LEARNED HAND, District Judge
(after stating the facts as above). [1] We think that the collapse of the wharf created a presumption of negligence under the rule correctly stated in Hastorf v. Hudson River Stone Supply Co. (D. C.) 110 Fed. 669. That presumption does not change the burden of proof, strictly speaking, since the libelant, though it makes a case by showing the collapse, does not put upon the respondent the duty of satisfying us that it was not negligent. When the respondent once put in proof that the wharf and embankment were well made and well maintained,' it had done all that was required of it under the presumption. The libelant must convince the court of the truth of all its allegations in this as in every other case. In so far as Hastorf v. Hudson River Stone Co. means more than this, we do not agree. The better form of expression appears in the kindred case of inevitable accident. Inland & Seaboard Coasting Co. v. Tolson, 139 U. S. 551, 11 Sup. Ct. 653, 35 L. Ed. 270. It is true that sometimes the phrase “burden of proof” is spoken of in cases of inevitable accident (The Edmund Moran, 180 Fed. 700, 104 C. C. A. 552); but this is to be taken only in the sense that the respondent in such cases must by evidence exclude all the possibilities of neglect. Such cases do not profess to lay down a rule touching the duty of finally satisfying any doubts upon the crucial issues. The analogy applies, we-think, of bailments (The Genessee, 138 Fed. 549, 70 C. C. A. 673), where this distinction is generally observed (Stewart v. Stone, 127 N. Y. 500, 28 N. E. 595, 14 L. R. A. 215; Claflin v. Meyer, 75 N. Y. 260, 31 Am. Rep. 467). Therefore, as the respondent put in proof to show that the wharf and the embankment were made in accordance with proper principles of. engineering, and that the embankment went out without warning, because of circumstances which they could not control, it had discharged any duty arising from the presumption. The burden of proof remained on the libelant. Wigmore, §§ 2485, 2487.
[2] The first question is whether the wharf was made according to proper principles of engineering. No challenge is made of its construction, except the-length of the piles, the depth of the excavation in front, and- the embankment on which lay the tracks. The best evidence of the length of the piles is that of Bierd (folio 4720), who built the dock. • He was corroborated in this respect by Hunt (folio 726). The longest piles which would naturally be, and are stated to have been, those at the front of the dock, were about 75 or 80 feet long. As they ran back to the shore, they naturally decreased this length, [451]*451so that some of them were probably 60 feet long. The testimony of Clark that they were from 50 to 75 feet long (folio 651) does not materially vary and was not intended to be exact. If so, they were certainly driven below any possibility of exposure by whatever dredging took place. To take the figures most favorable to the libelant, we should assume 10 feet of them above high tide — a tide of 20 feet — and a depth of 30 feet close to the wharf at low water. This leaves 20 feet of the. pile driven below the surface of the 'mud. These figures are too high. We should rather take 9 feet above high water below low tide, which leaves 27 feet of grip upon the mud. It is of no great consequence which of these figures we assume. By no possibility could ilie toes of the piles be exposed by any dredging which took place. If'the dredging was a contributing fault, it was for sonic other reason than this. There is other conclusive testimony: Many of the piles were broken off at the time of the slide. Hull says that all were broken off which he examined, and we read his testimony as meaning all those at the edge of the wharf. Mears says (folio 2295) that 25 per cent, v ere broken, which is enough to include those at the edge, though he does not say which lie means. Those, which were not broken off bent outward, after the slide, which showed that their toes could not have been detached. Those which were broken off just below the mud line, the broken feet of which swung outwards as the slide carried away the wharf, give an opposite impression at first, but not when one considers that their broken toes pointed west. There is nothing in tlic suggestion diat the teredos had weakened them, for the testimony is uniform that all broke below the mud line, where the teredo does not attack.
