Brown, J.
At a little past 12 o’clock on the night of December 2, 1883, when the steam-ship Alaska, bound for New York, was about 12 miles S. ¡3. E. from Eire Island light, the pilot-boat Columbia, in. preparing to put a pilot on the Alaska in answer to her signal, was run down and sunk, and all on board perished. The libel and supplemental libel were filed by the representatives of the principal owners of the Columbia, and by the personal representatives of four pilots and the cook, who wore on board at the time of the accident, and by the widows of the pilots and of the cook, to recover for the loss of the pilot-boat, the loss of personal effects, and for the loss of support. The circumstances proved are sufficient to identify the boat run down by the Alaska as the pilot-boat Columbia. Neither the boat nor the men wore ever heard from afterwards, and the pilots and the cook named in the libel and supplemental libel are proved, to have been aboard. All the evidence in the case as to the circumstances of the disaster is derived from the officers, seamen, and passengers on board of the Alaska. The principal facts are as follows:
At 11: 40 p. m., the Alaska then heading about W. by N., the pilot-boat’s torch was observed, bearing about S. W. The steamer answered with a blue light, indicating that a pilot was desired. The helm of tho Alaska was thereupon starboarded, so as to approach the pilot-boat, until she headed W. by S. S., when her helm was steadied. The wind was blowing a gale from N. W. The night was dark, [706]*706but not thick. The pilot-boat shaped her course to the northward and eastward, so as to intercept the course of the steamer, and at the proper time to launch a yawl, as customary, to carry the pilot to the steamer’s side. A ladder and light were placed at the steamer’s gangway on the port side, which was the lee side, as a signal to the pilot where he would be received. This was about 200 feet abaft the stem. When the Alaska steadied at W. by S. S. the pilot-boat bore about two points off the Alaska’s port bow. The master ordered the. helmsman to keep the pilot-boat at least two points off the steamer’s port bow, and to mind his port helm accordingly. Under these directions, as the pilot-boat hauled to the westward, the Alaska’s course was correspondingly changed until at 12: 06 she headed W. J S., and at the time of the collision, about 12:10, she headed, as the wheels-man testifies, W. \ N., nearly her original course.
When the pilot-boat was first sighted, and for some 12 minutes afterwards, the Alaska was making about 14 knots per hour. Her speed subsequently, and at the time of the collision, is one of the controverted questions in the case. The narrative of the log is as follows :
“At 11:40 observed pilot-boat’s torch bearing S. W.; 11:52 proceeded half speed. At 11:57 proceeded slow. At 12:06 stopped engines, ship’s head being W. 1/2 S., pilot-boat’s light bearing about S. W. by W. 1/2 W. At 12:08 observed the pilot-boat attempt to cross our bows, lteversed engines full speed, and in about two minutes she came into collision with our stem, sinking almost immediately. T2:13 stopped engines, and used every means of saving life with life buoys, lines, and also sending away a boat at 12:20. Cruised about in the vicinity of the disaster until daylight; then steamed around with a lookout at the mast-head, and, seeing nothing, proceeded on our course at 7:30.”
The entries in the engineer’s log agree with the above. The evidence of the pilots and others, called as experts, showed that there are three different methods pursued by pilots intending to board steamers when approaching in front of them in a strong head-wind. The first is for the pilot-boat to sail down into the lee of the steamer, and there launch her yawl, as Capt. Murray expected would be done in this ease; the second, for the pilot-boat to sail across the steamer’s bow, pass down on her windward side, wear around her stern, and launch the yawl as she comes up on the steamer’s lee quarter; third, to launch the yawl ahead of the steamer, so as to let the yawl go down upon her lee side, while the pilot-boat crosses her bows, and goes down to windward, and rounds her stern to pick up the yawl. The latter was the course pursued in the case of The City of Washington, 92 U. S. 31, and the course manifestly intended by the pilots in this case. One of the most experienced pilots called as an expert testified that the proper course for the steamer, after signaling the pilot, under circumstances like the present, is to make towards the pilot-boat, and come substantially to a stop,—that is, not exceeding half a knot or a knot an hour, when off the pilot-boat’s lee bow, and [707]*707a few hundred feet distant; and that in that situation the pilot-boat may properly pursue either of the last two methods; hut that the latter is not prudent or justifiable at night, in a strong wind, when the steamer is evidently in motion, and when her two colored lights have not been seen, and she appears to be keeping off to windward. The last two methods are deemed preferable to the first as a general rule, because the pilot-boat is thereby better enabled to keep clear of the yawl, and to keep control of her own motion. 92 U. S. 40.
