FRIENDLY, Circuit Judge (sitting by designation).

Farrell Lines, Incorporated, owner of a pier at the foot of 33d Street in Brooklyn, N.Y., fronting on Gowanus Bay, brought this libel against the M/V Birkenstein and the tug Pauline L. Moran in rem and North German Lloyd, a West German corporation, hereafter NGL, owner and operator of the Birkenstein, and Tug Agnes A. Moran, Inc., and Moran Towing and Transportation Company, Inc., New York corporations, the former being the owner and the latter the operator of the Pauline L. Moran, in personam. The libel sought recovery for damages caused to the pier by the Birkenstein while the latter was being maneuvered by a Moran pilot with the assistance of the Pauline during the afternoon of February 23, 1961. The Birkenstein and North German Lloyd denied fault but also asserted that if there was any, it lay at the door of the Moran defendants whom they impleaded. As might be expected, the latter claimed innocence on their own part and fault on that of the Birkenstein. The trial was expedited by the cooperation of the proctors, and their briefs have been pointed and helpful.

At the opening of the trial proctors for the three interests announced a settlement whereby Farrell Lines released all the defendants in consideration of $42,500, paid by NGL without prejudice to the Court's ultimate determination of its rights against the Moran defendants; Moran agreed to the reasonableness of the amount. Since the Birkenstein suffered no physical damage, the sole issue became where the cost of this settlement and of another compromise reached by NGL, relating to a personal injury claimant on the pier, should ultimately fall.

There is less disagreement as to the facts than is usual in a collision case — probably due to the circumstance that only one flotilla was in motion. At this point I make the following findings; I shall make additional ones in the course of the subsequent discussion:

1. The Birkenstein is a motor vessel of 5,798 gross tons, 3,350 net tons, 498'1" in length, and 60'4" in extreme breadth. She has a single righthanded screw. Her slowest speed is 5.3 knots. She is equipped with a Raytheon 10-centimeter Model 1402, 16-inch scope, radar. Her draft on the day in question was 15'6" forward and 17'6" aft.

2. The Birkenstein is one of a number of freighters operated by NGL into New York harbor. She and the other NGL freighters regularly dock at Prospect Terminal (also called Continental Piers) near the upper end of Gowanus Creek in Brooklyn.

3. The entrance to Gowanus Creek is through Gowanus Bay, which lies northeast of the Bay Ridge Channel. Gowanus Bay is some 750 yards long from 35th Street to 27th Street; Gowanus Creek, from 27th Street to the Prospect Terminal, is some 840 yards long. Between the 35th St. pier and Erie Basin, Gowanus Bay is about 400 yards wide; at the 33d St. pier it narrows to about 300 yards. Gowanus Creek is about 100 yards wide.

4. Moran Towing and Transportation Co., Inc. and NGL were parties to a Towing Contract. Under this contract, a printed form, Moran agreed "to furnish tugs for, and to attend to, all the towage requirements at the Port of New York" of all NGL vessels, at stipulated rates per tug for base periods and for additional time. The contract contained the following clause:

"PILOTAGE — When the captain of any tug furnished to or engaged in the service of assisting a vessel which is making use of her own propelling power, goes on board such vessel, or any other licensed pilot goes on board such vessel, it is understood and agreed that such tug captain or licensed pilot becomes the servant of the owner of the vessel assisted in respect to the giving of orders to any of the tugs furnished to or engaged in the assisting service, and in respect to the handling of such vessel, and neither those furnishing the tugs and/or pilot, nor the tugs, their owners, agents, charterers, operators or managers shall be liable for any damage resulting therefrom."

5. The practice of the parties was that NGL's New York agent, United States Navigation Co., would advise Moran of the name of an arriving vessel, her estimated arrival time, and the pier to which she was to proceed. NGL did not specify the number or characteristics of the tugs to be furnished. The practice was for Moran to furnish at least two tugs, sometimes three. There was no general practice that the tugs so furnished would be equipped with radar. Some 71% of Moran's New York harbor tugs have radar; all have radio-telephone. The radar generally used on Moran tugs is the RMCA Model CR 103 3-centimeter set.