The next question is of the; depth of the excavation. In view of the depth to which the piles were driven, this is a relevant question only because of the libelant's other theory that so much of the lateral support had been removed from the bank that it: swept between the piles and broke them off. Mears says (folio 2214) that they attempted to maintain for normal purposes a depth of 30 feet at low tide, but it is doubtful whether they ever succeeded. Later (folio 2313) he says that (hey did not get 30 feet, but only about 28 feet, which corresponds exactly with, the calculation reached by Judge Hough from the dimensions and position of the Newport after she sank. Clark says (folio 196) (liat “they were supposed to have 30-odd” feet 12 feet off the piles, but before the Cardenas dredged vessels of 22 feet or 23 feet draught lay aground (folio 197), and he thinks that ¡he dredge never came nearer than 25 or 30 feet of the piling (folio 290). This contradicts Geenziet’s version of dredging within 4 feet of the piles and McKay and Scliuber’s estimates, both of whom Judge Hough saw and discredited. Hunt merely says that the dredging was “inside of 12 feet” (folio 792), and that they dredged to 30 feet at low water. Yet they handled vessels of a maximum draught: of 25 feet 6 inches (folio 794). Moreover, some allowance must be made for 3 months’ fill after the Cardenas left. The contours upon Sartor’s map, Exhibit A, May 11, 1915, are at mean sea level, and the deepest of these, 35 feet, runs only under the stern of the Newport. The 30-foot contour at the bow, and further forward on the stern, is equivalent only to 22 feet. The [452]*452soundings near the collapse are, of- course, of no value to this inquiry. Sartor’s soundings of January, 1911, if at mean tide, give 28 feet, 30 feet off the pier end, and about 23 feet close to them. Drennan’s figures ■reduced to low tide, Exhibit 8, do not lead to the belief that as much as 28 feet was obtained at the pile ends, or over 30 feet at the next soundings, which are at a distance out from the piles of say 30 feet. They, like Exhibit A, May 11, 1915, are valueless over the area of the collapse. The contour lines of Exhibit T probably are not to be taken literally, though they are the basis of Spooner’s calculations. For what they are worth they show a mean tide depth opposite the Newport’s forward hatch of only 25 feet — 30 feet off. the -pile ends.
It is certainly impossible to ascertain exactly what the depth -was. Judge Hough made an elaborate calculation from the photographs of the Newport and her known dimensions, and reached the conclusion that under the Newport’s stern there was 28 feet at low water. This is as near accuracy as is possible. There is no reason to suppose that the foundations had been weakened by undue dredging. We can make nG sure deductions from Roquebert’s soundings. Probably the original dredging lowered the bottom over 20 feet as Bierd says, but the evidence shows that it preceded the driving of the piles. What the depth was close to the piles we cannot say, but we incline to put it at not more than 25 feet, though this is necessarily hardly more than a conjecture. These calculations are relevant only to a consideration of the general stability of the embankment, and therefore lead directly to a consideration of that feature of the case. There had been no change in this since the wharf was completed, about seven years before the collapse. The old tracks, two in number, leading to the old French fill, were carried on an embankment opposite the collapse, and three new tracks, about 18 inches below, and all substantially on a level, had been added to the west. Nothing more had been done to incumber the underlying strata from that time, on.
We do not forget so much of the fill for the new Isthmus Canal Commission’s railroad across the lagoon as was in place on August 17th. This can be laid out of consideration at the start, although some-tiring was made of it. It only touched the old embankment at one spot, and that at the extreme northern edge of the slide, and there it was about 80 feet back from the line of the break. The only possible effect could have been to push down some of the underlying mud, and we dismiss it as too speculative. Frear (folio 4293) was unwilling to commit himself upon its effect, and the point was not pressed with the other witnesses. There is no evidence that the two upper or eastern tracks had ever given any signs of movement, and the same is true of the middle track — that just west of the shed. The only part of the embankment which did show any signs of giving way was under the two tracks within the. shed, and it is true that these had from time to time prior to the collapse required to be resurfaced, as they had sunk a little, so that the floor of the car did hot come up to the floor of the dock. Clark says this was only in the roadbed for berth No. 1 (folios 206, 519), and the whole fall was not over 8 inches (folio 204); and while Hunt makes it extend across the back of berths 1 and 2 (folio 457), he says [453]*453it began in 1905, and, as we infer, continued off and on from time to time until 2 or 3 months before the accident (folio 758). Slayback (folio 1904) says that the only subsidence after Christmas, 1911, that he knew, was back of berth No. 1. Tysinger, who ought to have known all about it, says that the only resurfacing opposite berth No. 2 was in the early part of 1911 (folio 3594). McKay cannot give the date that the roadbed sank (folio 3963), though he does say it was for the most part opposite No. 2 (folio 3965), aud it seems likely from Slayback’s testimony that this was before 1912. We conclude that the total sinking had been in the aggregate 8 or 9 inches, and had been progressive, but that there had not been any opposite berth No. 2 for nearly 18 months before the collapse. e It seems undoubtedly to be the fact that in all structures of this sort there is a subsidence more or less progressive lor an almost indefinite time. Bierd, certainly a competent man, put it as high as 20 years (folio 4742) for piles under railroad bridges. Whether this applies to embankments is not clear.