Although both these latter methods have been long practiced by pilots, the master of the Alaska testified that he had never before known such an attempt at night. His testimony, and that of the officers of the Alaska, leave no doubt that the pilot-boat was expected by them to sail down upon the lee side of the Alaska, and there launch her yawl; and that it was not until the Alaska’s engines were reversed at 12:08, “about two minutes,” as the log says, before the collision, that the officers of the Alaska had any idea that the pilot-boat was intending to cross her bows. Their account of the disaster is, and the answer states, that immediately before the order to reverse was given, the pilot-boat, being then two points on the Alaska’s port bow, was observed suddenly to close in rapidly across the steamer’s course, when apparently only a short distance ahead. It is supposed that the pilot-boat then luffed into the wind, so as to reduce her speed, for the purpose of launching the yawl. Her previous speed is estimated at some seven or eight knots. The carpenter testified that he saw the pilot-boat suddenly luff when within 50 or 100 feet of the Alaska’s stem, and that he then saw, as he thought, one end of the yawl resting upon the pilot-boat, and the other end in the water; which, if true, would indicate that something unusual had happened in attempting to launch the yawl. The hull of the pilot-boat, after striking the steamer’s stem, was not again seen. It began to sink immediately, and passed along the starboard side of the steamer. The master ran down the ladder from the bridge, and saw from the starboard rail only the light, the mast, and the peak of the pilot-boat’s sail above water; and she disappeared altogether when about amid-ships. The yawl was capsized, and passed along the steamer’s port side. Several men were seen clinging to it, and were heard calling for help. Another man near it was clinging to a spar. The lines thrown out to the men from the steamer failed to reach them as they passed astern, and none were afterwards found.
On the part of the steamer the contention is that her headway was substantially stopped; that the pilot-boat, upon luffing, and while waiting to repair the supposed accident that happened to the yawl, was blown against the starboard bow of the steamer by the strong wind, upset, and stove in. The libelants contend that the steamer was under considerable headway, and ran upon the starboard quarter of the pilot-boat while she was engaged in launching [708]*708her'yawl, and cut her in two, before the pilots, who had a right to suppose the steamer to be stopped, or nearly so, discovered their mistake. •
The case, in many of its aspects, resembles that of The City of Washington, 92 U. S. 31, in which the respective duties of the steamer and the pilot-boat were fully and carefully discussed. In that case, as in this, the steamer was bound for New York, and the wind, was strong from the north-west. The pilot-boat, however, approached from the northward instead of from the southward. The City of Washington, after first porting her helm so as to approach towards the pilot-boat, when the latter was near to the line of the steamer’s course, starboarded her helm so as to turn her bows somewhat to the southward, towards the direction in which the pilot-boat was moving. In these respects the two cases are perfectly analogous. An examination of the “Apostles” also shows that in that case, as in the present, the engine had been first slowed, and then reversed full speed; and it was claimed there, as here, that the quick-water of the reversed propeller had reached amid-ships before the collision. The court found the steamer in fault for not sufficiently stopping her headway, and for starboarding so as to put herself in the way of the pilot-boat’s crossing her bows. The court also found upon the proofs that the crossing.of the steamer’s bows was justified by the custom of pilot-boats, and was not a fault on the part of the pilots. In the district court, (6 Ben. 138,) Benedict, J., said, (page 140:)
“The starboarding is admitted in the answer, and, under the circumstances disclosed by the evidence, I consider it negligence. The course of the pilot-boat was known to be crossing that of the steam-ship, the breeze was fresh, and it was known to the steam-ship that, at her request, the pilot-boat was endeavoring to place a pilot on board her. This maneuver the pilot-boat was entitled to be permitted to accomplish without embarrassment from the steamer. Certainly the steamer, by starboarding and giving herself a course across the course of the pilot-boat, while the yawl, which was to be picked up by the boat after the pilot was placed on the steamer, was in the act of passing to the steamer, attempted a maneuver which cast upon her the risk of its success. I think, also, that it was the duty of the steamship to stop still before she reached the pilot-boat; instead of which she was kept moving ahead,—slowly, it is true, but yet with a momentum which, with the star-boarding, brought her upon the pilot-boat, and sank her.”
In the circuit court, (11 Blatchf. 487,) Woodruff, J., said, (page 488:)
“The steamer was in fault in not slowing, and, if necessary, stopping, at an earlier moment, and before coming into such dangerous proximity to the sailing vessel. Indeed, upon the proofs, it was the duty of the steamer to stop to receive the pilot attempting to board her in the night season.”
The supreme court confirm these positions. 92 U. S. 38-41.
The same faults are charged upon the Alaska in this case that were established against the City of Washington, viz.: (1) Failure to [709]*709stop sufficiently; and (2) veering, up to the last moment, across the pilot-boat's course. That the Alaska did turn to the northward about a point during the last three or four minutes is shown by the log and the testimony; but I am of opinion that this would have been immaterial had the speed of the Alaska for a minute before the collision been reduced to the minimum consistent with holding her position,—say to half a knot, or even a knot, an hour. The libelants’1 evidence shows that that reduction of speed is regarded, even by pilots, as a substantial stop, and all that is required. In this case, therefore, the fault of the Alaska turns mainly, if not wholly, upon the question of her speed during the two or three minutes before the collision. If she was under a headway of several knots, that speed, and her veering to the northward, were both material faults, which combined brought about the collision.
