance over the vessel. Neither Skarvelis nor Vasilakis took any steps to follow up on the August 1976 warning regarding the gyro-compass. As already described, the testimony of the officers is that, commencing at about 6:00 P.M. December 14, trouble with the gyrocompass, or at least with the steering repeater, developed in the form of swinging over an arc of about 12-14°. Helmsman Roach testified that what occurred at this time was a little more play in the steering wheel from wave action, but it was no problem for him, and he recalled no change in the steering procedure. This testimony of a highly credible witness such as Roach tends to throw doubt on whether there was the specific gyrocompass malfunction on the evening of December 14, as testified to by the officers. Nevertheless, there is still the question of whether there was a basic problem about the gyrocompass which had persisted through the entire voyage. Continental introduced extensive expert testimony at the trial in an attempt to prove that the gyro-compass was experiencing a cyclical error— or “wander” — and that this error was unknown to the vessel’s personnel and misled them. Although this evidence was helpful in some matters of detail, I cannot accept the theory of “wander” presented by Continental. However, the evidence as a whole does compel the conclusion that something abnormal and improper was occurring with the gyrocompass during the final voyage of the Argo Merchant and that whatever it was antedated the alleged trouble of 6:00 P.M. December 14. One possibility is that the gyrocompass, or at least the gyro repeaters, were not operating at all during the final voyage. There was not record of gyrocompass courses. The helmsmen were steering by the magnetic compass when steering by hand. The gyrocompass and repeaters had a history of trouble, and a technician had recently declared the reliability of the gyro-compass to be very poor. Also, the officers’ testimony shows that the gyro repeater on the forward bulkhead of the wheelhouse was probably inoperative, and that the repeater on the flying bridge was certainly inoperative. As to the steering repeater— the one in the wheelhouse by the helmsman — the officers’ testimony is that this was operating until 6:00 P.M. December 14. However, under all the circumstances, there is clearly a question as to the veracity of this testimony. Another possibility, arising from the evidence, is that the steering gyro repeater had been adjusted in an improper manner so that it was not functioning as a. normal compass. This emerges from an analysis of the captain’s testimony. Before discussing the evidence about an improper adjustment, it should be recalled (see p. 447 supra) that one possible interpretation of the captain’s testimony is that he adjusted the steering repeater for the 1.5° east or west error. Such a correction would have been normal and proper, so that the degree readings on the steering repeater would correspond with the true directions on the earth’s surface. To be more specific, the “degree mark” on the repeater would show the true course being steered by the vessel. However, certain portions of the captain’s testimony justify a conclusion that an adjustment of an entirely different nature was made by the captain to the steering repeater — i. e., that he periodically adjusted the repeater so that the degree mark showed, not the course actually steered, but the course the captain hoped to make good after the captain’s estimate of current and weather. Thus, if the captain intended to make good a course of 036° and estimated that the current and weather were pushing the vessel to the right, the vessel would steer something to the left of 036° to compensate. The normal procedure, in using a properly corrected gyrocompass, would be to direct the helmsman to steer a course, for instance, of 033° or 034° on the gyro. However, the captain’s testimony, at least at some points, is that, instead of following this normal procedure, the captain manipulated the steering gyro repeater. Thus, in the above example, he would adjust the steering gyro repeater, by turning the repeater card to the left or counter-clockwise, so that the vessel would appear to be steering the 036° — i, e., the course intended to be made good — while it was actually steering the 033° or 034°. This is what the captain testified was done on December 14, when the “True” course entries in the log were 036°, whereas the courses actually steered were to the left of 036° (Thebes Ex. 134 D pp. 125-26). Column (4) of Table I shows that the courses actually steered during this time ranged from 033° to 035°, and were never 036°. The testimony about adjusting the steering repeater for courses hoped to be made good was not developed as much as one would wish. Thus the record does not contain any precise and complete statement as to the times when the captain made these adjustments, how many degrees each adjustment was, or how each adjustment was arrived