Spooner’s criticism presupposes that the original slope as shown in Exhibit T was stable, and figures upon the angle of repose as therein shown. All his conclusions are based upon supposed subsequent excavation close to the pile ends of 50 feet below the top of the embankment where the slide began; i. e., the two western tracks. This involves a recent excavation at that point to a depth of 32 feet at low water. He concedes that there is no a priori angle of repose, but that it is contingent upon the material. Now it is obvious that the libelant may not blow hot and cold with Exhibit T. If the ground lines do not represent the situation on August 17, 1912, there is no warrant for assuming that they represent anything. Spooner has proceeded upon the assumption in his diagram, Eibelant’s Exhibit C of June 3, 1915, that the stable conditions had been changed for the worse by the Cardenas. Elis whole conclusion hangs in the air for this. Thus we must assume that the embankment, which was designed by a competent engineer, had shown substantial stability for 7 years, barring the other indications of movement which we shall consider later. We conclude that both wharf and embankment were proper in design and that nothing has been proved of their being weakened by dredging.
There remains only the question whether the libelant has succeeded in proving that there were such warnings of the slide as required precautions. The libelant alleges a number of such warnings, of which the sinking of the tracks just west of the shed has been mentioned. In addition to these are the bulging of the‘string-piece, the sinking of the iloors, the jamming of the doors, the slanting of the roof posts, aud chiefly and especially those movements which resulted in the tying hack of the angle wharf and the coal wharf some years before, and of berth No. 1 about a month before the accident occurred. Bierd, Mears, and Goethals all testified that most of those symptoms of disturbance were not indications of serious insecurity. Piles will sink, as we have already said, 15 or 20 years after they have been set, and even when driven to a “refusal.” One sinking of the floor took place before 1907, for it was in Bierd’s time (folio 4750); indeed, Hunt (folio 740) says it was shortly after the pier was made in 1905. it cannot be taken as an [454]*454indication of sinister weakness. There seems to have been another depression in 'the floor some 3 or 4 months before the accident (folios 346, 347) of an inch or more .(folio 349). In August, 1911, Geenzier says he found at one place the floor 10 inches down (folios 2833, 2835), but Slayback puts this in the northern half of the dock (folio 2004) abreast of where the bracings were afterwards put. Geenzier was certainly less accurate in details than the other witnesses, and although he may have been quite honest, he is not to be relied on as much as Slay-back and Tysinger for exactness. There is no doubt that there was such a “mashed or soft spot” in the summer of 1911, due to an overload on the dock at that time (folios 3457, 3458). But these were all direct depressions of the piles, and did not indicate any sagging outward of the wharf as a whole, whether they were in the northern part or about'the center, as Tysinger and Geenzier recall. The jamming of-the doors and the slanting of the uprights were proved, and occurred from time to time; but these can all be accounted for, as they were accounted for, by the movements of the tops of the posts or piles, which are apparently not serious. In so far as these were a part of the bulging of the piles toward the west, as they may have been, they properly come in for consideration with that; but, taken alone, they seem not to have been necessary signs of a disaster. The same thing is true of the curvature of the crane tracks to the west. This was due probably to the bulging out of the stringpiece near which the crane tracks lay, or perhaps to the vertical, sinking of some piles at such places as they dropped.
We therefore come to the movements to the west, the symptom which was the most important of all. That the piles were themselves straight, Hull, the diver, testified, who had been over them about a month before (folios 1468, 1471). On June 25, 1912, Tysinger went along the whole bottom of the dock (folio 3468) and found the piles perpendicular (folio 3471). The only trouble he found was that in one spot some piles had dropped 12 inches (folio 3472) caused by a drop in loading, the crane. Slayback examined under the docks from month to month (folio 1846) until the 1st of July, 1912, and he found them all upright (folios 1902, 3574). He thought there might have been a movement outward (folio 2008) of the piles, but it was accompanied by no inclination. There certainly was a movement of the whole wharf outward in June, 1912, and upon this feature the libelant properly enough relies very much, for it indicated a slipping of the whole substratum, even below the toes of the piles.
