The libel is by a shipowner to recover the amount fixed as due from the cargo in a general average adjustment for sacrifices made under the following circumstances: The Italian steamer, Ettore, was chartered to Johnson & Turner to carry a cargo of cork from Lisbon to Philadelphia. She was laden in August, 1926, at Lisbon, and the master gave three bills of lading, two to the order of one consignee, the third to that of another, neither being the charterer. When about, four days out in a gale of wind, about one hundred and seventy miles south of the Azores, her tail shaft broke just forward of the screw, which dropped off, leaving the ship helpless. It was just after midnight; the master and engineer were asleep, and the cause of the accident was, and still is, unknown, The ship got help from a nearby vessel which towed her to Ponta Delgada, where after four surveys she refitted temporarily, and was- allowed to complete her voyage, which she did. At Philadelphia she claimed a lien on the cargo for general average, and the respondents guaranteed payment in consideration of its release. Adjusters were appointed who made a general average statement, the amount of which is not in dispute, and which found the share of the cargo for the expenses of salvage and temporary refitting at Ponta Delgada to be something more than twenty-three thousand dollars, which is the subject of the suit. The principal issue was as to the seaworthiness of the ship when she broke ground at Lisbon; much evidence was taken in and out of court, and the judge found that the burden of proof, as the case stood, was on the ship, and that she had not carried it. He also found that the screw had not struck anything afloat, and indicated, though he did not expressly find, that she was unseaworthy in three particulars: she was not in proper ballast; there was a flaw in the metal of the shaft; the shaft was out of alignment.
Nobody knew, or could positively ascertain, whether the screw had struck a floating timber or the like. The hull showed no marks except two slight dents on the plate of the screw box, assumed to be caused by the flukes after the serew'had broken off. From this negative evidence the judge inferred that it had not struck anything, and eliminated that as a possible cause of the aeeident. The gale had been on for about a day, though the wind had been rising for several days before. At least after it had reached its force, the ship was pitching violently, and the screw often racing; it is a reasonable inference that while it was doing so, or more probably upon its submersion, the shaft broke. The issue thus arose as to the fitness of the ship for her duties, upon which a large record'was built up. It does not appear to us necessary to make a finding upon the exceptionally baffling evidence before us, but to an intelligent understanding of the case something must be said.
The tail shaft had been renewed in 1916, and withdrawn in 1924, and the ship was overhauled generally shortly before the voyage. We shall assume with the judge that all was done which diligence required to make her seaworthy, except in respect of ballast, and that the only question is whether she was in fact fit. When the broken shaft was unshipped, a vein of dross of undisclosed size was found near its centre, the casting being pro tanto imperfect. A large amount of the testimony was taken as to the effect of this upon its strength. Being in the centre, the libellant urged that the fault did not weaken the shaft, since the torque Was least at that spot. A certain finding as to whether this, contributed to the result is not possible. The respondents also maintained that the ship was not in ballast, particularly in that it did not appear that the deep tank was filled. The record is silent on the issue, except for a general statement in the log that she was well ballasted, and testimony that the peak and bottom tanks had been filled. The master and first officer said that she was high in the water, and the log, which should have shown her draft, though brought to New York, the *771libellant did not produce. On the evidence the issue is not proved, but the failure to make proof must count against the libellant.
The most important question was, however, as to the misalignment of the tail shaft, which could have caused the break. Unhappily it is the most confused issue in the record. After the screw dropped off, the engines raced until the steam was shut off, and for some time thereafter. There is an ambiguity in the translation of the Italian depositions which admits as a possible interpretation that this run on mere momentum was for fifteen minutes; but this is so palpably absurd that we accept the possible alternative—fifteen seconds—and conclude that such wear in the bearings as was disclosed at the surveys, could not have been substantially increased after the aceident. We also accept the testimony of the engineer, Pinto, that the tail shaft passed through the first bearing f orward of the stem tube. While there is other testimony which throws doubt on this, it is not conclusive, and the engineer certainly ought to have known the truth. Thus it seems probable that the tail shaft was running in a worn bearing, which could put it out of alignment. The crank shaft bearings were also much worn, and parts, though possibly not vital ones, were deeply corroded. We conclude that the shaft may have been ont of alignment; but, as we have said, we do not find it necessary to go further.
