.PER CURIAM.
This is an appeal by the claimants, owners of a cargo in part destroyed by fire while in transit upon the steamship Strathdon, and in part saved,, but damaged in extinguishing the fire, from a decree in a proceeding instituted by the shipowners for limitation of • liability. The decree below adjudged that the fire was. not caused by the design or neglect of the owners of the vessel, and the loss was occasioned without; their privity and knowledge, and exonerated and discharged them for all loss and damage *602arising from or growing out of it, except contribution in general average. The Strathdon sailed from Java, with a full cargo of sugar, boiind for the port of New York. The cargo was shipped pursuant to a charter party by charterers who loaded it themselves or through agents, and received bills of lading reciting that it was carried pursuant to the terms of the charter party, and was deliverable to order. The charter party provided that the carrier should not be liable for loss or damage “occasioned by causes beyond its control; by the perils of the sea or other waters; by fire from any cause and wheresoever occurring; by collision, stranding, or other accidents of navigation of whatsoever kind, even when occasioned by the negligence, default, or error in judgment of the pilot, master, mariners, or other servants of the shipowner, not resulting, however, from any carelessness or want of diligence by the owners of the ship, or any of them, or by the ship’s husband or master.” After stopping at Point de G-alle and Perim for coals, the steamer arrived at Suez, and entered the canal, bound for Port Said, on the afternoon of October 31, 1893. About half past 2 in the morning of the next day .it was found that the cargo in the starboard between-decks was on fire. This fire spread rapidly forward in the between-decks, and to the lower hold, notwithstanding the most energetic efforts to arrest it were made by those in charge of the steamship, and was not extinguished until the morning of November 5th. In the meantime, owing to the large quantity of water which had been pumped into her, the steamship settled on the canal bank, listing over to starboard until more than half of her main deck was under water. She remained in this position until the water was pumped out of her, November 5th, and her cargo was shifted and partly discharged, when she was enabled to proceed in tow towards Port Said. No water entered any of the holds through the hatchways except hold No. 2. But the water pumped into No. 2 hold passed through the coal bunkers and engine room through a two-inch drain pipe into hold No. 3, and into No. 1 hold through openings in the bulkhead, caused by fire. Water reached No. 4 hold through the captain’s cabin. The bathroom and toilet pipes were submerged by the listing of the ship, so that water flowed through them to this cabin, and passed from there through a two-inch drain pipe leading into the bilge of hold No. 4. Large portions of the sugar in all the holds were melted, and totally lost. The portion of the cargo which was not destroyed, was discharged at Port Said, this work being completed -December 18, 1893. Duly-constituted surveys held on the vessel at Port Said held that she was unfit to make an Atlantic voyage without first effecting permanent repairs. The steamship remained in Port Said until January 21, 1894. At first it was supposed repairs could be made in Port Said, and some were inade, beginning on November 28th. It was subsequently found that permanent repairs could not be completed, nor could a certificate of seaworthiness be obtained, without docking the ship, and that, as there was no dry dock at Port Said, she would have to proceed for repairs to some other port. A further survey recommended that *603temporary repairs be made in Port Said, and that the vessel should proceed to Trieste to make permanent repairs. Temporary repairs were accordingly made at Port Said, and the steamship sailed thence for Trieste January 21st, arriving there January 27th. Work on the permanent repairs was promptly begun, and a large force was employed continuously making them, but the damage to the vessel proved to be so extensive that the repairs were not and could not be completed until April 26th. The steamer sailed from Trieste on that day, returned to Port Said, and took on board the cargo which had been stored, and then proceeded on her voyage towards New York, arriving there June 1, 1894, and thereupon delivered the remnants of the cargo to (.he claimants.
