Standard Oil Co. of Louisiana v. Roach

DeWITT, J. On January 11, 1933, Margie Roach, an unmarried woman twenty-five or twenty-six years of age, was fatally burned from an explosion in the kitchen of rooms which she and two other-young ladies rented and occupied in a frame residence in McMinn-ville. The explosion occurred when the deceased was in the act of making a fire in the cooking stove by pouring from a can upon some-live coals a liquid which had been purchased as coal oil from H. M. Green, a retail merchant, who had purchased it from the Standard Oil Compaq of Louisiana. In this action the father of Margie Roach, as administrator, obtained a -verdict and judgment against said company for $5,000 for negligence in selling and causing to be placed upon the market a liquid containing* gasoline or some other highly volatile and dangerous liquid which exploded with fire and caused the death of the young lady. The merchant, Green, was also sued, but the action was dismissed as to him upon motion for a directed verdict in his favor. The sole question presented upon this appeal in error is whether or not the motion made in behalf of the Standard Oil Company of Louisiana at the close of all the evidence should have been sustained upon the ground that there was no material evidence upon which a verdict for the plaintiff could be predicated, taking the most favorable view of it for the plaintiff. The evidence has been carefully examined and considered. It appears without dispute that the three young ladies worked at a silk mill; that they rented the house and subrented rooms in it to Mr. and Mrs. Simon and their daughter; that about 2 o’clock in the afternoon the deceased came home from her work, went into Mrs. Simon’s room, took a shovel, and got some live coals to make a fire in her stove, but stood there and talked a few minutes, then took the coals from the front room through the middle room into her own kitchen; that two or three minutes later Mrs. Simon heard *664 a noise like that of an explosion, which shook the house, accompanied by a, scream; that upon hurrying to the kitchen she found it all afire; that the smoke in the room was dense, but the noise of the flames could be heard; that city firemen came and found the kitchen and the next room in a solid flame, and the body of the young' lady was found in the front room with her clothes burned' nearly off and her body burned over her chest and abdomen; that the cooking stove was lying on its side on the floor; that this was a little box stove that stood on pine blocks about three feet from the floor, having four “eyes,” a door across the front with small holes in it, an apron in front fastened to the stove with a pin, the door having small holes in it and being fastened with hinges and pins; that after the fire Mr. Simon found the stove with one leg broken off, one side burst, the caps and crossbars scattered to different parts of the room from four to six feet; that he also found a five-gallon oil can about six feet from the stove, about eighteen inches from the door from the kitchen to the bedroom in the line from the stove to the front window where the body of the deceased was found; that the can had a large spout and a small spout; the large spout was closed and the small spout was open. Mr. Simon testified without contradiction that the can was in the same condition at the trial as it was when thus found, and was the same can the young ladies used and the same in which he saw the liquid placed on the Monday afternoon preceding the explosion on Wednesday. This can is exhibited with the transcript as a part of the record before us. The can is of galvanized iron. On the middle of the top is a small cap (for reception of liquid) screwed on tight. The can is mashed into an oval shape. It has a dent in one side. The bottom of it is torn loose about halfway and hangs down about eight inches. Whether this condition was due to an internal explosion of its contents, and, if so, when it occurred; or it was due to external force when the stove collapsed — are questions difficult to determine from any evidence. It was on Monday afternoon, January 9th, that the liquid one gallon was purchased by the young ladies and poured into the can by an employee of Green’s — the employee a boy named Hale, who was dead at the time of the trial. It was the liquid obtained by Green from the Standard Oil Company prior to that Monday afternoon that contained this one gallon. On Saturday, January 7th, the company delivered to Green in his tank fifty gallons as kerosene. The can was found after the fire, among débris. The witness T. J. Keenan, a technician of the gasoline inspection unit of :the state, who visited the house the day after the fire, testified without contradiction that he found the can under débris, and that “possibly” it was under pieces of timber; that he brought the can to the laboratory at Nashville. *665 Tbe jury manifestly found that tbe defendant company bad furnished to tbe merchant as kerosene a liquid containing gasoline or some other highly volatile and dangerous substance; that it was some of this liquid which caused the explosion; and that the deceased was not