These inconclusive results are all that we need for a disposition of the case. The Ettore broke her shaft in a sea which was not unexpected even in August; gales are likely at all seasons in the Atlantic, and this was at most not more. She should have been able to withstand it, else she was not reasonably fit for the duties she had undertaken, and was therefore not seaworthy. The Silvia, 171 U. S. 462, 464, 19 S. Ct. 7, 43 L. Ed. 241; The Southwark, 191 U. S. 1, 8, 9, 24 S. Ct. 1, 48 L. Ed. 65. That a ship may be reasonably fit and still break down under ordinary strains, may indeed be true, but when she docs, it rests upon her to show her fitness. This is well-settled law in 'suits upon contracts of affreightment. The Edwin I. Morrison, 153 U. S. 199, 14 S. Ct. 823, 38 L. Ed. 688; The Southwark, supra, 191 U. S. 1, 13, 24 S. Ct. 1, 48 L. Ed. 65; The Wildcroft, 201 U. S. 378, 389, 26 S. Ct. 467, 50 L. Ed. 794. We can see no reason for a distinction in cases of contribution. Some of the language in Klein v. Lindsey, L. R. (1911) A. C. 194, seems to look the other way, but we are not clear just what was the final conclusion. Lord Shaw said that normally the burden was on the cargo to prove unseaworthiness, though the breakdown of machinery soon after breaking ground, coupled with the ship’s history, in that case turned the scale. At the same time he quoted with apparent approval the language of the Lord Ordinary below, to the effect that a breakdown under ordinary use shifted the burden of proof back upon the ship. Just what the evidence must be to effect this, we are not clear; nor do we quite understand what is meant by the shifting of the burden. The same ambiguity pervades Broadnax v. Railroad, 157 Pa. 140, 27 A. 412. Apparently, the doctrine there accepted was that a breakdown created a presumption of unseaworthiness which shifted the burden, but if so, it must be confessed that it was scarcely applied. Judge Ward decided flatly that the burden was on the ship in The Lewis H. Goward (D. C.) 34 F.(2d) 791, following the analogy of suits on contracts of affreightment. We reserved the point in Spang Chalfant & Co. v. Dimon S. S. Corp., 57 F.(2d) 965, but we cannot do so hero. It seems to us that the rule should be the same as in suits for cargo damage, and that the burden is upon the ship, and remains so throughout. That she may start with a presumption in her favor, we need not deny. Franklin Sugar-Refining Co. v. Funch, Edye & Co., 73 F. 844 (C. C. A. 3). If so, the failure of so vital a part of her machinery in weather which it should withstand, if staunch, is all that the cargo is called upon to prove. Thereafter she must prove her general fitness, and will fail upon the issue, if the evidence is in balance. This is in accord with the rule in marine insurance. Watson v. Clark, 1 Dow, 336. And in view of her exclusive command of the facts, it appears to us the reasonable doctrine. We decide the case on the assumption that the shaft and bearings were not in good order, and the ship not in ballast.
The charter had expressly covenanted that the ship should bo “staunch, strong and fit,” but this was later limited by clauses both in the charter and the bills of lading, which if applicable to the time of breaking ground, wo will assume to have excused the defects here in question, though in fact it is difficult to see how they could excuse improper ballasting on any theory. The respondents insist, indeed, that these limitations, not being expressly referred to the warranty of seaworthiness, should be confined to the voyage under the rule in The Caledonia, 157 U. S. 124, 15 S. Ct. 537, 39 L. Ed. 644,, and The Carib Prince, 170 U. S. 655, 18 S. Ct. 753, 42 L. Ed. 1181; and so it would be, if the ship *772had been a common carrier. But she was a private carrier, the cork occupying the whole ship, and as such she was free from liability anyway except for lack of care. Not being under the absolute duties of a common carrier, the libellant argues that the clauses must refer to the covenant of seaworthiness, if they are not to be a nullity, which is not to be inferred', and that we so decided in The G. R. Crowe (C. C. A.) 294 F. 506. Therefore, it concludes, the ship was not warranted seaworthy, save as diligence would make her so, and her defects should not bar'her claim to contribution.
Though we accepted this argument, so far as concerns the ship’s excuse for a failure to perform her contract, it would not follow that she need not be seaworthy in fact, when she claimed contribution. It is indeed a plausible position that that right should run pari passu with liability. This was the argument unsuccessfully urged in The Irrawaddy, 171 U. S. 187, 18 S. Ct. 831, 43 L. Ed. 130, and it got the assent of an impressive minority of the court. But the law was then settled otherwise, though perhaps it resulted in no more than introducing the “Jason” clause into most bills of lading. The Jason, 225 U. S. 32, 32 S. Ct. 560, 56 L. Ed. 969. It is true that in The Irrawaddy it was the Harter Act (46 I7SCA §§ 190^195), not the contract, which excused the ship, and by which she tried to supplement her right of contribution. But we can see no difference. The doctrine, as we understand it, is that what determines the obligations of the ship under the contract, is not material to her affirmative right to contribution; that an excuse for non-performance will not supply a fact essential to that right. Therefore, we think that the charter must be disregarded in determining the suit at bar.
