Plaintiffs and appellants here are the widow and son of Leo H. Hubbert, who, as well as one Cerezo, died of burns consequent on an explosion on the premises of defendant and respondent, Aztec Brewing Company, in San Diego, on August 8, 1935. Respondent is engaged in the business of manufacturing and brewing beer. Its plant includes a vat building in which is an extensive cellar known as cellar “G”, which is considerably longer from east to west than it is wide from north to south. In this cellar there are 29 reinforced concrete vats, approximately square in ground plan and about 12 feet by 12 feet each in size, arranged in rows, with aisles between the rows, and in some instances aisles also between rows of vats and the walls. Each vat has an opening, consisting of a manhole sufficiently large for a man to crawl through, at the side of the vat, about three feet above the floor of the adjacent aisle or corridor. There are no other openings into the vat except such as are required for the penetration of pipes and refrigerating equipment. The most northerly of these rows of vats consists, in order from west to east of vats 13 to 24, both inclusive. This row is broken only by a cross aisle or corridor between vats 18 and 19. There are, however, further aisles or corridors between this row of vats and the north wall of the cellar and also along the west wall of the cellar, and also between the row of vats just referred to and the next row of vats to the south, which is, going from west to east, made up of vats 12, 11, 10, 9, 8 and 7, and after the interruption of a widened continuation of the cross aisle above referred to, by further vats numbered 29, 28, 27, 26 and 25. Vat 13 is thus just north and across the intervening aisle or corridor from vat 12 and, similarly, vat 14 is north across the same corridor, and opposite vat 11, which in its turn is just east of vat 12. It is upon this last aisle or corridor that the manholes of the vats which line it open.

The method adopted for waterproofing vats was first to “prime” their interiors with asphalt, which was then to be “cut back” with some liquid or solvent applied to the walls, after which the walls were mopped with melted asphalt, and, lastly, an asphalt material known as “brewer’s pitch” was applied in sheets to the wall's, and the seams between the sheets welded together. A binder had to be spread upon these sheets of “brewer’s pitch” to enable them to be made fast to the previously applied coats of material. This binder was heated and prepared on a gas plate or stove fueled with bottled gas. Such a gas plate or stove was found in the aisle last referred to after the explosion. In some vats other water-proofing material was applied as a preliminary to the use of any asphalt.

Respondent, in April, 1935, contracted with one A. 0. Miller, doing business as the A. 0. Miller Water Proofing Company, to prime and mop-coat 22 of the vats, saying in its order to Miller: “We will furnish material for mop-coat and prime coat and charge you for solvent used in prime coat.” It was respondent’s purpose to itself do the work of applying the final coat of “brewer’s pitch” which is about y% inch thick and was prepared in its own laboratory by its own chemist. The work contracted for with Miller went on intermittently for some months after the arrangement with him, respondent carrying out the work of applying “brewer’s pitch” within vats within which the preliminary part of the water-proofing work had been from time to time completed by the Miller Company at a proper interval after the completion of such preliminary work on a given vat.

Leo M. Hubbert was employed by Miller to carry out the work which the latter had contracted to do, on the basis of a daily wage plus a share in the profits. Cerezo was another employee of Miller who worked with Hubbert. On the morning of the accident the two were in cellar “Gr” for the purpose of proceeding with the work. About the middle of the morning an explosion occurred in vat 14 of sufficient violence to break a hole about six feet in diameter in the concrete top of the vat, leaving the reinforcing steel exposed, to cause a “V” shaped bulge in the vat wall opposite the manhole and to cause a bulging and buckling in the walls common to vat 14 and vat 13 and to vat 14 and vat 15.

There is a good deal of discussion in the briefs about where the decedents Hubbert and Cerezo were at the time of the explosion. An employee of respondent brewery, one Falkenberg, testified that just before it occurred he had been work ing in vat 13 cleaning a door but left immediately prior to the explosion; that in the morning something like ten of the brewery’s employees were working inside of the vats, but he saw nobody there when he left except the deceased Hubbert, who was “out in the hallway just outside a tank” (vat), though he could not be sure just what he was then doing. He did not then see Cerezo. This, however, as we shall hereinafter notice, was by no means all that Falkenberg had to say. Another of respondent’s employees, one Moir, an assistant electrician, testified that just before the explosion he was going west in the aisle or corridor herein last referred to, which was lit up by ceiling lights, between the two rows of vats, one of which, as we saw, included vats 13 and 14 with others, and the other, as already stated, vats 12 and 11 with others. He was hunting for a ladder that was at the extreme west end of this aisle in order to use it in installing two temporary drop cords at the opposite or east end of the same aisle or corridor. He testified that, as he proceeded westerly along the aisle, he saw Hubbert leaning-on the manhole of vat 12 which was then open, and that he noticed that the inside of the vat was lighted. Moir laid his drop cords down on the floor, took a beer hose off the ladder, and was stooping down and about to pick up the drop cords again when the explosion occurred. Moir said that the fire, following the explosion, was east of where he then was and right in his face when he looked in that direction, a sheet of fire, in the aisle along which he had come. He tried to escape in some other direction. He first turned south in the corridor that ran along the west wall of the cellar but found some obstruction and so turned around and went on north along the west side of vat 13 to the northwest corner of the cellar. Thence he turned east in the corridor along the north wall with vats 13 to 18, both inclusive, to the south of him. Finding his path again obstructed he turned back west and met Hubbert, who said: “I am done for” and addressed Moir as “Shorty”, apparently mistaking him for Cerezo, who went by that name. Hubbert’s clothes were on fire and Moir helped him off with them, and, being cut off from escape, • the two remained in the corridor just north of vat 16 for about three-quarters of an hour, when the fire department arrived and cut a hole in the roof through which they got out. Moir’s hands and face were burned but, according to the fire chief, Hubbert was burned from head to foot.