at, although the captain did testify at one point that he adjusted the repeater six times a day, which would accord with the number of regular course entries in the log (Thebes Ex. 134 I p. 766). However, the general testimony about making this kind of adjustment exists, and can hardly be ignored. Thebes’ navigation expert, Captain Patterson, interpreted the captain’s testimony to mean that he adjusted the steering gyro repeater for courses hoped to be made good (SM pp. 1774, 1778, 1781). Captain Patterson further testified (SM p. 1774) that this practice was “. . . about the greatest violation of good common sense seamanship I have ever heard of. I cannot understand a master of a ship adjusting a gyro, correcting it periodically to equal what he wants to make that ship steer. I have never heard anything like that in fifty years at sea.” Again, Captain Patterson testified (SM p. 1790) that: “. . . they were doing a very, very hazardous operation of changing the gyro every time they felt like it, to make it show up something that was not following the master gyro.” If the steering gyro repeater on the Argo Merchant was in fact adjusted as here described, this meant that the repeater was not functioning as a compass — i. e., it was not serving the purpose of showing the directions on the surface of the earth. The repeater was instead serving as a kind of pointer, set according to some sort of estimate made by the captain. There is no evidence that the captain made any record of these gyrocompass adjustments — the times when made or the amounts of the adjustments. There is no evidence that the captain had any procedure of informing the other navigating officers of the details of the adjustments. To recapitulate: The Argo Merchant had a gyrocompass and magnetic compasses. The logbook recorded magnetic compass courses, and the helmsmen testified that they steered by the magnetic compass. The gyrocompass had a history of failures and had been declared unreliable by a technician. There was no record of actual gyro-compass courses in the log. Yet the captain and the other navigating officers testified that the vessel steered by the gyrocompass until shortly before the grounding. The captain testified that he made some kind of adjustments to the steering gyro repeater to show courses hoped to be made good, rather than courses actually steered. An experienced navigation expert testified at the trial that he has never heard of such a practice in his fifty years at sea. The officers testified that the steering gyro repeater started fluctuating over an arc of 14° beginning at 6:00 P.M. the evening before the grounding. The one helmsman who was asked about the latter subject could recall no such incident. In view of the contradictions and anomalies in this record, the question naturally arises as to whether or not there has been a fabrication by the officers, or some of them, intended to conceal a more serious state of facts than negligence. The possibility suggests itself that the shipowner, through the agency of the captain and perhaps the other officers, intentionally grounded the vessel, or at least took an intentional shortcut to the shoal side of Nantucket Lightship. However, no party to the present litigation asserts intentional grounding or short-cutting, and, although the matter is hardly free from doubt, the probabilities tend away from any such conclusions. For one thing, certain of the deficient navigational practices occurred prior to the fatal lap from Cape Hatteras to the grounding. In other words, the indications are that the captain and officers of the Argo Merchant had on prior occasions sailed with a malfunctioning or improperly adjusted gyro-compass, steered imprecise courses, and kept irregular course records in the log, although these circumstances had not previously resulted in a serious casualty. It is necessary to return to the question of the precise nature of the problem with the gyrocompass. In my view, there are two possibilities reasonably suggested by the evidence — first, that the gyrocompass, or at least the gyro repeaters, were not operating during the final voyage; second, that the steering gyro repeater was being adjusted in an improper manner so that it was not functioning as a normal compass. Although there are numerous uncertainties presented by the evidence, I find that the probabilities favor a finding that the gyrocompass and the steering repeater were operating during the final voyage, but that the steering repeater was being improperly adjusted by the captain, on the basis of estimates of factors including weather and current. Thus the steering repeater was not showing the course actually steered, but was showing something else such as the course hoped to be made good. I further find, again on the basis of the probabilities presented by somewhat doubtful evidence, that the steering gyro repeater experienced a mechanical malfunction commencing about 6:00 P.M. December 14 in the form of rapid fluctuatio