Ecpeated consideration of all the evidence has failed to satisfy me that the Alaska, at the moment of collision, had come to a stop, or to a speed not exceeding half a knot or a knot an hour, i. e., the minimum headway sufficient to keep her in position. On the contrary, I am satisfied from some direct testimony, and from many circumstances of the case, that she was going at least four knots, and that the best-established facts of the case cannot be reconciled on any other view.
On the part of the libelants, the principal direct evidence of considerable headway in the Alaska, up to the moment of collision and afterwards, is derived from the testimony of several witnesses, who testify that, within a few seconds after the collision, parts of the wreck were seen near the bridge, some 200 feet aft of the stem, passing astern at considerable speed. Mr. Worcester and Mr. Challoner, two highly intelligent and competent passengers, who saw the capsized yawl and the spar, and the men clinging to them, estimate that they went astern along the side of the ship at the rate of six or seven knots. Mr. Worcester says it was about 10 seconds, and could not have exceeded 20 seconds, from the jar of the collision to the time when he was at the port rail near the bridge, 200 feet from the stem, and saw the capsized yawl and spar about abreast of him, and perhaps 50 feet from the steamer’s side; and Mr. Challoner, who saw the same from the port quarter, confirms this estimate. On the starboard side the' sinking mast, and the light, and the peak of the sail, were seen by the master and by the purser at about the same distance from the stem/ The master estimates the time to have been about a minute and a half after the collision; but no such interval is accounted for by any acts of his in the mean time. It is not probable that an energetic and alert commander like Capt. Murray would wait any considerable interval before stepping a dozen paces to the rail to see what was hap-'' poning to a sinking vessel along-side. The purser says the interval-was not over five seconds after he felt the slight jar of the collision; when, rushing from his room, he reached the rail to seo what was the [710]*710matter. The second officer, standing on the port side, 75 feet from the stem,felt the collision; and he testifies that he saw the capsized yawl abreast of him, with two men on it, “almost immediately”after. These witnesses state just what they did between the moment of collision and seeing those parts of the wreck abreast of them. While the precise number of seconds is not pretended to be stated accurately, the time it took for the wreck to reach the bridge, if they tell the truth as to what they did and saw, must have been short,—apparently not over one-third of a minute at most. This would indicate six knots as the combined motion of the drift and of the Alaska’s headway. Besides this direct evidence, there are several other circumstances indicating considerable speed in the Alaska, to which I shall presently refer.
The evidence that the steamer was stopped, or nearly so, is derived (1) from computations showing that from the length of time that the engines were at “half speed,” and “slow,” and “stopped,” and “re.versed,” as stated in the log, and from the number of revolutions at “half speed,” and “slow,” as stated by the engineer, there could not have remained any headway at the time of collision; (2) from the testimony of a number of the officers and men, who say that, in their judgment, the Alaska was stopped, and had sternway on at the moment of collision; (3) from their testimony that the wreck when seen abreast was going only slowly astern; (4) that the quick-water from the propeller was at the same time seen amid-ships, indicating a complete stop. The passing of the wreck astern is ascribed by the claimant’s witnesses solely to drifting in the high wind and sea.
In considering the weight of proof in support of these views, the libelants are entitled at the outset to the benefit of the natural probabilities arising out of the circumstances of the case. Pilots are among the most skillful seamen in the world. These were trained and experienced men. The Columbia was schooner rigged, but 88 feet long over all, and capable of being maneuvered with great quickness and dexterity. It is improbable, in' a high degree, that if the Alaska were substantially stopped, or moving at the rate of half a knot only, the Columbia, in launching the yawl ahead of her, with a full complement of skillful seamen, should either have sailed down, or been suffered to drift down, upon the Alaska’s bow while the latter was at rest. This is so improbable as to be almost incredible, unless she were disabled. Mere difficulty with the yawl could not account for it, and her filling away just before she was struck shows that she was not disabled.
Again, the natural bias of the Alaska’s officers and men in her behalf cannot be disregarded. Constant experience illustrates the effect of this influence, however upright the intentions of the witnesses.' Where the narrative of both sides is heard, its effect may be said to be neutralized; but where the lips of all on one side are closed, great caution is obviously necessary. When all on one vessel [711]*711are lost, it is not too much to require that an account derived wholly from the other shall be in its essential features consistent, rational, and probable; and that, in so far as it involves serious departures from these conditions, it should not be accepted, unless sustained by proof about which there could be no mistake.
