This case arose out of a collision in the East River between a carfloat in tow of the Baltimore & Ohio’s tug William J. Dickey and a carfloat in tow of the New York, New Haven and Hartford’s tug Transfer No. 15. The New Haven brought a suit against the B. & O. in admiralty to recover for the damages resulting from the collision. Judge Galston found the Dickey at fault and gave judgment for the libellant for the entire damages. From this judgment the respondent appeals.
Judge Galston found the facts as follows:
The Transfer No. 15 left the vicinity of the float bridges of the Central Railroad Company of New Jersey in Jersey City about 5:20 A.M. on October 19, 1947 bound for Oak Point in the Bronx with two carfloats in tow. The carfloats were secured to the tug, one on each side, by headlines, tow lines and stern lines. The bows of the carfloats pointed at each other ahead of the tug and their sterns flared out behind it. Each carfloat extended about 200 feet ahead of the tug.
The Dickey had come from the Long Island Railroad Terminal in the East River and was bound for the float bridges of the Central Railroad Company of New Jersey. It was towing two carfloats in substantially the same manner as the Transfer No. 15. Both tugs and their floats carried all the usual navigation lights.
There was a light fog, the visibility lowering from a half a mile at 5:00 A.M. to a quarter of a mile at 6:00 A.M. The tide was ebbing at about four miles per hour. As the Transfer No. 15 approached the Brooklyn shore she turned upstream *230against the ebb tide and proceeded nearly parallel with the Brooklyn pier ends. Her captain saw something about 1,000 feet upstream and soon distinguished a tow with carfloats showing a green light. The Transfer No. 15 was at that time 600 to 700 feet off the Brooklyn piers on a northeasterly course; the Dickey was about 300 to 400 feet off the Brooklyn piers on a southwesterly course. Thus each vessel showed its green light to the other and each was in a position and on a course which would naturally result in a starboard-to-starboard passing.
When the Dickey became visible to the Transfer No. 15 the Transfer sounded a two-whistle signal, indicating a starboard-to-starboard passing, but received no answering signals from the Dickey. Transfer No. 15 continued on her course with no further signals until the Dickey started to bear to starboard across the Transfer’s projected course, and at that time the Transfer sounded a two-whistle signal a second time. The Dickey answered the Transfer’s second two-whistle signal with alarm whistles. Upon receipt of the Dickey's alarm whistles the Transfer No. 15 sounded backing whistles and alarm whistles and reversed her engines. The Dickey continued to swing to her own . starboard until the collision took place ; the Transfer’s heading continued to be northeasterly. At about 5:40 A.M. the starboard bow corner of the Dickey’s starboard carfloat struck the starboard side of the Transfer’s starboard carfloat about 20 feet from the bow.
On the basis of these facts Judge Gals-ton concluded that when the Transfer and the Dickey first became visible to one another their legal obligation was to pass starboard-to-starboard, that the Transfer carried out its obligation but that the Dickey violated its obligation by attempting, after receipt of two-whistle signals, a port-to-port passage, and that this maneuver by the Dickey was the principal cause of the collision. He concluded further that when it became apparent that there was danger of collision, the Transfer No. 15 took prompt and proper avoiding action, but the Dickey did not, and her failure to do so was the cause of the collision. He therefore awarded judgment in favor of the libellant for its full damages of $2,572.57 plus costs.
We accept the trial judge’s findings of fact. The stories told by the witnesses for each side were squarely in conflict on many points. The judge, however, accepted the libellant’s version. He was justified in doing so and there was sufficient evidence before him to support the findings he made. We therefore reject the contention now made by the appellant that the vessels were approaching one another in such fashion that a port-to-port passing was appropriate. If, as Judge Galston found, the vessels showed their green lights to one another, it was their duty to make a starboard-to-starboard passing. The Hygrade No. 12 v. The Talisman, 2 Cir., 1946, 153 F.2d 52. The Dickey was clearly at fault in failing to heed the Transfer’s two-blast signal. W. E. Hedger Transp. Corp. v. Hart, 2 Cir., 1947, 163 F.2d 90. The appellant, however, contends that the Transfer was also at fault.