Respondent’s master mechanic and plant superintendent, Jaeger, having been informed of the explosion, went to the vat building and into the cellar and found Cerezo in the center of a “hallway” about halfway between vats 13 and 19. These vats are some distance apart in the same row and there are two aisles, as already noticed, one on either side of this row of vats, that is, the aisle along the north wall of the cellar where, as we saw, Hubbert and Moir took refuge and the parallel aisle next south of it. The “hallway” referred to in Jaeger’s testimony was the more southerly one of these and the same on which, as we saw, the manholes of vats 12, 13, 14, etc., open. Jaeger says that from this aisle he and an employee of respondent named Boroff helped Cerezo out. Notwithstanding Cerezo’s burns, his face was largely free from them. Cerezo’s clothing was gone. Jaeger was under the impression that his shoes were gone too. One Herzog, the respondent’s assistant general manager, testified that a pair of shoes was found in the aisle “just about at the division line between tanks 11 and 12”. The witness Porter, an insurance man, who visited the cellar shortly after the accident, says: “About here at approximately the division between” (vats) “11 and 12 was a pair of shoes. Between the manhole 11 and up beyond the shoes possibly two or three feet were fragments of clothing. There were some fragments of dollar bills on either side of the corridor.” The shoes were identified by Cerezo’s widow as the work shoes of her husband. The evidence shows beyond substantial dispute that the explosion occurred in vat 14, though the resulting fire burned in the aisle or hallway adjacent to it on the south, and various other parts of the cellar suffered a greater or less degree of damage.

Appellants’ original theory was that Hubbert and Cerezo were working in or about vat 14 at or just before the time the explosion occurred. This theory would be unfavorable to their view of the case to the extent of tending to dispel the inference on which they rely that respondent brewing company was in the exclusive control of the agency that, whatever it may have been, caused the explosion, since, if Hubbert and his companions were working in vat 14 it might as readily be inferred that the explosion resulted from some thing done by them or one or the other of them. During the progress of the trial, however, on hearing Moir’s testimony, appellants’ counsel took the position that the evidence showed the vat in which the men who were killed were actually working not to have been vat 14 at all, but vat 12, and now contend that they had not, for weeks prior to the explosion, had anything to do with vat 14, from which counsel claim further that whatever agency caused the explosion, since the latter occurred in vat 14, such agency causing it must have been under the control of the brewery rather than of the Miller concern. Having sought and obtained leave to amend their complaint accordingly, they finally presented their case on this latter theory. A bill from the Miller concern to respondent dated May 15, 1935, and approved by respondent on May 20, 1935, was produced. This purported to be for priming and mopping tanks (vats), 5, 8, 10, 13, 14, 15, 19, 20, 21, 22, 23, and 24, but is inconclusive as to what particular ones of these vats were completed as to these priming and mopping coats since it specifies that 12 tanks (vats) were primed and mopped as to sides and bottom and 11 primed as to ceiling only. From this it might seem that on 11 of the vats no mopping coats had then been as yet applied to the ceiling and as to one vat neither prime nor mop coat applied to the ceiling. However that may have been, the evidence, according to appellants’ counsel, indicates that when the explosion occurred Hubbert was in the aisle opposite the manhole of vat 12, and Cerezo either in vat 12 or just on his way out of it, and that both were for the time cut off from escape to the east along the aisle that they were then in, by the fire to the east of them, which filled that aisle opposite the manhole of vat 14; that it was from this fire that they sustained their ultimately fatal injuries; but that after it had subsided Cerezo made his way east along the aisle where the fire had been to the point where he was found. We take it that their theory as to Hubbert’s course must be that he finally reached the north aisle of the cellar in the same way that Moir testified that he had reached it.

The record exhibits no direct testimony as to the cause of the explosion in vat 14. Either, therefore, appellants made out a case for the jury under the doctrine of res ipsa loquitur or else they failed to make out one at all, and the trial court was right, notwithstanding the jury’s verdict in their favor, in entering judgment for respondent. The transcript includes 944 pages of matter and it is manifestly impracticable in this opinion to recite the evidence in detail. The following, however, as wc understand appellants’ counsel, are, according to their construction of the testimony, the salient features in it, which, in addition to those which we have already mentioned, they contend render the doctrine of res ipsa loquitur applicable:

Hubbert and Cerezo were the only employees of the Miller concern in the cellar. All of the others who worked there were employed by respondent. There tvas nothing, according to appellants’ claim, in the materials used by Miller’s men that was inflammable. Miller testified that acetylene torches could not be used in the application of the priming and mop coats called for in his contract but that these coats could only be applied with brushes and mops. On the other hand, appellants direct our attention to evidence which, as they claim, shows that, “there were certain highly explosive materials which were owned and used by, and were under the exclusive management and control of defendant brewery company in the aisle in the immediate vicinity of vat 14’’. These are specified as bottled gas, acetylene, gasoline, solvent and kerosene, with none of which, according to appellants’ claim, did either of Miller’s men have anything to do.

As respects the bottled gas, the testimony is chiefly that of Herzog who, as we saw, is respondent’s assistant manager. He said that there was in the aisle a container of bottled gas with a capacity of 100 pounds or possibly 20 gallons, used by respondent’s men in handling gas torches. He thought, however, that the container was just about empty. It had been full when placed in the aisle and he did not, after the explosion, examine it. This aisle, however, extends the whole length of the cellar or between 100 and 150 feet, and our attention has not been called to any evidence indicating that this gas container was anywhere near vat 14. The evidence seems to be that it was in the aisle at the southeast corner of vat 18, which would place it something like 45 or 50 feet away from vat 14, and Herzog claims that respondent’s men had not been using gas torches on the day of the explosion, though, so far as he was concerned, he was not in the cellar but in the office and, therefore, could not know positively that they did not. Jaeger testified that his men were putting on “sheets” in vats 17 and 18 which may argue, as appellants claim, that torches were on that day used in those vats, but both of these were at a distance from vat 14.