This was, moreover, not the first of such movements. The first of them occurred in Bierd’s time, before 1908 (folio 3538). It seems to have been due to putting in some fill to the east of the angle -dock at that time (folios 3539, 4899). O.ne of the caissons of the old French pier or steel dock also moved at the same time (folio 3543), and was apparently part of the same movement, perhaps caused by the same fill. The angle dock was then reinforced in the same’ way that berth No. 1 was reinforced in July, 1912. The coal dock also showed some movement (folio 751), probably some time in 1911, al[455]*455though the date does not appear. This was also tied back and gave no further trouble, like the angle dock.
In June, 1912, those movements, which the libelant insists were premo uitory of the eventual collapse, occurred, and it is of critical importance in our judgment to ascertain just where they were. This was one of the sharply contested issues in the ease, but, before entering upon it, it is well to remember that Tysinger and Slayback put in the deadmen iu berth No. 1 with the idea of correcting the disturbance where they had observed it. It is extremely improbable that they should have put in the bracing at any other place than where the disturbance was greatest. We may suppose that they did not cover all the area of the difficulty, as Mears was not there at that lime; but it is impossible to believe that, even though they did not go as far as they should, they selected for their bracing any other place than that which showed the greatest movement. There is a strong antecedent, probability, therefore, that the testimony of Schuber, McKay, and Geenzier that the maximum bulge was opposite where the accident afterwards happened is not true. In the first place, tve throw out Schuber’s testimony, not: only because Judge ITough disbelieved him, which we are bound to accept, but because, after reading it ourselves, he does not impress us even on paper. Nor are we disposed to put reliance on McKay, for the same reasons. Tysinger got word from the timekeeper that “part of the dock was going to sea” (folio 2465) in June, 1902, and on the 25th he went over and made an examination. He found about 200 feet of the track at the north end of berth No. 1 had moved 9 inches toward the west, and that, the piling had dropped down from the cap about 12 inches. Before acting he seems carefully to have watched for subsequent movement. The disturbance affected the -wharf for a distance of about 130 feet; and increased. It appeared to him as though the whole mass of the pier was moving, for the piles themselves did not incline; it is clear that iu fact the whole mass must have been in motion. Four days later the dock had moved out to 12 inches in all (folio 3476). There was no change on July 5th (folio 3477), or on July 8th, 13th, or 14th, on each of which he observed the area; but on July 19th he found that the total movement had become 14 inches, at which point it stopped (folio 3482). There was no change or disturbance adjacent to berth No. 2 (folio 3483); all of it was in the middle of berth No. 1, extending a little to the south towards berth No. 2. This was the occasion of the bracing, like that which had proved effective at the angle and coal docks, and which was completed on the 28th of July (folio 3485). Here, also, it checked the movement.
The libelant repeatedly relies upon Tysinger’s statement (folio 3572) that the greatest movement iu the lumber dock was “in the center,” seeking to infer that he had considered that the greatest movement was at 1 he. juncture of berths 1 and 2. But this is not the fair interpretation of his testimony. What he means is that the greatest outward movement was in the center of that part of the wharf which in fact did move. In other parts of his testimony he is perfectly dear as to the location of the movement as a whole. It was about 200 feet on the north end of berth No. 1 (folios 3466, 3467), which almost exactly cor[456]*456responds with the center of the area braced. Clark’s testimony (folios 216, 219, 270, 303, 368, 416, 417, 627) located the bulge at a different place, but all he saw was a bulge of 2 inches, which, if he is right, could hardly have referred to the bulge accurately observed by Tysinger and McKay. He placed this generally between berths 1 and 2 at about the center, but this location he first gave in response to a leading question (folio 216), and it is quite clear that his testimony is not to be taken as an attempt to fix the spot accurately. Hunt’s testimony is like Clark’s (folios 742, 743). McKay places it more accurately, and it must be admitted that he was in a position to know (folios 3961, 3962). The total bulge according to him was five hundred and twenty feet in length (folio 3391), starting 50 feet from the coal dock, and the greatest bulge of 24 inches was about the middle of berth No. 2 (folio 3993); over a distance of 150 feet the bulge was a foot or more (folio 4015), all of it within the area of the collapse (folio 4060). This, of course, contradicts Tysinger; but it is to be noted that, although McKay’s notebook gives a maximum of 24 inches, it does not give either the dates or the spot at which the bulge took place. Slayback says (folios 1904, 1925) that before July 3d, when he went away, there was one place in berth No. 1 that showed settlement, but that berth No. 2 was all right (folios 1932, 1933). His estimate of 6 or 8 inches of berth No. 1 (folio 1936) was probably about correct. Geenzier’s testimony is confused. At folio 2830 he says that the bulge began 350 feet from the angle dock and worked north. This would make the bulge begin about on the north end of the collapse, which corresponds with Tysinger’s testimony. He after-wards said that the greatest bulge was about 500 feet from the north end of the lumber dock (folios 2995, 2998), which, if taken literally, was way below the area of the collapse. Finally, under examination of his own counsel, he located the greatest bulging (folio 3055) as where the collapse occurred. He was confused as to his dates and his testimony is much too uncertain, particularly when taken in connection with his memorandum, which we shall refer to later. Schuber’s testimony is somewhat inexact, though he certainly placed the bulge upon No. 2; but he is discredited by Judge Hough and we do not rely on him.
