The appeal in this case is taken from a decree in a proceeding to limit and determine the liability of the Michigan Steamship Company for injuries arising from the explosion of a fuel oil tank on its steamer Progreso, on December 3, 1902. During the summer of that year the steamship company, under a contract with the Fulton Iron Works, whereby the latter had agreed to convert the Progreso from a coal-carrying and coal-burning steamship into an oil carrier and burner, moored the Progreso at the dock of the Fulton Iron Works in San Francisco. The contract called for a complete reconstruction of the interior of the vessel, whereby it was to be divided into transverse partitions, consisting of a series of six tanks, of which the hull of the vessel was to form the bottom and the sides, and the tops were to constitute the main deck. It was during the progress of the work on this contract that the explosion occurred. The space between the tops of the tanks and the upper deck was dark, and was lighted during the repairing by both electric lights and candles, all of which were furnished by the iron works. One of the tanks was intended for fuel oil. It was the fourth tank from the bow, and was just forward of the engine room. It was constructed to be filled from the upper deck through an extension of the tank, which was called the 'extension trunk,' which extended from the lower deck to the upper deck from the center of the forward end of the top of the tank. Five days before the explosion the fuel tank had been tested for leaks and the leaks had been stopped, and the tank had been approved by the United States inspector as to the tightness of its walls. The ventilators for the tank had been ordered by the inspector, and they were provided for in the plans and specifications for the work. The form of these ventilators and the precise places where they were to be attached to the tank does not appear from the testimony. A part of the work to be done on the ship consisted in putting in new boilers and an oil-burning furnace. The Fulton Iron Works furnished the parts of this machinery, and the employes of the Michigan Steamship Company installed them several days before the explosion. The steamship company desired to test this new machinery. It accepted the fuel tank from the Fulton Iron Works before the ventilators had been put in, and three days before the explosion it placed in the fuel tank about 400 barrels of fuel oil, partly filling the tank and leaving 1,200 barrels air space above the oil in the tank. By the contract of the steamship company with the iron works the latter was required to erect a stanchion to sustain the upper deck, which was to rest on the top of the fuel tank and to be secured to the under side of the upper deck and to the top of the tank by means of tap bolts. This required drilling through the top of the tank for the insertion of the bolts. The fuel oil was put in the tank before this was done. On the afternoon before the explosion Ferguson, the underforeman in the employment of the Fulton Iron Works, who knew that the oil was already in the tank and had examined the tank before the oil was received, sent McGinley, a driller, to drill the holes for the tap bolts of the stanchions. He partly completed the holes that afternoon, using for light a candle which hung from a string on the drill frame. The following morning he continued his drilling, and it would appear that on drilling through the top of the tank and withdrawing his drill a stream of oil gas from the tank was ignited by the candle, whereby the gas in the tank exploded, killing McGinley and about a dozen more, and injuring a large number. The claims presented to the District Court were for injuries to person and loss of life, and were of two general classes: First, those arising from killing and injuring sailors on the ship; and, second, those arising from killing and injuring employes of the Fulton Iron Works. The decree of the District Court limited the liability for all claims to the fund of $15,020 and interest. It allowed the claims of the heirs of sailors, but disallowed the claims arising from the death or injury to the employes of the Fulton Iron Works. The appeal is taken from that decree by certain of the employes of the iron works and their personal representatives.

William Denman, J. A. Watt, Sullivan & Sullivan, and F. R. Wall, for appellants.

Nathan H. Frank, for respondent.

Before GILBERT and ROSS, Circuit Judges, and HAWLEY, District Judge.

GILBERT, Circuit Judge, after stating the case as above, .