negligent, because she thought that she was using kerosene in the ordinary'way to kindle a fire. This verdict was based upon circumstantial evidence, for no living person could testify as to just how the explosion occurred. It is strongly urged that the verdict was based on mere guess or conjecture. There must be more than a mere probability that the defendant was negligent. But the plaintiff is not bound to exclude the possibilty1 that the accident might have happened from some other cause than that alleged. He is only required to satisfy the jury by a fair preponderance of the evidence that it resulted from the cause alleged. The facts must tend to exclude any other cause, but the inference of exclusion of any other cause than that alleged need not be urged beyond mere doubt; and where, after a fair consideration of the circumstantial evidence, the more reasonable probability is in favor of negligence, the case is for the jury. 45 C. J., p. 1267, sec. 835; Nashville Railway & Light Co. v. Harrison, 5 Tenn. App., 22. A theory cannot be satisfactorily established by circumstantial evidence in a civil case, unless the facts and circumstances shown are not only consistent with such theory, but inconsistent with any other reasonable theory, and such must be shown by the preponderance of the evidence. See Crowe v. Birmingham & N. W. Railroad Co., 2 Tenn. App., 634. However, the rule as to guess or conjecture does not contemplate that the evidence shall exclude the possiblity that the injury might have resulted from a cause other than that alleged. Such possibility is not to be allowed to defeat a recovery, where the evidence discloses sufficient facts and circumstances surrounding the occurrence to justify a reasonable juror in concluding that the thing charged was the prime cause. Nashville Railway & Light Co. v. Harrison, supra. In view of the issue before us, all of the evidence admitted by the trial court has been carefully examined, for in the disposition 'of the assignment that it was error to overrule the motion for a directed verdict, it is our duty to take into consideration the evidence on behalf of the plaintiff and such parts of the evidence introduced on behalf of the defendant as are not in conflict with 'the evidence for the plaintiff, disregarding all of the defendant’s evidence that is contradicted by or inconsistent with the evidence for the plaintiff. Nashville Gas & Heating Co. v. Phillips, 17 Tenn. App., 648, 69 S. W. (2d), 914, and cases there cited. Taking the evidence more strongly tending to support the verdict, and all reasonable inferences therefrom favorable to the successful party below, we must simply determine whether or not it affords ground for more than mere guesswork or conjecture as *666 a basis for satisfactory conclusion by tbe jury. When circumstances in evidence point to a conclusion of fact in such manner as to justify tbe finding of such fact, it cannot be said that a jury acting on such circumstantial evidence, and finding facts therefrom, was left to guess or speculate as to probabilities. Louisville & N. R. Co. v. Hall, 5 Tenn. Civ. App., 491, 498. But what is known as tbe “scintilla rule,” that where there is a mere spark or glimmer of evidence the case must be submitted to the jury, does not obtain in Tennessee. There must be some evidence of a material or substantial nature to support the plaintiff's case, for the court to refuse to undertake to determine its comparative value or weight. Brenizer v. Nashville, C. & St. L. Railway, 156 Tenn., 479, 3 S. W. (2d), 1053, 8 S. W. (2d), 1099. The evidence relied on to support the verdict must in itself have fitness to induce conviction, even if un-eontradicted. DeKalb County v. Tennessee Electric Power Co., 17 Tenn. App., 343, 350, 67 S. W. (2d), 555, and cases there cited. The contentions as to facts and circumstances relied upon to support the verdict are as follows: (a) The company hauled both kerosene and gasoline in its tank trucks, using alternately the same compartments, and depended alone on flushing with kerosene to remove gasoline from the kerosene compartment. This was the custom followed in its delivery to Green, which afforded opportunity for mixture of the two substances when the delivery was made to Green. Its driver, Locke, did not remember when changes were made from one to the other, nor when the change was made before this particular delivery. Also, both gasoline and kerosene were drained by faucets which were side by side at the bottom of the end of the truck. Locke testified without contradiction that between the compartments there were two walls; that the liquid in one compartment could not possibly get to the other by any leak, but would run out on the ground; that the gasoline faucet always carried a red tag and had a red nipple; that the kerosene faucet had only a tag distinguishing it from the other. From this evidence as to open space between the compartments it cannot be concluded that there was any leak from one compartment into the other. Locke said that in the compartment in which he hauled the liquid which was delivered to Green on January 7th, Saturday, he might have hauled gasoline several hours before; that