In disposing of these causes it is apparent that, where the testimony conflicts, as it does on many points, and where there is no, internal evidence in the narrative of the individual witness that he is for any reason untrustworthy, the court can only decide the disputed questions in accordance with the probabilities, and that in finally disposing of the cause its conclusions must be largely influenced by the operation of the rule as to the burden of proof. The witnesses from both sides agree that the Iroquois ported, and changed her course somewhat to starboard. The witnesses from the Iroquois assert that at the time she ported, and prior thereto, she had seen the red light of the Powell about a point and a half on her port bow. If this were the situation, she ported to a red light, which was a proper maneuver. The witnesses from the Powell, however, insist that that vessel was showing her green light to the green light of the Iroquois. If this were the situation, the porting of the Iroquois would be a faulty manuever. Moreover, it would be more than that; it would be reckless, apparently inexplicable, and highly improbable, since the ship must have known, and admits she knevr, that she was encountering a sailing vessel closehauled, to which it was her duty to yield the right of way. We concur with the district judge, therefore, in discrediting the story of the witnesses from the Powell that the vessels were green to green when the Iroquois ported, and in accepting the story of the witnesses from the Iroquois that the vessels were at that time red to red.

The Iroquois contends that the Powell changed her course to the eastward; that while the vessels were red to red, as we have found they were, the Powell starboarded to the red light,—a faulty maneuver, bringing her across the course of the ship. This maneuver, however,, would be quite as reckless, inexplicable, and improbable as the other. The Powell knew she was closehauled on the starboard tack, knew the approaching vessel was free on the port tack, knew that she (the Powell) was entitled to the right of way over any vessel she was to encounter. That, under these circumstances, she should have deliberately yielded her privilege, and changed^ to starboard, is almost inconceivable. We therefore accept the story of the witnesses from her deck that she made no such change, rather than that of the witnesses from the Iroquois to the effect that she did. The captain of the Powell was called on deck, from below, when the vessels were almost together. *175He jumped to the wheel, saw the red light of the Iroquois about a ship’s length distance, took hold of the wheel, and gave it a turn, and saw that he was cornered; that it was no use. He thinks he turned the wheel, what little he did, to starboard., We are not satisfied that this action of the captain altered the barkentine’s heading at all before collision; but, even if it did, it was an act in extremis when he was dismayed at suddenly finding himself “'cornered, so he could go neither way,” and not a fault for which the barkentine is to be held responsible. AVe agree with the district judge in the conclusion that collision was brought about by the leeway the barkentine was making. Each side claims that the other’s navigation was improper in not being so conducted' as to avoid the effects of such leeway. The Powell contends that the Iroquois, assuming that, the vessels were approaching red to red, did not allow a sufficient margin for passing the vessel out of whose way it was her duty to keep. Counsel cites from The Star of Scotia, 2 Fed. 591:

“It is evident that it is not enough merely to bring the red light on the port bow in order to pass in safely a vessel which is passing on the windward side closehauled. Every vessel elosehauled on the wind will yaw more or less. She is kept by the wind by the constant but slight movement of the wheel as she tends to fall cf. or to come up. The experts in this case estimate half a point each way as the ordinary variation from her course by the wind which must be generally expected from this cause. Then, also, the actual course of every vessel sailing by Hie wind is likely to be a little to the leeward of the apparent course as indicated by her lights, varying with circumstances, the weight of her cargo, her trim and sails. In judging, therefore, of the case presented by the Star of Scotia, it is necessary to take these points into consideration. If a vessel thus passing another had not made due allowance for these things, and has not given a safe margin to allow for the possible leeway of an approaching vessel, and for her possible yawing while doing all she can to keep by the wind, she is liable to be surprised, as the Star of Scotia was, by the unexpected disappearance of the red light, and the appearance of the green light of the approaching vessel under her bows when it is too late to avoid a collision.”

The same rule was enforced in The City of St. Augustine, 15 C. C. A. 488, 68 Fed. 393, where this court affirmed the district court in the conclusion that the City of St. Augustine “did not, in her maneuvers, allow a sufficient margin for passing the schooner, nor for the usual and necessary variation in her course” through yawing or leeway. In the case at bar, however, it is alleged in her pleadings that when the vessels were about 400 yards apart, and the red light still bearing on the port bow, the ship's helm was ported, to give more room, and the evidence indicates that this maneuver was made even sooner. It seems apparent that if the leeway had made a difference in the actual course of the Powell of from half a point to a point from her heading, as it did in The City of St. Augustine Case, supra, there would have been an ample margin of safely. It is conceded that the leeway made by the Powell was sufficient to change her actual course a point and a half. The cuse is barren of any evidence to show what is the normal amount of leeway fairly to be anticipated from vessels encountered under like circumstances to those attending this collision. Certainly, in the absence of knowledge or of sufficient indications of the extent of an approaching vessel’s leeway, the vessel charged with *176the duty of avoiding her ought not to he held responsible for the results of some abnormal or excessive leeway. Therefore, upon this record, we cannot find the Iroquois in fault because she ported to a red light in time to leave a margin of safety sufficient to provide for a change of one point from the apparent course, through leeway, and nevertheless came into collision with the other vessel because the latter’s actual change was a point and a half.

On behalf of the Iroquois it is contended that the Powell, being a sailing vessel, closehauled, and bound to hold her course, must be held in fault because “she did not overcome her leeway.” We are advised of no practicable method whereby a sailing vessel may continue on her course, and at the same time overcome her leeway. She may, from time to time, luff v. to the wind, and for a space greater or less, dependent upon the wind, her sails, her speed, and other factors, may recover a part of what she has lost. It may well be that the incessant shifting of her lights consequent on the repetition of such maneuvers would in itself add to the uncertainties of the situation. But we are referred to no authority which holds that the privileged vessel must, while in sight of an approaching vessel, perhaps for several minutes, persist in this maneuver when her leeway is only the ordinary amount which is to be expected as inseparable from navigation under the existing circumstances, and which approaching vessels are expected to provide for by allowing a reasonable margin in passing as the rule of the road directs. In The A. P. Cranmer, 8 Fed. 523, “the schooner was-making very great leeway, more than the tugs had any reason to understand she would make.” In The Agra and The Elizabeth Jenkins, L. R. 1 P. C. 501, there was no question of leeway. The privileged vessel had altered her course by shifting her helm, seeking to justify such change under the nineteenth rule by contending that the burdened vessel so long delayed porting her helm and giving way that it was feared she was trying to cross the bows. The court says:

“Even if the Elizabeth Jenkins had, from apprehension of danger, altered or interrupted her course, she should have done so by luffing v. to the wind, thereby stopping her way, and mitigating, as far as possible, the effects of a collision.”

In the absence^ of any evidence to show whether the leeway which the Powell was making was abnormal or excessive, or greater than was to be expected by approaching vessels, we cannot hold her in fault as the A. P. Cranmer was held, viz. on the theory that for all practical purposes the effect of her very great leeway in regard to the legal relations of the approaching vessel to her was the same as a direot change of course towards such vessel. The decrees in both actions are affirmed, without costs.