Full opinion text
WILLIAM C. COLEMAN, District Judge. The question in this case is as to the liability for an explosion which occurred on the French barque Richelieu while lying at the Baltimore & Ohio Railroad Company’s coal piers, Curtis Bay, Baltimore, on the late afternoon of January 4,1927, while she was being loaded with a cargo of pitch. As a result of the explosion a fire was propagated throughout the ’tween-decks and lower hold of the vessel, resulting in the death of seven stevedores, engaged at the time in trimming the cargo, and in very severe injury to a number of other stevedores and members of the crew. The vessel herself was damaged to a great extent and sank at the pier. A large part of the eargo was destroyed or damaged. The vessel was being loaded by electrical trimmers owned and operated by the Baltimore & Ohio Railroad Company, which operation was supplemented by hand trimming performed by the stevedores, who were employed by and under the direct supervision and control of the railroad company. As a result of the disaster, the owners of the Richelieu libeled the railroad company for the damage to the vessel. Three of the vessel’s crew and eight stevedores, who were injured, and the representatives of four of the stevedores, who were killed, also brought libels against the railroad and the vessel and cargo as well. The railroad company filed a cross-libel against the vessel and cargo, at the same time impleading the F. J. Lewis Manufacturing Company, manufacturers and shippers of the cargo, as respondent, under the fifty-sixth admiralty rule, claiming that, whatever liability might exist for the disaster, the latter company, and not the railroad company, was responsible therefor. Lastly, the owners of the Greek steamship Emanuel Stavroudis intervened against the railroad company, the Lewis Company, and the ship and cargo, claiming reimbursement for salvage of the Emanuel Stavroudis, allowed by this court, and occasioned by the fact that this vessel was lying at the same pier in close proximity to the Richelieu, at the time of the disaster, and was towed to safety. See The Emanuel Stavroudis (D. C.) 23 F.(2d) 214. All of the suits (except that for salvage reimbursement) were consolidated and heard together. The interrogatories and answers thereto are many and multifarious. The trial covered 29 days. Testimony of some 100 witnesses was taken, covering over 6,000 pages. In addition, there are depositions of 38 French sailors and others, who were aboard the Richelieu at the time of the explosion, and many photographs, drawings, and other exhibits. A model of the pier and trimmers, electrically operated, showing the Richelieu in the course of being loaded, was exhibited in the courtroom during the trial. At the conclusion of the testimony, court and counsel visited the pier, where a demonstration was given by the railroad officials of one of the trimmers loading coal into an open barge. Certain tests were also made, as to arcing and sparking of one of the trimmers, with samples of coal and also of pitch taken from the Richelieu. The managing owner of the Richelieu at the time was the Soeiété des Armateurs Franjáis, the French republic having an interest in the vessel as a training ship of the Soeiété des Navires Écoles Fransaises. For this reason there were a number of cadets aboard. By the terms of the charter party, the vessel was under contract to carry a cargo of pitch, consisting of not less than 3,900 nor more than 4,400'tons, to Lorient, France, where it was to he used as a binder in the manufacture of briquettes. The manufacturers and sellers of the pitch, the F. J. Lewis Manufacturing Company, had a separate contract, with which the vessel had no connection, with the buyers, Robinson, Butler; Hemingway & Co., of New York, for Walter H. Brown & Co., London. Summarized, the basis of the claim on behalf of the vessel and the various other libelants is that the railroad company was negligent in the means which it employed and the manner in which it loaded the cargo, or, more specifically, that the railroad company (a) undertook as stevedores to load the pitch safely into the vessel; (b) that in the loading it reduced much of the pitch to dust, rendering it highly inflammable and explosive; (c) that the reducing of the pitch to dust was by means of electrical trimmers, which, by reason of their imperfect design and run-down cofidition, produced ares and sparks; (d) that either these arcs and sparks, or the open lights or torches which the railroad company’s stevedores used in the dust area, or both, were the igniting agency for the explosion. The railroad company’s reply is that all that was required of it was ordinary care in the handling of the cargo, and that it exercised such care; that, as a, result of this disaster, pitch is now known to be highly inflammable, much more so than coal, but that such was not previously known to be true; and that, therefore, conceding that both the trimmers and the open lamps are dangerous instrumentalities to use with pitch, and conceding, further, that the open lamps were the igniting agency in the explosion, nevertheless, the railroad company is without liability, because the danger was neither known to it at the time, nor to the lay or scientific mind; and therefore that, if there was any negligence in relation to the disaster, sneh negligence was that of the vessel, or of the shippers of the cargo, or both — of the former on the ground that the vessel was not properly constructed, fitted, and supervised by her master at the time of loading; and of the latter, because the pitch, when delivered, was inherently dangerous and vicious, and contrary to the railroad company's contract. The Lewis Company, on its part, denied that the product which it delivered was not pursuant- to contract, or that it was inherently vicious and dangerous, and that whatever liability might exist for the disaster was the liability of the railroad company because of improper handling of the cargo. When libelants and the railroad company had concluded their testimony, the court found that the evidence was legally insufficient, for reasons hereinafter fully discussed, to warrant a finding that the Lewis Company was in any way responsible for the disaster, and so dismissed that company. At the outset, a clear understanding of the exact nature of the cargo which was being loaded into the Richelieu at the time of the disaster is essential, as is also a clear understanding of the type of the vessel herself and the method employed in receiving and loading the cargo. First, as to the cargo: This was coal tar fuel pitch, which is a by-product of coal. Crude coal tar, the raw material from which pitch is made, is produced by the carbonization or destructive distillation of high volatile coal, either in by-product coke ovens or in gas retorts. It appears that about 80 per cent, of the coal tar made in the United States is produced in by-product coke ovens, the principal products of these ovens being metallurgical coke; the tar being "the nonacqueous portion of the vapors that are distilled off the coal. The crude tar thus obtained is then placed in large storage tanks, where the free water separates and is drawn off. From these storage tanks the tar is pumped to the. still, where the oils are condensed and taken off; the first of these being known as light oils, containing benzol and other solvents of that series. Next in the distillation