The effect of this natural bias is apparent, I think, throughout the claimants’ case. A single entry in the log illustrates the subsequent changes that the same matter undergoes in the testimony. The log says: “At 12:08 observed the pilot-boat attempt to cross our bows. Eeversed engines full speed, and in about two minutes she came into collision with our stem, sinking almost immediately.” This entry was made shortly after the occurrence; even then presumably not wholly free from the tendency to excuse the ship. The expression adopted, “in about two minutes,” there naturally signifies “nearly two minutes,” or between one and two minutes. But on the. trial all the officers, except the master, who in his first statement calls it two minutes, call the interval three minutes. Again, the time of collision is not stated, evidently because the clock was not at that moment observed; but on the trial the third officer says he did observe the clock, and that it was 12:11, i. <?., making just the three minutes. But if the time had been observed to be three minutes, that would certainly have been stated in the log, as most favorable to the ship. Considering that the various other entries are made according to the time by the clock, it is highly improbable that 12:11 should not have been entered if it had been observed. Again, the log says the collision was with the stem; the answer and the testimony represent it as with the starboard bow, and not with the stem, —an important difference in several relations. Again, there is no intimation in the log of any change of course by the pilot-boat, nor of any sudden direction across the Alaska’s bows. The pilot-boat’s course was from the first necessarily directed across the Alaska’s bows. It was known to be so, because she wished to intercept the steamer. She was expected, however, to haul down upon the steamer’s lee side when she had approached near. Instead of doing as expected, she kept her course, as the log naturally imports, and tried to cross the Alaska’s bows; but in the answer and the testimony we have a “change of course,” a “sudden luff,” a “sudden closing in” of the light. In the log nothing of this kind is intimated. As respects each of these four particulars in a single entry, the officers were fully as able to state accurately, and quite as likely to do so, when they made up the log as afterwards. I must regard the log as the best evidence, where special reasons do not appear for departing from it.
1. The Alaska’s narrative as respects the occurrences of the few minutes preceding the collision, and her claim that her headway was then either fully stopped, or reduced to the minimum of half a knot, involve so many improbabilities, difficulties, and inconsistencies, as to prevent its acceptance.
[712]*712(a) Her witnesses say that at 12:08, two or three minutes before Ihe collision, according to their reckoning, there was a sudden luff by the pilot-boat,—a sudden “closing in” of her light across the Alaska’s bow. Most of them say that this took place when the pilot-boat was only 50 or 100 feet distant. The master and some of the officers call it 250 feet. The third officer, in one passage, calls it 500 feet. All, however, agree that when the light suddenly closed in the pilot-boat bore about two points off the Alaska’s port bow, as she had done all along. No one estimates the speed of -the pilot-boat in approaching .the Alaska at less than seven knots, until her supposed luff. In luffing to launch the yawl, pilot-boats do not come to a stop, but only check their speed. When her light was seen to be suddenly “closing in,” her course must necessarily have been directed across the Alaska’s course, and at first her speed would be but slightly checked. Bearing less than two points off the port bow at 12: 08, (because closing in rapidly,) if not then over 500 feet distant, she faould have had ■less than 200 feet to travel; and she would therefore have crossed the Alaska’s track in much less than a minute,—if only 100 feet off, in less than 10 seconds. But the interval sworn to is three minutes,— the log says “about two minutes.” . If it be said that the pilot-boat might have come to a stop when ahead of the Alaska, on account of the supposed mishap with the yawl, it is incredible—whether such an accident happened or not—that the pilot-boat should have lain still until the Alaska ran .over her, or until she had drifted down upon the Alaska’s stem with the latter at rest. But the evidence does not .admit.of any such stop; for the passage of the light across the .Alaska’s course must in that case have stopped also, and that fact would have been noticed.. The evidence indicates that the light, without stopping, ■drew directly across the Alaska’s bows until the moment of collision.
(6.) That the pilot-boat had been kept about two points off the Alaska’s port bow until she closed in to less than two points at 12:08, is one of the most certain parts of .the claimants’ evidence. But bearing-less than two points at 12:08, the pilot-boat, in order to reach the place of collision, had much less distance to travel than the Alaska. Whatever the pilot-boat’s speed, the Alaska’s must have been considerably greater. Whether the interval of time after closing in was more or less than two minutes, the pilot-boat’s speed, as she did-not stop, couíd not have been -less than two or three knots, and if she luffed “suddenly,” it must have been more. The necessary inference is that'the Alaska’s speed exceeded that.
