At about half past 5 on the morning of December 7,1890, a little before day-break, the libelants’ steam-ship La Champagne, outward bound on one of her regular trips from New York to Havre, came in collision with the British steam-ship lisbonense, inward bound, a little outside of Sandy Hook, and very near the point where the track of the main ship channel intersects the axis of the swash and south channels. The stem of La Champagne struck the port quarter of the Lisbonense about 20 feet from her stern, penetrating about 2 feet, cutting her down nearly to the water’s edge, and tearing off her plates aft, while the stem of La Champagne, and a number of plates from her port bow. *123were carried away, and left sticking fast in the side of the Lisbonense. Each charges the fault upon the other, and the above cross-libels were filed to recover their respective damages.

La Champagne is a steamer of the first class, 504 feet long, of about 10,000 tons displacement, and drawing at the time of collision 25 J feet. The Lisbonense was 270 feet long, of about --— tons burden, and drew 19 i feet. The former, on account of her deep draft, was necessarily proceeding out around the South-West Spit, by way of the main ship channel, and Gednev’s channel; the latter was coming in by way of the south channel and the swash, the axis of which crosses the main ship channel at an angle on the S. W. side of 101 points. At the time of collision tho night was clear, but dark; the tide had been running ebb 2 hours; the wind was fresh from the north-west. La Champagne, after leaving her dock the previous day, had been detained by fog near quarantine until about half past 4 a. m. of the 7th, when she got under way, and passed the meridian of the Sandy Hook lights at 5:20 A. si. At that time she burned her private signal torches, and soon after burned a blue light for the pilot-boat outside to be in readiness to receive her pilot. These signals, as well as her green and white lights, were observed by the Lisbonense, which was then on the usual course up the south channel, about N. W. 'i N. The red light of the Lisbonense was seen by La Champagne at about the same time, bearing 23 degrees, as indicated by the alidade, on her starboard bow. This would make the two vessels about two miles apart at that time, and La Champagne about one and onc-hal: miles, and the Lisbonense about one mile, from the point of collision. The speed of the former derived from her engines was about 10! knots; of the latter, as stated by the master to the pilot when he came aboard, 9 knots. Making corrections for the tide, which, according to the coast survey reports, would run at the rate of about 1.75 knots at that time, and which soon alter parsing Sandy Hook ran about 3$ points to starboard of the course of La Champagne, and about 2 points off the port bow of the Lisbonense, and allowing something for the fresh north-west wind, La Champagne would be making a little less than 12 knots over the ground, and the Lisbonense a little less than 7£ knots. Those conclusions as to speed harmonize with the other facts of the collision, and with the bearings as testified to by Capt. Boyer, and they so far confirm his narrative. They are not compatible, however, with the estimate of several of the witnesses for the Lisbonense that La Champagne bore about four or four and one-half points off her port bow after she had steadied up the south channel; and her bearing, until collision was near, could not have boon more than from three to three and one-half points off the port bow of the lisbonense. Capt. Boyer, of La Champagne, upon sighting tho Lisbonense, observed her carefully with the alidade, to see whether her bearing changed or not in reference to tho necessity of taking precautions against collision. After observing her a short time, not seeing any material change in her bearing, he informed the pilot, who replied that the Lisbonense would give way. Soon after, still seeing no change of bearing, and when the pilot was about to give a signal of two whistles *124to the Lisbonense, signifying that he would go to the left and ahead of her, he received a signal of one whistle from the Lisbonense, which was immediately answered by one whistle from La Champagne. There are no means of determining the precise time when these signals were exchanged, or the exact distance the two vessels were then apart. Most of the witnesses for the Lisbonense estimate the distance as about a mile, and I see no reason to believe it less than three-fourths of a mile. La Champagne ported her helm, so as to change her heading about one-quarter of a point to starboard, which was all that was safe for her in that channel, in an ebb-tide and north-west wind. The Lisbonense ported, so “as to change her heading half a point or a point only, whereupon she steadied her helm upon the proper course to go up the swash channel,” and thereby resumed substantially the same course she was on before. Capt. Boyer, after carefully continuing to watch the Lisbonense, and judging further maneuvers necessary, stopped his engines, and a few seconds afterwards reversed at full speed, repeating the order, “back as strong as possible,” and at the same time gave the Lisbonense a signal of three whistles, which announced his reversal to the Lisbonense, and soon after repeated the same signal. The Lisbonense was all the time going at full speed, and she continued at full speed until collision.