6. On February 22, 1961, NGL's New York agent advised Moran that the Birkenstein was expected at Ambrose Light at 13:00 on the following day, would proceed to Quarantine, and would then dock at Pier 3, Prospect Terminal.

7. The Birkenstein anchored at Quarantine on February 23, 1961, at 14:50 and departed for the inner harbor at 16:10. There is some variation in the estimates of visibility at the time of her departure from Quarantine; it seems likely that this was approximately one mile — probably less rather than more.

8. As the Birkenstein proceeded up the harbor, visibility worsened. Chief Officer Weisskopf testified that as she passed Anchorage 21-B at the entrance to Bay Ridge Channel, visibility was 400 meters. By the time the Birkenstein came opposite Bush Terminal, visibility had so decreased that the officers on her bridge could hardly see the bow 250' off. During the passage from Quarantine the Birkenstein was using her radar; this was functioning perfectly.

9. About 15:30 Moran's despatcher instructed the Pauline L. Moran that she was to meet the Birkenstein if the vessel left Quarantine. About 17:00 the despatcher advised that he thought the Birkenstein had left Quarantine and that the Pauline was to go out and get her and bring her into Prospect Terminal. He also advised that the James Moran was clearing the berth and that when the James was through with this, it would come out and help the Pauline dock the Birkenstein.

10. The Pauline L. Moran is a Diesel-electric powered tug with 1200 horsepower. She is 100.2' long; her gross tonnage is 211. She had no radar, as was apparent by inspection from the bridge of the Birkenstein. Her crew consisted of captain Cray, chief officer Smith, a deckhand, a cook, an engineer, and an oiler.

11. The Pauline had proceeded to a position off Pier 4, Bush Terminal Docks, when she heard a whistle which she thought came from the Birkenstein. The Pauline went out to find the vessel and did, first seeing her at a distance of 150' to 200'. The Pauline landed alongside to starboard, put a line up and put Cray aboard. This was at 17:16. Before leaving, Cray ordered Smith to put a line up in the first chock forward of the Birkenstein's bridge and to await further orders. The Pauline's engines were stopped. Smith took over command of

the Pauline and the deckhand manned the lines. No one was posted on the bow.

12. Cray went immediately to the bridge of the Birkenstein. The others there were Captain Conrad, the Sandy Hook pilot, chief officer Weisskopf, officers' assistant Koepke at the engine telegraph, second officer Stindt at the radar, a helmsman, and a bridge lookout. Third officer Goldenstein, carpenter Meyer and another lookout were on the bow.

13. There are some differences, although not real contradictions, in the testimony as to the conversation when Cray reached the bridge. Conrad recalled only telling Cray the ship's draft, the range at which the radar was set, that the engines were stopped, "and where we intended to go", and asking which pier was ready. Cray's recollection was that he and Conrad "discussed the weather conditions, advisability of proceeding further, or any such action as he would care to take I would be glad to assist him in doing so"; Cray said he "only outlined the possibilities". He "made no recommendation" since "any action would have been hazardous". Cray testified that the master made the decision to proceed. Although Cray's version may be somewhat on the pat side, I see no sufficient reason to disbelieve him; indeed, the master's own testimony suggests that he then had no thought of doing anything other than to proceed to the pier unless someone strongly advised against it.

14. Cray's first engine order on the Birkenstein, at 17:20, was "dead slow ahead." His course was along Bay Ridge Channel toward the entrance to Gowanus Bay and then into the Bay; entering the Bay would require a change from 35°-40° to 55°-65°.

