The opinion1 in this case states the evidence in detail, together with the substance of what the judge found, and we shall assume an acquaintance with it in what we have to say. The appeal raises three main questions: (1) whether the finding should stand that at 4:30 when the master, Beale, and the first mate, Thompson, went aft with their binoculars, they decided that the dredge was on an even keel; (2) whether, if that finding should not stand, Beale was justified in doing no more than reducing to half speed; (3) whether the tug has the burden of producing evidence that will support an inference that nothing could have been done between 4:30 P.M. and 7:30 P.M. to save the dredge. The second question as we have repeatedly held2 is not one of fact in the sense that we must follow it unless we think it 'clearly erroneous'; it concerns the duty- standard of care- appropriate to the occasion.

Although we accept the judge's commendation of the bearing and appearance of Beale and Thompson, a consideration of the record as a whole compels us to reject their testimony that the dredge had no observable list at 4:30. We start with Beale's own testimony that he chanced to look back at the tow and there seemed to be such a list. He was then in the pilot house and called Thompson to go aft with him for a more careful scrutiny. On his direct examination Thompson swore that he went first to the wheelhouse to get his binoculars before following Beale aft; and it was then that he made the entry in the log: '4:30 P.M. Dredge starting to list to Port. Reduced speed.' On his cross examination he swore that he went aft as soon as he got Beale's order, that they both looked at the dredge, and that neither could see any list. Beale already had his own binoculars, but Thompson wished to get his and went back to the wheelhouse; and it was then that he made the entry we have quoted. Even without Beale's letter of May 6th, four days later, though nevertheless ante litem motam, Thompson's story is incredible, whether we take the first or second version. According to the first, he made the entry in the log before he had made any personal examination himself and as he was preparing to go aft to make one after picking up his binoculars. All he had heard was that Beale thought that the tow was not 'behaving just right'; that 'he seemed to feel that the dredge * * * was showing some- a little slight list.' That was the moment he chose to write that the dredge had started to list to port and that Beale had slowed down the tug. We are to believe that he made this entry while he was preparing to join Beale in order that he might help Beale to find out whether Beale's impression was true, or what Beale afterwards called it- an 'illusion.' That appears to us an extremely improbable story, however, convincing the narrator's bearing and appearance. When we turn to the second version (which incidentally discredits the first which it contradicts), it is nothing short of impossible. Thompson went directly aft to where Beale was, when Beale said that he 'thought that the dredge was showing a little list'; and standing alongside of Beale, he 'could not determine any list in it'; nor did Beale. Nevertheless he went back for his binoculars; and when he got them, though without using them, he made the entry that there was a list. Surely nothing could be more patently absurd. Moreover, we should not forget the form of the entry itself; not only did it speak catagorically, but it declared what was being done to correct the list; and nothing was ever done to correct the error. On the contrary, Handran, who in contradiction to both Beale and Thompson swore that there had been a list, not only accepted the entry, but filled it out by adding '150 rev.'- 150 revolutions- to indicate what the reduced speed was. It is plain to us that the entry could only have been made after both Thompson and Beale had come to a conclusion that the dredge had listed, although, for the time being anyway, all that was necessary was to go on at half speed.

All this we conclude before adding the confirmation of Beale's letter of the 6th, the crucial language of which was: 'At 4.30 P.M. * * * dredge B 1 was commencing to list to port, so speed was reduced to half.' That exactly conformed to the log entry, and as flatly contradicted the testimony that no list had been observed and that the speed was reduced only as a 'precautionary' measure. The judge refused to receive the letter for two reasons: (1) it was introduced at the very end of the trial after Beale had left New York; and (2) the signature was not proved to have been Beale's. It is of course true that, if evidence is introduced so late that any reply to it will require an undue prolongation of the trial, the judge has discretion to refuse it, though it is a discretion more often honored in the breach than in the observance. Be that as it may, in the case of so critical a document as this and in a litigation already so long delayed, any further postponement that might have resulted was not important. As for its authenticity the letter was an official report, produced from the custody of those to whom the law directed it to be delivered. It was official, because it was prepared and filed in compliance with Sec. 137.3 of Title 46 of the Code of Federal Regulations (1st ed.), whose statutory warrant is to be found in 49 St.at L. 1383, R.S. Sec. 4450(j), 46 U.S.C.A. § 239(j); and, as such its authenticity was proved by its production from proper custody, for it contained nothing but what its author had himself witnessed. United States v. Grayson, 2 Cir.,166 F.2d 863, 869. True, this doctrine ordinarily applies to reports made by persons who are themselves public officials and not to private persons charged by law with a duty to report. Moreover, the decisions are not altogether uniform that reports of private persons are within the rule. However, there is substantial support for saying that they are, and this accords with the modern acceptance of reports made in the due course of business. The First Circuit twice so ruled,3 as did the early English case of Richardson v. Mellish, 2 Bing. 229, 240, which the Supreme Court quoted with approval in Buckley v. United States, 4 How. 251, 258, 11 L. Ed. 961. Wigmore, Sec. 1633(a), thinks there should be no difference, and we agree; nor is there anything to the contrary in United States v. Grayson, supra, 2 Cir., 166 F.2d 863, despite a note to Sec. 1633(a) of Wigmore in the 'Pocket Supplement, 1951.' On the new trial Beale may of course contradict the letter, but it will be admissible as such, if it is again produced by the Coast Guard.