Upon the testimony of the Lisbonense alone there would be much uncertainty and inconsistency as to the time when these signals were given, or were heard by her, as well as to what was done by her. The master and others of her witnesses say that when those signals were heard La Champagne bore abeam of her, and was only about 100 feet off. Her pilot at first said that when he heard the blast of three whistles she bore “fully.4J points on the port bow; may be 5.” Afterwards he says“She was pretty near right angles, — very nearly abeam, — 100 feet away.” Her other witnesses say the same. But this is so clearly erroneous as much to discredit these witnesses on other controverted points. The fact that the Lisbonense was penetrated only about 2 feet (the pilot says 15 feet) with the other’s great momentum (10,000 tons) is proof to my mind that her rate of motion at the moment of collision could not have exceeded 2 knots, (The Martello, 34 Fed. Rep. 71, 73,) though the master’s computation assumes over 3 knots. At that moment both wore subject to about the same effects from the tide, which in that respect is therefore disregarded. From the experiments made with La Champagne, as well as those made with similar vessels, {La Normandie, 43 Fed. Rep. 159, 161,) a reduction from 10 knots to 2 through the water (a half knot would be lost in stopping 20 seconds) would have required at least 2 minutes’ time with the engine backing at a power of only 16 knots, and an advance of at least 1,150 feet through the water, to which should be added 200 feet for difference of tideway between the two vessels. ■ When the three whistles of La Champagne were heard, the Lisbonense put her helm hard a-port, and must have changed from three to four points, since the collision was at about right angles, and La Champagne must have turned to starboard about one and one-half points while reversing. If, as her master says, La Champagne at collision was angling *125a little forward, the Lisbonense must have changed more than four points. This, however, I think improbable. Her change of about four points to a right angle with the course of La Champagne agrees with the testimony of officer Dupont, who observed the heading of the Lisbonense at collision. In making this change to starboard, the Lisbonense must have gone at least 1,000 feet, and probably more, so that the distance of the vessels apart when the three whistles of La Champagne were blown must have been from 1,500 to 2,000 feet. This makes probable the testimony of the master and pilot of the Lisbonense, who say that they had no apprehension of collision until these three blasts were heard. It is incredible that they could have had no fear of collision until I,a Champagne was within 100 feet of them, as they allege; or, if that were the case, that they should not have noticed her cabin lights, as they say they did not, to indicate a,n approach so near. The master also says he hard a-ported when those whistles -were heard, and, had they not been heard till La Champagne was within 100 feet, the lisbonense would not have swung a point to starboard before collision.

On the part of the Lisbonense, it is claimed that, in accordance with the ordinary rule of the road, La Champagne, having the lisbonense on her starboard hand, was hound, under article 16 of the international rules of 1885, to keep out of the way.

For La Champagne it is contended: (1) That it was the duty of the lisbonense to give way, under article 23, in the present case, because the navigation of La Champagne, on account of her great draft, was hemmed in by winding channel ways, and by shoals, which would not admit of the application of the rule for the open sea; that those difficulties were increased by a strong ebb-tide and a fresh north-west wind, in consequence of which it was impossible for La Champagne to keep out of the wav of the Lisbonense by going to starboard, without the certainty of running upon shoals before she could recover the channel, while the much smaller draft of the Lisbonense permitted her to go in any direction without hindrance; and that it was, therefore, the duty of the Lisbonense to suffer La Champagne to pursue her course on the same principle on which a boat, descending with the tide an obstructed channel-way or river, has the right of way as respects an ascending vessel, because the latter can perfectly control her movements, while the other cannot. The Galatea, 92 U. S. 439; Mars. Coll. (2d Ed.) 59. (2) That, without regard to the foregoing, the Lisbonense, after giving a signal of one whistle, under article 19, was bound, by the express command of that article, to continue her course to starboard, and that it was her failure to do so that produced the collision.