So far as the acetylene was concerned, the testimony of one Courser, the assistant fire chief of the city, who was on the ground just after the explosion, refers to two acetylene tanks “about opposite the staircase or stairway”. This would, according to a floor plan known in the record as plaintiffs’ exhibit 11, place them something over 50 feet to the east of vat 14. The type of tank is “a regular welding tank or acetylene . . . and combination oxygen”. According to this witness, when he visited the cellar after the explosion, a hose ran out of the manhole in vat 14 easterly along the aisle and was connected to both of these tanks. The fire marshal, Newton, describes the tanks as “acetylene and oxygen tanks” but he thought that they were in the vicinity of vat 9 which, if true, would place them only from 12 to 15 feet from vat 14. He did not notice whether the tanks were turned on or not. Mr. Porter, the insurance man above mentioned, referred to these tanks as being “an acetylene tank and an air tank” and said they were “tied together”. One Trussing, who is a safety engineer for the Union Oil Company, and who examined the cellar a day or two after the accident, refers to these tanks as “one tank of dissolved acetylene and one tank of compressed oxygen” east of but very close to vat 14. Miller, who also examined the premises after the explosion, says that he “saw one acetylene torch tank and a gas tank”.

As to gasoline, Fire Marshal Newton says that he noticed the' black appearance of asphalt and a strong gasoline smell in an overturned can in the aisle at a position which, according to the floor plan, would be in front of the easterly part of vat 16, or a distance of possibly 20 feet from vat 14. This witness says that he also observed directly in front of the opening (manhole) in vat 14, an overturned bucket with the same kind of liquid and the smell of gasoline. The witness Trussing, who noticed this latter pail and said he critically examined it, stated that the upper part of the pail showed carbon, but that there was material in the pail in a liquid state and that to some extent the interior of the pail had been splashed with some white material. The liquid referred to was black and sticky and smelled of gasoline. The man hole of vat 14 was, according to Trussing, directly above the pail. The pail was “adjacent to the manhole within a few inches of it. The wall was smudged with smoke or carbon to a height of perhaps the height of my head. The smudge started above the pail and continued to about six feet in height. ’ ’

Of the material known as Stoddard’s Solvent, the witness Jaeger testified that the company kept on hand never over 10 gallons at a time and usually 5 gallons. This was purchased from the Union Oil Company. The 5-gallon can of the solvent, according to Miller, was seen by him in front of vat 7. This would, according to the floor plan, place it 40 feet more or less from vat 14.

Jaeger testified that respondent keeps kerosene on the premises in 50-gallon drums. Our attention is not called to any statement as to just where it is kept.

There is, of course, no doubt that the acetylene and oxygen tanks referred to were intended to provide for acetylene torches. These, as the evidence shows, were intended for use in the welding together of the sheets of “brewer’s pitch’’ to be applied as the final step in the waterproofing of the vats. As respects gas torches, there is no dispute that it was respondent’s practice to use them in softening the sheets of brewer’s pitch to make easy their application to the mopping coat. In these torches, also, compressed air was used, this time in connection with gas. The process by which the appliance was conducted is thus described by the witness Jaeger.

The further process of bringing the gas into the vat, as described in the evidence is, we think, correctly epitomized in appellant’s opening brief as follows:

The compressed air for use in the gas torches was, according to Jaeger, “run directly from our reservoirs which are located on the roof of the engine room from 180 to 200 feet away from this cellar door”. The compressed air is carried in a % inch galvanized pipe line running along the outside of the rows of vats and tapped at each vat.

Upon producing a conflux of compressed air and bottled gas at the torch the manner of lighting the latter is to use a small wick, the wick being enclosed by a small piece of pipe. The wick is initially lit out in the aisle with a match and carries a flame about the size of a lead-pencil. It is customarily allowed to continue burning and is laid down on the temporary scaffolding within the vat between times when it is used to light torches, and remains there until the workmen are ready to go out of the vat. The testimony is that bottled gas, with the right proportion of air, will explode on coming in contact with a flame. It is established beyond controversy that, after the explosion here involved, a number of burnt matches were found in an aisle in the neighborhood of vat 14. A package of “Wing” cigarettes, the brand said to have been smoked by Cerezo, was also found under some of the debris left by the explosion in this locality, together with some cigarette butts. One of appellants’ theories as to what might have caused the explosion is that some of respondent’s employees, at the time working in or about vat 14, might, on leaving for lunch, which seems to have been served in the establishment about 10 o’clock, have left a wick burning, which might have set off some escaping bottled gas as the latter became mixed with the right amount of air. Appellants’ counsel call attention to “other instrumentalities than blow torches, wicks and matches ’ ’, with which the defendant’s employees may have set off the explosion. To these we will proceed to refer:

(a.) One Gallagher, a deputy county coroner, visited the cellar on the afternoon of the day of the explosion and found a piece of lamp cord or wire, consisting of two strands fas tened together, partly in vat 14 and partly extending through the manhole to the outside in and along the aisle. This cord or wire had an offshoot about 7% feet long. When found, it was badly charred, attached to nothing at either end, but had, at the inner end some sockets and at the outer end, a plug from automobile equipment which was smaller than the regulation plug designed to fit into a regulation socket. Along the north line of the aisle which ran south of the row of vats 13 to 18, six and a half to seven feet above the concrete floor level, was a temporary electric line used not for power but for furnishing light to the vats, and there were sockets, also obtained from automobile equipment, in this line, into which the plugs attached to cords that were to run into the vats could be thrust. Nearer the sockets were pipe chases or holes penetrating through the vat walls through which the cords could be run, though, as just noticed, this particular cord was not in fact run through a chase but instead through the manhole. No pipes had yet been inserted in these holes or chases. The explanation of using the small automobile plug instead of standard equipment is that it could be thrust through nipples in the wall. The particular cord found by Gallagher was identified by the witness Moir as his work and as looking like one of a half dozen that he had made under the order of respondent’s chief electrician, though he does not remember the offshoot; and there is testimony that three or four such cords or wires had, previously to the explosion, been used by respondent’s employees. The city of San Diego had adopted a safety ordinance of the Industrial Accident Commission of the state of California. This ordinance, in section 7-3201-b, under the heading “Flammable Liquids, Gases, and Other Mixtures’’, classified certain locations and includes the following :

In section 7-3203-1, it is, among other things, provided with reference to locations of classes I-A and I-B that:

In section 7-3203-m it is provided that:

The cord or wire found by Gallagher is in evidence and is neither of the type S nor the type PA referred to in the above section. These types of cord enclose two or three wires in an outer covering of rubber. In fact, according to the testimony, the type of cord found is not one permitted by the ordinance of the city for uses of a portable cord anywhere, but is a sort of wiring intended to be installed either in electric conduits or in open wiring for the installation and control of electric current.