15. Shortly before Cray boarded, Stindt had put the Birkenstein's radar on the 1-mile range. For a time the radar continued to work well; Stindt made reports to the master and to Cray. But "the more we came to Gowanus Channel, the more the narrows came together * * *. It was everything filled, we couldn't see the entrance." Such a loss of useful functioning by a Model 1402 radar when a vessel is entering a narrow body of water with metal buildings on either side is not uncommon; this is known as the side-lobing effect. To attempt to overcome that effect it is desirable to reduce the volume or "master-gain" and also to turn the STC gain control counterclockwise. Admittedly Stindt did the former. But he testified on deposition that he "increased the S.T.C. gain", and it is disputed whether by this he meant that he turned the control counter-clockwise, as he should have done, or clockwise. If it were necessary to resolve this, I would find, reading the entire testimony in context, that he meant he had acted to decrease sensitivity of the receiver and thus had turned the control counter-clockwise. Such adjustments of the controls are not uniformly successful even when correctly performed; Stindt's did not succeed.

16. Stindt testified on direct examination that he reported the loss of a useful radar picture to the master and to Cray "at once". However, I read his cross-examination to say that the report was made only after his efforts to improve the picture by adjusting the controls had failed, a span of perhaps two minutes.

17. As soon as Stindt's report was received, the engines were ordered stopped. This was at 17:24. Cray and Conrad looked at the radar screen; both found that the channel entrance could not be distinguished. Conrad refused to let the Birkenstein go further into Gowanus and wanted to drop anchor at the nearest safe place; Cray said that there was too much congestion to anchor where the ship lay and that she would have to go further out. Conrad placed this conversation immediately before the failure of the radar; I find it more consistent with the known facts that it occurred immediately thereafter.

18. There is a dispute as to what, if any, helm orders were given. Conrad testified that Cray, believing the Birkenstein must first go to starboard in order to turn, ordered a starboard rudder, and that the helmsman answered, announcing every 10° of the turn, until a time came when Cray ordered the helm to port. Weisskopf corroborated this to the extent of testifying that he heard the helmsman calling off the degrees — "70, 80 degrees, and 90 degrees", that he went into the wheelhouse to see what was going on, and that he protested to the captain. Cray said his last order to the helm, prior to the time with which we are now dealing, was to come from 50° or 55° to starboard to 65°. In view of conclusions that I have reached on other aspects of the case, it is unnecessary for me to determine whether, as NGL alleges, Cray wrongly ordered the rudder hard to starboard, whether, as Moran alleges, the helmsman wrongly took such action on his own, possibly misunderstanding Cray — either of which would constitute a fault attributable to the Birkenstein, or whether the rudder was not put so far to starboard at all. What is clear is that the vessel's head dropped off to starboard, a phenomenon explicable by her loss of steerageway.

19. Immediately after the Birkenstein's engines were stopped, Cray ran over to the starboard wing of the bridge, ordered Smith to shift the Pauline forward and make fast, and then directed the Pauline to "come ahead half a speed, then full speed ahead".

20. About this time third mate Goldenstein called from the bow of the Birkenstein, "There is something approaching the ship. It looks like a roof." The engines were ordered half astern, at 17:27, and slow astern at 17:27 1/2, and were stopped at 17:28. Meanwhile, at 17:27 the collision with the pier occurred. Cray testified that at some time he had ordered the starboard anchor dropped — whether such an order was given and, if so, whether it was before or after collision, cannot and, on the view I take of the case, need not be determined. If it was given, it was not transmitted. The carpenter, on the bow of the Birkenstein, acting on his own, let go of the starboard anchor. This fell on the pier.

As noted at the outset, the Birkenstein and her owner and operator, proclaiming their own innocence, level numerous claims of fault against the Pauline and her owner and operator, whereas the latter respond in kind. The parties are also in controversy whether if both sets of interests were faulted, the share of the Moran interests would be 75% (as NGL contends), 33 1/3% (as Moran contends), or 50% (for which neither contends). The latter issue, interesting as it is, does not arise unless Moran is found to be liable to the Birkenstein and NGL; if not, the loss, paid by the Birkenstein, doubtless because of the "presumption" of negligence when a moving ship collides with a shore installation, Wilmington Ry. Bridge Co. v. Franco-Ottoman Shipping Co., 259 F. 166, 168 (4 Cir. 1919), must lie where it has fallen. Although the discussion uses terms such as "fault" and "negligence" as to Moran, these are convenient shorthand for "breach of warranty of good workmanship" — the essence of NGL's claim, as its brief makes clear.