he drained it out before putting in kerosene; that he was always careful to do this. He testified positively that the liquid which he delivered on that day to' Green was kerosene, but as this statement is in issue we must disregard it. As there is no direct evidence whatever that Locke delivered any gasoline as kerosene to Green, the claim that he did must depend upon inference from other evidence. (b) There was no opportunity for mixture of gasoline and kero *667 sene after its delivery to Green, as he stated that he kept them in separate tanks and made deliveries of them in separate containers. This was evidently the basis for dismissal of the suit as to Green. It merely excludes from consideration any notion that he was guilty of negligence in selling gasoline instead of kerosene to the deceased. The delivery boy Hale cannot be presumed to have made any change in the contents of the can after it left Green’s store. Either in some way the company delivered to Green and he sold to the deceased a liquid, most likely gasoline if at all, much more volatile and dangerous than kerosene, and for which the company would be liable; or what the deceased purchased and used was merely kerosene, but she used it in a negligent and dangerous manner by pouring it on live coals, causing the explosion and fire — and for which the company would not be liable. There is no evidence pointing to any third cause. (e) The liquid used by the deceased was delivered on Monday night, placed on the rear porch in her can, and a small snuff box full of it was used by Mr. Simon, soon after its delivery, for the purpose of kindling a fire in his stove. He testified that he poured it on some sticks of hard oak and some soft wood in his stove, then stepped back and threw a lighted match on it. He said that “it flamed up something like a bunch of birds, flew up something like that and did not burn;” that it did not jar the stove; that it acted more quickly than the coal oil that he had been using, which -would burn more slowly; that it roared louder than a common fire; that he had never used kerosene that would flare up like that. There is also substantial evidence that the sticks of wood on which the deceased poured the liquid were found to have been scarcely scorched and not burned, but the testimony of the plaintiff’s witness Prof. J. M. Breekenridge was that when either gasoline or kerosene, poured on wood and lighted, would explode, it might not burn the wood. There is no evidence to the contrary. Therefore, the fact that the sticks of wood were not burned or seriously scorched is not substantial evidence for, as aforesaid, this might happen in either case. Mr. Simon admitted that he gave no warning to the young ladies about the use of that liquid, as he did not think the matter serious. He said that his two-gallon can containing one gallon of kerosene went through the fire without exploding or injuring the can, but when found it was empty. (d) On Tuesday night, January 10th, Miss Ruth Allison, one of the three young ladies, used some of this liquid to kindle a fire in the heater. Her testimony is that “it just flashed-up, sort of shook and flared;” that she paid no attention to it. Simon said that at that time he heard them rattling the heater in their room and heard a puffing and roaring of the fire. (e) The explosion occurred on a cold day in January. We see no *668 materiality in this circumstance, as tbe liquid was poured on live-coals in the stove — only a very few minutes after they were obtained from the other room. (f) The coals had sufficient time to cool enough so as not to generate kerosene vapors. The deceased borrowed two live coals, held them in a shovel for two or three minutes, was cautioned by Mrs. Simon that the coals would go out, then walked forty-eight to fifty feet with them in the open air to her kitchen, and it was two or three minutes before the explosion was heard. The testimony is that she went from Mrs. Simon’s room into the hall, then into her bedroom, and back to the kitchen. Dr. J. M. Breckenridge, head professor of chemistry at Vanderbilt University, testified that in his opinion the coals would then not heat the fire box in the stove sufficiently to bring it up to a temperature to explode kerosene; and that it was something more volatile than kerosene that exploded. His expert testimony will later be discussed. (g) The coals were placed in the stove box together with wood,the can exploded and the deceased was burned simultaneously; the explosion scattered the contents of the can over the room and ignited all inflammable parts at once, as they were easily inflammable; it was a very fast burning fire. The question is whether these and other circumstances warrant the conclusion that the explosion was from kerosene or gasoline or some other liquid than kerosene. It can only be inferred from the evidence that the young lady was standing close to the stove, holding the can somewhat above it close to her body, and pouring the liquid on the wood and live-coals. The front of her body was very badly burned. There is no evidence that any one heard noise as of a second explosion. The contention is made that the can exploded