(b.) What has just been said has mainly to do with the detached cord or wire referred to, found after the explosion projecting from vat 14 out through the manhole. Mention has, however, been made of the sockets in the temporary wire outside the vat. Appellants stress the claim that the

connection arrangement was defective. On that subject the ordinance mentioned provides:

One Gill, an expert in the electrical business, testified as to various possibilities of a flash in various connections which might be caused either by the pulling from the socket of the plug on such a cord as the one mentioned, if the cord were attached to a lamp on the other end, or by the cord’s being in contact with something where abrasions might occur in such insulation as it did have, and said “there is a safe, portable extension cord which may be safely used in an enclosed area with inflammable gases”.

(e.) Respondent had, as part of its equipment for supplying air to the vats while work was going on within them, a number of blowers, each consisting of an electric motor with a fan on the end of it for removing air. Witness Gill says that he took the end-plate off one of these and looked inside and that it was what is known as a brush type motor; that when it is in operation there is a spark depending upon the load that it has on it. The load on a blower, according to the witness, is constant, so it is fully loaded and there is a spark there from the brush to the commutator when it is in operation. It cannot be operated without throwing a spark which will ignite inflammable gas if it comes in contact with it. The spark, however, only occurs in starting the motor. The motor is not what is called a “vapor-proof” motor nor an “explosion-proof” motor. The ordinance above referred to provides, with respect to motors and generators, as follows:

The testimony is that there are two large air-conditioning units in the hallway between tanks 7 and 29 (which would place them between 50 and 60 feet from vat 14) and that these circulate the air in the entire corridor and create quite a draft. There is testimony that they were operating at the time of the explosion. The purpose of the blowers just referred to is to pick up the air in the corridor and blow it into the tank and for that purpose the practice, according to Herzog, was to place the blowers right in the manhole and turn the current on to start the fan going. Our attention has not been called to any evidence, however, that there was any blower in vat 14 on the day of the explosion, either in use or otherwise.

(d.) Besides the cord or wire above referred to as having been found projecting out of the manhole from vat 14 there was, after the explosion, found lying on the floor of this vat, a piece of lamp cord of another description, that is, one which was rubber covered, detached except for the remnant of a lamp on one end of it. There is testimony that respondent had furnished this type of cord to Hubbert and Cerezo. Appellant claims that this cord, as well as the other, was defective for the reason, among others, that it failed to conform with the ordinance provision that in such a locality the flexible cord should “contain one extra insulated conductor which shall be properly connected to some grounding connection for metal guards”.

(e.) It is further claimed that the entire temporary lighting line above described was installed in violation of the ordinance of the city, although the provisions of the ordinance referred to in that connection are principally those already quoted. It is further said, however, that it was installed without a permit from the city of San Diego. This latter claim is based on the testimony of one Johnstone, a city electrical inspector, to the effect that he was aware of no permit for any temporary work issued subsequent to July 14, 1934. It is questionable, however, whether that contention is correct, inasmuch as the vat building was still unfinished and a permit for the wiring of the building generally had been issued, which appears to have included a permit to use temporary wiring in and about construction, if incidental to the construction.

The foregoing statement in a general way summarizes those parts of the evidence as introduced on which, together with the presumption that Hubbert and his companion used due care for their own safety, appellants must, of necessity, mainly rely, to sustain their view that the case is one in which the doctrine of res ipsa loquitur is applicable.

In many details this evidence is undisputed in the record. In others it is sharply controverted. Herzog says that Hubbert and Cerezo were to work in vat 14 that day and that none of his own men were to do anything in that vat at the time. Jaeger says that he directed Hubbert and Cerezo to work that day in vat 14. The witness Falkenberg, already referred to as having been working on the morning of the accident in vat 13 and as having seen Hubbert in the aisle as he left, also testified definitely that Hubbert and Cerezo had been working that morning in vat 14 (which was the next vat to him on the east). One Egan, an employee of respondent, testified that during the morning of the day of the explosion he saw Hubbert using the “gas plate” or “hot plate” in the hall, heating his tar or asphalt. We have already noticed that the asphalt used in the mop coat was melted asphalt. He says that while such a plate was from time to time used by respondent’s employees, they did not use it in the hall. Jaeger gave it as his impression that vat 12 had been filled with water on August 3d. Witness Porter says that on his inspection of the premises just after the explosion on the same morning, the burner of the gas plate was not in operation. He says he noticed that the valves of the acetylene tank were closed. He did not look in the morning at the valves of the gas tank but did notice that afternoon that they were closed. In addition to the tar bucket just under the manhole of vat 14 he noticed a smaller bucket or pail and between the two was laid a brush. Just east of the manhole of this vat there was a tin can, oblong, with a gas mask lying on it, and a piece of hose coming from it over the tar bucket. The portion over the bucket was very badly charred and burned. The end of the hose ran into the manhole above and there was a piece of the other end of the hose that protruded out again from vat 14 through the pipe chase for about six inches into the hall. This hose the witness later referred to as a “garden hose”. He says that the manhole of vat 12 was closed and locked. According to him, there was no hose connected with the acetylene and air tanks at all. Asked whether he saw any hose connected to the bottled gas tanks or other tanks located at or about the southeast corner of vat 18 (where, as we saw, the record apparently shows these tanks to have been located) and whether he saw hoses running along the hallway and into tank 14, he said :