NGL levels the following specifications of fault against Moran, all of which are alleged to fall outside the pilotage clause, the validity of which NGL does not dispute:

"1. Moran furnished only one tug to assist the Birkenstein into Gowanus Bay, and failed to call for the assistance of a helper tug when its need was apparent.

"2. Moran improperly positioned the Pauline L. Moran alongside the Birkenstein.

"3. Moran failed to recommend to the master of the Birkenstein that the passage of Gowanus Bay and Creek should not be attempted under the circumstances prevailing.

"4. Moran was at fault, and the tug Pauline L. Moran unseaworthy, because she lacked radar.

"5. The tug Pauline L. Moran had no lookout."

Considering items 1 and 4 to be the most serious, I shall first deal with items 2, 3 and 5.

Improper positioning of the Pauline.

NGL's claim is that, before boarding the Birkenstein, Cray should have ordered the Pauline to the Birkenstein's starboard bow. Assuming without deciding that an erroneous order given the tug before boarding the assisted vessel is outside the pilotage clause, the claim assumes a degree of prescience that Cray was not required to attain. Of course, if he had known before boarding that the first service to be required of the tug would be to turn the Birkenstein sharply to port, he should have placed the Pauline on the starboard bow, where she could best do this — as, indeed, he conceded. However, he could not know at that time what course of action would be decided. The Pauline had come amidships to put Cray aboard. In that position she could be used to push the Birkenstein away from the piers, as well as to kill headway or for backing, and she was in the best possible position for receiving orders or for communicating with the Moran office and other Moran tugs. Even if there was undue delay by Cray in later ordering the Pauline to the starboard bow, which I do not find, that would be within the pilotage clause.

Failure to recommend against proceeding.

It does seem to a layman that planning to sail the Birkenstein nearly a mile up a narrow bay and channel, with visibility only some 200', was a hazardous business. This view is reinforced by the frank admission of NGL's radar expert, in answer to a question from the Court, that "the average officer would realize that with this type radar on that ship that he is pushing it to the very limit of his [sic] capabilities in that spot, and if he is able to use it, he is fortunate." On the other hand, Moran cannot be blamed for anything done before it became involved, and by that time the Birkenstein was in a position where any course of action had its hazards. Staying where she was or a return to Anchorage 21B would have involved the danger of collision with other vessels, whose presence in the area was attested by the blowing of fog signals. If she anchored where she was, there was no telling when the fog would lift. Proceeding to Anchorage 21B would have required a 180° turn or a course passing Erie Basin, either of which involved distances greater than to the Prospect Pier although neither entailed the navigational hazards of a narrow channel and the danger of loss of useful pictures on the radar screen. While, particularly with the inestimable benefit of hindsight, going to the Anchorage would seem to have been the wiser course, its superiority has not been shown to be so evident that Cray can be found to have been negligent in not making a positive recommendation to the master, who was quite familiar with the topography of Gowanus Bay and Creek and the traffic likely to be encountered there and had better knowledge of the reliability of the ship's radar than Cray.

Moreover, if Cray was at fault as claimed, the fault was within the pilotage clause. NGL does not dispute this would have been so if visibility had been, say, 500' when Cray boarded the Birkenstein but had suddenly decreased to 200' while he was getting to the bridge and talking things over with the master; its claim is that under the circumstances here he should have made up his mind before going "on board such vessel." This places the duty of decision too early; there was no reason why Cray should have formulated a fixed view until he had all the facts. Moreover, his failure to formulate a view was without consequence; what mattered was what he did when he boarded the Birkenstein. If Cray then neglected to make sound recommendations, it was a failure, after he went on board, "in respect to the handling of such vessel," and thus within the pilotage clause.