But if so, it may be asked, by what is the claim to be determined? It is the creature of equity, and it is unfair to charge an, owner with defects which due diligence would not have detected. The answer, we think, is that when a man offers a ship to carry goods at sea, the shipper may assume that she is fit for the purpose. Though the owner do not expressly so agree, the law of the sea imposes the undertaking upon him. If he is a private carrier, he can indeed rid himself of his obligation, as the libellant has done here, but in so doing, as we have shown, he only shields himself from attack. If he would provide himself with a weapon, he must secure a stipulation that, notwithstanding his failure to provide a fit ship, he may recover. This he may do by any appropriate language; but unless he does, he has failed to make góod a declaration which stands at the basis of every maritime venture, and he cannot charge his cargo with the costs of a common sacrifice. In short he is to be treated as at fault. If it be asked whether this does not imply a covenant in addition to the express covenant of. the charter, we agree. The same argument can be pressed against the result in The Irrawaddy. The' obligation of careful navigation was excused; as to the contract of carriage it did not exist. But it existed when the question arose as to contribution; it could so exist only because the law implied it in respect of that right. When therefore the libellant invokes the “equity” on which contribution rests, we can assent only in ease among the facts which make it up, is the fulfillment of the undertaking implicit in every carriage by sea.
We have not been able to find many cases on the point, for usually the owner is privy to the unseaworthiness of his ship, and such instances we disregard. In Schloss v. Heriot, 14 C. B. (N. S.) 59, the successful plea did not allege that the unseaworthiness was the result of negligence; and we read the brief remarks of Earle, C. J., as meaning no more than that the owner’s failure to fulfill the covenant was necessarily negligence. The same is true of the language of Blodgett, J., in Nine Hundred and Twenty-Eight Barrels of Salt, Fed. Cas. No. 10,272, a case where the owner was certainly not shown to have failed in care. In Bowring v. Thebaud, 56 F. 520, we held in effeet that seaworthiness was a condition to any recovery. The chief part of the discussion did indeed concern the time from which the covenant spoke, but the defect was one against which the shipowner could not have provided, and the. result presupposed that “fault,” in the sense which the libellant here uses it, was not a material issue. These authorities are not perhaps very impressive, but on the other hand, we have found nothing to countenance the libellant’s position, and we are free to hold as seems to us most in accord with maritime notions. Accordingly we say that in the absence of some reservation of the right, an unseaworthy ship may not have contribution against the cargo, when her defects occasion the sacrifice, though they are not discoverable by due diligence, and though she has excused herself for breach of her contract.
A subsidiary question arises over the form of the bills of lading. The charter-party had provided for general average un*773der the York-Antwerp Rules of 1890. It alsoi contained a clause that the owner’s bills of lading might be employed “conjointly with this Charter and their clauses to hold good, provided they are not contrary to those of the present Charter-Party.” The master’s bills of lading provided that general average should be payable according to the York-Antwerp Rules of 1924. The rules of 1890 make no provision as to sacrifices made necessary by the fault of one of the parties, but those of 1924 allow contribution notwithstanding. We do not find it necessary to consider the question decided by Wright, J., at nisi prius in Tempus Shipping Co. v. Louis Dreyfus Co., L. R. (1930) 1 K. B. 699, for it is clear that the bills of lading did not in any event control; they were “contrary” to the charter-party in respect of general average. Though the earlier rules did not deal with faults, we cannot read both codes together in so far as they do not clash. The charter-party selected the rules of 1890 as the measure of contribution. It was as much a part of these rules that they should not control the obligation to contribute as that they should define the subjects of contribution. Ralli v. Troop, 157 U. S. 386, 15 S. Ct. 657, 39 L. Ed. 742. Their last clause is substantially to that effect. It is not therefore possible to piece both together and make a new code, comprising so much of them as do not conflict inter se. Again, though the bills of lading came into the hands of consignees, who might possibly invoke the more favorable code, if they were suing, any such right would rest in estoppel; when the ship makes claim against them, she is bound by the charter.
A last point is that the agreement to enter into the general average adjustment was a contract of arbitration; and the statement, an award. Rebora v. British & Foreign Marine Ins. Co., 258 N. Y. 379, 180 N. E. 90. Without considering the vital differences between the contract at bar and that in the decision cited, it is enough to say that this is an eleventh hour suggestion, not even intimated at the trial. The libel is the usual one for contribution under the maritime law, without a syllable to suggest that the adjusters’ statement was an arbitration award, though the assignments of error do indeed raise the point. We would not reverse a decree after a trial conducted on an entirely different theory, even though the facts gave more excuse for doing so than they do.
Decree affirmed.