The appellant argues first that the Transfer was at fault in failing to give an alarm signal when she first sighted the Dickey and in failing to give the alarm and reverse her engines when the Dickey’s sheer to starboard was first observed. We are not persuaded, however, that the Transfer was in these respects guilty of a fault which contributed to the collision.
It is true that Captain Kristen-sen did not give any signal when he first saw the “loom” of the Dickey through the mist at a distance of about 1,000 feet. Instead he waited until he could make out her running lights and then gave his two-blast signal. He testified that the Dickey was at this time about 600 feet away. We are not persuaded that Captain Kristensen was at fault in so delaying his signal. It is true that a signal for port or starboard passing should be given as soon as is practicable when vessels are approaching at close quarters. National Motorship Corp. v. Pennsylvania *231R. Co., 2 Cir., 1947, 160 F.2d 510. Here, however, the vessels were on parallel courses far enough to starboard of one another that they could, without changing course, pass in safety. In such a situation the approaching vessels need not reach an agreement before proceeding with a starboard-to-starboard passing. The Bellhaven, 2 Cir., 1934, 72 F.2d 206; The Delaware, 2 Cir., 1933, 66 F.2d 467. Under these circumstances we think that the Transfer was not at fault in delaying its signal until the Dickey’s running lights became clearly visible. Appellant, however, relies on Rule III of Article 18 of the Inland Rules, 33 U.S.C. A. § 203, which provides as follows:
“If, when steam vessels are approaching each other, either vessel fails to understand the course or intention of the other, from any cause, the vessel so in doubt shall immediately signify the same by giving several short and rapid blasts, not less than four, of the steam whistle.”
Appellant argues that since the Transfer must have been in doubt as to the Dickey’s course when she first appeared through the light fog, an alarm should have immediately been given. Since the Transfer did sound a proper signal for a starboard-to-starboard passing at a time when it was possible to execute that maneuver without a change of course by either vessel, it would be difficult to say that the failure earlier to sound an alarm contributed in any way to the collision.
The appellant argues, however, that in any event the Transfer should have given the alarm and backed when the Dickey began to veer toward her. At that time it became clear that the course of the Dickey might lead her into collision. The Transfer indicated her doubt as to the Dickey’s course by repeating the two-blast signal. We have often held that in such a situation it is improper to repeat the passing signal; the only proper course is to sound the alarm. E. g. James McWilliams Blue Line, Inc., v. Card Towing Line, Inc., 2 Cir., 1948, 168 F.2d 720. But although the Transfer was at fault in not immediately giving the alarm, we think it clear that this fault did not contribute to the collision. The Dickey responded to the Transfer’s second two-blast signal with alarm whistles of her own. She was apparently made aware of the danger by the reiteration of the starboard passing signal; alarm whistles from the Transfer could have produced no different effect. Socony Vacuum Transp. Co. v. Gypsum Packet Co., 2 Cir., 1946, 153 F.2d 773, 777; Seaboard Tug & Barge, Inc., v. Rederi AB/Disa, 1 Cir., 1954, 213 F.2d 772, 774-775.
A more difficult question is whether the Transfer was at fault in failing to reverse her engines until she received the alarm whistles from the Dickey. If she had done so a few seconds earlier when she first became aware of the Dickey’s change of course her speed might have been sufficiently diminished to minimize the collision.