It is the contention of the appellants that the Michigan Steamship Company was negligent in using oil of the quality of that which it used, and especially in putting oil into the fuel tank before the installation of ventilators in the tank, and that, if the tank had been supplied with the ventilators universally used in such fuel oil tanks, there could have been no explosion under the circumstances disclosed in the evidence, and that the employes of the iron works had the right to rely on the belief that the tank was properly ventilated, and that otherwise oil would not be placed in it. The undisputed evidence is that an inflammable gas rises from all of the crude oils used for fuel purposes, and that such gas will rise more or less rapidly, irrespective of the temperature of the body of the oil, or of its flashing point; that the gas unmixed with air is not explosive; that, when mixed with a certain proportion of air, it is highly explosive; and that as that proportion is departed from in either direction, whether by excess of air or excess of gas, the combination of gas and air is still explosive, but with diminishing force until a point is reached where the air is so much in excess of the gas, or the reverse, that no explosion will occur. The evidence shows, also, that the explosive combination of gas and air will not explode unless it is brought in contact with a spark or a flame, or is heated to a temperature of approximately 800 degrees. The testimony shows, further, that the gas rising from fuel oil is slightly heavier than air, and that ventilator tubes are ordinarily fixed in the top of closed fuel oil tanks, so as to permit the overflow of the gas after it has risen and occupied the space in the tank above the oil. But the testimony leaves it extremely doubtful whether the presence of such ventilators in the tank in question would, at the time of the accident, have rendered the gas in the tank nonexplosive. The oil had been in the tank about three days when the explosion occurred. During a portion of that time the cover to the trunk of the tank was off, leaving an opening about 3 feet 6 inches long by 1 foot 6 inches wide exposed to the air. The cover was off, according to the testimony, from six to eight hours on the day preceding the explosion. The ventilators which were to be placed in the tank, and which were approved by Lloyd's surveyor and by the government inspector of steam vessels, consisted of tubes rising from the top of the tank through the upper deck and extending above it.

The testimony is, without contradiction, that there is no danger of explosion in a tank tightly closed and unsupplied with ventilators. Prof. O'Neill, who is shown to be a qualified expert, testified that the removal of the top of the trunk of the fuel oil tank would probably not ventilate the tank so as to avoid the accumulation of dangerous gases, and that such a ventilation would not be safe. Mr. Ransome, an expert witness, testified that as long as the tank does not leak 'you do not need ventilation.' Capt. Metcalf, who was Lloyd's surveyor, testified that ventilators were not used to prevent explosions, but for utility. He said:

'There is no place safer in a ship than for the gas to be retained in the tank itself; and, if it was not necessary to admit air and allow air to be expelled in filling and emptying, you could not have it better than to have it air tight absolutely.'

Prof. O'Neill suggested a method of ventilation which would be absolutely safe-- a method by which a current of air might be forced into the tank by the use of a centrifugal blower or fan connected with a pipe leading from the blower into the tank. This device, however, does not appear to have been used on board any of the numerous ships that use oil for fuel. We do not overlook the testimony in regard to the adoption of such a blower on board the San Pablo after the accident to the Progreso; but that blower was not connected with the fuel tank. The fuel tank was placed in the hold, and it was in order to prevent the accumulation in the hold of gases escaping from the tank that a pressure fan was installed in the bottom of the hold, connected with a canvas hose to one of the portholes, through which it blew the gas, not from the tank itself, but from the hold. The use of that device was not required by the supervising inspector, and, according to the testimony, that officer has not at any time required fuel tanks to be ventilated in that manner. We do not think that the steamship company was bound to adopt this device, or any device not dictated by their own knowledge and experience, or that of others, or that its failure to do so was negligence. Nor does the evidence in regard to the probable utility of ventilator tubes rising from the fuel tank through the upper deck and discharging gases into the air carry the conviction that, had such ventilator tubes been installed in the fuel tank in question, the explosion could not have occurred.

The whole question of the explosibility of the contents of the tank above the oil, with or without ventilation, must necessarily rest in mere conjecture. It is conceivable, it is true, that the presence of such ventilators might have permitted all the air in the tank to escape, leaving nothing but the gas above the oil in the tank. It is equally conceivable that, with the presence of such ventilators, enough air might have remained in the tank to have rendered the explosion more violent than it was. There is no evidence in the record that the purpose of ventilating tubes, such as were to be installed in the fuel tank in question, was to prevent explosion in the tank. On the other hand, the evidence seems to indicate that such ventilators were to serve the double purpose of permitting the inflow of air to facilitate the pumping of oil from the tank for fuel purposes and to furnish a vent for the escape of gas at a place where it could work no injury.