Lack of a lookout by the Pauline.

NGL's contention as to lack of a lookout on the Pauline fails to take proper account of the varying capacities in which tugs serve. When a tug is operating independently, as in The Supply No. 4, 109 F.2d 101 (2 Cir. 1940), and Moran Towing Transp. Co. v. United States, 80 F. Supp. 623, 629 (S.D.N.Y. 1948), she may well have the same duty to maintain a lookout as any other vessel. When she is in charge of a flotilla of barges, she is bound to maintain a lookout where he can look out, whether on the tug or on one or more of the barges if they are obscuring the view; that is the lesson of The Madison, 250 F. 850 (2 Cir. 1918); Martin Marine Transp. Co. v. Jakobson Peterson, Inc., 135 F.2d 325 (2 Cir. 1943); and Gulf Oil Corp. v. The Socony No. 16, 162 F.2d 869, 870 (2 Cir. 1947). But NGL has cited no decision that a tug which is merely assisting a vessel in docking must maintain a lookout; in such cases the responsibility would seem to lie on the assisted vessel, which in this instance is the leader of the flotilla just as the tug is when it has barges in tow. With Cray on the Birkenstein, Smith in the wheelhouse and the deckhand at the lines, none of the above-deck crew of the Pauline was available to act as a lookout. Yet there is no contention, as there was in Moran Towing Transp. Co. v. United States, supra, see 46 U.S.C.A. § 223, that the Pauline was undermanned; Smith's testimony that the Pauline's was the normal crew for a tug was uncontradicted. Hence I do not reach the issue whether, if the absence of a lookout on the Pauline had been a fault, she could meet the heavy burden of negating the possibility of a causal relation, The Pennsylvania, 86 U.S. (19 Wall.) 125, 22 L.Ed. 148 (1874).

Failure to furnish a second tug.

This contention has several facets; although certain of them seem at first to be plausible, analysis shows each to have a fatal flaw.

As previously indicated, it was undisputed that the Birkenstein was never docked with less than two tugs and sometimes with three. It was equally undisputed that the tugs did not normally reach the Birkenstein at the same time. The evidence is not so clear as to where the second tug would join. Captain Conrad testified that the second tug usually arrived "near the entrance of Gowanus Channel * * * sometimes inside, sometimes outside," and that he inquired as to its non-arrival shortly before the Birkenstein's radar picture became useless. Weisskopf placed the usual rendezvous "between Bay Ridge and the entrance to the Gowanus Channel". Cray testified that normally the second tug would join "in the mouth of Gowanus Channel" — outside if the second tug had been working outside Gowanus, inside if she had been working in the channel. There is also some dispute what "entrance" and "mouth" mean. In answer to a question from the Court, Cray identified the "mouth" as an elliptical area whose horizontal axis would run from Erie Basin to the 33d Street pier. Conrad placed the "entrance" off the 35th Street pier. The uncertainty may well be due to the fact that the second tug was not thought of primarily as an aid to navigation but rather as an aid to docking. Cray testified "two tugs are used to dock the Birkenstein because of the slip and piers that they use. It is a very congested area. The piers are narrow; there are always lots of barges; quite a bit of traffic in Gowanus Canal, and it expedites the job to have a second tug available."

NGL contends, relying on Publicker Industries, Inc. v. Tugboat Neptune Co., 171 F.2d 48, 49 (3 Cir. 1948); American South African Line, Inc. v. Sheridan Company, Inc., 1936 A.M.C. 287 (E.D.Pa. 1935); Gypsum Queen v. Tug Peerless, 1953 A.M.C. 2071 (E.D.Va. 1953); and People of State of California v. The Jules Fribourg, 140 F. Supp. 333 (N.D.Calif. 1956), that undertaking a docking operation without providing the number of tugs needed for reasonable safety in the light of foreseeable hazards is a breach of the warranty of workmanlike operation implied in the towage contract and falls outside the pilotage clause. Moran answers that there was no evidence that under normal conditions two tugs were needed to assist the Birkenstein in navigating a bay 300 yards wide, or even a channel which, at its narrowest point, was five times as wide as her breadth. I accept that answer so far it it goes — under normal conditions, the need for a second tug, as Cray testified, was rather for the docking at the pier, from which the Birkenstein was still far away, and certainly not for the navigation of the Bay where the accident occurred.

