LACQ1JBE, Circuit Judge.
Several of the witnesses, including some of the more important ones, gave their testimony in the presence .of the district judge, and in such cases this court, for reasons which have been frequently given, rarely disturbs the findings of fact. This case, however, presents exceptional circumstances. It seems quite apparent from the opinion that in finding (hat the tugs did not use the reasonable caution required of them the court gave much weight to the fact that storm signals had been displayed at the barge office in this city, and was largely influenced by the circumstance that one of the tugs was not sem. ahead as a scout. The court says: “In approaching exposed situations with boats of this character, not able to withstand heavy seas, it has often been shown before me that tugs are accustomed to go out ahead, and make examination, before taking tows out int.o an exposed situation. Bee The Bordentown, 40 Fed. 682.” Reference to The Bordentown shows that the navigation then under consideration was through the Bay of New York. There is no evidence before us, however, of any such practice of sending scouts in advance prevailing on the Hudson river, where the conditions of navigation are very different. Not a witness produced by the libelants makes any such suggestion; on the contrary, the competent navigators called by libelants unite with those called by claimant in testifying that in navigating the Hudson river with tows they continue on and come through unless the weather is something extraordinary,, in which case they seek the nearest harbor. This particular storm was the most violent, for that season of the year, within the memory of any of the witnesses, several of whom had been navigating there for 20 years and upwards; *186but, with the ordinary heavy blow which is to be expected on those waters in August, a method of navigation which does not contemplate the use of advance scouts seems always to have prevailed, and to have proved ordinarily safe. We are not prepared, by affirming these decrees, to change the practice of years, nor to require navigators on these narrow inland waters to adopt expedients' which may be in use in bays and harbors opening into the Atlantic Ocean. Of the three tugs which took hold of the tow in the Highlands, one or more, about the time of starting up the river on the afternoon of the 28th, had been in such a position as to be able to observe the cautionary signals for coming storm (high southeast wind) which Were displayed at the signal stations in New York City, but paid no attention to them. The weather bureau had also made public dispatches announcing “a storm off the east Florida coast [August 26th]. The West India hurricane is now central off the ¡Georgia coast. * * Danger is not so threatening as yesterday [August 27th]. The hurricane has moved northward near Charlotte, N. C. Expect southeast gales on the Middle Atlantic coast, with rain in the Middle Atlantic states [August 28th, 9:30 a. m.]. Expect southeast gales, with rain, tonight [August 28th, 2:10 p. m.].” The navigators of the tugs did not notice the cautionary signals displayed, nor did they advise themselves of the dispatches received at the New York weather bureau station before their departure to get their tow. The district judge holds them in fault for not “regarding the previous public notice of a coming violent S. E. storm, with which they must stand chargeable.” We are unable to assent to this proposition. We do not find in the record any evidence that the movement of a tropical storm up the Atlantic coast will produce such a condition of affairs on the Hudson river, 40 miles from its mouth, as to make it unsafe for navigation by tugs and tows; and, in the absence of such evidence, are not inclined to hold navigators responsible for failure to keep in touch with the weather bureau reports, as they do on the seacoast, or in Long Island Sound. So far as appears, navigators upon the Hudson river have never regulated their movements by the reports as to gales or hurricanes upon the Atlantic coast, nor tied up their tows awaiting the fulfillment of weather bureau prophecies. It is suggested in proof that, if they did, transportation down the river would be seriously interfered with, and we see no reason to interfere with what seems to be the universal practice of navigation on those waters. If navigation in conformity to weather bureau prognostications is to be required of these tugs which happened to have left New York City the afternoon before, and to have been near enough to the signal to make it out, it should equally be applied to tugs which haul tows the whole distance from Albany to New York, although no weather bureau signals of any kind are displayed on either bank for the entire stretch of 150' miles, and the navigator could learn of the bureau’s prognostications only by stopping at one' place or another during his 48-hour trip, and getting into telegraphic communication with the nearest station, or buying the latest edition of the New York *187papers. Certainly no such method of navigating these waters with tugs and tows has ever been practiced before.