comes the creosotes — that is, the coal tars — and, depending upon the grade of pitch desired, the distillation is prolonged or stopped, the amount of oil varying in proportion to the hardness of the pitch; that is to say, the harder the pitch, the greater the amount of oil that must be taken off the tar in the form of a distillate. When the proper grade of pitch is thus obtained, it is run into a cooling tank, and then into bays or vats, where it is seasoned. By this time the pitch has changed from a viscous liquid to a solid mass, and is ready to be broken up for shipment. For the manufacture of fuel briquettes, in which this particular pitch was to be used, the European process consists of grinding the pitch, then mixing it with pulverized coal in the proportion of approximately 93 per cent, coal and 7 per cent, pitch, and the mixture is then heated and by means of presses run out into briquette form. The briquetting industry abroad is very extensive, more so than in this country — a great-deal of such fuel being used in the larger sizes in Europe for railroad locomotive fuel; in the smaller ones for domestic fuel. The exportation of briquette pitch from the United States to Europe is small, relatively speaking, and such as has been exported has, to a very large extent, passed through the port of Pensacola, Fla. England supplies the bulk of the export business to the European continent. The Baltimore & Ohio Railroad Company's tariffs, and the tariffs of carriers generally, provide for this commodity under the generic term “pitch.” The tariff under which this particular cargo was handled provided that it would be accepted-only after prior arrangement which, as will be hereafter referred to, had reference, not to hazards in handling, but because it was desirable to give coal shipments preference-at the piers. Pitch was not then classified in any of the effective tariffs or bulletins issued by the carriers, or allied agencies, as being dangerous. The melting point of this particular pitch was approximately 180° Fahrenheit, and its content of volatile combustible matter approximately 69 per cent. Next, as to the type of vessel into which the cargo was being loaded: The Richelieu was a steel barque, built in 1919 in Hamburg, Germany, 323 feet long, 47 feet beam, 25 feet draught, with a gross tonnage of 3,-216 tons. Her lower hold was entirely open from stem to stem bulkheads, broken only by three webs, and by a temporary wooden bulkhead placed between No. 1 and No. 2 hatches, 'tween-decks. Between the main deck and the 'tween-decks was a clearance of approximately 7 feet, and from the 'tween-decks down a depth of approximately 20 feet. There were four hatches, varying in size from 8 feet 3 inches by 8 feet 5 inches to 24 feet 10 inches by 14 feet, which latter dimensions are those of No. 2 hatch, with which, as will hereinafter appear, we are primarily concerned. The vessel also had 23 trimming -hatches, 3 feet by 3 feet 6 inches, and 8 ventilators, 4 of which ventilated the lower hold, as well as the ’tween-decks, and. 5 manholes. Shifting boards were provided for this cargo in the ’tween-decks, but not in the lower hold. The vessel arrived with sand ballast, most of which had been removed at the time of the disaster. It was admitted that the general construction of the Richelieu was modem and of approved design for a vessel of her type. Steam vessels loading at respondent’s piers usually had some 16 trimming hatches, as against 23 in the Richelieu. Next, as to the method employed in receiving and loading the pitch by the Railroad Company: It seems desirable that a description of the entire operation should be given; that is, what happens to the pitch from the time it is received at the piers until it is put into the vessel. It arrives in carloads at the piers, and each car is dropped by gravity to the “Barney” pits, where the “Barney,” which is a small cable car operated by steam, pushes the loaded car onto a car dumper (of which there are two), which is an elevated platform or cradle, so hinged that it rotates on a horizontal axis. As this cradle turns over with the ear in it, the pitch runs out of the car into a pan, thence into hoppers from which the pitch is received on feeder belts, which, in turn, feed it to any one -of three conveyor belts. One of these conveyor belts leads to concrete storage bins, and the two others to the piers. These car dumpers have a capacity of forty 280,000-pound gross weight cars per hour. These conveyor belts are continuous from one end of the pier to the other; that is, they rim up an incline to the towers on the pier, back down the incline out to the end of the pier, and back to the point of starting. There are four movable loading towers, approximately 45 feet above the piers proper, two fed by each ear dumper. These towers may be moved lengthwise along the pier, and they support shuttles approximately 115 feet long, within which are located endless belts. These shuttles may be moved longitudinally and transversely in relation to the pier, and also may be raised or lowered. ' At the end of each shuttle is attached a mechanical trimmer, operated by a boom or hoist structure on top of the tower. The trimmer itself consists of a telescopic chute, supported and stayed laterally by pipes. At the lower end of the telescope is the trimmer proper, which consists of a curved elbow, designed to deliver the contents of the chute to a small belt, located at the extreme end, or mouth, of the trimmer. This entire terminal device may be rotated on a vertical axis 360° in either direction. The speed at which the coal or other cargo is run through the trimmer is regulated to 45 feet per second, or 2,700 feet per minute, which is approximately the same rate at which the same cargo is calculated to fall from the end of the shuttle belt when the operation known as “free dumping” is employed; that is, by use of the mechanism just described without the mechanical trimmer, an operation whereby the coal, or other commodity, is run over the belts onto the shuttle, and from' the end of the shuttle deposited by gravity into the vessel. In other words, the underlying principle embodied in the mechanical trimmer is not to increase the speed with which the commodity is loaded into the vessel, but simply to change the direction of its flow; that is, from a vertical to a horizontal direction, thereby depositing it, not merely in one place, but at any desired spot in the vessel’s hold, and to minimize the breakage of the coal, or other commodity, as much as possible, consistent with rapid operation. The entire operation at and upon the pier, with the exception of the dumping of the cars, is electrically driven, and all of the belts leading to any one tower are interlocked by a system of wiring, so that only one belt can be started at a time. Each of the towers is supplied with direct current, 550 volts. Each tower, considered as a unit, has 8 motors; two of these motors are installed inside of each trimmer, one of 50 horse power, the other 13 horse power; the purpose of the larger motor being to drive the belt, and of the smaller motor to swivel or revolve the trimmer. The larger motor is not of dust-proof design, nor are the resistors or other accessories; whereas the smaller motor is completely inclosed. The supply of electrical energy for both motors is transmitted by way of a pair of angle iron collector rings, mounted on the trimmer, and a pair of cast