Upon the construction which seems to me necessary of the present rules of navigation, I find the last contention of La Champagne well founded. Article 19 says that—

“A steam-ship may indicate her course to any other ship which she has in Bight by the following signals of her steam-whistle, namely: One short blast, to mean * I am directing my course to starboard; ’ two short blasts, to mean •I am directing my course to port;’ three short blasts, to mean ‘I am going *126full speed astern.’ The use of these signals is optional; but, if they are used, the course of the ship must be in accordance with the signal made. ”

The meaning of, the last clause seems to me unmistakable. It is not complied with by navigating in the prescribed direction at the moment of giving the signal, and then changing it the next moment, or the next. The purpose of the signal is to apprise the other vessel of some maneuver the former is making, and to let the other know what she may or must count upon. The obligation, therefore, is not a mere momentary one. How long a port wheel must be continued in such a case it is not necessary to consider; but the rule requires at least some substantial change of direction to be made and preserved, provided the circumstances of the navigation present no obstacle thereto, until the danger is over, or some new maneuver becomes necessary. On any other construction, the signal would be virtually a false signal, and a snare; since the other vessel could never tell for how long or how short a time the signaling vessel would continue to act in accordance with her signal,, nor upon what to count. In the present case, the only change of course to starboard in accordance with the signal claimed to have been made by the Lisbonense is “from half a point to a point,” which was accomplished in going some 200 or 300 feet, when the vessels were three-fourths of a mile apart. The Lisbonense then resumed substantially her former course. There was nothing in the circumstances of her navigation to compel her to do so. The swash channel was nearly a mile and a half distant, and there were no obstructions to prevent her going to starboard ad Kbitwm,, in accordance with her signal. But in fact no substantial change of her course to starboard was made. She had been for some time previously heading up the swash-channel course, and, after going “a half point or a point” to starboard, she again headed up the swash-channel course, as before. The change in her position effected by her slight porting was perhaps 25 feet, more or less, to the eastward, and then her former heading was resumed, only a trifle more to the westward even than before. Such a resumption of her former course substantially after going to starboard “a half point or a point” was not only a violation of the obligation of article 19, but a violation of article 22 also; because it changed her course a half point or a point towards La Champagne, after she had once effected at least so much change away from her. The least she could do under articles 19 and 22 was to preserve unchanged such a direction to starboard as she had already made under her signal. Her change of a half point or a point back towards La Champagne, which she must have made in order to head up the swash channel, was of itself far more than enough to produce the collision; while a reasonable continuance of her port helm long enough to effect a substantial compliance with her signal, and an adherence to the change of course thereby made, would have carried the Lisbonense still further to the eastward.

Assuming that the Lisbonense had the right of way, and that it was the duty of La Champagne to keep out of the way, article 19, in my judgment, did not authorize her to take the initiative, or to give any sig*127nal at all to La Champagne at the time she did, when three-fourths of a mile away from her. She had no such right until constrained to change her course. That article authorizes signals only when “taking any course authorized or required by the regulations.” But the regulations do not authorize any change of course by crossing vessels, except by the vessel bound to keep out of the way; the other must “keep her course.” Article 22. The system of signaling under article 19 is essentially different in this respect from the system authorized and required by the rules of the supervising inspectors in river and harbors. The latter rules, for the most part, do not prescribe or always require any change of course; they are to indicate only on which side of the other the signaling vessel proposes to pass. They are compulsory. An answer is required, that a common understanding may he had, and tho answer does not necessarily demand any change of helm. Whether such a change is required or not depends on the circumstances. Under article 19 no answer to another’s signal is required or provided for. The signal is an announcement of a change of course to the right or to the left. It is optional; hut, if given, the corresponding change must he made. It is said (.hat the maneuver of the lisbonense only aided La Champagne in her duty of keeping out of tho way. That would have been true had the obligation imposed on the lisbonense by article 19 been performed, and the maneuver adhered to; but (lie obligation being broken, and the maneuver revoked without notice, La Champagne was misled to her obvious prejudice. Instead of going off to starboard, as her signal promised, the Lisbonense, after turning a trifle to starboard, came hack on La Champagne’s course, thus doubly deceiving her. The fault of the Lisbonense in this respect was the same as in the more common cases of an unexpected change of course across the other’s how. The Roanokc, 45 Fed. Rep. 905; The Newport, 44 Fed. Rep. 445.