The cases at bar depend upon the question whether or not the master of the Pocahontas (deceased before trial), who had charge of the navigation of the ílo tilla, a man of large experience in Hudson river navigation, and bearing the reputation of a careful and prudent navigator, committed an act of negligence either in proceeding on below Stony Point: instead of making a harbor above Verplanek’s Point, or in not making a harbor after he had entered Harerstraw Bay, and when he found the wind so high and sea so rough as to imperil the safety of the tow. As was stated in Sewall v. La Champagne, 8 C. C. A. 624, 60 Fed. 299, this court is reluctant to substitute its judgment upon appearances it has not seen for that of experienced and competent navigators. Several propositions hearing upon the question at issue are abundantly settled by the evidence. The flotilla passed Stony Point no later than 1 a. m. according to the overwhelming weight of evidence. The district judge so finds, and the libelants so contend. Probably the time was somewhat earlier, but for the purposes of the argument it may be taken as 1 a. m. The concurrent testimony of all the competent navigators called on both sides is that, after the flotilla had once entered the rough water which it found in Harerstraw Bay, it could not turn around, and retrace its course to a harbor above Stony Point or Yerplanck’s Point. To undertake to do so would be to throw the tows into the trough of the sea, and imite inevitable disaster. Bo, too, the preponderance of the testimony given by the competent navigators called by both sides sustains the proposition that the magnitude of a, storm prevailing on Harerstraw Bay cannot always be determined by a navigator while he is coming south out of the Highlands, nor until he has got around Verplanck’s Point, and by the upper end of the bay itself; in other words until he is actually in the rough water. Moreover, the same witnesses practically ail agree Unit from Stony Point down to the Tappan Zee (here is no harbor in a southeast gab; unless Rockland Lake landing can be held to be one. It is apparent, then, that the first disputed question to be settled by the court is as to the condition of the weather during the time the flotilla moved from Cauldwell's, under the shoulder of the Dunder-berg, at the gateway of the Highlands, down to Blouy Point, where it entered the broader waters of Haverstraw Bay. As is usual in all such cases, then; is a violent conflict of evidence between the interested witnesses, die officers and crew of the tugs on one side and the owners and occupants of the canal boats on the other. The testimony of some of the eight interested witnesses called by the libelants, however, is entitled to little weight. Collins, on the Taylor, turned in to bed about 8 or 9 p. m., and did not turn out until after 3 a. m., when, as lie says, they were near Rockland. Peter Weber, on the P. W. Webber, went to bed about 9:30 p. m., when they were going by Cornwall, at the upper gateway of the Highlands, and got up, as lie says, at 4 a. m., “just breaking day.” John Weber, on the Webber Bros., went to bed about 9 p. m., and *188woke about 3 a. m. His evidence rather corroborates the testimony of claimant’s witnesses that the storm broke suddenly upon them after they had proceeded some distance into Haverstraw Bay. He says that he “hadn’t heard any noise before, but was wakened up about that time with a noise, and supposed it was blowing.” It is true he says they “must then have been just out of the Highlands,” but he fixes the time at 3 a. in., and “an hour before daylight,” and the evidence on both sides clearly fixes the location of the flotilla at that time a considerable distance below Stony Point. Minch, on the Sternburgh, is so reckless in his statements that no credit can be given to any part of his testimony. He says that they came out of the Highlands after daylight; that it was blowing very hard in the Highlands; that the sea was very rough there, and water came into the bow stables while they were still in the Highlands; that it was not safe for the tow even in the Highlands; the water was loading them down so deep there was danger of sinking them. All these statements are flatly contradictory of the concurrent testimony of the other witnesses on the same side. Hines, on the E. A. Culver, turned in at 9 p. m., and turned out again at 1 a. m., “or a little after.” He says the tow was then about Peekskill, in which he is manifestly in error, — a confusion as to locality which he evidently carried on during the rest of his progress down the river, for he testifies that “about an hour and a half after he got out” two boats of the tow broke away, whereupon the tugs were signaled, and the tow made up again, all of which, Hines insists, took place before they got into Haver-straw Bay at all. Two boats did thus break away, but all the other evidence in the case fixes the place as somewhere near Teller’s Point, or at least quite a distance down Haverstraw Bay. If a correction for error equivalent to the distance between Peekskill and Stony Point be applied to Hines’ entire narrative, it will be found rather to corroborate the story told by claimant’s witnesses. In view of the conflict between the three remaining interested witnesses called by libelants and the ten interested witnesses called by claimant, we must turn to the disinterested witnesses for a solution of the question. Five of these were called by libelants. Shook was master of the Leonard, which, with the Hudson as helper, was towing 16 canal boats south. His flotilla is referred to in the testimony as the “Beverwyck tow.” He says he found the weather very moderate when he came out of the Highlands and passed Stony Point. That what little wind there was was from the southeast. That it was calm. He “couldn’t fell whether there was any wind or not when [he] was just above 'V'erplanck’s.” That he left the “mouth of the Highlands” about 10 or 11 p. m., which would make it about 12 m. as they rounded Stony Point into Haverstraw Bay. That by the time they got two or three miles below there, they began to get wind, and the tide had got setting down. In reply to a question as to the prudence of going on he said: “We didn’t have nothing to