iron shoes located over the trimmer operator’s cab, these shoes revolving on the angle iron collector rings as the trimmer is swiveled; that is, they turn with the trimmer, and are always over the trimmer operator’s cab, and never over the mouth of the trimmer. The collector rings are not completely inclosed, but covered by a metal cone or canopy, open at the bottom and with doors on the side. The mechanical trimming just described is supplemented by hand trimming; that is, the stevedores are sent below to assist in the trimming of the cargo, by shoveling it about over spaces not adequately or easily served by the trimmers. To facilitate this operation, the stevedores are divided into gangs of eight men each, and each gang is supplied with four torches, or open flame lamps. These lamps consist of metal cans, 6 or 8 inches tall, 4 or 5 inches in diameter, with a neck in the top through which is run a wick providing an exposed flame. Kerosene oil is used. The lamps have wire handles, some 15 or 18 inches long, bent at the end, so that they may be hung up on any convenient protruding surface, or hooked into a crevice or hole. The explosion was described by most, if not all, of the eyewitnesses to it, as a dull, suppressed boom. Some said it was preceded, others that it was followed, by a flash of Are, and then came dense smoke, and hot, molten pitch. The exact part of the vessel where it occurred will be hereinafter discussed. At the time of the explosion, two gangs of eight men each were in the lower hold, and eight men were in the ’tween-decks. The men in the lower hold had in use at the time four or more lamps, and probably no less than four lamps were burning in the ’tween-decks. There is conflict and uncertainty in .the testimony with respect to the exact number, and their location, but the foregoing represents the most reliable evidence. No. 3 trimmer was operating in No. 2 hold, and No. 4 trimmer in No. 3 hold. Approximately 3,500 tons of pitch had been loaded into the vessel over a period of some 10 days, the loading being frequently interrupted and the vessel removed to another berth in order to accommodate other vessels, thus leaving approximately 1,000 tons still to be placed in the vessel. It was not until the day of the explosion that hand trimming was necessary. The filling of the lower hold had just then been sufficiently completed, and it was not until about one hour and a half before the explosion that the hand trimmers for the first time went to work, and they worked concurrently with the electric trimmers now loading into the ’tween-decks. The same two trimmers had, however, been working simultaneously in the same hatches the previous day. The great preponderance of evidence and all reasonable presumptions are to the effect that the explosion originated in No. 2 ’tween-deck. This is conceded by both sides to be true. The uneontradicted evidence of diver Bums is alone sufficient to establish this fact, because of the character and extent of the damage to vessel and cargo which he found at that point. Also it, is accepted and the court finds, that what occurred was a dust explosion. Further, there is no claim by any one that the proximate cause of the disaster could have been other than the trimmers, or the lamps, or both, which were in use at the time at or near that point. That is, all other theories, such as spontaneous combustion, or a fire whose origin was totally disconnected with either of the above-named agencies, are discarded. Beyond this, however, there is no concession in the ease. In fact, there is a great deal of conflicting testimony, even among witnesses on the same side, as to the origin of the explosion. There are, therefore, two questions to be determined: First, the exact source of the ignition, that is, whether the trimmers, or the lamps, or both; and, second, whether there was any negligence on the part of the railroad company or the vessel with respect to the creation of the explosion and the propagation of the resultant Are. The ease is narrowed to these considerations because the impleaded respondent, the shipper of the cargo, has been dismissed for reasons which it is proper, at this point, to state. It appeared that the agreement which the railroad company made with the Lewis Company was that the forme* would transport the pitch and load it into the vessel under two conditions: First, if it was like the sample submitted; and, second, if it would load like coal. For the purposes of the present ease, we may treat the ordinary common carrier function, namely, the transporting and the loading or trimming function,'as one and the same. Both were performed under tariffs duly filed as required by law. Assuming that the understanding was that both of the aforesaid conditions should be met respecting the entire operation of carriage and also of loading through the trimmers into the vessel, there is still no affirmative evidence that this pitch was not like the sample, as understood and adopted at the time, because, although the volatile content was admittedly somewhat higher than the sample, there was no understanding that identity of chemical analysis was a prerequisite. The railroad company admitted that, in submitting the sample, there was in mind no thought of hazard or inherent danger, but simply facility in loading and the avoidance of delays and possible damage to the trimmers by clogging along tbe conveyor belts and otherwise, as had been experienced on a previous occasion. It did, however, appear affirmatively that some of the pitch would not load like coal; .that is, that it would not run properly through the trimmers, because it was congealed and too bulky. However, this failure to come up to the second specification of the warranty was waived by the conduct of the railroad company in accepting the Lewis Company’s offer to bear the extra expense of breaking up the pitch, so that it would load properly, and in loading the greater portion of it, which operation extended over a period of some 10 days, without further complaint. It is admitted that, even though the railroad company may be said to have waived the second prerequisite of performance of the contract with the Lewis Company — that is, the requirement with respect to the facility with which the pitch could be loaded — it does not, of course, follow that, if the pitch was still latently vicious and dangerous, such waiver could be said to amount also to a waiver of any claim, for damage which the railroad company, whether we now consider it as a common carrier or as a stevedore (that is, as a mere bailee), might suffer by reason of such unknown and unforeseeable characteristics. That is to say, irrespective of the provisions in the bill of lading that the shipper of dangerous goods, who has not fully disclosed to the carrier their nature, shall be liable for and indemnify the carrier against all loss and damage caused by such goods, we might have a state of facts where, either at common law or in admiralty, such shipper would be liable. But here the burden of proof has rested upon the railroad, it being in the position of plaintiff with respect to the Lewis Company, to show that such inherently vicious and dangerous characteristics existed above and beyond the extent to which they existed in the sample, according to which sample the agreement to carry and load was made, and also above and beyond any'reasonable construction of the agreement, to be derived from all