NGL responds that, whatever the nature of the need, there was a custom or practice with respect to a second tug, which had become in effect a term of the contract and on which it justifiably relied. However, NGL has failed to show, by a preponderance of evidence, a custom that the second tug would rendezvous at least as far to the southwest as the area opposite the 35th St. pier; I accept Cray's testimony, which, indeed, is rather corroborated by Conrad's, that the practice was not that well defined. Moreover, the Birkenstein had not yet reached 35th St. at 17:24 when her engines were stopped. Stindt testified she was still below 35th St. at the time, and this checks with Conrad's estimate of the place where Cray boarded and the undisputed facts of speed and time. Neither can Moran be held liable, on a similar theory, on the sole basis of the James' failure to appear during the three minutes between the stopping of the engines and the collision; the Birkenstein was still below the northeastern rim of the "ellipse" within which, on Cray's testimony, the rendezvous commonly took place.

Although I have accepted Moran's position that its obligation of workmanlike performance did not require it to furnish a second tug simply for navigation of Gowanus Bay under normal conditions, conditions on February 23 were not normal, as Moran knew. Before his second conversation with the Pauline, about 17:00, Moran's despatcher was aware that visibility at Quarantine was only 1/10th of a mile and Cray, on the Pauline, knew just how bad the weather was in Gowanus Bay. A second tug would thus have surely been a desirable aid to the navigation of the Birkenstein through the Bay under the circumstances. There is thus force in another facet of NGL's claim, namely, that under the particular circumstances of February 23, Moran ought to have provided a second tug at a time sufficiently preceding the accident to have prevented it, even though this involved greater effort than was required under normal conditions. However, particularly because of the pilotage clause, it is not enough to say that "Moran" should have done something different than it did; it is necessary to determine who in "Moran" failed and when.

One possible basis of liability would be that Moran's despatcher should have instructed Cray not to undertake the navigation of Gowanus Bay under the conditions prevailing without having the James Moran actually at hand or at least being certain that she or some other tug, in addition to the Pauline, would be available by the time the Birkenstein reached 35th Street, where the Bay narrows and radar difficulties might be anticipated. It is now easy to see that things would have been better if the despatcher had urged greater expedition by the James or, if that was not possible, had summoned another tug, which was not engaged in clearing the slip as the James had to be, or had ordered Cray not to go beyond a certain point without a second tug although, as previously indicated, this last course, too, had its dangers. However, the Towing Contract does not indicate by its terms that Moran undertook so heavy an obligation and the record is barren of evidence of any custom or practice that the despatcher, or Moran's head office generally, assumed, or was required by express or implied warranty to assume, responsibility for such detailed supervision of docking operations of which supposedly competent masters and tug captains were in charge.