as the young-lady was holding it. It is not reasonable to infer that no explosion occurred in the stove; how else could it have been thrown down and its parts scattered over the room? There being only one explosion, if the can exploded, then there must have been a general explosion of can and stove, all at once. There was a conflict between the opinions of the experts, Dr. Breckenridge for the plaintiff and Amis for the defendant, as to whether or not the can could have been exploded simultaneously by kerosene vapor taking fire. Mr. Amis, a petroleum chemist, testified that kerosene so poured on a fire could cause vapors that would cause the can to explode or blow out the bottom, by forming an explosive' mixture in the can. Dr. Breckenridge was of a contrary opinion. He said: “The explosion would have to come from the embers, come up and meet, travel along the stream of kerosene as it was coming out of the can, a little fire that would follow the kerosene poured out in this direction and the flame might travel back *669 toward this. Now, since in this case we do not have the explosive mixture we have been talking about, there would not be much chance for that flame to go back into the can.” He said that the kerosene would have to be raised to a high temperature, the boiling point, iir which case the fumes would come out of the top, and if there was no air in the can the flames could not jump back; in other words, the kerosene would burn in the can and come out. We think that there was evidence from which the jury could reasonably infer that the can did explode. The bottom of it is half torn loose and the loose part extends down about eight inches. The inference might be made that the partly mashed condition and the dent resulted from objects falling on the can on the floor after the explosion. Now was there evidence from which something more than guess or conjecture could be made as to what was in the can? The record contains much testimony of- expert chemists and other persons familiar with the properties of these liquids. These witnesses agree that both gasoline and kerosene are refined from crude oil; that gasoline is more volatile than kerosene, igniting and exploding at a lower temperature by reason of a quicker vaporization; that when they vaporize, respectively, they form the same gas, which is equally explosive and with equal violence. Dr. Breekenridge testified that kerosene, when sufficiently heated, throws off this vapor in much smaller quantity than does gasoline, at room temperature. The vaporizing point of kerosene is at a temperature of not less than 115 degrees Fahrenheit, sometimes a very few degrees higher. Dr. Breekenridge said that if you throw kerosene in a red hot coal fire, it will explode and come back at you; but he did not believe that two live coals put under wood in the stove, as described to him in a hypothetical question, would have heat enough to vaporize kerosene to produce an explosive mixture to have the force for such an explosion. He thought that it would be difficult to get a temperature of 115 degrees in a fire box with two big lumps of live coal in it burning. However, he further said (and it may be readily understood) that whether this is possible or not depends on the size of the fire hox and the size and temperature of the coals. Much is made of the time element — the minutes that elapsed between procuring the coals from Mrs. Simon’s stove and the explosion — as evidence tending to show the likelihood that the heat was insufficient to vaporize kerosene; but it is not possible, of course, to say how hot those coals were. Mr. Keenan was of the opinion that live coals carried fifty feet were still giving off sufficient heat to produce inflammable vapors. Dr. Breekenridge said that it will take more air to explode with kerosene than with gasoline; but it must be inferred that the top of the fire box was open while the liquid was being poured in. As aforesaid, kerosene vapor will explode like gasoline vapor —only the conditions under which the explosions occur are different. *670 If keros.ene would explode and cause the results shown in this record, under the conditions shown — if this was substantially a probable cause — then the judgment is based on conjecture. After explaining that both kerosene and gasoline are refined from crude oil as their base, and that it would take a higher temperature to explode kerosene than to explode gasoline, Dr. Breckenridge was asked and answered an elaborate hypothetical question. As his answer must have been of great weight with the jury, and as the correctness of the question is seriously challenged, we here copy the question and answer: "Doctor, I will ask you this question; Assuming that any person wishing to kindle a fire with kerosene, or with a substance thought to be kerosene,- first borrowed two live coal embers about the size of a woman’s fist, and waited about two or three minutes after taking those out of the stove, then carried them the distance of about forty-eight feet from where they were taken out of the stove and placed them in a cooking stove of the box type variety, on three or four pieces of hard wood, seasoned red oak wood, and