the surrounding circumstances. The further fact is important that the railroad company assumed by agreement entire charge and control of the loading. The Lewis Company exercised no supervision over it. The court has dismissed the Lewis Company, because it has found that the railroad company has failed to sustain this burden o'f proof. Its three principal witnesses, Mr. Church, and Drs. Patrick and Adams, all testified that, from the safety point of view, they could not distinguish between the sample and the pitch actually shipped; that they were both merchantable pitches; that it was dangerous to use either with open lights or with electrical trimmers, but that, whereas the 10 per cent, greater volatile combustible matter in the Lewis pitch than in the Barrett, or sample, pitch indicated that the former was the more inflammable, the exact degree of greater inflammability, and therefore the greater hazard, they also stated was impossible of determination, so far as their investigations and those of any one else had gone. Furthermore, these same witnesses declared it was not until this explosion took place that any one knew anything definitely about the combustible character of pitch; that prior thereto they all would have said that pitch dust was no more dangerous than coal dust, and that they would then have said, as they now say, that coal could have been loaded with complete safety under the conditions as they existed on the Richelieu. They also testified that most of the various bureaus constantly investigating such matters never knew that pitch dust was explosive prior to.this disaster, and 'therefore it would be improper to charge the railroad company with knowledge which no one else, not even these bureaus, possessed. With these admissions, the question with respect to the Lewis Company, therefore, resolves itself into this: Was there any warrant for charging the Lewis Company with a greater degree of knowledge of the chemical analysis of the pitch, and its explosive qualities, than that degree to which the railroad restricts itself? Should such greater knowledge be demanded, because of the fact that the Lewis Company is not merely the shipper, but also the manufacturer, of the pitch? In other words, was it fair to say, because of this latter fact, that the Lewis Company must be presumed to have a more intimate knowledge of the product, because it made it, and that, therefore, it puts its product on the market at its own risk, regardless of the uses to which it may be put; that is, that it must be presumed to know how such a product will act under all conditions, even though some of those conditions, as, for example, loading in the presence of igniting agencies, are not conditions which the manufacturer subjects' it to, or should be expected to subject it to? If pitch, instead of being generally known under the generic term and put to a great variety of uses, had heretofore been unknown to commerce — that is, had it been an article entirely new to the Baltimore & Ohio Railroad — there might be some force in.this argument; but the facts are that the railroad company, without attempting to differentiate between grades, in so far as chemical analysis or combustibility was concerned, had moved pitch over its rails, had loaded it through its trimmers three times previously, and had recognized it as a proper subject for carriage and loading into vessels by express provisions in its tariffs, and had charged and. still does charge a higher rate for this service. It is admitted that the tariff proviso to the effect that pitch would only be accepted subject to prior arrangement had nothing to do with hazard in handling, but was inserted for the greater accommodation of the carrier in handling coal shipments, to which it desired to afford the greatest expedition at its piers. Also there is evidence that pitch had been loaded into vessels, infrequently, to be sure, at one or more other points on the Eastern seaboard. Further, it is in evidence, and not contradicted, that nowhere is pitch of this or any other character classified in the tariffs or bulletins of common knowledge as being dangerous. The present situation is governed by the well-established principle that one who delivers goods to a carrier for shipment, whether he be the manufacturer thereof or not, is obligated, if the goods are known to him to be of an inherently dangerous character, or if he, tested by the standard of the average prudent man, ought to have had such knowledge, to warn the carrier of such inherent danger, unless the carrier itself knew, or might, by the exercise of ordinary care, have known, of the same. The fact that the shipper is also the manufacturer of the goods does not, in and of itself, impose a peculiar and greater obligation upon him, but he is governed by the principle just stated. Brass v. Maitland, 6 El. & Bl. 470; Acatos v. Burns, 3 Exch. Div. 282; Nitroglycerine Case, 15 Wall. 524, 21 L. Ed. 206; Bamfield v. Coole & Sheffield Transportation Co., Ltd., [1910] 2 K. B. 94; International Mercantile Marine Co. v. Fels-Naphtha Soap Co. (C. C. A.) 170 F. 275, 18 Ann. Cas. 18. Under the facts in the present case, the court finds no proof that the Lewis Company possessed any peculiar knowledge of the pitch which the railroad company did not possess, or that it was its-duty to have acquired and to have imparted such knowledge. At the time the Lewis Company, for the aforegoing reasons, was released, the separate cross-libel of the railroad company •against the Richelieu and the Lewis Company was also dismissed as to the latter, because what the railroad company thereby claimed was damages on account of stevedoring services rendered and supplies furnished, injuries to its pier property, claims made and suits filed, or that might be filed, for the death or injury of any of the stevedores or other persons; that is to say, the cross-libel, in so far as it affected the Lewis Company, involved questions which have either been answered by the dismissal of that company, or which axe of a nonmaritime character, and, therefore, non-cognizable in this proceeding. Turning now to the question as to what the evidence shows as to the source of the ignition, we may properly divide the testimony into two classes: First, that of fact witnesses; and, second, that of expert witnesses. The fact witnesses may be subdivided into (1) eyewitnesses and (2) noneyewitnesses. The eyewitnesses include at least six French sailors, the preponderance, if not the entire weight, of whose evidence attributes the source of ignition to, the trimmers. But upon analysis of their testimony it appears that they base their conclusions as much, if not more, upon what they heard in the way of cracking or sparking noises, as upon what they actually saw, or at least what they saw which can be accepted as actually the initial stage of the explosion or ignition. It is further to be noted that none of these members of the vessel’s crew were engaged in the loading operation in any manner whatsoever, nor were any of them below the main deck, or in as close proximity as were other eyewitnesses to hatch No. 2, where, as heretofore explained, there is a general admission that the ignition originated. The other eyewitnesses —that is, eyewitnesses at least to some early, if not the earliest, phase of the ignition— were 24 stevedores, 7 ealled to testify on behalf of the ship, the remainder on behalf of the railroad company; the operator