Taking all this evidence into consideration, and especially the probability that the bracing was put in where the bulge is greatest, it seems to us likely that the only bulge thought of consequence was that opposite the braced part of the dock. It is quite likely that Clark and Hunt were right in saying that there was a bulge of 2 inches in March or April about at the division line of the two berths. S'uch a bulge, located as generally as that, was certainly not an object of alarm.- We conclude, therefore, that the bracing which was put in answered any alarming developments, and also that it was sufficient, as judged by the results of similar action in the past. This aspect of the case, therefore, comes down to this: Were the prior evidences of movement in the angle dock, the coal dock and the lumber dock sufficient to show that the structures taken together were instable, and, if so, what was the reasonable requirement under the circumstances? This, of course, is a question between experts, and on the one hand we have Goethals and Mears, who thought that the precautions were sufficient, and Frear, [457]*457Crary, and Spooner, who thought lhat either the bracings should have been carried further along or that riprap should have been put at. the bottom of the piles. Upon that question the libelant has the burden of proof, and we are not disposed to disturb the finding of the court below. While it is true that the dock had moved in the.three places mentioned, it is equally true that in the two earlier precisely the same remedy had checked any further movement, and it is also true that the movement had been stopped for over two weeks in berth No. 1; at least Tysinger says so, and his observations were accurate. It is easier to attribute a failure of foresight after the event than at the time.
But there is another consideration which seems to us of controlling consequence. In all three cases the movement, when it did develop, had been very slow, and the engineers had had ample time to correct it. If any other part of the docks showed signs of movement to the extent even of 14 or 24 inches, the engineers had no reason -to think there would have been a sudden collapse before they could remedy it. We are willing to agree that they were hound to be very watchful for movements elsewhere, but that they were bound to anticipate any such sudden collapse as took place seems to us entirely outside of the facts, and directly contradicted by the facts. Indeed, the witnesses of the respondent do not, at least expressly, go so far. Therefore we do not charge the respondent with the collapse for failure to extend the bracing further than it did, even assuming, what is in some doubt, that the bracing would have done any good, had it been put in. As to this the respondent’s witnesses are somewhat cautious. None of them would assert that, the bracing would have done more than retard the slide, while Goethals and Mears are very decidedly of the opinion that it would not.
The only remaining charge of negligence is the appearance of the cracks on the day in question, and perhaps on the 13th, according to Geenzier, and this brings up the memorandum which Geenzier says he made on August 13, 1912. Both he and Tysinger say that at that time he was at work in the lagoon — he says, repairing launches; Tysinger says he did have a talk with him that day, but that he went over into the lagoon and talked about other matters. Whatever the original form of the memorandum, it has certainty been changed at some time; and on his own statement, in December, 1914, his mind was at least open to question as to when it was made. Now in August, 1911, he was at work on the docks, and it is possible that there was a bulge of two inches. We agree with Judge Hough, however, that the whole issue was of small importance, for it would have been absurd for Tysinger to keep vessels off the berths because of any such bulge.