then stood by that stove and poured some substance, thought to be kerosene at the time, from a five-gallon ordinary can, as of the variety exhibited here, from which there was one or two gallons in the can, and assuming that of that same kerosene or substance a small snuff box full had also been used in kindling a fire prior to that time, and in the kindling it was poured over soft wood and a match was thrown on it, after the liquid had been poured on the wood and there was a flash and a small noise and the flame jumped up two feet above the wood and then went out, and the wood was only scorched, and not burnt; and assuming the same kerosene from the same can, or the same substance thought to be kerosene, was poured by this party in this box stove onto this wood, and within a period of two or three minutes after the coals were borrowed, or after the lady left the room after borrowing the coals, there suddenly was a loud explosion, the entire room burst into flames immediately, and the receptacle from which this substance was poured was this particular receptacle here which has been exhibited to Mr. Simon’s testimony, and it caused, or that condition of that can was caused there by the explosion that followed; and assuming that that explosion was so loud it could be heard at least seventy-five yards away, across the street, inside of another house, and that the house jarred, in which the explosion occurred, the dishes rattled, the windows shook, the stove fell over, the caps and eyes went in different directions, the apron was torn off the stove, the door was tom away from the stove, the stove was cracked, and afterwards the wood was found in that stove, the same wood, or afterwards wood was found in that stove which was not burnt, but merely scorched; assuming all these facts, state whether or not in your opinion that substance in that can used to kindle that *671 fire, was kerosene? A. Before answering the question, I would like to get some more information about the size of the fire box in that stove, how big the fire box was. “Mr. Hatcher: If your Honor please we are reserving our exception. “Q. 46. Just the language of that question; assuming that the fire box in the stove was the ordinary fire box in the ordinary box stove of four caps on top, four eyes, it being a fiat top stove? A. In my opinion the material was not kerosene.” The value of the opinion of an expert witness upon a supposed state of facts must depend upon the agreement or non-agreement of the supposed case with the actual facts of the ease on trial. Haskins v. Howard, 159 Tenn., 86, 16 S. W. (2d), 20; Fisher v. Travelers’ Insurance Co., 124 Tenn., 450, 138 S. W., 316, Ann. Cas., 1912D, 1246; Woodward v. Iowa Life Insurance Co., 104 Tenn., 49, 56 S. W., 1020. Mere opinions cannot prevail over actual facts. However, it is sufficient if the hypothetical question fairly states such facts as the evidence fairly tends to establish, and fairly presents the claim or theory of the party offering it. Tenn. Central Ry. Co. v. Gleaves, 2 Tenn. App., 549. There was evidence tending to show the facts assumed in the hypothetical question. The objection to it was properly overruled. The answer stated merely the conclusion of the expert. Upon cross-examination he was first asked and he answered as follows: “Q. Now, Dr. Breckenridge, what was there in the hypothetical state of facts that was given to you by Mr. Davenport which is inconsistent with the substance being kerosene? A. The fact of the explosion. ’ ’ Dr. Breckenridge said that he had not assumed that only the can exploded. He said, “The stove may have exploded.” The can was exhibited to him and he said that its condition could have been caused by an explosion — basing it, as he said, on “the very shape of the blow out on the end.” He was then asked and he answered as follows: “Q. It could have been caused by something else besides an explosion, could it not ? A. I do not see any evidence of the can being abused, but it looks like a straight blow. “Q. Don’t you see the sides of the can? A. I see the sides, but I do not see any indentation as if somebody walked on it, or something fell on it, it has the same base on that can, which would take a straight pull to get it out. “Q. You are speaking of the bottom of the can? A. Yes, sir.” As aforesaid, the can shows some indentation on the side, which the witness did not see. He was not told that the can was found on the floor under debris. The opinion of an expert as to the operation of physical *672 laws is receivable in evidence wben it relates to matters wbieb are not witbin the common observation and experience of jurors and they are not equipped to draw inferences from the facts shown, the weight and value of such opinion to be determined by the jury. The answer of Dr. Breckenridge to the hypothetical question went far toward purporting to determine the ultimate issue before the jury; but the objection related to the assumption of facts in the question, and not to asking for an opinion whether or not the substance used to kindle a fire was kerosene. The opinion of an expert may be reduced to mere conjecture by proof of physical facts completely inconsistent