of No. 4 trimmer, which was working in No. 3 hatch; and also the 2 operators in Nos. 3 and 4 towers. To this list should be added the name of Cantler, the operator of No. 3 trimmer working in No. 2 hatch by reason of his declaration, admitted in evidence because made the evening after the disaster in the hospital, and shortly before he died of the burns which he received. This declaration was made to the priest who attended him, was very brief and simply to the effect, ■in response to a question as to what caused the disaster, that “it was electricity.” , It is not easy to summarize the testimony of the stevedores. Many of them had been through, a frightful ordeal and still bore, and •will always bear, the scars and disfigurements of their terrible bums. They were colored men, for the most part extremely illiterate, but also, it is only fair to say, for the most part gave the appearance of trying to be truthful, and to give as best they could an accurate story of what they witnessed. Unquestionably the preponderance of their evidence is to the effect that the original explosion, whatever its direct cause, occurred in the ’tween-decks, just forward of No. 2 hatch. The court finds that the preponderance of their evidence also clearly indicates that the fire did not originate on the trimmer, but in the ’tween-decks, just forward of No. 2 hatch. This is home out by a number of things. First, those stevedores who were in the ’tween-decks between Nos. 2 and 3 hatches uniformly testified that the effect of the explosion upon them was very violent, more violent than stevedores in other locations testified its effect was upon them. Several of the first named were knocked down and their clothes took fire. It is trae a number of them testified that they saw the trimmer in No. 2 hatch ablaze, but the most reasonable inference from this part of their testimony, taken as a whole, is that they saw this after the first shock or explosion had occurred. Third, those stevedores who were in the lower hold, for the most part, at least, testified that what they first heard or saw was above them, and that the first flame which they saw came down through one of the manholes, which tends to controvert the theory that the fire originated on the trimmer itself in No. 2 hatch, because it is only reasonable to assume that, if the latter were true, the first onset of the flame visible to these men would have engulfed that hatch, or at least been most intensified "just at the mouth of that hatch. Admitting that it is difficult, if not impossible, to reconcile the many conflicting statements of the various stevedores as to just what they saw and heard, the court does not feel that it is fair to conclude from this that their testimony is to be given little or no weight, and that the testimony of the French sailors is to be accepted as clear proof of the source of the ignition. It is true the testimony of the latter is borne out by the near-death declaration of the operator of No. 3 trimmer, but his brief statement cannot be accepted as conclusive, in view of the circumstances under which it was made, and in view of the testimony of Price, the operator of No. 3 tower, who testified emphatically that No. 3 trimmer was not swiveling at the time of the first explosion. Additional weight is to be given to this testimony by reason of the at least partial corroboration of it by Mr. Hansell, the electrical foreman of the railroad company, who testified that the control lever for the swivel motor in No. 3 cab was shut off when examined by him after the disaster. Of course, this latter statement also cannot be taken as conclusive proof that No. 3 trimmer was not swiveling at the moment of the explosion, but simply that it had been stopped at some time. It is a fair presumption that it had been stopped before the first explosion, although it may also be a fair inference that Cantler, although fatally burned, stopped the swiveling before leaving the cab in an endeavor to save his own life. The testimony of the foreman and assistant foreman of stevedores is not helpful, because the former was on the castle deck, near No. 4 hatch, and the latter at the coaming of the same hatch, when they first heard the explosion; that is, neither of them was in a position to see the origin of the first ignition. For similar reasons, the testimony of the operator of No. 4 tower and of No. 4 trimmer is not helpful, nor is the testimony of the pier superintendent, Mr. Haver, on this point. The testimony of diver Bums which, standing by itself, would seem to be adequate proof of the fact, now admitted, that the fire started in the ’tween-decks, near No. 2 hatch, does not assist in proving the precise point now under consideration, namely, whether at that location the ignition was caused by the trimmer or by the open lamps. While the claim is made that the trimmers as a whole, and trimmer No. 3 in particular, were in a state of such bad repair and so unfit for this particular use that arcs and sparks emanated from points other than at the collector rings, the part of the trimmer towards which the ship’s argument is most seriously directed is the collector rings, through which the current was carried to the motors. There is no contradiction of the fact that this trimmer was running after the explosion, as was also trimmer No. 4 in No. 3 hatch. Add to this the inference, which is reasonable, because of the preponderance of evidence, that the trimmer was not swiveling at the time, it follows that the shoes on the collector rings were not moving, and, if they were not, it is not conceivable, from any construction of the testimony of the various electricians, that, except under the most extraordinary conditions, the collector rings would have arced or sparked sufficiently to cause a combustion. ; But let us seek further evidence of the exact source of the ignition from the testimony of the second class of witnesses, namely, the experts. The experts for the vessel, Drs. Whitehead, Penniman, Kouwenhoven,' and Porter, and Messrs. Weiss, Brueekman, and Whitson, all highly qualified technicians, uniformly agreed that the arching and sparking from the collector rings was the original source of ignition, although Dr. Kouwenhoven stated that he had earlier believed that the lamps, and not the trimmer caused the ignition. The experts for the railroad company, equally eminent, expressed different opinions. Dr. Patrick stated that the lamps were the cause. Dr. Adams attributed the ignition to a frictional spark from the trimmer belt. To these opinions may be added the opinions of Messrs. Davis, Hansell, and Haver, officials of the railroad company, the first two of whom attributed the ignition to a foreign substance or piece of pitch issuing from the trimmer and hitting a stanchion in the vessel, thereby creating a spark; while Mr. Haver was of the opinion that the oil content of the pitch, mixed with the great accumulation of dust, caused combustion; but this theory is obviously not convincing, because it omits all mention of an igniting agency. Mr. Church, consulting chemist, who also testified for the railroad company, declined to advance any theory as to the source of ignition. If we are to apply to the testimony of the experts a quantitative analysis, then we are forced to the conclusion that the contention of the libelants is sound that the source of ignition was the trimmer, just as we would be forced to the same conclusion if merely