Before L. HAND, AUGUSTUS N. HAND, and CLARK, Circuit Judges.
L. HAND, Circuit Judge.
This appeal is from a decree in the admiralty disposing of a petition for limitation of liability, filed by the New York Tank Barge Corporation, as owner of the barge, "T.N. No. 73". This company filed the customary petition and stipulation and the cause was referred to a commissioner to receive proofs of claim; only one claim was filed - that of the Commercial Molasses Corporation, the owner of the cargo at the time the barge sank. The facts as developed upon the trial were in substance as follows: On October 11, 1928, the petitioner, the Barge Corporation, and the claimant's predecessor, the Dunbar Molasses Corporation, entered into a contract for five years from February 1, 1929, to January 31, 1934, by which the Barge Corporation agreed to carry, and the Molasses Corporation agreed to ship, all molasses of the Molasses Corporation, or its subsidiaries, in New York Harbor from steamers or tide water terminals or refineries, to the Molasses Corporation's customers in the Harbor. The contract fixed the price for various points of destination, and contained the following two clauses on which the controversy turned in the district court: "The barge owners undertake and agree that barges owned and/or chartered are tight, staunch, strong and in every way fitted for the carriage of molasses within the limits above mentioned and will maintain the barges in such condition during the life of this contract." Several paragraphs further on in the contract occurred the other clause: "The Dunbar Molasses Corporation shall insure the cargoes carried by the New York Tank Barge Co. Inc., in its own, or chartered or operated barges, for the account of New York Tank Barge Co. Inc. and/or the owners of such barges; and * * * neither the New York Tank Barge Co. Inc., nor such barges shall be liable for any loss in respect of which insurance has been or could have been effected." This contract was extended to cover the year 1937, and was in force when the loss happened. On October 23rd of that year, the Barge Company delivered alongside the SS. "Athelsultan" in New York Harbor the barge, "T.N. No. 73" to be filled with molasses. She lay along the port side of the ship, made fast by four lines; and the molasses was pumped into her through a hose from the tanks of the "Athelsultan". The barge had a rake at either end, beginning 23 inches below the deck, and making forward and aft peak tanks, but the main cargo holds were four tanks, made by two bulkheads, one running fore and aft, and the other, thwartships. Between 8:30 and 9:00 on the evening of the 23rd, the "Athelsultan" began to fill the two forward tanks; this continued until some time between 11:00 and 11:30, when the valves to the forward tanks were closed and those to the after tanks were opened, and the molasses began to flow into these. The ship continued to pump into the after tanks until about five minutes after one in the morning of the 24th, when, for some unexplained reason, the barge unexpectedly lurched and shortly afterwards sank. The valve leading to the after tanks had never been shut off after the ship began to fill them.
The customary method of stowing the barge was as follows: molasses was pumped into the forward tanks until the barge had a fixed freeboard forward; then into the stern tanks until the stern had another fixed freeboard; then back once more into the forward tanks until she was trimmed fore and aft. Shortly before one in the morning, the stern had about the proper freeboard, and the mate went forward to tell the men on the ship to open the forward valve to trim the barge. He engaged in some talk on the way which delayed him, and the Barge Company attributed the sinking to his negligence in waiting too long. Both parties put in a great deal of evidence; both by experts, as to the theoretical capacity of the barge at various freeboards, and by lay witnesses as to the facts. The judge examined all the evidence with great care and thoroughness - of his own motion requiring the case to be reopened for further light - but in the end concluded "that the cause of the accident has been left in doubt". From this he held the Barge Company liable because of the "presumption" of unseaworthiness, arising from the barge's sinking without adequate explanation. However, as he concluded that the second clause of the contract quoted above, relieved the Barge Company of the covenant of seaworthiness contained in the first (the Molasses Company having failed to take out any insurance for the Barge Company) he dismissed the claim (1939) A.M.C. 673. The Molasses Company appealed.
The finding that the "cause of the accident has been left in doubt" means, we take it, that the evidence as to whether or not the barge sank because of unseaworthiness, was so evenly matched that the judge could come to no conclusion upon the issue. Though negative in form, it was as much a "finding" as an affirmative finding, and we are to respect it as such. Nashville Interurban Ry. v. Barnum, 2 Cir., 212 F. 634. Although Admiralty Rule 46 1/2, 28 U.S.C.A. following section 723, does not lay down the measure of conclusiveness of findings in the admiralty, it prescribes nothing different from Rule 52 (a) of the Rules of Civil Procedure, 28 U.S.C.A. following section 723c. The Niel Maersk, 2 Cir., 91 F.2d 932; Barlow v. Pan Atlantic SS. Co., 2 Cir., 101 F.2d 697. Therefore, since we cannot say that the finding was "clearly erroneous" in holding that no conclusion was possible upon the issue; we are to dispose of the case with that as datum.