Considering the claim of La Champagne’s pilot, that the practice of phots was that vessels coming up the south channel should give way to the big ships going down with the ebb-tide through the main channel, the initiative taken by the Lisbonense, without reason or necessity, and in violation of article 22, if the ordinary rule were in operation, when at least three-fourths of a mile distant, afforded the pilot of La Champagne some ground for the supposition that the Lisbonense intended to waive the ordinary rulo of the road, in accordance with the practice, as he claims to have understood it. But, on the other hand, if such had boon the intention of the Lisbonense, the more appropriate course would have been to go to the westward, with a signal of two whistles, passing under La Champagne’s stern. In fact, the signal of one whistle, given under the present international rules, and under the obligations of article 19, in the circumstances of the Lisbonense, and with her actual destination, was so singular a signal that I have some doubt whether the pilot of the Lisbonense, whoso experience was not of the best, was mindful at the time he gave that signal of the obligation that article 19 imposed, or had any thought of observing it. ’ That the master had no such thought is to be inferred from his testimony that they steadied to keep the ship “in *128the proper place for entering the swash channel.” The international rules of 1885 are, however, expressly made applicable to the “coast waters,” and have been repeatedly held applicable in this situation. The Aurarda, 29 Fed. Rep. 98; The Non Pareille, 88 Fed. Rep. 524. La Champagne, therefore, not only had the right, but was bound, to regard the signal given by the Lisbonense as notice that she was going off to starboard, and she did so understand it. This notice caused La Champagne to abandon the course to port which her pilot was about to adopt, and, after proceeding as far towards the southerly side of the channel as was safe, (which I am satisfied she did,) all that her master and pilot were required to do was to observe the movements of the Lisbonense, and, when risk of collision appeared, to stop or reverse in time to avoid it. This is precisely what they did. They had the right to count upon the course of the Lisbonense to starboard which her signal had promised. They had no reason to suppose she was doing otherwise; and, when La Champagne reversed, she reversed in time to avoid collision without aid from the Lisbonense, had the Lisbonense been going upon a course to starboard which her signal had promised, and which La Champagne supposed she was pursuing. Had the Lisbonense been on that course, clearly no collision would have occurred, for La Champagne would have been stopped before the Lisbonense was reached. The master and pilot of La Champagne had no notice, as I have said, that the Lisbonense was not performing her duty, or had not taken the course to starboard which her signal indicated; and they are not chargeable, therefore, with fault in not reversing sooner. As it was, two seconds more time would have saved the disaster, although the Lisbonense had precipitated collision by resuming her former course towards La Champagne before her last porting. This last maneuver of the Lisbonense was not, in fact, of any benefit, but disastrous. Her duty at a distance of 1,500 or 2,000 feet from La Champagne, when the three whistles were heard, was to stop and back, in accordance wfith article 18. Had she obeyed that article, a plot of the navigation shows that she would undoubtedly have thereby passed considerably astern of La Champagne. She was at that time too far off to plead a situation in extremis, for I wholly discredit her testimony as to her distance at that time; and, even if in extremis, the plea would not avail, because the peril was by her own prior fault in changing her heading towards La Champagne. The Elizabeth Jones, 112 U. S. 514, 526, 5 Sup. Ct. Rep. 468; The Maggie J. Smith, 123 U. S. 349, 355, 8 Sup. Ct. Rep. 159.

The primary cause of the collision, however, was the neglect or misapprehension by the Lisbonense of the obligation imposed by article 19, after her signal of one whistle was given. All that followed was the natural sequence of the disregard of that obligation. It was no fault in La Champagne that she did not anticipate or guard against the Lisbonense’s violation of her duty, for at night her fault could not be perceived. No other fault being established against La Champagne, the libel of the Lisbonense is dismissed, and that of La Champagne sustained. Decrees may be entered accordingly, with costs.