(c) The supposed accident to the yawl is not sustained by sufficient e.vidence to be accepted as a fact. It is supported by the testimony of the carpenter, Duffy, only. Of the half a dozen other men whose business.it was to watch the Columbia, and who were watching her, not one saw the yawl. Duffy was not watching her, but as he looked out, of the port bow to take the draught, he says he saw the Columbia luff about 50 or 100 feet off, show her side, and the yawl with one [713]*713fend in tlie water. Ho did not remain at the bow, but wont away. He estimates this as from five to ten minutes before the collision, and tells what he did in the mean time. If the interval was one-quarter of his estimate, the Columbia must have been at least several hundred feet áway, and he could not have distinguished in the night-time any such peculiar situation of the yawl upon a small schooner like the Columbia. It is certain that the Columbia did not lull so as first to show her side when within 50 or 100 feet of the Alaska; and if she had done so when -only two points off the Alaska’s port bow, she would have shot like an arrow across the Alaska’s path. The close proximity -of the Columbia is a necessary condition of any probability in JDuffy’s story; and the other particulars stated by him, as well as by ©titers, disprove any such close proximity. His cross-examination shows ike weakness of his testimony. The supposed accident to the .yawl is ¿nmatcrial,"except as a clue to explain the delay of the pilot-boat in getting out of the way. . But if the Alaska was then at a substantial stop, Hie explanation is inadequate; if she was not at a substantial stop, the supposed accident affords no justification of her fault.
(d) 'The corarse of the pilot-boat is put by the Alaska’s witnesses all the way from N. N. E. to E. N. E. The master judged that if she had carried colored lights she would have shown him her red light before her sudden closing in across his bow. As the Alaska was then .heading W. S., and the pilot-boat was only two points on her port bow, iia order- to show her red light her course must have been nearly E. N. E. Bat since at 11:40, 28 minutes before, she bore nearly. -S. W.., it is Hear that, upon a course of E. N. E., from 11:40 she would not have approached the Alaska at all. The tracing of their positions according to the bearings sworn to, and allowing the highest possible -speed of the Columbia, proves that the general course of the pilot-boat after 11: 40 could not possibly have been more to the, eastward than N. N. E. Her heading was probably about N. by E., and her course about N. by E. E.; or, if her angle of leeway was a full point, she probably headed N. E., with the wind—the dire.c-, tion of which is not certain to a point—N. W. ^ W. The fact that she approached gradually, keeping about the same distance pff the Alaska's port bow, shows that her course was probably not much changed until she luffed. Had she changed at any time three or,, four points to the eastward,—say at 12: 06,—so as to show her red light, she would have broadened off the Alaska’s bow to much more than two points, instead of closing in as she did. The more rapid sheer of the Alaska to the northward during the few minutes preceding 12:08 shows that the pilot-boat w'as hauling towards her., more rapidly than before, and not bearing off to the eastward. Her; hauling in resulted naturally from her nearer approach, though con-, t tinning the same course as before. The evidence shows that from--, 32:02 to 12:06 the Alaska sheered half a point to the northward; ” [714]*714from 12: 06 to the collision, a point; and prior to 12: 02, only half a point,—from her previous course of W. by S. S., as fixed some 15 minutes before. The evidence of the first officer and helmsman, who relieved the watch at 12: 02, show, I think, that the master is mistaken in supposing that the Alaska had ported to W. J S. before that time. The fact, moreover, that she sheered a point to starboard after 12:06, though her helm was not nearly hard over, and although the wind operated strongly against a starboard sheer, is further evidence of considerable headway.
(e) The general course that the pilot-boat must have taken from 11: 40 to the collision can be determined very nearly from the known course of the steamer, and the distance traveled by her, and from the bearing and speed of the pilot-boat. The average course of the steamer, during the whole interval, was about W. by S. The distance run by her, even upon the figures given by. her own witnesses, could not be less than four and one-half miles, as she ran at least two and three-fourths miles during the first 12 minutes. The speed of the pilot-boat was stated to be from seven to eight knots. During the whole interval of 30 minutes the pilot-boat would therefore run from three and three-fourths to four miles. If the pilot-boat bore S. W. or S. W. ¿ W. at 11: 40, her utmost speed of eight knots would not have been sufficient to enable her to reach the Alaska, had her general course been more than one and one-half points E. of N.
(/) The master estimated the pilot-boat to be two miles distant when he slowed at 11: 57, 11 or 12 minutes before the collision, and “nearly a mile” distant at 12: 06, when he stopped the engines. But at 12: 06 the distance must have been much less,—probably lesg than half a mile. As the pilot-boat during this interval of nine minutes must have been going within two or three points of a right angle to the course of the Alaska, the latter must have diminished the distance that separated them during those nine minutes by nearly a mile, or at an average speed of over six knots. So, if four minutes before the collision they were nearly a mile apart, even if they had been approaching all the time head on, if the pilot-boat could not make but the rate of eight knots during that time, the Alaska must have made nearly seven. But during the last two minutes, according to the Alaska’s own account, the pilot-boat had luffed up so as to draw rapidly across the Alaska’s bow. If, therefore, at 12: 06, the vessels had been only one-half mile apart, instead of nearly a mile, they could not have reached each other, upon the course the pilot-boat sailed, had ,not the Alaska made an average speed of at least four, knots in that interval.