Finally, in passing on the credibility of Beale's and Thompson's testimony, the reduction of speed is of itself important evidence that they had concluded that the dredge was listing. If, after examining her at a distance of only 700 feet, they had agreed that she was on an even keel, we cannot understand why Beale should have thought it desirable to reduce speed. Certainly that showed that he did not think it safe to proceed as before, and that is scarcely consistent with the premise that what he had seen had reassured him that the dredge was as seaworthy as she had been. In the face of this is it not most improbable that he really thought the list was an 'illusion'? We have examined the evidence with so much detail because we are never willing to reverse a finding of fact unless we cannot escape concluding that it is 'clearly erroneous.' On the other hand it must be remembered that in spite of such proper reluctance, we cannot escape some residual degree of responsibility for the facts.

Since we find that the dredge showed a perceptible list at 4:30, and since it is conceded that the tug did nothing but keep on at half speed, the next question is whether her navigation was actionably negligent. We need go for an answer no further than to Beale's own admission upon cross examination: 'Q. After you resumed towing at 150 revolutions, if you had seen that the barge had a list, would you have gone back to investigate? A. I certainly would, would have done some- taken some precaution to find out whether there was anything wrong or not.' With his assent the court then summed up his testimony as follows: 'if at any time while he was traveling at 150 revolutions the barge had listed, he would have thought that he should have gone back and would have gone back.' (Literally, this testimony relates to a putative list after the tug had started up again, but obviously that applies equally to a list at the time the tug had stopped.) There was no expert testimony that, if there had been a 'noticeable list'; e.g. one of five degrees, the tug need not have investigated it further. We do not understand that the judge meant to hold the contrary; he based his conclusions throughout upon his finding that the dredge did not begin to list til 7:10. Even if he had held otherwise, we should feel bound to decide that to reduce to half speed was not in these circumstances the measure of that duty of every tug 'to exercise such reasonable care and maritime skill as prudent navigators employ for the performance of similar services.'4

For the foregoing reasons we hold the tug at fault for her failure to inspect the dredge at 4:30- which was entirely feasible in spite of the boarding up of her sides. Yet it does not inevitably follow that the defendant became liable for her foundering three hours later; that depends upon whether the failure to inspect her was a cause of the disaster, and the record does not contain evidence which would support a finding that, if the tug had done her duty, she could have saved the dredge. On the other hand, if her fault threw on the tug the duty of producing evidence that the fault did not cause the loss, she did not discharge that duty, for the evidence does not prove the negative. We think that such a burden did fall upon the defendant under all the circumstances at bar. We do not forget that Stevens v. The White City, supra, 285 U.S. 195, 52 S. Ct. 347, 76 L. Ed. 699, established it that a towing contract does not make the tug a bailee of her tow; but it does not follow that she should not be charged with the burden of producing evidence that her fault did not cause the damage; and, quite regardless of whether there is any general rule, it is enough in the case at bar that the tow had no power to show what in fact caused the disaster; and that, if the tug had done her duty, the cause would probably have been disclosed. It is often a controlling factor in deciding where to throw the burden of producing evidence- and obviously it ought to be- that the proper party to charge is he who alone could discover the truth.5 It is another matter whether in the end the burden of proof should also have been upon the defendant; that question we reserve for the time being. Some decisions do indeed hold that once a tug is shown at fault, the burden of proof in the sense of satisfying doubts also falls upon her;6 and although in The Carbonero, 122 F. 753, 755, the First Circuit spoke to the contrary, in the end it held that the tug had been free from fault, so that the language was at most only a dictum. However, on the next trial the first question will be whether the defendant puts in enough evidence to support an inference that, no matter what an inspection at 4:30 might have disclosed, nothing could have kept the dredge afloat until she was brought to a place of safety. If the defendant succeeds on that, the second question will be whether the plaintiff has the eventual burden of satisfying the court that the tug's fault in fact caused the loss, and, if so, whether it has done so.

Judgment reversed; new trial ordered.