On the 4th of January, 1887, at about 10 o’clock in the morning, as the large side-wheel passenger steamer St. Johns was making her daily trip from Sandy Hook to New York, she came in collision with the libelant’s canal-boat S. A. Pyatt, off Castle Garden, striking with her stem the starboard side of the canal-boat about 10 feet from her stern, and making a deep wound, which caused largo damage to the boat and cargo; to recover which this libel was filed. The canal-boat was in tow along the starboard side of the steam-tug Gen. Rosecrans, bound from near the Pennsylvania Feny dock, Jersey City, to pier 34, East river. The tide was ebb, and the tug and tow going at the rate of four or five knots. They were moving across the river, and were probably heading about two or three points down river, so as to round the Battery within about 400 or 600 feet of the shore. The tug had the St. •Johns on her starboard hand, and was therefore bound to take proper steps to keep out of the way. She had given a signal of two whistles, to which the St. Johns had responded with two, importing that the tug should go ahead of the St. Johns. For the tug it is contended that the steamer, contrary to her signal, improperly sheered to starboard and across the course of the tug, which the steamer denies. The latter claims that the collision was wholly owing to the tug’s stopping and backing shortly before the collision, instead of keeping on her course, as she should have done.

The case has to be considered, however, with reference to the presence of another steam-tug, the Delaware, which, soon after the Gen. Rosecrans left Jersey City, also left the dock next below, with a car-float about 200 feet long upon her starboard side. She also was bound for the East river, and followed a course about parallel with that of the Gen. Rosecrans, and from 100 to 300 feet below her in the river. Pier speed was greater than that of the Rosecrans, so that at the time the last signals Avere exchanged her float had got nearly abreast of the latter’s stern. When the two tugs were about one-third of the way across the river from the Jersey shore, the master of the St. Johns, then abreast of Fort William, and probably about one-fourth of a mile from it, observed them from two to three points on his port boAv; the Rosecrans being then considerably in advance of the Delaware. When.he Avas a little Avay above the fort, the Rosecrans bleAV him two whistles, and then the DelaAA'are blew to him two. He replied to the DelaAvare Avith one, to Avhich the DelaAvare ansAvered Avith *765one; thus agreeing that the St. Johns should pass ahead of the Delaware. The Gen. Itoseerans then repealed her former signal of two whistles to the St. Johns, to which the latter replied with two, and received an answering signal of two whistles from the Itoseerans; and it was thus agreed that the Itoseerans was to go ahead of the St. Johns. By these signals, about which there was no misunderstanding, the St. Johns was to pass between the two tugs, i. e., ahead of the Delaware and astern of the Rosecrans. She was heading nearly up the river, probably a little to the right; and, being bound for her usual landing place, pier 8, North river, her course was, I have no doubt, directed so as to pass as usual within 100 yards of pier 1. The St. Johns, until she passed Fort William, was going at her full speed of about 13 or 14 knots. Upon receiving the single whistle from the Delaware, and when about 300 yards distant, as the master estimates, her engines were slowed, and her speed was soon reduced to about 5 or 6 knots. The Delaware reversed her engines as soon as it was understood that the St. Johns was to go ahead of her, and Iher head swung around somewhat to the northward, so that the St. Johns passed about 100 feet to the eastward of her. The pilot.of the Delaware testifies that the St. Johns sheered a little to the eastward in order to pass him, and -that that was necessary to clear the Delaware. The pilot of the Koseeraus also testified that the St. Johns sheered to the eastward, so as to point to his bows when only 200 or 300 feet below him, in consequence of which, believing collision to bo inevitable, he ordered bis engines reversed, and gave notice to the persons on the canal-boat of the danger, leaving the pilot-house and going aft for that, purpose. The master of the St. Johns denies that the St. Johns sheered to the eastward at all, or that any sheer was necessary in order to avoid the Delaware; and, as the latter was herself swinging to port, the St. Johns would seem to those oil the Delaware to go to starboard, though she did not change at all. No reliance can be placed on tbo supposition of the pilot of the Delaware under such circumstances. Any such sheer is positively denied by the witnesses for the St. Johns, and is not corroborated by any further testimony, or by any apparent necessity or probability; and it must therefore be regarded as not proved. As soon as the Kosecrans was observed to be reversing, the St. Johns reversed full speed, gradually ported her wheel, and must have boon nearly stopped at the collision. She struck the canal-boat, as above stated, about 10 feet from her stern. The pilot of the Kosecrans testifies that shortly before the collision the engines were again ordered ahead, and the engineer says he got one or two turns ahead before he felt the blow. The master of the St. Johns testifies that the bells to go ahead were rung on the Gen. Kosecrans only at the moment of collision, and that, when the vessels struck, the tow was substantially at rest; and that his stern remained imbedded in her side for some little interval, until he backed out, without any further breaking or tearing of the wood-work, as would have happened had the tow then had any forward motion.