•(g) Again, if the Alaska’s speed when her engine stopped at 12:06 was only three and one-half knots, and her headway was stopped at the collision five minutes afterwards, as the Alaska’s witnesses contend, she would have gone during this interval about 1,050 feet only, and the movements of the pilot-boat during these five minutes could [715]*715not be rationally accounted for. At 12: 08 the pilot-boat bore a little less than two points off the Alaska’s port bow; and had she been coming up on her general course, N. by E. E., and the Alaska in the moan time moved as slowly as claimed, the pilot-boat would have borne, at 12:06, two minutes before, nearly four points off the Alaska’s port bow, instead of two, as the log states. Had the pilot-boat, however, during these two minutes, come upon a course of N. E. by E. so as to preserve her bearing of two points off the Alaska’s port bow, then, having the wind aft, her speed would have been increased, and on luffing suddenly just before 12:08, within 500 or 600 feet of the Alaska, she would have come round so rapidly as to cross the line of the Alaska’s path, not 250 feet distant, in less than half a minute, instead of three minutes. Luffing suddenly from such a course, her speed would not be much cheeked, and she would continue at rapid speed till she headed north; nor could she have been kept for three minutes from crossing the Alaska’s path without being brought to a stop for a considerable interval. Not only does the testimony, as I have said, show that there was no such stop, but such a mode of navigation for a pilot-boat designing to launch her yawl ahead of the steamer would be in the highest degree improbable. She would naturally cheek her course gradually as she approached the steamer’s head by a gradual, not a sudden, luff.
All the difficulties above referred to proceed really from two assumptions on the Alaska’s part: First, too little speed; second, too much time after the engines were reversed. Correct these, and all the difficulties vanish.
2. As regards the time, I have already observed that the entry in the log, “about two minutes,” was evidently an estimate only. The evidence shows that when the engines were reversed collision was apprehended. Experience proves that persons expecting disaster, and not busily employed, overestimate the time. It is natural, therefore, that this interval, then estimated at “about two minutes,” should have been much overestimated. It is probable that the time was not over one minute. The fact that the pilot-boat did not stop, and was loss than two points off the port bow at 12: 08, and yet did not clear the steamer, makes it difficult to believe the interval was over a minute. If it was .not, the order to reverse would have had little effect. Other circumstances confirm this view.
At 12:08, when the officers first perceived that the pilot-boat meant to cross the Alaska’s bow, the commander exclaimed: “My God! what is that man trying to do?” The first officer exclaimed: “J3y Jove! that man will be into us; I never saw such a thing in my life !” These exclamations show apprehension of immediate collision; but as the pilot-boat was then less than two points off the Alaska’s port bow, and seemed to those officers to be from 250 to 500 feet distant, and had to run only half the distance to roach the point of intersection that the Alaska had to run, it is difficult to see why they [716]*716should have had such apprehension of collision if the Alaska at that moment was not running over two knots, as they estimate. The officers were not then aware that the pilots were intending to launch the yawl ahead. I think it was the Alaska’s own speed that made the danger, and the apprehension, of collision; and hence the immediate order, “Full speed astern.”
'Again, the log shows that the Alaska continued backing at full speed from two to three minutes after the collision. Had her headway been stopped, or nearly so, at the moment of collision, there would have been no need of backing so long; for that would have carried her far astern of the capsized yawl sought to be reached. But it was precisely what she would have done had her speed at the moment of collision been some four knots.
Again, when the lines thrown to the men failed to reach them, a seaman ran almost the whole length of the ship, nearly to the stern, where the buoys were kept, and lighted one, and threw it over. This must have occupied at least a minute from the time of the collision. The master estimates from one to two minutes. The yawl and the men had already gone astern, and the evidence shows that the buoy also w'ent first astern, proving that the steamer had headway even a minute after collision, and after two minutes backing. The buoy was soon afterwards overtaken and passed, through the continued backing of the ship, so that it was then seen abeam or forward of abeam.
3. The drifting of the wreck on each side of the Alaska at the rate of from four to six knots is in my judgment far too great to be ascribed tó the winds and waves in an ordinary gale. The hull began to sink immediately after the collision, and no rapid drifting was possible after the hull was under water. The capsized yawl, and the spar with the men clinging to them, could not drift at such a rate, even if exposed to the wind; but they were out of the wind in the lee of the ship. The evidence also shows that the different parts of the wreck passed along on both sides of the ship at about the Bame rate, and appeared opposite the bridge at about the same moment. Only one cause can account for this, viz., the forward motion of the Alaska.
4. Nor can I accept the Alaska’s suggestion that instead of encountering her stem the pilot-boat was capsized and smashed in by drifting or being tossed sideways against the Alaska’s bow just aft of the stem.- The log says the collision was with the stem. One of the officers says'the pilot-boat at the collision was right ahead; though another says her light, which was probably on her mainmast, was a little to starboard. When seen a few seconds afterwards only one mast was visible; the other had already disappeared. That must have sunk almost instantly. This does not seem to me likely to have .arisen from the pilot-boat’s being merely thrown on her beam ends on being tossed by a wave against the Alaska’s bow; but from being [717]*717cut through by the Alaska’s stem when the latter was under considerable headway.