The fire marshal, Newton, agrees with Porter that an inspection following the explosion disclosed that there were no hoses attached to the acetylene tank. The gas plate or stove, according to him, was not even warm. One Klein, a timekeeper for respondent, testified that Hubbert had, on previous occasions, used gasoline as a solvent. Frank Cramer, another employee of respondent, who says he had from time to time been assigned to help Hubbert and Cerezo, testified that some two weeks before the accident Hubbert had had him buy gasoline as a solvent for him and that Hubbert had paid for it; that the witness had in the past helped Hubbert and Cerezo mix gasoline with asphaltum in the yard. Respondent’s superintendent, Jaeger, said he had seen Hub bert and Cerezo use both kerosene and Stoddard’s Solvent as a “cut back” and that on one occasion he had seen them using gasoline and had forbidden it. Newton testified that the flash point of gasoline is just above zero, while that of either kerosene or Stoddard’s Solvent is above 100 degrees Fahrenheit. Gasoline vapor, according to him, mixed with air, is explosive if the proper percentage of vapor is reached. Bespondent’s chemist testified that a spark from the blower would not cause an explosion of either Stoddard’s Solvent or kerosene and that it would not even cause an explosion of gasoline if the blower were kept going so that the gas would not accumulate. There is testimony that Hubbert had at some time before the explosion declined the use of a blower to ventilate such vats as he worked in, saying that he had his own “system”. Bespondent’s theory about the gas mask and garden hose referred to attached to it, is that they were a device which Hubbert employed to obtain sufficient air while working in vat 14.

The court addressed to the jury a special interrogatory as to whether Cerezo and Hubbert used gasoline as a solvent to “cut back” the material used and to be used in vat 14, to which the jury answered “no”.

It would prolong this opinion to inordinate length to further detail the evidence and we think the foregoing statement sufficient as a basis for disposing of the ease. Since the motion for an instructed verdict was necessarily made after all the evidence was in and the inferences which appellants seek to draw are not confined to those founded alone on the testimony introduced by themselves, it has not been practicable in stating the evidence to exclude from the statement all that is favorable to respondent. At times appellants rely on statements of witnesses, as notably in the case of witness Falkenberg, favorable to themselves, even though further statements of the same witness are unfavorable. The circumstance, however, that we have referred to evidence favorable to the respondent does not, of course, mean that we are unmindful that in reaching our decision we must resolve every real conflict in appellants’ favor and allow appellants the benefit of every inference that may legitimately be drawn, whichever side may have introduced the evidence on which it is founded.

The principles of law that must govern the action of a trial court in passing upon a motion for judgment non obstante veredicto are thoroughly settled. Such a judgment is in order “when a motion for a directed verdict which should have been granted, has been denied, and a verdict rendered against the moving party”. (Code Civ. Proc., sec. 629.) The state of the evidence that calls for such a judgment, therefore, is the same which would have called for a directed verdict. That is to say, as stated in Barthelness v. Cavalier, 2 Cal. App. (2d) 477, 480 [ 38 Pac. (2d) 484 ] :

As said in Estate of Lances, supra (p. 400), quoting from Newson v. Hawley, 205 Cal. 188, 190 [ 270 Pac. 364 ] :

Expressions to this effect have become so trite that it is needless to even refer to the numerous other authorities in which they appear.

In the instant case, however, all possible theories about the cause of the explosion and fire that resulted in the death of Hubbert and his companion, must necessarily be built up from the circumstances. Appellants’ counsel strongly emphasize the undoubted principle of law that everyone is presumed to have used reasonable care for his own safety. (People v. Milner, 122 Cal. 171 [ 54 Pac. 833 ]; Sarraille v. Calmon,. 142 Cal. 651 [ 76 Pac. 497 ]; People v. Siemsen, 153 Cal. 387 [ 95 Pac. 863 ] ; Pabst v. Shearer, 172 Cal. 239 [ 156 Pac. 466 ] ; Thompson v. Davis, 172 Cal. 491 [ 157 Pac. 595 ] ; Gilmour v. North Pasadena Land etc. Co., 178 Cal. 6 [ 171 Pac. 1066 ]; Olsen v. Standard Oil Co., 188 Cal. 20 [ 204 Pac. 393 ] ; Mar Shee v. Maryland Assur. Corp., 190 Cal. 1 [ 210 Pac. 269 ] ; Smellie v. Southern Pac. Co., 212 Cal. 540 [ 299 Pac. 529 ] ; Pacific Portland Cement Co. v. Reinecke, 30 Cal. App. 501 [ 158 Pac. 1041 ] ; Grantham v. Ordway, 40 Cal. App. 758 [ 182 Pac. 73 ].) It is settled also that this presumption is evidence and may, in certain cases, outweigh even positive evidence produced against it. (Smellie v. Southern Pac. Co., supra, p. 549, citing the same cases last above referred to.) Taking this presumption, however, at its uttermost value, it can, in any event, have no greater effect here than to exculpate Hubbert and his companion from any charge of having been themselves guilty of negligence. It has not of itself any tendency or force to impute negligence to respondent. The burden was still with appellants to establish not only such negligence on respondent’s part, but also that respondent’s negligence was the proximate cause of the explosion and the resulting fire in which the fatal injuries to Hubbert and his companion were received.

To be sure, these elements of appellants’ case may be established by circumstantial evidence, but the question here is whether the circumstantial evidence adduced is legally sufficient for that purpose.

As laying the ground for applying the doctrine of res ipsa loquitur, counsel for appellants (if we may somewhat abbreviate their statement) urge: that the explosion and fire occurred at vat 14 and that vat 14 was under respondent’s exclusive control and that concrete vats do not in the ordinary course of things explode or burn; that Hubbert and Cerezo were not at the time of the explosion in or at vat 14 but had been applying in vat 12 water-proofing material supplied by respondent, with mops and brushes and not using any flammable materials; that respondent kept stored and used on its premises highly flammable substances and on the morning in question respondent’s employees were using torches fueled with bottled gas lighted by matches and wicks and that the acetylene-oxygen and bottled gas tanks were connected with vat 14; that various elements of wiring and electrical connections were maintained by respondent in and about vat 14 which might readily produce arcs or sparks when in use or even when not in use and especially in the work being carried on that morning by the respondent’s workmen; and that though Hubbert and his companion, both now deceased, were the only two employees of the independent contractor working in the cellar that day, and there were some ten of respondent’s employees working there, these were for the most part not produced; from which, according to appellants, it must be assumed that if produced their evidence would have been unfavorable. In the light of this situation it is claimed that the application of the doctrine res ipsa loquitur makes out a case for appellants, in the face of which the trial court could not reasonably have said that the verdict was unsupported by substantial evidence.