a quantitative test were applied to the testimony of all of the eyewitnesses. But obviously the weight which is to be given to evidence is not governed merely by mathematical computation, but by its soundness; that is to say, by its true probative value, viewed in the light of all the surrounding circumstances. Credibility does not depend on numbers. If it did, decisions would turn upon a quantitative and impersonal test applied to the testimony. Thus, there is no mere rule of thumb which says that positive evidence of witnesses, who hear a sound or see an object, is entitled to greater weight than the negative evidence of those who failed to see or hear the same. It may be in certain cases. See The Charlotte (D. C.) 124 F. 989. In others, positive testimony, even though totally uncontradieted, should not'eontrol. The rule admits of many exceptions. See Quock Ting v. United States, 140 U. S. 417, 11 S. Ct. 733, 851, 35 L. Ed. 501. All of the expert witnesses agreed on the general proposition that a dust explosion is nothing more or less than the result of the mixture of dust with air, in such proportions as to cause inflammability or combustion, when brought into contact with sufficient igniting source. Purther, they all agreed that the principal conditions influencing the explosion of combustible dust are nature of the material from which the dust comes; density of the dust, that is, its weight per unit volume of air; its fineness; the ignition, which includes temperature, size, and duration of the igniting source; nature of the atmosphere surrounding the dust particles; confinement of the dust; and, lastly, turbulence. The experts on behalf of libelants stressed the temperature of the igniting source as being the important condition in determining combustibility, and laid little emphasis upon the size and duration of the igniting source. However, Dr. Patrick, on behalf of the railroad company, took a contrary view, which the court believes to be the soundest, and which inevitably leads to the conclusion that the lamps, and not the trimmers’ ares or sparks, or a friction spark, were the source of ignition. Dr. Patrick explained the importance of the size and the duration of the igniting source by the illustration of a mixture of gasoline and air, which forms the explosive mixture for the common internal combustion engine. It is a fact that there may be an electric are or spark in such mixture so small that the mixture will not explode, and yet the temperature of this are or spark is the same as that of any other electric are or spark; the fact that it does not explode the mixture of gasoline and air being due primarily to lack of a sufficient number of particles simultaneously heated. Dr. Patrick then explained that the nature of the .atmosphere surrounding the dust particles was, in his opinion, the most important -factor of all of those which influenced the explosion. He stated that, according to his theory, volatile matter, coming off the pitch, formed an atmosphere of inflammable vapor around the dust particle, changing that particle from a solid, almost noninflammable, particle, to an almost gaseous particle, and that, in the case of coal, such an atmosphere or “shell” could not, under ordinary conditions, exist — -never under conditions which could be found in the loading of any vessel. He explained that coal, chemically, is essentially a different substance from pitch; that it has a very high molecular weight, with no tendency to give off vapors; that-it contains only a trace of hydrocarbon, while pitch is essentially composed of hydrocarbons and fixed carbons. In other words, coal may be compared to a piece of wood, while pitch is like camphor or naphtha in moth balls. These latter substances are known to give off vapors, and they disappear, although they are solids, while wood never gives off any vapor. The gas liberated from coal, namely, methane or fire damp (CH4), light carbureted hydrogen, the product of decomposition of organic matter in mines, is admittedly entirely different and in and of itself a serious hazard in mines, and the one commonly referred to in connection with safety measures to be taken. This inflammable gas is liberated chiefly while the coal is being mined, and its escape diminishes constantly thereafter, and in the present ease there is no proof, if, indeed, any serious contention, that any of the coal while loaded over these trimmers gave off any of this gas of-appreciable amount or danger. While the evidence is somewhat obscure and contradictory as to how many lighted lamps there were in use at the time, and particularly how many there were in the ’tween-decks, we have the testimony of the captain that he saw four lighted lamps at No. 3 hatch, ’tween-decks. Certainly the preponderance of evidence is that there were several lamps there. It further appears that the direction and velocity of the wind (about 22 miles per hour) and the position of the vessel were such as to blow the dust generated from the pitch, as it was thrown out of the mouth of the trimmer, forward of No. 2 hatch. There was a temporary bulkhead between No. 1 and No. 2 hatches, and No. 1 hatch was covered, all of which impeded the circulation of air and created an excess accumulation and confinement of pitch dust within a limited area. The confinement was further intensified by the presence of water tanks forward of No. 2 hatch. It is not controverted that this area was the most confined space in which the dust had an opportunity to accumulate. Thus, since there were one or more open lamps burning in this space at the time, the flame or flames of which were obviously of much greater duration than any arc or spark which was generated from the trimmer, and since the size of this flame or flames must also have been larger than any such arc or spark, it necessarily follows that, if these factors are more vital than temperature, we are forced to but one conclusion, namely, that the conditions" immediately surrounding any one or more of these lamps were far more conducive to a combustion than those surrounding any part of the trimmer. We need not concern ourselves with the precise number of lighted lamps. One under such conditions would be sufficient. This conclusion is strengthened by the further uneontradieted faet, heretofore referred to, that at no time prior to about one hour and a half preceding the explosion had any of the trimmers been in operation simultaneously with the use of the open lamps. If the lamps were not the igniting agency, why, the railroad company asks, did not the catastrophe happen the day before, when both trimmers were operating in the lower hold of the Richelieu, without the concurrent use of the open lights; or why did it not occur early in the afternoon of the day of the explosion, prior to the time when the lamps were first put into use simultaneously with the use of the trimmer, at which prior time, according to the testimony of the French sailors, they saw sparks and flashes, and heard cracking and other electrical noises at the trimmers? A possible answer is that the volume of dust in the vessel at such times had not reached that density, which is unknown and can merely be estimated, necessary to cause it to explode. So the faet that no explosion occurred at either of those times is not conclusive that the lamps were the source of the ignition, but it is not an unreasonable basis for such an inference. Having determined that the