The claimants alleged that the fire and consequent damage to their sugar were caused by the neglect and default of the shipowners, and that: the latter were liable for the damages. They also alleged that the shipowners negligently delayed making the repairs necessary to enable the steamship to proceed after the fire, and carry the cargo to its destination, and should have caused the cargo to be transhipped and forwarded by another vessel; and that in consequence of their default, the market value of the cargo having declined, the claimants sustained large damages. Their claim for damages to the cargo was based upon the theory that the tire was caused by heat transmitted from the flue of the donkey boiler; that the steamship was not equipped with proper preventive means; and that the vessel was not provided with proper appliances for preventing the access of water to the cabin, which entered No. 4 hold from the cabin, and occasioned the injury to their cargo stored in that hold. The claimants also alleged that the steamship and freight moneys were liable to contribute in general average for the value of the sugar damaged and destroyed by the water poured into the steamship in order to extinguish the fire. The court below decided 1hai: the steamship was in all respects seaworthy; that (lie construction of the donkey-boiler and flue complied with all the known demands of skill and safety; that the repairs to the vessel were proceeded with as speedily as possible, that the claimants had acquiesced in having the cargo forwarded to its destination by the steamship rather than by another vessel; and decided as conclusion of law that the petitioners were not liable for any damages to the claimants. The decree appealed from proceeded upon these conclusions.
In disposing of the claim for general average, the court found as a fact that the Are was comimlnicated to the cargo by the heat disseminating from the donkey boiler flue in consequence of the carelessness of the employes of the steamship in causing the flue to become overheated; and decided as a conclusion of law that in arriving at the general average contribution the adjustment should be made as if there had been no negligence on the part of the shipowners. This conclusion proceeded upon the ground that the shipowners were exonerated from liability by the statutes relieving them from liability for losses caused by Are occurring by the default of *604their servants without their own neglect; and that, notwithstanding the shipowners could not compel the cargo owners to respond in general average for losses which would not have arisen except for the default of their own servants, when the cargo owner invokes a recovery for general average in such a case the shipowner is also entitled to contribution as though innocent of fault; otherwise the cargo owner would recover by selecting his form of proceeding for losses for which the shipowner was not responsible. We think that the court below made a correct disposition of the case, and concur in the main with the conclusions of fact, and fully with the conclusions of law, set forth in the very th.orough and satisfactory opinion of Judge Thomas in deciding the cause; and we should deem it unnecessary to add anything to his opinion if we did not differ with him in respect to one question of fact. A careful study of the proofs leads us to conclude that it was not established that the fire originated from the donkey boiler, or was caused by negligence of those in charge of the boiler. Although this -conclusion does not affect the correctness of the decree appealed from, it is proper that our reasons for it should be stated. The theory of the learned judge was that by the carelessness of those in charge of the donkey boiler the flue became overheated, or red-hot, and generated sufficient heat to set fire to the cargo stowed in the between-decks. This theory was largely based upon the circumstance that the fire could not be attributed, except conjecturally, to any other cause. It is true that the fire was first discovered in the cargo in the between-decks in the vicinity of the donkey boiler; and it is also true that, unless the fire was communicated to it by the overheated flue, no definite producing cause can be found, and it can only be conjectured that it may have been caused by a spark entering through the ventilator, and igniting the baskets in which the sugar was packed, or by the friction of the baskets, or by a match dropped among the baskets by the stevedores in loading and ignited by attrition, or by some other unknown agency. We are not satisfied that the flue was suffered to become red-hot during the night of the fire, or that, if it had been overheated to any degree for any considerable period before the fire, the heat would have sufficed to ignite the cargo; and the burden of proof was upon the claimants to establish both of these propositions, The donkey boiler occupied a room below the between-decks. It stood in a recess in the stoke hole formed by the stoke-hole bulkhead. The roof of the recess wras the iron main deck forming the floor of part of the between-decks. The boiler was 14-| feet high. The crown of the boiler was about 34 feet below the iron roof. The flue arose from the dome, was of wrought iron, was about 18 inches in diameter, and led