A more plausible variation is that Cray, before leaving the Pauline, ought to have bound himself not to proceed, or not to proceed beyond a certain point, without a second tug, or at least ought to have done something more to get the James on the scene than merely to rely on the despatcher's report, made around 17:00, that the James would come down when she had finished clearing the slip. I have little difficulty in concluding that Cray was not negligent in failing to bind himself as to his course of action before boarding the Birkenstein, even if such a non-feasance be assumed to fall outside the pilotage clause. Nothing would have been gained by self-preclusion at so early a stage; it was better to postpone decision until Cray could talk things over with the master and others on board — moreover, as indicated above, what mattered was what Cray did, not what he had decided or failed to decide to do. The close question relates to his failure to do something about the James while the radio-telephone on the Pauline was so readily at hand. But here again one must steel one's self against the temptation to substitute hindsight "in the peace of a quiet chamber", Hellenic Lines, Ltd. v. Brown Williamson Tobacco Corp., 277 F.2d 9, 13 (4 Cir.), cert. denied, 364 U.S. 879, 81 S.Ct. 168, 5 L.Ed.2d 102 (1960), for the prospect facing a tug captain in Bay Ridge Channel on a foggy winter afternoon. Cray had no reason to doubt that the James would appear where second tugs normally had in the past; when he left the Pauline, he did not know what course of action the Birkenstein would take; conditions, although bad, were not so bad that the second tug was needed immediately and in all events; and the Pauline's radio-telephone could still be used, with only a few seconds' delay, by oral communication between the Birkenstein's bridge and Smith. Doubtless Cray thought the best single thing to do was to get to the Birkenstein's bridge just as fast as he could; apart from the problem of causation discussed below, I cannot say that was so wrong a decision as to constitute a breach of Moran's warranty of good workmanship.

There remains the contention that, once Cray had boarded the Birkenstein, he should still have made certain of the availability of the James before proceeding up Gowanus Bay or, not having done that, should have stopped the Birkenstein when a second tug did not appear. The alternative contention is rather clearly lacking in force. In fact, the Birkenstein was stopped at the channel mouth; in any event negligent failure to stop her sooner would be within the pilotage clause. The primary contention has greater merit. We are dealing now with a time when Cray knew that the navigation of the channel was to be attempted, and he ought to have known the presence of a second tug would help. Still the case is not quite like People of State of California v. The Jules Fribourg, where the pilot, misjudging the strength of the tide, wholly failed to order a second tug which he admitted he would have ordered if he had judged aright. The James was not that essential and Cray thought she was coming in any event. Moreover, there is no proof that a telephone call would have gotten the James or any other tug to the scene soon enough to help; the record is remarkably vague with respect to the orders to and the movements of the James. Hence, in order to meet this problem of causation, NGL's argument has to be not merely that Cray should have called for help but that he should not have allowed the Birkenstein to proceed beyond a certain point until he could be certain the James or some other tug would meet her outside the channel mouth. But even if we should assume arguendo that Cray's failure to do something more about getting a second tug was outside the pilotage clause even though the failure occurred while he was on the Birkenstein, the accident was caused not by that failure but by his decision to allow the Birkenstein to go ahead without being sure another tug would meet her outside the mouth, and this constitutes action "in respect to the handling of such vessel" within the pilotage clause. If the alternative ground of decision in the Fribourg case, 140 F. Supp. at 340, should be to the contrary, and I am not at all sure that it is, I would not be able to agree with it.

It may be said that the foregoing is equivalent to holding that although Moran's conduct may have been negligent under the circumstances when viewed with ordinary vision, it ceases to be so when examined under a microscope. The answer is that Moran's over-all obligations under the Towing Contract were to assist the docking, not to manage it; that, accordingly, if something more should have been done than was done, the failure must have been by some person; and that when I examine the behavior of the various persons, as the pilotage clause requires, I find that NGL's case against Moran hinges on a fault by Cray — proceeding without assurance of the second tug — that comes within it.

Lack of radar on the Pauline.

NGL's final claim is that the Pauline was unfit for her assignment because she was not equipped with radar. Much of the expert testimony was directed to whether, as claimed by NGL, there is an essential difference in types of radar equipment — with the Raytheon Model 1402 10-centimeter set superior for work at sea and the RMCA Model CR 103 3-centimeter set superior at close range, so that the radar of an ocean going vessel requires supplementation by a tug's. I find this particular contention not to have been established.

However, that does not end the matter. Just as two eyes are better than one, so are two radars, particularly in view of the evidence that two radars of the same model, both operated properly, might respond differently under the same conditions. Of course, this is necessarily subject to the qualification that the second radar would not be working under conditions that would render it useless. Moran contended exactly such would be the case with a radar in proximity to the side of a steel ship. Although the evidence showed that this would be a seriously impairing factor, it did not go to the point of excluding all possibility of usefulness in radar on a tug assisting in the docking of a ship.