Under the circumstances here, however, we find that the Transfer was not at fault. It has many times been held that a vessel is ordinarily under a duty to stop and reverse rather than driving ahead into a dangerous or doubtful situation. E. g., National Motorship Corp. v. United States, 2 Cir., 1948,171 F.2d 413; The Quogue, 2 Cir., 1931, 47 F.2d 873; The Bern, 2 Cir., 1934, 74 F.2d 235. But this is not an invariable rule and sometimes it is safer to go on. Construction Aggregates Co. v. Long Island R. Co., 2 Cir., 1939, 105 F.2d 1009, 1012. It may even be a fault to stop if it is safer to proceed. The Sidney M. Hauptman, 2 Cir., 1929, 34 F.2d 622. The Transfer was a single screw vessel with right-turning propeller. Backing would have tended to throw her stern to port and her bow to starboard. See Knight’s Modern Seamanship, pp. 280-84 (11th ed. 1945). Her master testified that this was her behavior. Thus if the Transfer’s engines had been reversed, her bow would have been thrown toward the Dickey and the likelihood of a safe starboard-to-starboard passing would have been diminished. Cf. Construction Aggregates Co. v. Long Island R. Co., supra; *232The San Simeon, 2 Cir., 1933, 63 F.2d 798, 801. By going ahead there was still a chance of effecting such a passing if the Dickey would stop her sheer to starboard. We cannot say under these circumstances that proceeding was any less safe than stopping would have been. A few seconds later when the Dickey’s continued sheer to starboard made a safe passing impossible, the Transfer properly reversed her engines.
The Dickey was clearly and obviously at fault. Her belated and ill-advised attempt to pass across the Transfer’s bow in disregard of the Transfer’s signals was inexcusable. Since the Dickey’s navigation was responsible for creating the predicament in which these vessels found themselves, there is little warrant for making a nice analysis of the Transfer’s efforts to avert the disaster which was thrust upon her. The Ludvig Holberg, 1895, 157 U.S. 60, 15 S. Ct. 477, 39 L.Ed. 620; The Hallgrim, 2 Cir., 1927, 20 F.2d 720, 723. We therefore conclude that the Transfer was not at fault in failing to stop and back sooner.
The appellant contends finally that the Transfer was at fault in failing to post a lookout in the bow of one of her carfioats rather than in the pilothouse. This would have placed the lookout some 200 feet ahead of the pilothouse and in a better position to see any approaching vessel. We have held that this is the proper place to post a lookout, and that the failure so to post a lookout must be presumed to have contributed to the collision unless the offending vessel shows that it could not have contributed. Poling Russell, Inc., United States, 2 Cir., 1952, 196 F.2d 939; United States v. The Adrastus, 2 Cir., 1951, 190 F.2d 883. We think a sufficient showing was made here. When the vessels sighted one another they were in position to pass safely. It is true that if the Transfer had seen the Dickey a few minutes sooner she could have sounded her two-blast signal a little earlier. But the Dickey did not heed the signal when it was sounded and there is no reason to suppose that an earlier signal would have had any greater effect. As the Dickey was in a position to effect a safe starboard passage when the Transfer gave her initial two-whistle signal, it is irrelevant that she was in a similar position earlier and that a two-whistle signal could have been sounded sooner. There is therefore no reason to suppose that sighting the Dickey a few seconds sooner could have prevented the collision. Cf. W. E. Hedger Transp. Corp. v. Hart, supra; Compañía de Maderas de Caibarien, S. A. v. The Queenston Heights, 5 Cir., 1955, 220 F.2d 120, certiorari denied sub nom. Esso Shipping Co. v. Compania de Maderas de Caribarien, S. A., 1955, 350 U.S. 824, 76 S.Ct. 52. It was the Dickey’s maneuver to starboard which created the hazard. This maneuver was made when the Dickey was well in view from the Transfer’s pilothouse. The trial judge found that the Transfer reiterated her two-whistle signal as soon as the Dickey started her move to starboard. Since the master of the Transfer, standing in the pilothouse, was aware of the movement as soon as it occurred, a lookout farther forward could have given him no additional information or warning. The proper posting of a lookout would not have affected the Transfer's navigation and the failure to post the lookout could not have contributed to the collision. Rice v. United States, 2 Cir., 1948, 168 F.2d 219. The Poling Russell and Adrastus cases, supra, are distinguishable. In the Poling case it appeared that the pilot of the one vessel was not aware of the other’s change of course and that a proper lookout might have given him warning. In the Adrastus case the approaching vessel was not observed until too late to avoid a collision.
Since the Dickey’s fault was the sole cause of the collision, the trial judge properly allowed recovery by the libellant of the full amount of its damages. The judgment must therefore be affirmed.