We do find, however, that the steamship company was negligent in putting in the fuel tank oil of the quality of that which it used, and fuel oil, no matter what its quality, during the progress of the work on the tanks, and at a time when it knew that work remained to be done in drilling holes into the tank for the insertion of tap bolts to secure the stanchion, without any warning of danger to the workmen of the Fulton Iron Works.

The question then arises whether the negligent act of the steamship company was the proximate cause of the injury. Many definitions of proximate cause have been formulated, but probably no fixed and definite rule can be applied to all cases. Said the court, in Insurance Co. v. Boon, 95 U.S. 117, 130, 24 L.Ed. 395:

'The proximate cause is the efficient cause, the one that necessarily sets the other causes in operation. The causes that are merely incidental or instruments of a superior or controlling agency are not the proximate causes and the responsible ones, though they may be nearer in time to the result.'

In Milwaukee, etc., Ry. Co. v. Kellogg, 94 U.S. 469, 475, 24 L.Ed. 256, 259, the court said:

'Did the facts constitute a continuous succession of events so linked together as to make a natural whole, or was there some new and independent cause intervening between the wrong and the injury? It is admitted that the rule is difficult of application, but it is generally held that, in order to warrant a finding that negligence or an act not amounting to wanton wrong is the proximate cause of an injury, it must appear that the injury is the natural and probable consequence of the negligent or wrongful act, and that it ought to have been foreseen in the light of the attendant circumstances.'

Applying these principles to the present case, we inquire whether the officers of the steamship company ought to have foreseen that probably an explosion of the gas might occur from the work of the employes of the iron works while putting in the stanchions between decks. Evers, the superintending engineer of the steamship company's vessel, was its officer to inspect the work and to accept it on its completion. He had been on the vessel nearly every day during the progress of the work, and 'saw every bit of it done. ' He knew, when he put the oil in the tank, that work still remained to be done in putting a stanchion between the fuel tank and the upper deck. He knew that the contract contemplated the bolting of the stanchion to the upper deck and to the top of the tank. He knew that this involved drilling holes through the top of the tank. He must have observed that the employes of the iron works were plentifully supplied with candles and were using them, as well as the electric lights. We think it a fair inference from all these facts and circumstances that he ought to have foreseen that an accident was likely to occur, and we hold that the injury was the natural and probable consequence of his act. It is true that the negligence of the steamship company resulted in injury only through the concurrent act of another. But the rule is that where the concurring act, though to a certain extent unusual, might have been foreseen and its consequences provided against, the first negligent act is the proximate cause of the injury. The Joseph B. Thomas, 86 F. 658, 30 C.C.A. 333, 46 L.R.A. 58, and cases there cited; Scott v. Hunter, 46 Pa. 192, 84 Am.Dec. 542; McDonald v. Union Pac. Ry. Co. (C.C.) 42 F. 579; Texas & P. Ry. Co. v. Carlin, 111 F. 777, 49 C.C.A. 605, 60 L.R.A. 462.