It is well settled that the owner of a vessel always impliedly covenants that she is seaworthy unless he expressly stipulates the contrary (Cullen Fuel Co. v. Hedger Co., 290 U.S. 82, 54 S. Ct. 10, 78 L. Ed. 189) and unless there is somthing exceptional in the situation at bar - of which more hereafter - the promisee has the burden of proving the breach, by which we mean that, if the judge is unable to make up his mind upon the issue, the promisee fails. It must, however, be owned that the matter has become somewhat embroiled by the loose use of the phrase, "burden of proof", and the word, "presumption" - the particular offender being the second. It has for long been the custom of judges to say, when a ship founders, or developes some defect, shortly after her voyage begins, that there is a "presumption" that she was unseaworthy when she broke ground. The question here is whether this "presumption" has any effect upon the "burden of proof".
In England at any rate the law is plain, though at the start there was confusion. Watson v. Clark, 1 Dow. 336, arose on a maritime insurance policy; the insured ship had turned back, clearly unseaworthy, within a fortnight after leaving Belize, Honduras, and the underwriter alleged that she had been unseaworthy at the outset of the voyage. Both Lord Eldon and Lord Redesdale said that when a ship becomes unable to go on with her venture "a short time" after leaving port, there is a "presumption" that she was unseaworthy when she started, and that "the onus probandi" is thereafter upon the assured to prove the contrary. This may have only meant that the owner must put in some explanatory evidence on penalty of having the issue taken against him; but it must be confessed that the phrase, "onus probandi", in all probability meant more than that. The point was cleared up in Pickup v. Thames & Mersey Marine Insurance Co., 1878, L.R. 3 Q.B. 594. In that case the judge at nisi prius had answered a question of the jury in the language of Lord Eldon, and both appellate courts held that this was wrong; they declared that while the development of unexplained defects shortly after the ship breaks ground, will justify relating them back to that time, yet the assurer must satisfy the jury that she was unseaworthy, and that it was an error to tell the jury anything else. The Privy Council made the same ruling in Ajum Goolam Hossen & Co. v. Union Marine Insurance Co., 1901, App. Cas. 362, 366; and the last word is Lindsay v. Klein, 1911, App. Cas. 194, 203-205. In that case also the ship broke down soon after the beginning of the voyage and had to put back; and the owners sought to charge the cargo with a contribution in general average which the shippers resisted on the ground that the ship was unseaworthy. So the House of Lords held on the evidence; and while it is true that the speech of Lord Shaw of Dumfermline was dictum, nevertheless at some length he discussed the question and overruled Watson v. Clark, supra (1 Dow. 336), so far as the House of Lords can ever be said to overrule its own decisions. There can be no doubt that by the law of England (and of Scotland as well) although the development of a defect early in the voyage is proper evidence on which to base a claim of unseaworthiness, it is nothing more, and the promisee loses, if the evidence as a whole is too evenly balanced to admit an affirmative conclusion.
The law is not so clear in this country. Before 1813 (when Watson v. Clark, supra (1 Dow. 336), was decided) several courts had indeed held that if a vessel became unfit shortly after she broke ground and no explanation was offered, she would be "presumed" to have been unseaworthy when she started out. Barnewall v. Church, 1803, 1 Caines 217, 234, 235; Talcot v. Commercial Insurance Co., 1807, 2 Johns. 124; Cort v. Delaware Insurance Co., 1809, 2 Wash. C.C. 375 (semble). And since 1813 the same ruling has been made again and again. Paddock v. Franklin Insurance Co., 1831, 11 Pick. Mass., 227, 237; Walsh v. Washington Marine Ins. Co., 1865, 32 N.Y. 427, 436, 437; Work v. Leathers, 1878, 97 U.S. 379, 24 L. Ed. 1012; The Arctic Bird, D.C., 1901, 109 F. 167; Forbes v. Merchants' E. & T. Co., D.C., 1901, 111 F. 796, 800, affirmed 2 Cir., 120 F. 1019; Oregon Round Lumber Co. v. Portland & Asiatic SS. Co., D.C. 1908, 162 F. 912, 920, 921; Sanbern v. Wright & Cobb Lighterage Co., D.C. 1909, 171 F. 449, affirmed 2 Cir., 179 F. 1021; The Kathryn B. Guinan, 2 Cir., 1910, 176 F. 301; The Loyal, 2 Cir., 1913, 204 F. 930; United States Metals Refining Co. v. Jacobus, 2 Cir., 205 F. 896; The Jungshoved, 2 Cir., 1923, 290 F. 733; S.C. Loveland Co. v. Bethlehem Steel Co., 3 Cir., 33 F.2d 655; The Harper No. 145, 2 Cir., 42 F.2d 161. (Dupont v. Vance, 19 How. 162, 15 L. Ed. 584, has been several times erroneously cited as in accord with this. The mistake apparently arose because the quotation marks in The Arctic Bird, supra, which correctly indicate the end of the quotation from Dupont v. Vance, supra, were misplaced in Oregon Round Lumber Co. v. Portland & A.S.S. Co., supra.) We have not tried to collect all the cases; but there can be no question that they amply establish the "presumption"; and a number of them say that it shifts the "onus probandi" or "burden of proof" to the other side. We need not, however, take these statements as necessarily intended to mean that it is the owner which must persuade the court of the ship's seaworthiness; in none of the cases was exactness of expression as crucial as it is here, and was in Pickup v. Thames & Mersey M.I. Co., supra (L.R. 3 Q.B. 594). And indeed it is notorious that judges often speak of the "shifting" of the "burden of proof" when they mean no more than that, if the trial had ended with the evidence that is said to "shift" the "burden", that party would lose to whose shoulders it had been shifted. Even in cases like The Harper No. 145, supra, 2 Cir., 42 F.2d 161, where we did indeed use the phrase in a way that seemed to mean that the shipowner must satisfy the court that the barge was seaworthy, it is impossible to be sure that we meant more than this. At any rate, if we did, it would certainly be wrong to apply it here.