Although I thus cannot find that a radar on the Pauline might not have helped, the undoubted limitation on the usefulness of a tug's radar under such conditions greatly impairs the case that there is a duty to have one. NGL relies on Judge Learned Hand's famous opinion in The T.J. Hooper, 60 F.2d 737, 740 (2 Cir.), cert. denied, Eastern Transportation Co. v. Northern Barge Corp., 287 U.S. 662, 53 S.Ct. 220, 77 L.Ed. 571 (1932), holding tugs liable to their tow for a failure to carry radio receiving sets, although "there was no custom at all as to receiving sets; some had them, some did not; the most that can be urged is that they had not yet become general," and on Judge Medina's prophecy, in Afran Transport Co. v. The Bergechief, 274 F.2d 469, 474 (2 Cir. 1960), "that a rule requiring radar, subject to some limitations and qualifications, will sooner or later be formulated." If this were a case in which the Pauline on a journey of her own down Gowanus Bay had damaged another ship or a shore installation, or, as in The T.J. Hooper, had lost barges she had in tow, and lack of radar was proved to have been a cause, she might well be faulted on that score; such a case would indeed remind of The T.J. Hooper, in which Judge Hand imposed a standard of conduct beyond any practice that had become truly general where it was evident that the practice ought to become so. See also Mr. Justice Holmes in Texas Pacific Ry. Co. v. Behymer, 189 U.S. 468, 470, 23 S.Ct. 622, 47 L.Ed. 905 (1903). However, in such cases, the tug's radar would have been the only eye available. A duty to supply an eye does not inevitably carry with it a duty to supply what would here be a second one, especially when the second would have only limited usefulness and would be needed only in the abnormal event that the dominant vessel got into a position where her own eye became useless and the second eye might also be. Hence any obligation to equip the Pauline with radar did not attach to her in the role of an assisting tug; if Moran preferred to avoid the added cost of radar for a vessel which was largely engaged in that activity and take its chances on liability for lack of one when she was not, that was a choice Moran was free to make.

It is true that, if the Pauline had been carrying a radar for purposes of navigation when operating independently or towing barges, the radar would have been on hand on the fateful afternoon. But that is not decisive — for two reasons, either sufficient. One is the serious doubt whether a radar on the Pauline would have served any really useful purpose unless another crew member had been on hand to scan it. Here, as on the issue as to a lookout, account must be taken of the limited personnel complement of a harbor tug and of the reduction inevitably effected in that when the tug master boards the assisted vessel — a factor also not without relevance to the question of duty discussed above. To be sure the evidence does not exclude the possibility that a radar could be placed where the chief officer could view it at times when he was awaiting an order, but NGL had the burden of showing that this was a feasible procedure and did not sustain it. Yet I am unwilling to find that an assisting tug must not only have a radar but an extra crew member to man it as well. The other reason is the established proposition that breach of duty to one class of persons does not give rise to a claim by another class even though performance of the duty to the first class would have prevented the loss suffered by the second. Gorris v. Scott, L.R. 9 Exch. 125 (1874); Garland v. Boston M.R.R., 76 N.H. 556, 86 A. 141, 46 L.R.A., N.S., 338 (1913); Lang v. New York Central R. Co., 255 U.S. 455, 41 S.Ct. 381, 65 L.Ed. 729 (1921); Irwin Savings Trust Co. v. Pennsylvania R.R., 349 Pa. 278, 37 A.2d 432 (1944); American Law Institute, Restatement of Torts 2d, Tentative Draft No. 4, § 281(b) comment c. Although examples of this rule are generally in tort, the same principles must be a fortiori applicable when, as here, the claim sounds in contract and the claimant could so easily have made its own law by stipulating for whatever aids to safety it thought needed.

Accordingly the claims of the Birkenstein and NGL for indemnity for amounts paid or payable as a result of the accident are dismissed. If a form of decree cannot be agreed, let it be settled on notice.