If it was, at the time of the accident, a matter of common knowledge, of which the court will take judicial notice, that all crude fuel oils give off gas at a low temperature, then it must be said that the workmen of the iron works shall be deemed to have possessed that knowledge, and that the use of a burning candle by McGinley was a negligent act, intervening the negligence of the steamship company and the injury. The tank containing the oil was immersed in cold water. The skin of the vessel was the bottom of the tank, and forward of the tank was a compartment containing water. The water of the bay was of a temperature below 60 degrees Fahrenheit. Is it matter of common knowledge that gas arises from crude petroleum under such circumstances? Things of common knowledge must be facts of such universal notoriety, and so generally understood that they must be regarded as forming a part of the common knowledge of every person. It may be said to be of common knowledge that crude oil is inflammable, and that all petroleum oils, whether crude oil or the refined oil used for household purposes, when heated, generate inflammable and explosive gas. Further than this we do not think common knowledge of the danger of petroleum goes. The very act of McGinley is persuasive argument, at least, that he did not know that he stood upon a mine of explosive gas. Ferguson, the underforeman in charge of the gang of men at work on the repairs, who gave McGinley his orders, saw him drilling into the tank by the light of a candle, and perceived no danger. He testified that he saw no occasion to warn the man that there was danger. In Brown et al. v. Piper, 91 U.S. 43, 23 L.Ed. 200, the court said that the power to take judicial notice 'is to be exercised by courts with caution. Care must be taken that the requisite notoriety exists. Every reasonable doubt upon the subject should be resolved promptly in the negative. ' In Norwalk Gas Light Co. v. Norwalk, 63 Conn. 527, 28 A. 32, the court took judicial notice that the operation of blasting is intrinsically dangerous. In Jamieson v. Indiana Natural Gas Co., 128 Ind. 555, 28 N.E. 76, 12 L.R.A. 652, it was held to be a matter of common knowledge that natural gas was a dangerous agency. The court said: 'We know, as everyone knows, that natural gas is in a high degree inflammable and explosive. ' But in Mississinewa Mining Co. v. Patton, 129 Ind. 472, 28 N.E. 1113, 38 Am.St.Rep. 203, the court refused to take judicial notice that natural gas will not pass through the soil from a leak in a street main. In Cherokee, etc., Coal Co. v. Wilson, 47 Kan. 460, 28 P. 178, the court said: 'If dry, fine coal dust is an explosive, it is not so universally recognized to be such that the court will take notice of the same without proof. ' And in Taggart v. Newport St. R.R. Co., 16 R.I. 668, 19 A. 326, 7 L.R.A. 205, the court took judicial notice that electricity, developed to a high degree of intensity, is exceedingly and even fatally dangerous, but the court said it had no judicial knowledge of the fact that it is dangerous when used by a street car company in the ordinary way as a motive power. There is no evidence in the case that the workmen engaged in the repairs had had any experience with fuel oils, or any peculiar knowledge of their properties. The only evidence in the record as to the general knowledge of the public on the subject is that of Charles Towe, the fire marshal of San Francisco. He testified that the occurrence of gas and its explosibility was generally known to oil men, that it was one of the first things they learned about the oil, but that this knowledge was confined, in his opinion, to the experts in the business, and was not generally known among the common people.

We have this situation, then: The steamship company filled the tank with a gas which, because of its volatility, was intrinsically more dangerous than gunpowder, and the employes of the iron works were permitted to work on and about the tank, presumably without knowledge of the presence of the gas, and without a warning of their danger. How was McGinley to know that gas would rise through the small hole he was drilling in the tank and seek the flame of his candle? He knew, it is true, that there was crude petroleum in the bottom of the tank, but he knew that the oil itself could not explode. If he knew, as he probably did, the quantity of oil that was there, he was aware that there was an air space of 1,200 barrels above the oil. We are of the opinion that it was error to impute to him the knowledge that that space was filled with an explosive combination of gas and air, and therefore to hold that he was negligent in lighting his work with a candle.

But under the provisions of the act limiting liability is the steamship company responsible beyond the value of the vessel for the injuries resulting from the negligent act of its superintending engineer? The appellants contend that it is, and base their contention on two propositions: First, that the engineer was the alter ego of the company; and, second, that he was an incompetent agent, and that the company failed to show that it had used diligence to inquire as to his competency before intrusting to his charge the handling of fuel oil. We think that the first of these propositions cannot be sustained. Evers was not a general manager. He was an employe whose duties were confined to the engineering department of the company's vessels. He was the port engineer, with supervision of the motive power of the several ships. The repairs on the Progreso were to be approved in the first instance by Lloyd's surveyor, but Evers was there to see that the steamship company 'got what it paid for.' The second proposition presents a question of greater difficulty. Jerome, the general manager of the steamship company, was absent from the state when the oil was put in the tank and when the accident occurred. He testified that, so far as the care and custody of the oil and the repairing of ships were concerned, 'at that time Evers was the man who represented the company. ' Before his departure Jerome had ordered the oil of the Union Fuel Company, and had instructed Evers 'to tell them when to put it in. ' His instructions were: 'When the Fulton Iron Works are done with the tank, and it is ready to receive oil, you can put the oil in. ' Evers did not obey these instructions. He had the oil put in before the iron works was done with the tank, and before it was ready to receive oil. He testified that he noticed at the time of receiving the oil that it was light oil, but that he took no test of it, and that he did not regard it as a dangerous oil. He said, 'I did not regard that there was any danger there. ' In answer to the question, 'Were you familiar at that time with the handling of light California fuel oils, I mean, you, yourself, personally?' he answered, 'No sir; I was not, personally. ' The oil which was put in the tank was an extremely fluid oil which flashed at a temperature of 85 degrees Fahrenheit. By an ordinance of the city of San Francisco it was made a misdemeanor to use for fuel any oils flashing at a temperature below 110 degrees. It is unnecessary here to discuss the question whether the ordinance was applicable to vessels lying in the harbor. As to his acquaintance with fuel oils, Evers testified that about 11 years before the time of the accident he was for eight or nine months first assistant engineer on English steamers in the Black Sea fleet, on which steamers oil was carried and burned for fuel, and that he had to do with loading oil on these ships, which he did by opening the valves and admitting oil into the ship when needed, and that the oil was black and thicker than the California oil.