open lamps were the source of the ignition which caused the original explosion, it may still be asked whether, assuming this to be true, the arcs and sparks emanating from the trimmer in No. 2 hatch may not have been a contributing cause to the propagation of the fire which resulted from this initial explosion. This question is important as bearing upon the question of whether or not the railroad company was negligent in loading the Richelieu, to which question we may now turn, because, although it has been found as a matter of fact that the open lamps and not the trimmer started the explosion, the railroad company would be liable in either ease, and also in the event that either one of the mediums did not actually cause, but propagated, the fire, provided it appears that the railroad company was wanting in the exercise of due care in the use of either of these mediums, as pail; of the operation of loading'the vessel. [5, 6] "We must first direct ourselves to the question whether it was negligence per se on the part of the railroad company to use at all (1) the open lamps, and (2) the electric cal trimmers, separately or concurrently. The railroad company was not an insurer with respect to its stevedores, the vessel, or her crew or cargo. Thus, before recovery can be had by any of these parties, the burden must- be sustained of establishing negligence on the part of the railroad company with respect to any such party. What is negligence? It is defined as the omission to do something which a reasonable man, guided by those considerations which ordinarily regulate the conduct of human affairs, would do, or the doing of something which such reasonable man would not do. The Nitrogyleerine Case, 15 Wall. 524, 21 L. Ed. 206. That is, the test is reasonable care under given circumstances. We have, therefore, to determine what constituted reasonable care on the part of the railroad company, under the conditions as they existed on the Richelieu, with respect, first, to the railroad company’s own stevedores; and, second, with respect to the vessel herself, her crew and cargo. First, with respect to the stevedores, it was incumbent upon the railroad company to "afford them, throughout the given employment, a safe.place in which, and safe appliances with which, to work. Longshoremen or employees of stevedoring companies are within the definition of seamen, and entitled to the protection of the Jones Act (USCA tit. 46, § 688; International Stevedoring Co. v. Haverty, 272 U. S. 50, 47 S. Ct. 19, 71 L. Ed. 157), whereby'the common-law defenses of contributory negligence and the fellow-servant doctrine are taken away, if the optional remedy of an action at law, given by this act, is invoked. We may still have a situation, however, where stevedores have, assumed the risk incident to their work. But where they undertake such work at the direction of their superiors, as was done in this case, stevedores are held not to assume such risk (Shipping Board v. O’Shea, 55 App. D. C. 300, 5 F.[2d] 123), unless the risk is so obvious and dangerous that no man of prudence would have obeyed his superiors, which is not here claimed (Coast S. S. Co. v. Brady [C. C. A.] 8 F.[2d] 16). Therefore the relation between the railroad company and its stevedores was simply that of-master and servant. The master does not guarantee the safety of his servant while engaged in the discharge of his duties, for he is not an insurer. He is not bound to furnish him a place and implements that are absolutely safe, but merely those that are reasonably so; and this is true, no matter how hazardous the business may be in which the servant is engaged. Employers are liable for the consequences, not of all danger, but of only such danger as is the direet result of their negligence. The standard of reasonable care is the conduct of the average prudent man. The employer is not bound to furnish the best known, the most modem or approved, or the safest machinery or appliances, but only such as are reasonably fit and safe for the work in hand. It is culpable negligence which makes the employer liable, not a mere error of judgment. Thus summarized, the servant, in order to establish his master’s liability to him, must prove that at the time of the injury (1) the place or the implements which the master furnished were not reasonably safe for the work to "be performed; (2) that the master knew this, or by the exercise of reasonable care could have known it; and (3) that the failure to afford a reasonably safe place or implements was the proximate cause of the injury. Patton v. Texas & Pacific Ry. Co., 179 U. S. 658, 21 S. Ct. 275, 45 L. Ed. 361; South Baltimore Car Works v. Schaefer, 96 Md. 88, 53 A. 665, 94 Am. St. Rep. 560; Dulligan v. Barber Asphalt Paving Co., 201 Mass. 227, 87 N. E. 567; West v. Brevard Tanning Co., 154 N. C. 44, 69 S. E. 687; Bertha Zinc Co. v. Martin, 93 Va. 791, 22 S. E. 869, 70 L. R. A. 999. In the Patton Case, supra, the principle is succinctly stated thus (pages 664, 665 [21 S. Ct. 275]):, “ * * * While the employer is bound to provide a safe place and safe machinery in which and with which the employe is to work, and while this is a positive duty resting upon him, and one which he may not avoid by turning it over to some emplóyé, it is also true that there is no guaranty by the employer that place and machinery shall be absolutely safe. Hough v. Railway Co., 10 Otto [100 U. S.] 213, 218 [25 L. Ed. 612]; Baltimore & Ohio Railroad v. Baugh, 149 U. S. 368, 386 [13 S. Ct. 914, 37 L. Ed. 772]; Baltimore & Potomac Railroad v. Mackey, 157 U. S. 72, 87 [15 S. Ct. 491, 39 L. Ed. 624]; Texas & Pacific Railway v. Archibald, 170 U. S. 665, 669 [18 S. Ct. 777, 42 L. Ed. 1188]. He is bound to take reasonable care and make reasonable effort, and, the greater the risk which attends the work to be done and the machinery to be used, the more imperative is the obligation resting upon him. Reasonable care becomes, then, a demand of higher supremacy, and yet in all cases it is a question of the reasonableness of the care — reasonableness depending upon the danger attending the place or the machinery. ‘ ‘ The rale in respect to machinery, which is the same as that in respect to place, was thus accurately stated by Mr. Justice Lamar, for this court, in Washington & Georgetown Railroad v. McDade, 135 U. S. 554, 570 [10 S. Ct. 1044, 34 L. Ed. 235]: ‘Neither individ, uals nor corporations are bound, as employers, to insure the absolute safety of machinery or mechanical appliances whieh they provide for the use of their employes. Nor are they bound to> supply the best and safest or newest of those appliances for the purpose of securing the safety of those who are thus employed. They are, however, bound to use all reasonable care and prudence for the safety of those in their service, by providing them with machinery reasonably safe and suitable for the use of the latter. If the employer or master fails in this duty of precaution and care, he is responsible for any injury whieh may happen through a defect of machinery whieh was, or ought to have been, known to him, and was unknown to the employe or servant." Second, with respect to the duty owing by the Railroad Company to the vessel, her crew and cargo, it may be defined in its relation to all-three as the duty to use the same degree of care, whieh we have just defined, as that imposed with respect to the stevedores. There was no bailment of the vessel to the railroad company, and