There is considerable difference as to the place of collision. Most of the witnesses for the Gen. Kosecrans put it abreast of Castle Garden, and *766about 200 yards from it, i. e., inside of the line of the piers. The master locates it about 400 yards S. W. by W. from the end of pier A, some 400 feet outside of the line of pier 1, which would be-about 450 yards from Castle Garden. Several of the witnesses wefe not in a position to observe the distance from the shore with any accuracy. I have no doubt, as above stated, that the place was near the line of the usual course of the St. Johns in approaching the landing, and not far to the.eastward of the place stated by her master. Holding to that fact, the precise place is not further material.

Upon the above view it is clear that the immediate cause of the collision was the fact that the Rosecrans stopped in the water instead of keeping on in accordance with the previous understanding by signals. She was the vessel bound to keep out of the way. She selected her own mode of doing so. She adhered to this choice, and repeated it, after she knew that the St. Johns was to go ahead of the Delaware. It is plain that, had she kept on, she would have cleared the St. Johns on the latter’s starboard side by at least 100 feet, i. e., by as much space as the Delaware had on the St. Johns’ port side, and probably, more; and that without any change of the St. Johns’ helm. This was sufficient space. The mode agreed on for avoiding each other was a proper one. .Under this agreement, had the Rosecrans kept on, there would have been no risk of collision. The St. Johns, under such circumstances, owed no duty to the Rosecrans, except not to thwart her attempts to keep out of the way by going ahead as agreed, which the St. Johns did not do; and except that, after risk of collision appeared through the Rosecrans’ fault, she was bound to do what was possible to avoid her. City of Hartford, 11 Blatchf. 72, 75; The Nereus, 23 Fed. Rep. 455, 456; The Vanderbilt, 20 Fed. Rep. 650; The Governor, 1 Abb. Adm. 108; The Greenpoint, 31 Fed. Rep. 231. Even with this stopping by the Rosecrans, she lacked but 10 or 12 feet of clearing. There was no reasonable or apparent necessity for stopping contrary to the agreement under which both had been acting. The St. Johns had a right to rely on the Rosecrans keeping on as agreed; and stopping, instead of being “necessary,” was the maneuver that tended to bring on collision, instead of avoiding it. Rule 21 has no application in such circumstances. The Northfield, 4 Ben. 112; The Britannia, ante± 546. For stopping, the Rosecrans must therefore be held to blame.

I do not think any fault -is established in the St. Johns. The courses of all these three vessels were direct, straight, and capable of being determined with reasonable certainty by pilots of ordinary skill. For vessels of but moderate size, all going at such moderate speed as these were going, and upon direct lines, in the day-time, and in clear weather, I cannot find that shaping their courses so as to allow a space of say 100 feet on each side for one to pass between the two other tugs with tows along-side is unusually close, or is dangerous or culpable navigation in this harbor, so as to involve the St. Johns in fault, as well as the Rosecrans. The ordinary practice, and the necessities of the harbor, do not require and often would not admit of more room being-taken, without *767causing unreasonable embarrassments and delays in navigation. The question of a reasonable space and margin for safety is almost wholly a question of circumstances. The situation here was as simple and as little complicated as possible. If was quite di fferent from that in the case of The Active, 22 Fed. Rep. 175, and from that of the Britannia and the Beaeonstleld in the case above cited. There, the Britannia’s course, instead of being direct, fixed, and determinable, like the St. Johns’, involved a swing of five or six points, and was therefore largely indeterminable. The fault of the Britannia was in coming into the wrong part of the river; and the only question as regards her was whether that fault actively and proximately contributed to the collision. It was the same in the case of The Active. Here the St. Johns'was going where she had a right to go; she was pursuing her direct and proper course; she did not, as L find, materially, if at all, deviate from it; she was not guilty of fault in shaping her course, or in consenting to go between the two tugs, as the Roseerans proposed she should go; and the St. Johns had already slowed. The Roseerans had no right to expect more; and she would have had a reasonably sufficient margin of safety had she kept on, as she was hound to do, and as the St. Johns had the right to rely on her doing. There was no apparent necessity for her contrary maneuver, and when the St. Johns sa w this dangerous maneuver she instantly reversed, and hailed the Rose-erans to go ahead. It is urged that the St. Johns’ helm ought to have been put instantly hard to starboard, instead of gradually; but any difference in the rapidity of putting the wheel over, when so near as they then were, could not have avoided this collision. The whole fault, I think, was with the Roseerans. Decrees accordingly, with a reference to compute the damages.