5. The testimony as to the quick-water is of slight weight, because the time when it was observed is easily liable to be mistaken. A seaman testified that he saw the men drawn down by the quick-water amidships; but Mr. Challoner, some 200 or 300 feet further ait, saw the men still drifting astern rapidly. So the purser may have been mistaken as to the time of seeing the quick-water, or have mistaken for that the commotion of the sinking vessel.
6. The estimates by which the Alaska’s speed is arrived at by her witnesses are not convincing, -when opposed by so many circumstances and so many contrary indications. It is significant, moreover, that not one of the Alaska’s officers testified to her ordinary speed when her engines are working at “half speed” or “slow.” In the absence of all testimony on the subject, it is scarcely credible that the commander and all the officers are alike ignorant on this point. To omit direct evidence, and to resort to the computations of a mechanical engineer who never saw the ship, based on the engineer’s estimates of the number of revolutions per minute, that were not entered in the log, is a substitution of a very inferior kind of evidence, when much better was presumably in the claimants’ power.
In other cases the rates of steamers at “half speed” and “slow,” as compared with their “full speed,” has been often proven before me by the direct evidence of the officers who knew the facts. In no case that I recall have these rates been so much reduced comparatively as the estimates in the present case would make them. Usually what is called “half speed” is fully two-thirds of “full speed;” and “slow,” about half of “full speed,’’and that is so in other steamers of the size of the Alaska; and in fair weather, or in a moderate gale, the proportions remain about the same. These usual proportions, applied to the Alaska, would agree with all the other indications in giving her a speed of from six to seven knots at 12:06, and about four knots at 12:09.
Had the officers of the Alaska supposed that the Columbia intended to cross her bows, and go to windward, instead of porting her helm and going to leeward, as she approached the Alaska, no doubt the order to reverse the engines would have been given earlier, and her speed would have been brought down to the proper minimum. The master testified that the night was “a good one for seeing lights, hut had for estimating distances.” Mistake as to the pilot-boat’s distance through this cause probably contributed also to the delay in reversing the engines. The master’s estimate at 12: 06 that the pilot-boat was nearly a mile distant was from two to three times too great. The mistake was a natural one, as the, pilot-boat was a small object. It was partly from this cause, I think, and partly because the intention of the pilot-boat to keep her course and cross the Alaska’s bow was wholly unexpected, that the Alaska’s speed was not brought to a [718]*718substantial stop. Notwithstanding the emphatic testimony of the officers of the Alaska and others, the crossing of the steamer’s bows in this manner has been practiced by pilot-boats so long, and the duty of coming to a substantial stop for the purpose of taking on pilots is so well settled, that the Alaska cannot be held legally justified in assuming, up to nearly the last moment, that the pilot-boat would change her course, which was obviously across the Alaska’s bow, and on that ground be exempted from the duty of coming substantially to a stop. It was the duty of the Alaska to reduce her speed nearly to a stop, so as not to imperil the pilot-boat in any method of approach the pilots might think advisable.
It is not unlikely, on the other hand, that the pilots, at about 12:06, when about one-third of a mile distant from the Alaska, being then about 1,000 feet from the line of her path, in consequence of the great size of that vessel, made the opposite and equally natural mistake of supposing her to be only half the distance off she really was. Thus erroneously believing she was within some 500 feet of the steamer’s track, instead of about 1,000 feet from it,4 the pilot boat would luff .when too far away, in order to reduce her speed as usual, so as to launch her yawl. Having twice as far to go as estimated, and being therefore unexpectedly delayed in getting almost ahead of the steamer before she could launch the yawl, aided as this mistake would be through the steamer’s constant veering to the northward, the pilot-boat’s speed at length would become insensibly so much reduced that, when she did get nearly ahead of the steamer, the latter, coming on under moderate headway, instead of being nearly stopped, the pilot-boat probably had not speed enough remaining to admit of the usual and necessary dexterity in handling, so as to fill away quick enough to escape. The evidence shows that she did bear away, but not in time to clear the steamer. These views of the probable courses and speed of the two vessels fulfill all the conditions of the best data in the evidence; viz., the recorded bearings and times, and violate no natural probabilities. No other view that has been presented to me does this, and on careful study I have not been able to discover any other that does so.
Second. I must hold it a further fault in the Alaska that, having first starboarded so as to approach the pilot-boat in the proper direction, she afterwards ported, and under a port helm kept veering to the northward up to the moment of collision; thus delaying and thwarting the expectations and the maneuvers of the pilot-boat to launch the yawl, and then get away. Precisely similar was the course of the steamer that was condemned by the supreme court in the ease of The City of Washington. But for this latter fault the pilot-boat, notwithstanding the Alaska’s too great speed, would have gone clear. The fact that the pilot-boat’s general course was crossing that of the Alaska was. certainly known. It was the Alaska’s duty to come as near to a stop as practicable, and leave the rest to the pilot-boat.
