The appellee brought a libel against the appellant to recover damages resulting from a collision between the lighthouse tender Manzanita, a steam vessel, belonging to the appellee, and the dredge Columbia,' then being navigated on the Columbia river by the tug John McCracken, both of which latter vessels belonged to the appellant. The collision occurred at about 6 :15 o’clock p. m., October 6, 1905, at a point about 2.15 miles below Waterford, Wash. The dredge Columbia was a scow-shápéd craft 265 feet in length, and at the time of the collision had a draft of from 6 to 8 feet. She carried a cutter extending about 30 feet beyond her bow, for loosening the substance of the bed of the stream while dredging. The height of the dredge and its superstructure was 24 feet 8 inches, above the water line, and her pilot house extended 2 feet 10 inches still higher. She carried astern, for the purpose of discharging the spoil when dredging, a pipe of an aggregate length of from 900 to 1,000 feet, sustained upon a line of" 27 pontoons. While dredging the cutter was lowered to the bed of the stream, and the line of pontoons extended across the channel at about right angles, so as to deposit the spoil in shoal water near the shore. When the dredge was being towed, the pontoons extended in a line astern. The tug John McCracken was about 90 feet in length, and the top of her pilot house was about 22 feet ábové the water line. At the time of the collision she carried running lights in screens upon either side of the pilot l^otjse. The Manzanita was 152 feet in length and was drawing 11 feet’4 inches aft, and about 6 feet 6 inches forward at the time of the collision.

For some time prior to the date of the collision, the Columbia had been engaged in dredging at a,'point about >3% miles below the point where the collision occurred. About 4 p. m. on that day she was taken in 'tow by the tug, which was fastened to her starboard quarter, while a wood scow was placed in front of the tug. At that time the tide was ebbing. At the time of the collision it had commenced to flood slightly.' *868 Slow progress was made, a mile the first hour, a mile and a half in the second, and about "a mile and a quarter in the last three-quarters of an hour immediately preceding the collision. The Manzanita left Portland at about 1 p. m. on that day, with orders to proceed to Astoria. She made from 9 to IQ knots an hour until she arrived abreast the Waterford Post Light on the Washington shore. Prom that point the ordinary course pursued by navigators of the Columbia river is to proceed beyond the light from a quarter to a half mile, then to head directly for Westport Reach Light on the Oregon shore, a light which stands slightly higher than the 'surface of the water. The master of the Manzanita testified that, after passing the Waterford Post Light, he directed the vessel’s course directly for the Westport Reach Light, and almost immediately thereafter he and the mate sighted the dredge, which to them appeared to be at anchor or engaged in dredging. They saw the train of pontoons in the wake of the dredge on which were distributed seven lights, and they assumed that the starboard channel was thereby closed. They decided to pass the dredge on her starboard side. Owing to the height of the dredge, the running lights on the tug were not visible to a vessel coming down the river, and approaching» her port side. The dredge herself carried no running or navigating lights, but had an electric white light above the pilot house and two lanterns attached beneath the same, and her staterooms and working rooms were brightly lighted up. The point of collision was from 200 to 300 feet instream from the upper end of two fish traps located upon the Oregon side of the river, which extended into the stream about 500 feet. As to what occurred between the time when the master and the mate of the Manzanita discovered the dredge and the time of the collision, there is some uncertainty in the testimony. There can be no question that the officers of the Manzanita believed the dredge to be at anchor, and believed that her line of pontoons extended across the starboard channel. As the Manzanita approached the Columbia at a distance variously estimated by her own officers at from slightly less than half a mile to two ship lengths, she blew two whistles as a signal that she would pass the dredge starboard to starboard. There was no answer from the tug or from the dredge. The helm of the Manzanita was put astarboard, so as to change her course two points to port, and very, shortly thereafter the collision occurred.

At the time when the captain and the mate of the Manzanita first sighted the dredge, the latter was about two miles away, and they testified that the Manzanita had been put upon her course toward the West-port Light. They both testified, also, that, when they sighted the dredge, she was two points on their starboard bow. The court below rejected this testimony as incredible, and reasoned that if the Man-zanita, after passing the Waterford Light, was headed toward the Westport Light, to place the position of the dredge two points on the Manzanita’s starboard bow would be to place her near the Washington shore, a position which she could not possibly have occupied. We are not convinced that the testimony should be rejected on this line of „reasoning. It is not improbable that the officers of the Manzanita, who never had navigated the Columbia river at night, and had had no particular occasion to observe the shore lights, may have mistaken the *869 light on one of the fish traps for the Westport Light. In that case the dredge would have been, as they testified that it was, about two points on their starboard bow. That their testimony was not inad' vertently given is indicated by the fact that Capt. Byrne in that connection explained to the court that a point was 11 degrees and 15 minutes. There is corroboration of this view of the course of the Manzanita in the statement of the mate, who testified that he did not remember seeing the Westport Light until about the time when the Manzanita signaled the dredge, and that then he saw the light clear in front of or across the bow of the dredge. It was impossible for him at that time to have seen the Westport Light in that position, because the dredge had long since intercepted the course from the Waterford Light to the Westport Light, and, if he saw a light across the bow of the dredge, it must have been some other than the Westport Light.