The positions thus taken are, so far as the facts are concerned, only in part supported by evidence. It cannot, indeed, be doubted that the explosion and fire did occur within vat 14 and in the aisle adjacent thereto on the south and it is true enough that concrete vats do not of themselves explode. If it be true that Hubbert and Cerezo were not on the day of the explosion working in vat 14, then it may be believed that on that day that vat was under the exclusive control of the respondent. Despite Falkenberg’s positive statement that Hubbert and Cerezo were in fact working in vat 14 that morning it might be possible, if such statement were not believed, to infer from Moir’s statement that he saw Hubbert looking into vat 12 and that there was a light therein, coupled with the position of Cerezo’s shoes after the accident, that Hubbert and Cerezo had not worked in vat 14 that day, though the inference would by no means be a necessary one. Appellants, nevertheless, have not in their briefs pointed out anything in the record which actually shows that either acetylene or bottled gas had in fact been used at all in or about vat 14 on the day of the explosion. As to the acetylene, we have seen that the evidence conflicts with respect to whether or not its tank was, at the time, connected by hose with vat 14 or not, but if it be assumed that it was, still such testimony as there is on the subject is to the effect that the valve on the tank was closed. Nobody claims to have found it open. While the bottled gas was always available on the outside of the vats in the sense that there were pipes there leading from the tank where it was kept, the testimony of Jaeger is that the valve in the tank was found after the explosion to be closed and we have had no evidence to the contrary pointed out, and, moreover, we are aware of no evidence that there was, on the day of the explosion, any device attached and thereby made available at vat 14 for conducting it into that vat. Whatever was the function of the garden hose that witness Porter says he saw projecting into the vat through the manhole and coming out again through a pipe chase with the gas mask attached to it, there is nothing to show that the hose was connected with any gas pipes. Neither is there any testimony at all that respondent made any use of gasoline either on the day of the explosion or at any other time, nor any evidence of any sort to that effect, unless it be inferred from the presence in the aisle after the accident of the two asphalt pails which smelled of gasoline and from which it was manifest that somebody had used gasoline. As respects kerosene and Stoddard’s Solvent, there is, in the first place, no evidence that they were being used that day in vat 14 or anywhere else, unless it is inferable that the decedents were using them, and, secondly, while they might be set afire by a spark if there was one, neither would be likely to explode. Our attention is called to no evidence that in terms shows that any of the ten men whom respondent is said to have had at work in the cellar that day were engaged in applying “brewer’s pitch” otherwise than in vats 17 and 18 or which can rea .sonably be the basis for inferring that they were using any torches elsewhere, if,. indeed, they were using any there. Assuming it to be true that respondent’s electrical wiring and connections did not measure up to the requirements of the city ordinance and that the unattached double wire that was found was used by somebody that day for lighting up vat 14 and that the lighting arrangements might have caused some spark, it is yet manifest that such spark could have brought about an explosion only if it came in contact with something explosive and, as just observed, our attention has not been called to any evidence other than such as may be supposed to result from inference, that respondent was making use of anything explosive in or about vat 14 on the day of the accident, or that any of its own men were working there that day at all. Whatever evidence, then, to that effect, may have been conceived to exist, must amount merely to an inference drawn either from the fact that bottled gas and acetylene, as well as kerosene and Stoddard’s Solvent, were available for use generally in the work of water-proofing the vats, or from the fact that after the explosion the buckets smelling of gasoline were found. Respondent did have something like ten employees working in the cellar that morning, only three of whom, Moir, Falkenberg and one Walters, took the stand. Not any one of these three appeared to have been using any gas or acetylene torch, or gasoline, or anything else of an explosive nature that day. All of the ten employees except Moir had left for lunch before the explosion, and it is hardly possible to infer, from the mere fact that seven of them were not called to testify, either that they were handling explosives in proximity to vat 14 that day or that they had worked in a position that would have made their testimony important or have enabled them to shed any light on the accident.

What basis, then, does all this afford for applying the res ipsa loquitur doctrine? The generally accepted statement of the res ipsa loquitur doctrine laid down in Scott v. London etc. Docks Co., 3 H. & C. 596, 601; 159 Reprint, 665, repeated in Shearman & Redfield on Negligence, 6th ed., p. 132, and referred to in 45 O. J. 1193, as the one “most frequently adopted and applied in subsequent decisions”, is that:

This has, in many California cases, been stated to be the law (Judson v. Giant Powder Co., 107 Cal. 549, 556 [ 40 Pac. 1020 , 48 Am. St. Rep. 146 , 29 L. R. A. 718]; Harrison v. Sutter Street Ry. Co., 134 Cal. 549, 550 [ 66 Pac. 787 , 55 L. R. A. 608]; Kahn v. Triest-Rosenberg Cap Co., 139 Cal. 340, 344 [ 73 Pac. 164 ] ; O’Connor v. Mennie, 169 Cal. 217, 223 [ 146 Pac. 674 ]; White v. Spreckels, 10 Cal. App. 287, 292 [ 101 Pac. 920 ]). In Michener v. Hutton, 203 Cal. 604 [ 265 Pac. 238 , 59 A. L. R. 480], it was said to be “elementary that the maxim ‘res ipsa loquitur’ translated means simply ‘the thing or affair, speaks for itself’ ”. Appellant says that here the “thing” which caused the injury was the “explosion and resulting fire”. To say, however, that the “explosion and resulting fire” were under the exclusive control of the respondent would be to talk nonsense, and manifestly, as indeed appellant's concede, that which must be shown to have been under the respondent’s exclusive management is the instrumentality, whatever it was, that caused the explosion. There are, indeed, cases in which a showing that every instrumentality to which a given injury could with reasonable probability be attributed was under a defendant’s management has been accepted by the courts as, for practical purposes, the equivalent of a showing that the defendant controlled the particular instrumentality that did cause it. Such a ease is Judson v. Giant Powder Co., supra, where the plaintiff’s property was injured by an unexplained explosion in an adjacent dynamite factory under the exclusive management of the defendant. Cases of the same general description are Faras v. Lower Calif. Dev. Co., 27 Cal. App. 688 [ 151 Pac. 35 ], Lippert v. Pacific Sugar Corp., 33 Cal. App. 198, 208 [ 164 Pac. 810 ], Hallawell v. Union Oil Co., 36 Cal. App. 672 [ 173 Pac. 177 ], Phoenix Assur. Co. v. Texas Holding Co., 81 Cal. App. 61 [ 252 Pac. 1082 ], and Smith v. Southern Counties Gas Co., 89 Cal. App. 81 [ 264 Pac. 532 ]. Such, also, are many of the cases in which the doctrine res ipsa loquitur has been applied to fix the responsibility for injuries sustained on railroad trains or other conveyances.