There remains, therefore, only the testimony of Schuber and Arps. Schuber says that he saw one crack in the morning and one in the afternoon of the 16th, the. first of which he reported to Clark and the second of which he did not, as he was too busy. The only crack which Clark remembered was across all five railroad tracks and was 3 or 4 months prior to the accident (folios 398, 408), it was not much more than an inch deep (folios 401, 402), and he did not know whether it continued to the time of the accident or not. Goethals examined for it on the [458]*45818th and failed to find any such; his testimony is to be preferred, in case of conflict. Nothing can be based upon Schuber’s testimony as such after the finding of the court below, and we do not think we ought to assume that he did report it. That he saw one or both cracks, however, seems unquestionable. Arps corroborates him as to that in the afternoon, and it appears pretty clearly that he testified to seeing the crack at the investigation, when it can hardly be supposed that he had any motive to misrepresent. We must accept it that there was one crack that day, and that Schuber saw it, though the size and position depend only upon his word. There are two reasons, however, for not treating this as of critical consequence in’the case. In the first place; he wa.s only a dock clerk (folio 4473) under Clark (folio 4364), though he considered himself as a wharfinger (folio 4478), having general charge of the wharves. He tells what his duties were (folios 4366, -4367, 4368). He had absolutely nothing to do with ascertaining the condition of wharf, but had ‘simply to keep it in good condition, clean the tracks, etc. He reported the conditions affecting the handling of cargo, such as tracks, doors, etc. This evidence is somewhat contradictory. If he had no duty to ascertain the condition of the wharf, his knowledge would not bind the respondent; if he .was charged to report the condition of the tracks, the contrary would seem to be true. Clark’s testimony (folios 125, 128) does not clear up matters very much. Clark says that his and Schuber’s duties were to discharge and load ships, general supervision of the wharves, and “reporting the conditions.” We think that the libelant has hardly made out a case from this to show that Schuber was under any duty to watch the wharf, in such sense that his default charged them with responsibility. The fact that he was discharged for failing to report rests only upon his own word. The second reason is that, if he had noticed the crack, it did not become the duty of the respondent to warn the Newport away from the wharf that night. Such cracks had existed before, and they had never been followed by an immediate collapse. The structure had stood for nearly 7 years, and there had always been plenty of time to put in bracing, even when much more serious deformation had occurred. The appearance of such a crack would not have required (Goethals, folio's 1749, 1750)' — would, indeed, hardly have justified — warning to'the Newport to move from the wharf, which is all that could in any event have been done that night. This basis of negligence was not established, therefore, even assuming that notice to Schuber was sufficient to charge respondents.
We do not think it is necessary for us to express any opinion as to the character of the slide. Bierd, who knew the geology of the bottom, for he had built the wharf, does not substantially differ from Goethals in his testimony. That the strata may have slanted upwards at the place of the slide is, of course, a reasonable possibility. The collapse, when it came, certainly did not reach below the toes of the piles, for we have it that none of these were dislodged; they were either bent or broken. In this respect the' collapse was different from the movement at berth No. 1 in June and July. We cannot say with certainty what was the depth of the mud stratum or of the [459]*459Indiana day beneath; we may only say negatively that the slide did not go below the latter. There is some evidence of an upheaval under water, and so far as appears the conditions which Goethals mentioned were actually present. His conclusion and Mears’ are not to be lightly disregarded. The presence of the crack that afternoon, of which he knew at the investigation (folios 1806, 1810), did not change his conclusion. Indeed, his search for cracks back of the break was against the possibility that the slide might extend further (folios 1805, 1806). We cannot agree that the testimony of other engineers, however competent, is of so much weight as that of these gentlemen who lived lor years at the Isthmus anil who actually built the canal. However, we do not think that this issue is in any sense material to this case. While we should be disposed to accept the respondent’s position touching it, yet, if the slide was in fa.et of the other kind, we do not see that the libelant lias carried the burden of proof in showing that the respondent had been guilty of any imprudence either in the construction or maintenance of the wharf. That- movements were to be apprehended is true; they had occurred elsewhere, and the general conditions were the same all along the line. That is, however, by no means enough; the relevant inquiry here is whether such a sudden collapse was to be apprehended. That we can see no ground to expect; the prior history of the wharf had been all of the other kind. Whenever anything did move, there was ample time to stay it, long after any such disturbances went beyond the utmost that we can suppose existed opposite berth No. 2. The libelant’s case can only rest upon the assumption that the bracing done by Tysitiger and McKay for some quite unreasonable cause they put in where it was not'most needed.
The libelant failed to prove its case, and the decree is affirmed, with costs.