'Q. I will put the question to you again: At the time that the Progreso blew up were you familiar with the explosive character of gases arising from fuel oils? A. I knew that all oils generated a gas that would explode if it was the right mixture of air. I knew that much about it. Q. And you were aware that that mixture in the tank there was dangerous? A. No; it was not dangerous to my knowledge then. How could it be dangerous when it was closed up? There was no air that could get in their except what was in there before. * * * Q. You knew at that time that the gas would come off the oil and make an explosive mixture? A. No; I did not know that when it was put in there. Q. You say you were sufficiently familiar with fuel oils at that time to know that the gases were explosive when mixed with air? A. I said I knew that all fuel oils could be ignited when mixed with air; that is to say the gases of them.'

The acts of Evers in handling the oil, and his testimony in regard to his knowledge of its properties, are such as to carry the conviction that he did not have the knowledge and experience necessary to render him competent to have charge of the work involved in changing the vessels of the steamship company from coal burners to oil burners, and the care and protection of the oil. He had no knowledge of the properties of California fuel oils. There is no evidence that he thought it necessary to acquaint himself with their properties. He handled the oil with apparently no regard to the danger involved.

The right of a shipowner to limit its liability is dependent upon his want of complicity in the acts causing the disaster, and the burden

Page 796.

of proof rests upon him to show affirmatively that he was properly officered and equipped the vessel for the contemplated service. Parsons v. Empire Trading & Trans. Co., 111 F. 208, 49 C.C.A. 302; The Main v. Williams, 152 U.S. 122, 14 Sup.Ct. 486, 38 L.Ed. 381; The Annie Faxon, 75 F. 312, 21 C.C.A. 366; In re Myers Excursion Co. (D.C.) 57 F. 240; The Republic, 61 F. 109, 9 C.C.A. 386; Quinlan v. Pew, 56 F. 111, 5 C.C.A. 438; The Colima (D.C.) 82 F. 665. In The Cygnet, 126 F. 742, 61 C.C.A. 348, the Circuit Court of Appeals for the First Circuit held that the analogous provisions of the Harter act cannot be invoked to relieve a vessel from liability for a loss occurring from errors in navigation on the part of the master sufficiently negligent to raise a presumption of his incompetency merely upon a showing that the owners had no knowledge or reason to believe that he was incompetent. The court said:

'There is no evidence in the record that the owners of the tug, either the record owners or the owner pro hac vice, had made any particular inquiries as to his competency. The petitioners seem to think it is sufficient to maintain their case that the owner or owners had no knowledge or reason to believe that the master was not competent; but this form of statement is not sufficient, because it does not comply with the statute, which requires 'due diligence."

Referring to the negligent act of the master in failing to observe whether his tow was straightened out on its course, the court said:

'An omission so gross as this raises so strong a presumption of fact that the master was not competent as practically to throw the burden on the petitioners to establish the proposition that they used due diligence with reference to his selection, whether the statute does nor not impose such a burden.'

The language of the court in that case is, we think, applicable to the present case. The acts of Evers and his testimony are such as to raise a strong presumption of his incompetency. The steamship company has introduced no testimony whatever, either to show that he was competent or to indicate that at the time of changing its vessels from coal burners to oil burners, or at any time, it made any inquiry as to his knowledge of the properties of the dangerous agency they were about to introduce upon their vessels, or his fitness to handle it. The limitation of its liability must be denied.

The decree is reversed as to the appellants, and the cause is remanded for further proceedings not inconsistent with the foregoing opinion.