In the case at bar the Barge Company was acting as a private carrier and was therefore only a bailee; it is well settled that the burden rests upon the bailor to prove some breach of duty by the bailee other than his mere failure to return. Kohlsaat v. Parkersburg & M. Sand Co., 4 Cir., 226 F. 283; Alpine Forwarding Co. v. Pennsylvania R.R., 2 Cir., 60 F.2d 734; Gerhard & Hey Inc. v. Cattaraugus T. Co., 241 N.Y. 413, 150 N.E. 500. The Molasses Company acknowledged this by alleging in its claim as a fault that the barge was unseaworthy; it was obliged to do so; and, having done so, it had to persuade the judge or fail. It is true that the facts necessary to create any legal liability are what the law chooses to make them, and what we call a "defence" may theoretically quite as well be treated as part of the constituent facts of the liability; it is mere convention whether the plaintiff must show that it does not exist, or the defendant that it does; or vice versa. This is well illustrated by the controversy of contributory negligence. But once the law has established which party shall prove a fact, that party must do so, and the duty of doing so does not shift, save possibly in extraordinary circumstances of which we cannot at the moment recall an instance. In all cases, therefore, where the shipper must show the ship unseaworthy, that duty remains upon him throughout. It is enough on his case in chief if he shows that she developed an unaccounted for defect early in the voyage, because it is reasonable to infer from that that the defect must have existed before, but he must lay any doubts which remain when the whole evidence is in. Del Vecchio v. Bowers, 296 U.S. 280, 286, 56 S. Ct. 190, 80 L. Ed. 229; New York Life Ins. Co. v. Gamer, 303 U.S. 161, 170, 58 S. Ct. 500, 82 L. Ed. 726, 114 A.L.R. 1218. The question as to the burden of proof under § 3 of the Harter Act, 46 U.S.C.A. § 192, is not to be confused with this. That section gives the owner an excuse for a fault which by hypothesis the shipper has proved; and the owner must prove it, as anyone must prove an excuse. The Southwark, 191 U.S. 1, 12, 24 S. Ct. 1, 48 L. Ed. 65; The Wildcroft, 201 U.S. 378, 386, 26 S. Ct. 467, 50 L. Ed. 794; May v. Hamburg, 290 U.S. 333, 346, 54 S. Ct. 162, 78 L. Ed. 348. But the question here is whether the ship was at fault at all; and she was not unless she was unseaworthy.
We need not examine whether this was a "presumption of fact", or a "presumption of law". Strictly the first phrase is misleading, and has caused much confusion, though it accords with colloquial usage. Pariso v. Towse, 2 Cir., 45 F.2d 962, 964. What we are here dealing with is a rational inference, unlike a true "presumption" which presupposes that no inference can safely be drawn from the facts which make it up, and which merely relieves the party from proving the issue unless his antagonist moves. We do not understand that the judge in declaring that he could not decide whether the barge was seaworthy, did not consider her unexplained sinking as part of the evidence before him; but that, on the contrary, having weighed everything, including her sinking when she did, he was left in doubt. He then took recourse to the presumption which he supposed to persist as a determining legal factor. So far as we can see, that was the exact equivalent of putting the burden upon the Barge Company to prove that the barge was seaworthy.
The Molasses Company also argues that if the barge sank because her hatches were left open during the lading, she was not seaworthy for that reason. This wholly misapprehends our ruling in the Fred E. Hasler, D.C., 55 F.2d 919, which merely applied the well-settled doctrine that a ship is unseaworthy, if. When she breaks ground, some part of her gear is not in place which her officers suppose to be in place. We did not decide that negligence of the crew in failing to attend to the gear during lading makes the ship unseaworthy. Here the bargees deliberately kept the hatch covers unfastened, and if that was a fault, it was one of management.
Since therefore we think that the claimant did not prove its case, even if the clause in the charter regarding insurance did not cancel the warranty of seaworthiness, it is not necessary to pass upon that question.