ROGERS, Circuit Judge
(after stating the facts as above). The court below reached the conclusion that the damage to the cargo was caused by'the escape of fresh water from a pipe in the hold of the vessel, and that the water escaped from the pipe because there was no cock, 'plug, or stoppér in the pipe. The defense claimed that the stopper must have been knocked out of the pipe by the working of the cargo and that the cargo worked or shifted by peril of the sea. The court below reached the conclusion that no such sea peril was shown.
[ 1 ] The bills of lading provided that the carrier was not to be liable for loss or damages “occasioned by causes beyond his control, by the peril of seas or other waters, by collisions, stranding, jettison, or other accidents of navigation of whatsoever kind.” And it is claimed that the court fell into manifest error in holding that the burden of proof was upon the carrier of the goods to show that perils of the sea caused 'the damage to the cargo and in holding that the burden had not been sustained. It is admitted that the bales of hemp were received by the carrier in good condition and delivered in bad condition. That being so, there certainly is no question but that the carrier, in seeking to be relieved from liability for damages under the exceptions of perils from the sea, was bound to prove that the injuries were the result of such untoward circumstances as could not have been anticipated and guarded against by the exercise of ordinary care and prudence.
[2] But, conceding the burden to be upon the vessel, was the burden satisfied by the evidence? Perils of the seas are understood to mean those perils which are peculiar to the sea, and which are of an extraordinary nature or arise from irresistible force or overwhelming power, and which cannot be guarded against by the ordinary exertions of human skill and prudence. See 13 Cyc. 258.
[3] The master of the vessel testified that on December 25th he encountered very bad weather; that his ship was rolling badly and the seas coming on deck. He testified that the feeding tube to the conden- - ser broke, as did one of the valves in one of the boilers; that he was compelled to stop the ship and to lower the sea anchor, to have the ship drag; and that the gale was so strong that the pipe rope he had to the sea anchor broke. He also testified that the cargo was moved by the weather and that on his arrival in port he made a regular protest. The protest states that the ship proceeded on her voyage from December 12th to December 18th with variable winds and weather, and that on the afternoon of the latter day a strong northerly gale set in and the sea was high and rough, the ship rolling and pitching, with heavy seas breaking over the decks; that towards night the weather became worse, and the ship was fearfully shaken by the huge bodies of water dashing over and against the ship. The weather continued bad from that time until after December 26th, the ship rolling and pitching and *747having her decks flooded. On December 24th the protest refers to “furious storms,” with “high, tempestuous sea, ship laboring fearfully, and decks flooded until afternoon.” On December 25th “heavy sea, ship rolling heavily.” December 26th, “sea rough, and ship plunging fearfully, and taking seas over all.”
There was, however, no testimony in the case showing that the liemp had been damaged by sea water. But there was testimony to show that it had been damaged by fresh water. When the vessel reached New York, and while the cargo was still in the hold, the testimony showed that fresh water was running out of a fresh water pipe and into the compartment where the hemp was stowed. The hold in which the bales of hemp were stowed was one intended to be used when necessary for passengers, and on this particular voyage three-quarters of the whole cargo was in that hold.' In this hold there was a fresh water pipe, which the master of the ship described by saying:
“This pipe was coming out through the upper ’twoen-deck on the forward bulkhead between the two angle irons and down from the upper ’tween-deck to about five or six feet from the lower ’tween-deek where the cargo was.”
The pipe was connected to a tank placed in one of the highest places of the ship and was used to supply the fresh water to the passengers when the hold that on this voyage was used for cargo was occupied by passengers. The tank held between 2% and 3% tons of water, and was filled about five times a day from the big storage of water kept in the ballast tanks. When this hold was occupied by passengers a cock was attached to the pipe, which passengers used in drawing water, and when the hold was used for cargo the cock, was removed and a brass plug was screwed on in its place. The pipe was of galvanized iron and about three-quarters of an inch thick.1 The cock or plug was protected by an iron fender, which was fastened to the bulkhead of the ship with two three-quarter inch bolts. When the ship reached New York, it was found that this bracket was hanging down and one of the bolts holding the bracket was gone. The plug which should have been in the pipe was not there, and the claim is that it was there in place originally, but had been displaced by the alleged shifting of the cargo. But if the plug was originally fastened into the pipe, and had been worked out by the shifting of the cargo, it should have been found; but it never was found. This plug, which it was the custom to use, and which should have been used, but apparently was not, was a piece of brass threaded at one end and squared at the other, so that it could be screwed tight with a spanner. The testimony showed that the pipe had not been broken or injured and that the threads inside the pipe were not broken. It would seem that if the plug had been in place, and forced out by the shifting of the cargo, there should have been some indications on the pipe showing the strain to which it had been subjected. We are not convinced that the cargo ever shifted, or that, if it did shift, that it displaced the plug running to the fresh water tank, the water from which in our opinion damaged this cargo.
We have not overlooked the testimony of the master of the ship that a plug was screwed into the pipe at Trieste; but we are unable to accept that testimony, in view of the fact that the carpenter whose duty k *748would have been to insert the plug was not called and the log books were not produced. The conclusion to which we have arrived is that the ship was not seaworthy at the commencement of the voyage. There is nothing in the testimony to indicate that the plug was broken off or unscrewed in some way. It seems probable that after the cock was removed, when it was decided to put cargo into the hold, some one failed to screw in the plug. At least such a supposition is as probable as that some movement of the cargo in heavy weather unscrewed it, after it had been screwed into place tightly with a spanner. Presumably the tank would not be in service until the ship was ready to receive passengers, and if there was failure to insert the plug at that time no water would escape. Rater, when passengers went on board, and the tank was made ready for their use — for there were other pipes running from the tank to other parts of the ship — the cargo no doubt had been stowed to a point above the aperture, so that the absence of the plug was not noticed.
We are unable to find on this record that the plug disappeared through perils of the sea. The appellant has not sustained the burden of proof, and the decree appealed from is affirmed.