stop us until it came all of a sudden. Then we couldn’t turn around’ after it kicked up the sea. The wind kept breezing up then.” He adds that it was a *189very sudden increase of the wind from moderate to extraordinary, and that he does not think he ever saw' such a sudden increase oí wind in so short a time, lie determined to put into Rockland Lake landing, and made harbor there “between one and two.” In view' of the distance between Stony Point and the landing, and the conditions of wind and tide, it must: have been near 2 a. m. Lyle, the master of the Hudson, also called by libelants, testified: That he came on duty at about 12 m., when they were coming oat of the Highlands. That it was then raining a little and blowing a little; not much, just a nice S. E. breeze, no sea. That after they got out of the Highlands it kepi: breezing up, and by the time they got down to Snedicor’s, about a mile off Rockland Lake, It was “blowing pretty hard, and gelling precLv rough.” That they made harbor at Rockland Lake, about 2 a. in., and saw the Ronan tow (claimant's tow) pass them about an hour later (3 a. in.), which harmonizes with the other evidence in the case, and with the distance, the chart showing that ii. is 6^- statute miles in an air line from Stony Point Light to Rockland Lake dock. Dwyer, the engineer of the Leonard, agrees with these witnesses that they passed Stony Point about 12 m., but says that the weather “was very bad and threatening then.” Wilson, pilot of the Dean Richmond, came down the river so late, passing Stony Point after 5 a. m.. that his testimony throws little light on the condition of affairs there at 1 a. m. The same observation applies to Person, pilot of the Saratoga, who came out of the Highlands between 4:30 and 5 a. m. He infers from his observations during the time he came on duly at 1 a. m. that the storm was a sudden one, and that it caught the Ronan tow below the Highlands. One disinterested witness was called by the claimant, — Sirrine, master of the R. G-. Townsend. He came up the river on the evening of the 28th to Highland Lake, and points near by, to pick up a tow of four boats, carrying ice and concrete. He arrived at Tompkins Cove (opposite Verpimieles Point) about 12 m., and while there the Ronan tow passed him, going down into Haverstraw Bay. He says that the weather was then fair; quite a fresh southerly breeze, but the moon shining: that witness started with his own tow about 1:30 a. in., at which time he considered it fair to proceed, and had got between Stony Point and Grassy Point (about a mile below), “when,” he says, “the storm first hit us, — that is, in fury; * * broke on us in a squall like.”' He pulied ahead, just about holding his own against the wúnd and sea through the night, and reached Rockland Lake at 9 a. m., before which time the wind had veered around to south by west. Upon all this testimony, without now rehearsing the evidence of the masters and pilots of the claimant's tugs, we are of the opinion that there is not sufficient shown to warrant the court in finding that in failing to stop and seek a harbor north of Stony Point the master of the Pocahontas, who was in charge of the navigation, was negligent, oí* failed to exercise reasonable prudence.
The only remaining question is whether he should have put into Rockland Lake landing after the storm broke, assuming that it *190did so before he had passed through Haverstraw Bay. Concededly, the wind was then from the southeast. A reference to the chart shows that the course of the river is also southeast. Wilson, of the Dean Richmond, says that he knows of no place where a tow such as this could be anchored after getting into Haverstraw Bay; that in such a storm it is safer for a tow to be below Rockland Lake landing with a southeast storm because the east bank breaks the force of the storm, making smoother water. . Person, of the Sara-toga, says Rockland Lake landing is not a very good harbor for a southeast wind. He adds that for a tow situated as Ronan’s was, meeting a storm south of the Highlands, — high wind and high sea, with a southeast wind, — “there ain’t any harbor to make in a southeast wind. In a southwest wind you might possibly get into Rockland Lake. In a southwest wind that would be a kind of a harbor, but in a southeast wind it draws right up there.” With such evidence from disinterested witnesses of large experience produced by the libelants, it would be difficult for the court to hold that thé navigator of the flotilla was negligent in not trying to make a harbor there. 'It is true, as the libelants urge, that the Leonard and the Hudson took their tow in there, and rode out the gale in safety; but the evidence of the navigators of those vessels shows that they had not over 16 boats in tow, and that they were able to shelter these alongside of the docks there, keeping the tow on short hawsers, with the line fast to the steamboat dock, and sheltered by a light crib dock which the ice company had built there for the protection of their barges in southeasterly weather. Some ice barges that were lying at the latter dock helped to break the sea. Neither of these witnesses had ever made a harbor there before. It seems quite apparent from the testimony that while the wind continued southeast there was no place of safety at Rockland Lake landing for an additional flotilla of 32 canal boats. Subsequently the boats which survived the breaking up of the flotilla were towed in stern first, and remained there in safety; but they did not arrive until after the R. G-. Townsend’s tow, and before that time the wind had shifted to south by west, thus making the landing a comparatively sheltered harbor. We are of the opinion, therefore, that in deciding to keep on slowly in the hope of ultimately making a harbor under the easterly shore, below Rockland Lake, the navigator of the claimant’s tugs acted with a reasonable exercise of judgment, and should not be held negligent because he might have made the landing, and might have ridden out the storm there in safety, as the Beverwyck tow did, — a hypothesis which seems to us extremely doubtful. The decrees of the district court are reversed, with costs, and causes remanded, with instructions to dismiss the libels, with costs.