therefore no presumption of negligence. Does the master of the vessel share in this responsibility? Not under the existing circumstances. It is true that a duty rests upon the master to see that his vessel is seaworthy, and consequently he must see that her cargo is so stowed and trimmed that her seaworthiness will not be affected. Stevedores cannot be expected to know the trim of a vessel, as does the master; therefore the master has the primary responsibility of deciding how much cargo is^to be loaded and where. But this is not tantamount to saying that stevedores are relieved of responsibility for the proper handling and loading of the cargo into spaces thus designated, any more than they would be relieved of responsibility for damage to the cargo, due to their rough handling or their inadequate equipment. In short, it is well established that an independent contracting stevedore/-sueh as the Baltimore & Ohio -Railroad Company in the present ease, is responsible for the selection and proper use of the labor' and equipment whieh it employs, even when that equipment is supplied by the vessel itself, and so the officers of the vessel owe it no duty of supervision. Atlantic Transport Co. v. State of Maryland (C. C. A.) 259 F. 23; The Satilla (C. C. A.) 235 F. 58; The Persian Monarch (C. C. A.) 55 F. 333; Jeffries v. De Hart (C. C. A.) 102 F. 765. That is to say, the latter had the right to assume, not only that the trimmers were in proper condition and properly operated, but were a safe device for the use for whieh they were employed, and they had the further right to make a similar assumption with respect to the open lights or torches. It appears that the master of the Richelieu had no actual knowledge of the use of these lights in the tween-decks until a few moments before the explosion. But, even if he had had earlier information, it does not follow that he was obligated to protest against them, when their use was with. the apparent sanction and approval of those whom he had a right to assume were fully competent in what they were doing. A considerable portion of the railroad company’s testimony was devoted to an attempt to establish a custom for doing what it did, both with respect to the open lights and with respect to the trimmers. Custom, if not a negligent one, may often justify a given course of conduct; but no custom will justify an unreasonable practice. “What usually is done may be evidence of what ought to be done, but what ought to be done is fixed by a standard of reasonable prudence, whether it usually is complied with or not.” Texas & Pacific R. R. v. Behymer, 189 U. S. 468, 23 S. Ct. 622, 47 L. Ed. 905; The Charles Rohde, 1925 A. M. C. 1594, 8 F.(2d) 506. What does the evidence show, first, with respect to the use of open lights ? It appears that they have been in use at the port of Baltimore in connection with the loading of coal for upwards of 40 years, and that on this one pier of the Baltimore & Ohio Railroad Company 15,000,000 tons of coal have been handled. On-the other hand, there had been but 15,000 tons of piteh loaded at the same pier prior to this disaster, and this tonnage was loaded into three ships, in two of which open lamps were not used, but were used in the third. With respect to the practice at other places, it is in evidence that, at ten points along the Atlantic and Gulf coasts, it has been the practice for years, and still is, to use open flame lamps, candles, or other unprotected lights in the holds of all vessels wherein artificial lighting is necessary in connéction with, the loading and trimming of coal. It also appears, however, that at none of these ports, except at Newport News, Va., and Pensacola, Fla., were any bulk cargoes of piteh loaded or trimmed prior to the Richelieu disaster. The court excluded testimony offered by the fact witnesses from the various ports to prove that there had been no explosion, because there was a failure to show conditions sufficiently similar to those existing in the loading of the Richelieu. With respect to the use of electric trimmers, it appears that, at the time the Baltimore & Ohio Railroad Company trimmers were constructed, there was only one other trimmer of such general design in use, namely, by the Pennsylvania Railroad at Greenwich, Pa. Later, however, trimmers of the same general design, one or more of them with motor or motors located in the end of the mechanism, were put into operation by other carriers or coal companies at Baltimore, Norfolk, and Philadelphia; but there is evidence of only one or two instances in which piteh was loaded through these other trimmers, they being used primarily, and in most instances exclusively, for coal. The evidence is to the effect that, at these other ports referred to, it was common practice to use the open lights and trimmers concurrently in loading and trimming coal; but there is no evidence that at such ports there was ever such concurrent use with respect to piteh. It necessarily follows, from the evidence above summarized, that while there is proof of an established custom to use open lights and trimmers, and to use them concurrently in the loading and trimming of coal generally at Eastern seaboard and Gulf ports, there is no such proof with respect to pitch. Therefore such evidence as we have is at most merely persuasive of the reasonableness of a similar practice with respect to pitch. Whether it is in fact reasonable— that is, whether it is a reasonably prudent method to adopt — depends upon (a) what the Baltimore & Ohio Railroad Company actually knew about the inflammable or explosive tendencies of pitch dust at the time; (b) what it might have known from its own experience with coal dust; (e) what it might have known from the knowledge of pitch dust which the piteh manufacturers had; (d) what it might have known from such scientific or expert knowledge as was available; and, lastly, (e) the extent to which the railroad company was obligated to inquire, and thereby add to its own knowledge. We must therefore now take up these latter considerations in the order above named, and first concern ourselves with what the Baltimore & Ohio Railroad Company actually knew about the explosibility of pitch dust at the time it undertook to load the Richelieu. We have already seen that it had previously loaded only three cargoes of bulk piteh, in two of which instances trimmers were used, but without the open lights; in the third instance, the loading of the Mayaro, both the trimmers and open lamps were used, and used simultaneously. It is true that the railroad company’s officials admit having had trouble with flashes at the collector rings, due to accumulation of pitch dust, when loading the Conde Abasólo the previous December; also one of the stevedores testified, and it was not denied, that his foreman, on the afternoon of the disaster, shortly before the gangs went below to work, instructed them to be careful with the lights, saying, ‘.‘Remember this stuff is dangerous.’’ But, in the absence of further explanation of this statement, it seems as reasonable to presume that what was in the foreman’s mind was merely the burning qualities of pitch as a solid or m mass, as to presume that he was thinking of pitch dust — in fact, more so, because he probably had the common knowledge or recollection that piteh or asphalt would bum easily in mass, but no previous knowledge, so far as is disclosed, that pitch dust, as dust, would take fire