The instant ease is of a very different character. The mere fact that an accident has occurred does not of itself result in any inference of negligence as against a defendant (Depons v. Ariss, 182 Cal. 485, 488 [ 188 Pac. 797 ]). To justify the invocation of the rule res ipsa loquitur the instrumentality which caused the injury must have been under the exclusive management of the defendant. As was said in Biddlecomb v. Haydon, 4 Cal. App. (2d) 361, 364 [ 40 Pac. (2d) 873 ], in which a hearing was denied by the Supreme Court:

In the instant case we are forced by a process of elimination to conclude that there is no sufficient evidence to sustain any inference as to what instrumentality caused the explosion, or that it was' caused by any of the instrumentalities that have been pointed out, unless it be the inference that whoever had been using the tar buckets smelling of gasoline, one of which was found just below the manhole of vat 14, had that morning been using gasoline within that vat, and that some spark or flame without the vat ignited the gasoline in that bucket, which not only exploded but was so near the manhole as to result in a coincident explosion of the accumulated gasoline fumes within the vat. Assuming that the jury did draw that inference, as it may have done, we would, to sustain the verdict, next be required, in the light of its answer to the special interrogatory, to understand that it further inferred, and to hold that it could be justified in inferring, that neither Hubbert nor Cerezo was responsible for the presence of the gasoline, and then that the jury carried, and could be justified in carrying, that inference a step farther, by inferring that because Hubbert and Cerezo were not responsible for the presence of the gasoline respondent must have been; and further, that the jury inferred and could be justified in inferring that the flame which ignited the gasoline either originated in some piping or connection under the respondent’s exclusive control or else was something the presence of which, in the exercise of ordinary care, the respondent ought to have anticipated; all this as a preliminary to applying the final inference necessary to the application of the res ipsa loquitur doctrine, namely, that whatever negligence it was that brought the spark or flame into actual contact with the gasoline was respondent’s negligence.

Respondent, planting itself upon the authority of Puckhaber v. Southern Pac. Co., 132 Cal. 363, 366, 367 [ 64 Pac. 480 ], asserts that the argument that legal liability can be predicated because of an inference piled upon another inference or a presumption piled upon additional presumptions, has long been repudiated in this state. As substantiating the same proposition we are referred to Hernandez v. Southern California Gas Co., 213 Cal. 384, 387, 388 [ 2 Pac. (2d) 360 ], and to various explosion cases from other jurisdictions, including Schaller v. Independent Brewing Assn., 225 Ill. 492 [ 80 N. E. 334 ], Sowers v. McManus, 214 Pa. 244 [ 63 Atl. 601 ], Ammon v. Horn & Hardart Baking Co., 321 Pa. 49 [ 183 Atl. 786 ], and Courtney v. New York, N. H. & H. R. Co., 213 Fed. 388 , and Puget Sound T. etc. Co. v. Hunt, 223 Fed. 952 [ 139 C. C. A. 432 ], wherein it is said that:

Appellant argues that each of the group of cases last cited is for one reason or another distinguishable in its facts from the instant case. Doubtless some of the differences pointed out do exist. We do not think, however, that the instant case is one that justifies extensive research into borderline distinctions. There must be some point in the stretching of a thread at 'which it will break. In this case, in order to have reached its verdict, the jury must necessarily have based one inference on another until the filament of its reasoning was so extended and so tenuous as to have brought about nothing better than a mere guess as to the presence of circumstances on the basis of which it must, then, have undertaken to apply the doctrine res ipsa loquitur. A verdict of that description is not supported by substantial evidence. In these circumstances the respondent was entitled to an instructed verdict, and, therefore, the trial court’s action in granting a judgment non obstante veredicto was right.

The judgment is affirmed.

Barnard, P. J., and Marks, J., concurred.

A petition for a rehearing of this cause was denied by the District Court of Appeal on June 28, 1938, and the following opinion then rendered thereon:

We have given due consideration to the petition for rehearing filed by appellants. It is therein claimed that the statement of facts contained in our opinion is in various particulars incomplete and inaccurate. As respects its comprehensiveness, .while it would obviously be out of the question within any moderate space to recite at large the contents of 944 pages of typewritten transcript, we do not think that we have omitted anything essential to a statement of the evidence on which the conclusions to be reached necessarily depend. As respects asserted inaccuracies in the somewhat extended statement of facts there contained, we have carefully rechecked the opinion with the record and heretofore caused the correction of the only two inaccuracies which we are able to discover, neither of which in fact affected our reasoning. Most of the others asserted to exist amount merely to the omission in some particular connection of matter which is fully stated elsewhere in the opinion, or else to an apparent misapprehension of what we have said. An example of the latter is .counsel’s comment on our statement with respect to bottled gas, wherein, after noting that it Avas piped to the outside of each vat, we added: “We are aware of no evidence that there was, on the day of the explosion, any device attached and thereby made available at vat 14 for conducting it into that vat.” Counsel seems to think that we were in that connection talking, not about bottled gas but about acetylene, to which we had shortly before referred, and call our attention to the testimony of Assistant Fire Chief Courser (disputed by other witnesses) that a hose connected vat 14 with the acetylene tank, all of which is, as a matter of fact, fully noted and taken account of elsewhere in the opinion. Many of the complaints made of our review of the facts have to do not so much with what is stated as to the stress which we have placed on some phase of the case as compared with some other phase of it. As to these matters counsels’ objection is in reality related not so much to what we have said about the facts as to the view we take of the law.