diagonally across the ship to the funnel of the main boilers, and at the point nearest the roof was 19 inches distant therefrom. The furnace was at the bottom of the boiler, and from the top of the fire to the top of the boiler the distance was 10 or 12 feet. The entire shell of the boiler, including the dome, was constructed to contain .water, and there were four transverse tubes for water, each 12 inches in diameter, leading across *605the interior from side to side. The expert witnesses testified— and their testimony commends itself to onr judgment — that it would be almost impossible for the heat to rise in the boiler as constructed in sufficient volume to make the flue red, at least to any appreciable extent; and also that a fire intense enough to make the flue red would cause the steam to explode the boiler, if the safety valve did not lift, and thus give warning to those in the stoke’ hole and engine room. It was contrary to instructions to permit- the flue to get red-hot, because of the tendency to deteriorate the iron. The ordinary use of the donkey boiler did not require it to be forced so as to overheat the flue; and the preponderance of the testimony of those who saw the flue shortly before and after the time of the fire indicates that it was not, and had not been, red-hot. The only testimony to show that the flue was red on the night in question is found in the deposition of Love, a fireman, who was temporarily acting as third engineer. Before he testified for the claimants he was expected to testify for the shipowners, but, upon being discharged from the steamship for misconduct, he promptly put himself into communication with the claimants. He testified that after the fire was re-, ported he- went into the stoke hole to see what he could see, looked ‘at the flue, and it was red “top, bottom, and all around” for three or four feet from the top of the boiler. His testimony upon collateral facts tending to the prejudice of the shipowners was proved to be untrue. We regal'd him as a discredited witness, whose testimony was of no value. The builders, surveyors, and all other expert wit: nesses were unanimously of the opinion that it was physically impossible for sufficient heat, to be communicated from the boiler flue through the intervening spaces, baffle plates; and iron deck so as to set fire to the cargo. The cargo was stored in baskets, and these baskets rested on planks, and, before the baskets could have been set on Are, the heat in the iron deck must have been sufficiently intense to ignite them. The flue was in sound condition. Two curved, wrought-iron baffle plates or awnings, one-eighth of an inch thick, were attached to the flue, and extended oyer the whole of that part of it which was under the deck. The first one of the awnings was about 3 inches above the flue, and was attached directly to the flue by studs; and the second was about 8 inches above the first,; and was attached to the first awning by studs. Still another baffle plate, 2-1 feet wide, was attached to the iron beams upon which the deck rested, and extended over the whole of that part of the flue which was under the deck. There was an air space of 10-J inches between this baffle plate and the deck, and a space of 3 inches between it and the upper one of the two awnings which were attached to the flue. According to persuasive testimony, the air space of 19 inches between the flue and the iron floor would have protected the deck from becoming dangerously heated had there been no baffle plates. The utility of any of the baffle plates except the one attached to the deck beams was. doubted by some of the witnesses,' who were of the opinion that the freer circulation of air without them would have compensated for their absence. But all the wit*606nesses agree that the baffle plate attached to the beams would protect the deck from getting overheated under any circumstances, and that to whatever degree it could become heated from the flue the ample air space between it and the deck would adequately protect the deck. The witnesses were men of experience and great intelligence, and there is no reason to doubt the candor of their statements. If it should be assumed that the flue was red-hot, and should then be conjectured that the nearest baffle plate, and next the second baffle plate, became intensely hot, and that next the independent third baffle plate became intensely hot, there still remained to be bridged the air space of nearly a foot before the deck could become dangerously hot. The testimony is so cogent that it would be impossible to transmit enough heat from the flue to the deck to dangerously heat the deck that we cannot reject it. It is so inconceivable that the heat from the flue, even if the flue were red-hot, could have been communicated through the series of baffle plates and air spaces to the deck, and through the deck with intensity enough to ignite the planks, that the conjecture that it did so is no more prohable than the other conjectures as to the cause of the fire. The occult causes of fire are as numerous as they are mysterious. If the flue was red-hot on the night of the fire, it may or may not have been the producing cause. If, as we think, it was not, there is no satisfactory proof of the origin of the fire. The decree is affirmed, with interest and costs.