[719]*719Third. The evidence of the expert pilots in the present ease shows fault on the part of the Columbia. They testify that so long as the leeward light only of the steamer is seen, or if she seems to be keeping away, or if the steamer is perceived to he under any considerable headway,no attempt should be made at night to launch the yawl ahead, and cross to windward; that such a maneuver could not be justified; and that when the failure to make the windward light shows that the steamer is keeping off, the pilot-boat should also keep off safe to leeward, and not attempt to launch the yawl ahead. This is certainly reasonable, and I cannot doubt its truth. This evidence materially modifies the custom .relied on in the case of The City of Washington. Since the date of that case the usage may have been changed somewhat with reference to vessels of the class of the Alaska. The Alaska, upwards of 7,000 tons burden, is at least twice the size of that steamer. With the powerful electric lights now employed by such steamers, and with the general lighting up of the whole ship, when the Alaska approached within a quarter of a mile, I cannot resist the conclusion that the fact that the Alaska was under considerable headway ought to have been apparent to the pilot-boat. In my judgment the steamer must have been going at that time at the rate of five or six knots. A careful watch would have shown that she was not at rest, nor nearly so. In a gale, moreover, such as then prevailed, I should hesitate to find upon the evidence here that a vessel like the Alaska was required to come to a perfect stand-still in the water. The evidence shows that is not now expected. She would very quickly become unmanageable, and fall off' into the trough of the sea,—a situation that no pilot would expect her to assume. Neither the supreme court nor the circuit court, in the ease of 'The City of Washington, declare it to ho the duty of a steamer to come to an absolute stop, unless necessary; and the testimony here shows that an absolute stop was not necessary. In the case of McLaren v. Compagnie Francaise, 9 App. Cas. 640, referred to by counsel, the statement of the head-note is not sustained by the opinion; and that case, moreover, was one of ordinary navigation, not one in relation to the exceptional conditions arising between a steamer and a pilot-boat. T can have no doubt that the pilot-boat was designing to launch her yawl when ahead of the steamer; because there was plenty of time and space for her to cross to windward, and round the Alaska’s stern, and there launch the yawl, had that been her intention. Not doing that, nor sailing to the steamer’s lee, she must have designed the only remaining course of launching the yawl ahead. Such a course was dangerous and unjustifiable in the gale of that night, even if it be ever justifiable in the night-time. It was still further unjustifiable, and a fault, to persist in this design when she failed to make the Alaska’s green light, owing probably to the Alaska’s greater distance than supposed, and to her veering to the northward, until her own spood was so re[720]*720ducecl that she could not be handled with the dexterity that is usual and necessary in order to avoid the steamer in such a maneuver.
Fourth. Under the circumstances of this case, and the difficulties of obtaining evidence of the facts, I must hold that the libelants did not delay beyond a reasonable time before filing the libel on the thirteenth of November, 1884; and that there was no such laches as should discharge the steamer from the maritime lien acquired less than a year previous, in consequence of the transfer in the mean time of the title of the Alaska by Mr. Guión to Mr. Pearce, in October, 1884. It appears that Mr. Pearce, her builder, had claims upon the ship, secured by mortgage, for a great proportion of her value; In the final settlement upon the repurchase, as I understand, he advanced to Mr. Guión about £7,000 cash, besides canceling his other claims. Although the libel had not been filed at the time of this settlement, the accident was notorious; and the possible liability of the Alaska was a circumstance that could scarcely have escaped any reasonably diligent inquiry, had Mr. Pearce desired to ascertain all possible outstanding liens, and made reasonable inquiries in that regard. See cases reviewed in The Bristol, 11 Fed. Rep. 156; affirmed, 20 Fed. Rep. 800.
Fifth. As respects the right to recover damages in admiralty for the loss of life by the wrongful conduct of vessels on the high seas, some differences are found in the adjudications. It has been repeatedly discussed in its various aspects, and it is understood that the question is now pending in the supreme court. Awaiting the result of the determination of that court, and without referring to the common-law authorities, I shall hold in this case, as seems to me most consonant with natural equity and justice, that the pecuniary loss sustained by persons who have a legal right to support from the deceased furnishes a ground of reclamation against the wrong-doer which should be recognized and compensated in the admiralty. Cutting v. Seabury, 1 Spr. 522; Plummer v. Webb, 1 Ware, 75; The Sea Gull, Chase, 145; The Garland, 5 Fed. Rep. 924; The Harrisburg, 15 Fed. Rep. 610; The E. B. Ward, Jr., 17 Fed. Rep. 456; S. C. 23 Fed. Rep. 900; The Manhasset, 19 Fed. Rep. 430; The City of Brussels, 6 Ben. 370.
A decree may be entered for the several libelants to recover half their damages, with costs, and a reference taken to compute the amount.