sugar in bags, consigned to the libelant. The sugar in controversy was stowed in No. 3 and No. 4 holds, and some of it was found, while being discharged, to have been damaged by water. It does not appear precisely how much of the cargo was damaged. Apparently the bulk of it was sound. But at the top of No. 3 hatch, on the starboard side, the sugar was wet all across the hatchway to a depth of about eight feet. Under the starboard ventilator of the same hold a burrow, caused by water, extended down about eight feet. Under this there was a layer of sound cargo down to a point about three feet from the bottom. On the bottom of No. 4 hold also, as well as of No. 3, there was a layer of damaged cargo about three feet thick. There was no damage elsewhere in No. 4. In each hold the damage at the bottom was on the starboard side of the tunnel.

The damage at the top of No. 3 hold was caused by salt water that found its way into the hold on April 19th in the manner hereafter stated. The damage at the bottom of the holds, however, was caused by fresh water, the marks on the bulkheads showing that both holds had been flooded to a height of two to four feet. In order to determine how water, either salt or fresh, may have found its way into these holds, it is desirable to refer to the construction of the vessel in some respects, and also to the circumstances of the voyage. The Wildcroft has four holds, two on the fore side of the engine-room tank, and two on the after side. No. 3 and No. 4 holds are on the after side of the engine room, and are separated from each other by a grain-tight bulkhead. A water-tight bulkhead separates No. 3 hold from the engine room, and a water-tight bulkhead also separates No. 4 hold from the peak tank aft. The flooring under No. 3 and No. 4 holds is not water-tight, but is better than grain-tight. The average depth of the bilges below the flooring is 2 feet 6 inches. As the bulkhead between No. 3 and No. 4 holds is pierced by limber holes, these two compartments are practically one, so far as the passage of water is concerned. The vessel has five tanks. No. 1 is under No. I hold, No. 2 under No. 2 hold, No. 3 under the engine room, No. 4 under No. 3 hold, No. 5 under No. 4 hold; and the after-peak tank is aft of No. 5. All these tanks, excepting the afterpeak tank, are built on the double cellular principle. No. I and the after-peak tank were empty during all the voyage from Baltimore to Havana, and thence to Cardenas, Matanzas, and Philadelphia. Nos. 2, 3, 4, and 5 tanks were filled with water for ballast at Havana after the discharge of the cargo, but when the vessel arrived at Cardenas they were pumped dry, and remained dry during the rest of the voyage to Philadelphia.

After the discharge of cargo at Havana, the sluices, holds, and bilges on the ship were overhauled and cleaned. At Cardenas part of the cargo was loaded, the rest being taken on board at Matanzas. After leaving this port, each hatch was protected with a wooden cover and with three tarpaulins that were securely battened down with iron bars and wedges. On leaving Matanzas, the vessel drew about 21 feet 6 inches fore and aft, and was then in every respect seaworthy, and properly manned, equipped, and supplied. The hatches were not taken off until after the vessel reached Philadelphia. Dur

ing the voyage the weather was exceptionally severe, culminating on April 19th in heavy gales, tropical rain, and high seas, the result being that the vessel rolled and pitched heavily, and shipped large quantities of water on deck. About 3 p. m. on that day a heavy sea tore away the three tarpaulins that covered No. 3 hatch and the starboard ventilator cover at the forward part of the hatch. New tarpaulins were lashed temporarily over the hatch as securely as the violent weather would permit, and a spare cover was put over the ventilator, but the weather did not moderate sufficiently until 6 or 7 o'clock the next morning to permit the crew to fix the tarpaulins permanently over the hatch, and thus to make it again as secure as when the vessel left Matanzas. During this interval, the vessel shipped heavy seas, the downpour of rain continued, and a considerable quantity both of sea water and of rain water entered No. 3 hatch through the joints of the hatch and down the ventilator hole. Not much water reached the bottom of either hold at this time. Both holds were sounded night and morning during all the voyage, and no more than the usual amount of water was found-two to four inches. The vessel arrived at Philadelphia on April 22d, and was taken to the libelant's wharf to discharge her cargo. Before the discharge was begun, the sugar directly under the hatch of No. 3 was found to be damaged by water. All the bags across the hatchway were wet. Under the ventilator whose cover had been washed off there was a burrow about eight feet deep between the bags where the water had run down. The discharge began on April 23d, and was continued without further incident until April 29th. At 7 a. m. of that day, in accordance with the standing orders that soundings should be taken night and morning, the pumps connecting with No. 3 and No. 4 holds were started and were worked for 15 minutes, but nothing was pumped out. About 3 p. m., however, the stevedore that was discharging the vessel reported to the chief officer and the chief engineer that there was water in hold No. 4. Soundings were taken immediately, and between five and six feet of water were found. There was no water in the engineroom bilge, however, and the sluices in the tunnel connecting with No. 3 hold and the after-wells of the engine room were then opened, and the water and molasses were pumped out, both holds being dry by midnight. The chief engineer tasted the liquid in the bottom of the hold, and found it to be molasses and fresh water. In order to determine whether any of this water had come into the ship through a leak in the hull, or in any of the ballast tanks, the vessel was thoroughly examined by Lloyd's surveyors in Philadelphia immediately after the cargo had been discharged, and before repairs of any kind had been made. No leak or injury to the hull or tanks was found, and the surveyors thereupon gave the vessel a certificate of seaworthi

ness.

