This is an appeal from a final decree in admiralty holding solely at fault the American Tug Crowley No. 29, hereafter called Crowley, for damage done to the American Tug Milton S. Patrick, hereafter 'called Patrick, in a collision between the two tugs on November 10, 1938, in the main channel leading to the inner harbor of Los Angeles Harbor. The only claimed errors of the district court have to do with the navigation of the tugs. The Crowley was unharmed and the amount awarded the owners of the Patrick is not here questioned.

The two tugs were rivals in serving the vessels traversing and mooring in Los An-geles Harbor. The Crowley had sought and obtained the engagement to tow the inbound steamship Canonesa, and met her opposite the Southern Pacific slip when the steamer was proceeding under her own power in a general northerly direction in the main channel. The Canonesa was in the middle of the channel; the Crowley proceeded along the Canonesa’s port side. No question is raised as to fault on the Crowley’s part with reference to proceeding on her portside of the channel.

The maneuvers leading to the collision occurred between 5:30 and 5:45 in the morning, when the channel was smooth and the visibility good. It was not full daylight but the vessels and their lights were in plain vision of both the tugs throughout their maneuvers.

The Patrick, unaware of the engagement of the Crowley, was waiting in the channel to tender her services to the Canonesa, at a place some 1,600 feet to the northerly of the point where the Crowley began to proceed alongside the steamer. When the Crowley approached the Patrick she was free from the Canonesa, no towing lines having been passed between them. The Crowley took a sheer from the Canonesa’s bow wave which turned her towards the Patrick and with which she collided.

The Crowley’s contention is that the Patrick was in fault because her position was too near the course of the Canonesa and Crowley, and that this fault was the sole cause of the collision. The district court held that the Patrick’s position in the channel was at a proper distance from the Crowley’s course, that the Crowley’s sheer did not bring her to a contact with the Patrick but that the additional distance the Crowley traversed was caused by the Crowley’s captain giving a left rudder and not sufficiently checking her forward motion. This maneuver, though apparently in extremis, the court held was due to the prior and continuing absence of a lookout on the Crowley, whereby her captain had been so occupied in passing a heaving line to the Canonesa on his right that he had not been aware of the Patrick and did not see her until the bow wave of the Canonesa compelled him to look ahead on his course. This although the Crowley had proceeded for 1600 feet northerly up the channel approaching the Patrick at about a six knot speed — that is, the Crowley’s captain had not seen the Patrick *653 though she was in plain sight for over two minutes.

The experienced proctors for the Crowley make a perfunctory contention that, although the Crowley was free from and not lashed to the Canonesa, her captain was entitled to rely on the Canonesa’s lookout to guide him up the channel. Obviously, the Canonesa’s lookout, compelled to watch ahead, could not be looking back down to a free tug under the steamer’s portside and abaft her bow, and become responsible for the tug’s guidance. Here was gross negligence on the part of the Crowley proximately contributing to the collision, even disregarding the strict sanctions of the absence of lookout rule in The Ariadne, 13 Wall. 475, 479 , 20 L.Ed. 542 . The testimony offered no conflict with regard to this fault of the Crowley. Assuming the Patrick was stationed too near the course of the Crowley, it was a remote incident in the causal chain; the latter had for an ample period of time the opportunity and, in the circumstances, the obligation to drop far enough astern to free her of any sheer from a possible bow wave of the Canonesa, and thus to have passed the Patrick in safety. The Steel Inventor, 2 Cir., 43 F.2d 958, 960 ; The Pleiades, 2 Cir., 9 F.2d 804, 806 ; The Curtin, 4 Cir., 217 F. 245, 246 .

Concerning the Patrick’s position in the channel with reference to the Crowley’s course, the testimony is 'conflicting. The district court found there was abundant room for the Crowley to have passed in safety. All the witnesses on both tugs testified viva voce and the court accepted as more accurate the statements of those from the Patrick that she was lying dead in the water over a hundred feet from the Crowley’s course, waiting to offer her services to the Canonesa. It is obvious that the testimony of those on the Patrick, a stopped vessel, observing an approaching vessel to communicate with her, would be of more likely accuracy than of those on the Crowley, so .preoccupied with making fast to a vessel away from the Patrick that they did not see her until in the excitement of the impending collision.

Carefully worked out combinations of the estimates of distances and of the speed of the Crowley tend to show the Patrick nearer the Crowley’s course, but we are required to assume that the district court, hearing the witnesses, rejected the testimony of factors not comporting with its findings. An appellate admiralty court cannot redetermine the credibility of witnesses heard below, despite its broad powers recognized in The Ariadne, supra, and The Ernest H. Meyer, 9 Cir., 84 F.2d 496 . The evidence amply supports the finding that though there was abundant distance for the Crowley to pass in safety, her captain, in his unwarrantable surprise on suddenly seeing the Patrick, gave the Crowley a left rudder without sufficiently checking her speed.

The fact that the Patrick was dead in the water abundantly distant from the course of the Canonesa, and waiting to tender her towing service, created a situation of special circumstances, to be expected in any port offering towage to its visiting ships. Since the Patrick was dead in the water there is no merit in the Crowley’s contention that the Patrick was “approaching the Canonesa head and head and on the Canonesa’s port side, * * * [and] failed to pass the Canonesa port to port” in violation of Article 18 of the Inland Rules of Navigation, 33 U.S.C.A. § 203 . Nor was the Patrick, in such special circumstances, required to keep out of the way of the Crowley under Inland Rule 19, 33 U.S.C.A. § 204 . She was out of the Crowley’s way.

Nor is there merit in the Crowley’s contention that the Patrick violated Inland Rule 28 when she failed to blow three whistles on putting her engines half speed 1 astern when maneuvering to her stopping and awaiting the Canonesa. If this be fault it did not contribute to the collision in view of the abundant passing room. The three blast whistle is not a warning signal. That signal is four or more short blasts. Inland Rules, Art. 18, Rule III. A master cannot be charged with causatively contributing to a collision by failing to blow the full astern signal when, in the circumstances, his reversing would not affect the navigation of an approaching vessel, because, if blown, it *654 might have operated as a warning signal to that vessel which, without his knowledge, was violating the Ariadne rule.

The decree is affirmed.

“When vessels are in sight of one another a steam vessel under way whose engines are going at full speed astern shall indicate that fact by three short blasts on the whistle.” (Emphasis supplied).. Article 28 of the Inland Rules, 33 U.S.C.A. § 213 .