When we come, then, to the legal questions involved, the gist of the discussion centers about what application is to be given to the doctrine of res ipsa loquitur. Appellant insists that the case is governed by the authority of Judson v. Giant Powder Co., 107 Cal. 549 [ 40 Pac. 1020 , 48 Am. St. Rep. 146 , 29 L. R. A. 718], and the subsequent cases which have applied its principles, and objects that, though we have recognized that decision as stating the law, we have said that the instant case is a different one without pointing out the difference. The distinction seems to us entirely clear. Judson v. Giant Powder Co., supra, involved no situation in which it could be claimed either that the plaintiff or anyone else, except the defendant, had any control over any of the instrumentalities to which the explosion could be reasonably attributed. In these circumstances it was held unnecessary to show to what one of the various agencies controlled by the defendant the accident was attributable. If all agencies, in other words, to which the explosion could reasonably be attributed, were under the defendant’s control, the doctrine res ipsa loquitur would be equally applicable, whichever one actually caused the explosion. Contrariwise, in the instant ease it might readily have been believed from the evidence that the inflammable substance involved was gasoline and that both it and the agency which ignited it were under the control of the decedents or one of them. The jury indeed rejected this hypothesis, as it was its right to do, but the circumstance that, as an hypothesis, it was available and tenable seems to us radically to distinguish the case from the line of authorities represented by Judson v. Giant Powder Co., supra, and to throw it into an entirely different category from them. It is further insisted, however, that our opinion rests “on a discredited application of the doctrine res ipsa loquitur”, namely, that the doctrine “is to be applied only when the nature of the accident itself, not only supports the inference of the defendant’s negligence but excludes all others”. We are not aware that we have announced any such view, though we did quote from the comparatively recent case of Biddlecomb v. Haydon, 4 Cal. App. (2d) 361, 364 [ 40 Pac. (2d) 873 ], wherein Lucid v. E. I. DuPont etc. Powder Co., 199 Fed. 377 [L. R. A. 1917E, 182], is cited to that effect. There is, however, a wide distinction between demanding as a requisite to the application of the doctrine res ipsa loquitur that the instrumentality which caused the accident shall have been under the exclusive control of the defendant, and demanding that the evidence exclude any other inference than that the defendant negligently handled it. It is of course true that though an agency be under a defendant’s exclusive control and though the accident be one which in the usual course of things does not occur when ordinary care is used, there remains a possibility that the occurrence xvas not due to the defendant’s fault. As against that possibility, however, the doctrine res ipsa loquitur permits the jury to infer negligence. When, however, the instrumentality itself is not clearly within the exclusive control of the defendant the situation is different. Since the general rule is that the occurrence of an accident in itself raises no presumption of negligence, dependence must be had on some inference in addition to the fact that there was an accident, and manifestly it must be an inference additional to the further fact that the accident was one of a sort that does not usually occur where due care is used. That inference may be entertained though no specific act of negligence be proved if, in addition to the circumstances just noted, it appear that the instrumentality which caused the accident was under the exclusive control of the defendant. If it was not under his exclusive control manifestly the inference' may not be entertained. Such cases as Judson v. Giant Powder Co., supra, are undoubtedly authority for the proposition that the particular instrumentality that caused the accident does not necessarily need to be pointed out but, as said in our opinion, if it be shown that every instrumentality to which the accident could with reasonable probability be attributed was under the defendant’s control, that is for practical purposes the equivalent of a showing that he controlled the particular instrumentality that caused it. Our expressions in the recent ease of Wills v. Price, ante, p. 338 [ 79 Pac. (2d) 406 ], referred to by counsel, go no farther than that. The distinctions to which we have just adverted appear to us entirely sufficient to reconcile the quotations by counsel from such cases as Hilson v. Pacific Gas & Elec. Co., 131 Cal. App. 427 [ 21 Pac. (2d) 662 ], Mintzer v. Wilson, 21 Cal. App. (2d) 85 [ 68 Pac. (2d) 370 ], and Barham v. Widing, 210 Cal. 206 [ 291 Pac. 173 ], as well as the earlier case of Ley v. Bishopp, 88 Cal. App. 313 [ 263 Pac. 369 ], with what we have in the instant case said. It was, for that matter, laid down in Hilson v. Pacific Gas & Elec. Co. (p. 434): “that the doctrine of res ipsa loquitur can only apply when the defendant has exclusive control of the instrumentality causing the injury, and can in no event apply when the accident may as well have been due to one or more causes over which the defendant had no control”. This, the court says, “is a correct statement of the rule”.

Returning to the facts in the instant case, if, the situation being res integra, the jury were confronted with evidence requiring no resort to very doubtful inference to enable them to conclude that vat 14 and the adjacent corridor and all explosive materials as well as all instrumentalities capable of igniting them were at the time of the explosion under the exclusive control of respondent, then we would have no hesitation in saying that the inference res ipsa loquitur might be properly indulged.

We do not wish to be understood as holding that circumstantial evidence will never suffice to establish the control by the defendant of the instrumentality causing the accident so as to lay the foundation for the inference res ipsa loquitur. That would be going altogether too far and be contrary to the views expressed in such cases as Lejeune v. General Petroleum Corp., 128 Cal. App. 404, 416 [ 18 Pac. (2d) 429 ], and Hilson v. Pacific Gas & Elec. Co., supra, as well as to the reasoning of Barham v. Widing, supra, although that case did not, strictly speaking, involve the doctrine res ipsa loquitur. When, however, as here, the establishment of such exclusive control must itself rest on inferences, which are so far from necessary ones that the contrary inferences are at least as tenable as the ones adopted, we do not think the doctrine applicable.

The petition for rehearing is denied.

A petition by appellants to have the cause heard in the Supreme Court, after judgment in the District Court of Appeal, was denied by the Supreme Court on August 1, 1938.