The fresh water in the holds is accounted for as follows: On the morning of April 29th, No. 3 engine-room tank was filled with fresh water from the Delaware river, for ship's purposes. It began to flow into the tank at 10 o'clock and the flow was uninterrupted for three hours. A drawing showing the arrangement of pipes, connections, and valves involved in this explanation, is annexed to the claim

ant's testimony. This drawing was prepared by a consulting engineer who examined the vessel in London in June, 1901, before any repair, alteration, or change in arrangement had been made in any of the pipes, cocks, or valves in use on April 29th, when the damage was done. No. 3 tank, just referred to, is under the engine room. It is filled by opening a valve, A, in the ship's side, which admits water from the river into the tank-filling pipe, A, B. This pipe is connected, on its continuation below the point B, with valves in the tank distribution box. From this box another pipe leads to the tank top. (Neither the tank distribution box nor the pipe connecting it with the engine-room tank are shown on the plan, as they have no connection with or relevancy to the present controversy.) There is no direct connection between this tank and No. 3 and No. 4 holds. There is, however, a pipe connection, C, governed by a cock at D, that leads from the tank-filling pipe to the service donkey; and this in turn is connected with another distribution chest, from which a pipe is led to No. 3 hold for the purpose of pumping out the bilges of that hold. This pipe is indicated by the letters, E, F, G. At the head of this suction pipe in the distribution chest is a non-return valve, E. This valve and the seat on which it rests are made of gun metal. The valve apparently works on a hinge, so that it opens when suction is applied to it from the top by the pumps, but when the suction ceases, the valve falls back on its seat, and prevents anything from entering the pipe which it covers. If, therefore, any foreign substance, such as a small piece of wood, had been drawn up the pipe, F, G, when the pumps had been worked last on the limbers of No. 3 hold, and had become caught in the valve, E, so that the valve had been wedged open, and if at the same time the cock, D, in the donkey service or branch pipe were open, it is apparent that water flowing from the river into the tank-filling pipe, A, B, leading to the engine room, would also flow through C, D, to the distribution chest, and down the valve, E, and the suction pipe, F, G, leading to the limbers in No. 3 hold, and thence into No. 4 hold; for, as already stated, No. 3 and No. 4 are practically one hold, so far as water communication is concerned. If these cocks and valves, D and E, or either of them, had been properly closed during the process of filling the engine-room tank, it would not have been possible for any water to find its way into No. 3 and No. 4 holds.

I think, therefore, that the damage to the sugar was caused by the water that went down No. 3 hold during the storms on the voyage, and by the water that flowed into the hold through the pipe line on April 29th in the manner just described. Indeed, it is impossible that the damage could have occurred in any other way, unless some of the water that entered through No. 3 hatch on April 19th passed down the sides of the ship into the holds; and I do not think the testimony justifies such a finding.

It will be observed that nothing is said in the foregoing statement of facts concerning damage to sugar in No. I and No. 2 holds, for I agree with the respondent's contention that the libelant should not now be allowed to make any claim therefor. The libel claims in general terms for damage to "a large portion of said cargo of sugar,"

without specifying in which holds the injured portion was to be found.. The libelant's testimony was confined, however, to the sugar in No. 3 and No. 4 holds, nothing whatever being said concerning the sugar in No. I and No. 2 holds, except one question and answer in the examination of James W. Andrews, who was apparently a general utility man in the libelant's employ; his testimony concerning his duties being: "I don't know just what my place at McCahan's is. I have to look after things generally." The question and answer just referred to are these:

"Q. You have not made any demand for damage outside of Nos. 3 and 4?

"A. No, sir; there were 209 in No. 1 and 42 in No. 2."

It is now objected that the witness was a mere clerk, without authority to bind the libelant, and that his answer should be disregarded. To this objection I think it is enough to reply that the libelant accepted the answer without apparent demur, and could not, without obvious unfairness, be permitted to repudiate the witness at the argument of the case. Prompt notice should have been given that his answer was not admitted to be correct. Aside from this consideration, however, there is no testimony whatever (save the brief answer just quoted) concerning the damage in No. 1 and No. 2 holds; and the libelant has, therefore, failed to make out its case in this particular. Whether the sugar in these holds was injured by salt water or by fresh water, or by some other cause, does not appear; and it is manifestly out of the question to allow a claim for damages that is founded upon a single short answer that is so vague and indefinite as the few words to which I have already referred.

Turning to the damaged sugar in No. 3 and No. 4 holds, and inquiring concerning the vessel's liability for the injury, it is first to be noted that no claim is made for the harm done by salt water to the bags in the top of No. 3 hold. This damage was concededly caused by a peril of the seas, and this part of the claim was formally withdrawn by the libelant's counsel. Concerning the bags that were damaged at the bottom of the holds by fresh water, the defense to the libelant's claim rests upon certain provisions in the bill of lading and upon the third section of the Harter act. The libelant asks me to say that a presumption of negligence arises from the fact that at the end of a voyage merchandise that has been received by a carrier in good condition for safe conveyance is found to be damaged; and that the burden of proof is upon the carrier to explain the cause of the injury, upon penalty of being held liable if he is unable to clear his skirts of fault. No doubt this proposition is sound, and does not need the support of authority; but it does not establish the libelant's right to a decree in the present case, for the plain reason that the ship accepted the burden of proof, and has satisfactorily shown how the sugar in the bottom of the two holds came to be injured by fresh water. To my mind, as already intimated, the conclusion is irresistible that the water found its way into the holds on April 29th, while the ship was lying at the wharf discharging her cargo, and that the entry of the water could only have been made because the valves were improperly open. I adopt the theory pro