All efforts to harmonize tbe evidence have proved hopeless. Tbe testimony of tbe witnesses from tug and schooner is-most unsatisfactory when consulted in print, and the opinion of the district judge indicates that the appearance of both these groups upon tbe stand did not add to their persuasiveness. Under these circumstances it seems the wiser course to turn to witnesses not connected with either side, and see whether their story confirms the theory of tbe Garlick sufficiently to warrant a reversal of tbe district court. Borne propositions are clearly established by the proof. The colliding vessels were not coming together bead on. One was to the eastward of the other; and, if both bad continued up and down the river respectively, there would have been no collision. Both could not have changed their course to the westward, and produced the results described. It: follows that one of them must be in fault; the other not. Moreover, it is manifest that the one whose change of course to the west produced the collision must, before such change, have been the easternmost of the two. Unless, therefore, the evidence of the disinterested witnesses should he found to indicate that the Garlick, while overtaking it, was nearer the Niagara’s tow than the schooner was when passing it, the conclusions of the district judge should not he disturbed. One Terry, or Tyrrel, was the owner of the canal boat Bogart, — outside boat in the fourth tier. He says he was on deck, and noticed the Garlick as she lapped the tow, and again when she was about a length ahead of him, and noticed the schooner coming showing both lights. He says the tug passed up parallel with the tow about 150 feet off. The intelligence of the witness, as disclosed in his testimony, compares unfavorably with thaf of the other two disinterested witnesses. Moriarty, the pilot of the Komuk, testifies that she was tied up to the outside of the third tier from the rear of the tow when the Garlick passed him. The latter blew a whistle, which he took to be a salute to himself, and which he returned. The Garlick was then heading straight up the river, and about 500 to 600 to the eastward of him. After returning the supposed salute, he dropped back to the rear of the tow to take out two boats, and noticed nothing further until he heard the Garlick’s alarm whistles, when he looked up, and saw the latter heading diagonally across to the westward towards the Niagara’s tow, pulling over for the tow, and near the head tier. The Garlick *544sliut off the schooner’s lights, but witness, by the lights' in Yonkers, .could see the schooner herseif heading straight down, as near as he could see. Wood, the pilot of the Niagara, went into the pilot house about 12:20 p. m. According to his custom, he looked around, and saw the red light of the Garlick way in to the eastward. Subsequently he saw the schooner while still quite a long way off. He is confused and inexact as to the bearings of both vessels when first seen, but is very positive that the schooner was nearer to the Niagara and her tow than the Garlick was, and that the latter changed her course so as to go across the river in front of the schooner. Upon this state of the proof we do not feel warranted in reversing the decision of the district court.
The appeal in the proceeding to limit liability arises on the following facts: The collision occurred October 4,1894, and the Garlick •was attached November 16th. On November 20th two appraisers, and on November 23d a third one, were appointed by the court to appraise her value. They each made a sworn report NQvember 27th, agreeing on a valuation of $3,000. The claimants prepared bonds for that amount, but subsequently changed their minds, and did not execute them. The vessel was thereupon sold by the marshal December 27, 1894, for $1,790, and the proceeds paid into the registry. On August 26, 1896, petition was filed to limit liability, and on August 1, 1898, an order of reference was made to the United States commissioner to take evidence, and make due appraisement of the value at the end of the voyage, October 4, 1894. He reported $3,000. Exceptions were filed and argued before the district judge, and the commissioner was sustained. From that decree appeal is taken.
There was conflicting evidence before the commissioner. Claimant and his partner, and four of the bidders who attended the sale in December, and made a somewhat cursory examination of the tug, testified to valuations ranging between $1,500 and $2,000. On the other hand, besides the sworn appraisals made, one of the appraisers, who has made a very careful examination, testified to a valuation of $3,000 as of October 4th. The commissioner saw and heard these witnesses, and, in view of this conflict, his conclusion on a question of fact will, ordinarily, not be disturbed; and we do not find in the results of this sale at auction, held three months afterwards, in the depth of winter, and after the Garlick had been in another collision, sufficient to warrant a departure from the usual practice. The last clause of admiralty rule 57, “If the ship have already been libeled and sold, the proceeds shall represent the same for the purpose of these rules,” cannot be construed as providing that in the case specified the value of the vessel, freight, etc., for the purposes of limitation of liability, is to be assessed as of any later period than the termination of the voyage during which collision happened. The decrees of the district court are affirmed, with costs.