But whether or not they were mistaken as to the light to which their course was directed is not of particular importance. Whether it was the Westport light or some other light mistaken for it, it is obvious that the Manzanita had before her an unobstructed course at the beginning of which the dredge was on her starboard bow, and it is undisputed that the Manzanita, after heading for a light on the Oregon shore, changed her course but once before the collision. Both the captain and mate testified that, when the Manzanita had approached within a distance o f between a quarter of a mile and a half mile of the dredge, the Manzanita was slowed down and her engines were stopped, after which she blew two whistles as signals to pass the dredge starboard to starboard, and about two seconds thereafter starboarded her helm two points, so that the dredge then bore three or four points on her starboard bow, and that they continued on that course until the collision, expecting to pass the dredge at a distance of 200 or 300 feet from her bow. The mate testified that he discovered that there was going to be a collision when the steamer was about a ship’s length away from the dredge, and that nothing further was done to avert collision by reversing the Manzanita’s engines or otherwise. The testimony of several witnesses who were on the dredge and saw the approaching steamer is to the effect that the latter approached head on directly toward the dredge until -within two or three boat lengths, when she turned her course and came around directly across the bows of the dredge. The evidence that the Manzanita was very near the dredge when her course was changed is confirmed by the deposition of the captain of the Manzanita given on May 3, 1906, in which he deposed that he kept on his course until about half a minute before the collision, and that then, when he was about two ship lengths from the dredge, he attempted to pass her on the Oregon side. When, ten months later, he gave his testimonj' at the trial, he testified that he was in error in so testifying in his deposition, and that, upon refreshing his memory, he would state that, at the time when he put his helm to starboard to pass the dredge, he was something less than half a mile distant from the dredge.

We do not deem it necessary to attempt to reconcile this conflicting testimony. On either statement of the facts, and assuming that they were as the trial court found them, it seems clear that negligence must be imputed to the officers in charge of the Manzanita. It is true that *870 the dredge was navigating the river without running lights, and that-there was everything in her appearance, when the officers of the Man-' zanita first sighted her, to lead them to the conclusion that she was at, anchor. But if the Manzanita was in fact pursuing a direct course between the Waterford Light and the Westport Light, it is evident that, very soon after that course was begun, the dredge must have intercepted the line thereof. The dredge was at that time moving at a speed of a little more than a mile and a half an hour. The Man-zanita had started on that course at nine knots an hour, but from the time when she whistled she proceeded at about three knots an hour. The fact that it had become necessary to change the course of the Man-zanita was proof to her officers that the dredge had changed her position and was moving upstream. It became their duty to take notice of that fact and adopt all reasonable precautions to avoid collision.' And it can make no material difference at what point on her course it was found necessary to starboard her helm to pass the dredge. Whether it was at a distance of a half a mile or only 300 feet, if they had observed, as they should have done, that the dredge was moving, they could have avoided the collision.

We think that the Manzanita was at fault, also, in not turning to. starboard instead of to port shortly prior to the collision. The position of the dredge and the pontoons had been materially changed in the 20 minutes which ensued after the dredge was first seen. Assuming, as it was found by the court below, that at the time when the Manzani-ta sighted the dredge the latter was crossing the channel to the Oregon side and was about to intersect the Manzanita’s course, and that she did cross it about a half a mile from the place of collision, it must be evident that at the time of the collision the train of pontoons, which extended only 900 or 1,000 feet astern of the dredge, was well upon the Oregon side of the channel, and that, before the time when the Manzanita’s helm was put to starboard, it would have been apparent to the officers of that vessel, had they taken the precaution to look, that the starboard channel was open for their passage. It was their duty to follow that channel under article 25 of the Pilot Rules, which reads:

In Chamberlain et al. v. Ward et al., 21 How. 548 -567, 16 L. Ed. 211 , the court said:

We do not overlook the rule announced in The City of New York, 147 U. S. 72 , 13 Sup. Ct. 211, 37 L. Ed. 84 , in which it was said:

We think the fault of the Manzanita is well established by the circumstances disclosed in the evidence, and in the facts as they were found by the District Court. The belief entertained by the captain and the mate of the Manzanita that the dredge was at anchor was at the first justified. The dredge carried no red light upon her port side to indicate that she was proceeding upstream, and she was about two miles away. But we are unable to see how it was possible for those two officers of the Manzanita, who stood upon the bridge thereof from the time of passing the Waterford Light to the time of the collision, to fail'to discover that the dredge was in motion. After they left the Waterford Light, it was not so dark hut that they could see the Oregon shore where the Westport Light was. There was nothing to obstruct their vision of the dredge or of the shore lights. They started upon a course in which at the beginning the dredge was upon their starboard bow, and, if the dredge had been at anchor, they would have crossed her bow at a distance of at least a thousand feet. When, as they proceeded, they found that the dredge intercepted the light to which their course had beep directed, they had before them clear evidence that the dredge had changed her position. Again, by taking timely observation of the starboard channel, they must have seen, before they turned their vessel to port, that that channel was then sufficiently clear to allow them to pass the dredge and pontoons port to port. Their errors in these respects were not slight errors, nor were they errors committed in extremis. They were substantial errors in navigation which contributed to the disaster, and which we think are sufficient to justify a court of admiralty in imposing upon the Manzanita one-tliird of the damages which resulted from the collision. The Gray Eagle, 9 Wall. 505 , 19 L. Ed. 741 ; The Mary Morgan (C. C.) 28 Fed. 333 ; Briggs v. Day (D. C.) 21 Fed. 727 .

We find no ground for disturbing the amount of the damages which the court below found that the appellee sustained as the result of the collision. But we are of the opinion that the injuries received by the Manzanita from the collision and the damages resulting to her owner therefrom were caused by concurring and co-operating faults of the dredge, the tug, and the steamship, and that they should be borne equally by each.

The decree is reversed, and the cause is remanded, with instructions to enter a decree as herein indicated.