therewith; and an opinion which is mere conjecture is not evidence upon which a jury can properly rely. 22 C. J., 640. Now we must consider certain uncontradicted testimony which is inconsistent with the evidence relied on to support the verdict. Mr. Green, the merchant, introduced as a witness by the plaintiff, testified on cross-examination that he used for kindling fires in his stove some of this liquid which was delivered to him on January 7th and out of which came the liquid here in question; and that there was no peculiar action from it. He said that he had poured kerosene on live coals and found that it would then explode. His wife testified that she used some of this liquid for kindling fires by putting it in a stove and putting a match to it; and that it did not act differently from kerosene. She said that she had not for a long time tried kindling a fire by pouring kerosene on a fire after the fire was in the stove, for she tried it once and it exploded. Mr. Green testified that he sold to certain persons some of this same liquid that was delivered on the 7th, they being persons hereinafter mentioned who testified that they used it in kindling fires without anything unusual happening. This testimony of Mr. and Mrs. Green may be treated as insufficient in itself to take the case from the jury, on account of their interest in the litigation; but it is not contradicted. But certain other witnesses testified to facts which are not contradicted and which relate to this very subject. Their testimony in itself shows that between Saturday, January 7th, and Tuesday, January 10th, they bought from Green some of this liquid and used it without the happening of anything unusual or inconsistent with the behavior of kerosene. The substance of the testimony of these witnesses is as follows: Everett Ctrnningham testified that he bought it on Saturday or Monday; that he used it in kindling fires in a coal stove, putting in kindling, pouring oil on it, and lighting it with a match; also used it for cooking on an oil stove. He said that he had poured kerosene on a fire already started and it would “fame up, blow up.” Mrs. Moss Hale testified that on Saturday before the accident she bought a gallon from Green and used it for kindling fires in her cooking stove and in a lamp; that she had been accustomed to *673 kindling fires with kerosene; and that this liquid acted just like any other kerosene, giving no trouble. Mrs. J. C. Davenport testified that a few days before the accident she got oil from Green and used it in lamps and in kindling fires; that it burned just like any other kerosene; nothing unusual happened. It appears without contradiction that just after the accident a sample of this liquid, purchased and used by Mrs. Davenport, was taken and later tested by Mr. Keenan by scientific test and found to meet all the requirements for safety prescribed by the state laws. John Bess testified that two or three days before the accident he bought oil from Green and used it in making fires and in the lamps; that it acted like all other kerosene he had ever used. Mrs. J. B. Greek testified that two or three days before the accident she purchased kerosene from Green, used it in her cooking stove, and it acted like all kerosene which she had otherwise used, giving no trouble. Ruth Allison, one of the young ladies who occupied the rooms with the deceased, testified that after they got this one gallon of oil on Monday, she used it several times in kindling fires and she did not notice anything .peculiar about the way it burned from other oil which she had been using, except the one time on Monday evening when she said that when she used it in kindling a fire in the heater it flashed up, though she didn’t notice that it burned faster than any other kerosene. Mr. Green testified that he sold to all these persons liquid out of the fifty gallons which were delivered to him on Saturday, January 7th. He sold at least twenty gallons of it between Saturday and Tuesday, when it took twenty gallons to refill it. He produced a list of items of charge from his tickets or books and they purchased as follows: Everett Cunningham, five gallons on the 9th; Henry Hale, father of Mrs. Moss Hale, one gallon on the 9th; John Bess, one-half gallon on the 9th; Mrs. J. C. Davenport, one gallon on the 9th. This was on Monday, when Mr. Green sold out of the same supply the one gallon to the deceased young lady and which was being used by her when the accident occurred. Mr. L. H. Stanton testified that he represented the Standard Oil Company in McMinnville in January, 1933; that the fifty gallons of kerosene sold to Green on Saturday, January -7th, came out of a carload that was received by him in November, 1932; that before this was unloaded he took, placed in a brown bottle, and preserved, a sample of that liquid and delivered it to Mr. Keenan during the trial, to be tested. Mr. Keenan testified that he tested this sample and that it was kerosene, meeting all the safety requirements of the state; that it was not contaminated with gasoline or any other more volatile substance. During the trial, and before the jury, Mr. Keenan *674 made a demonstration of tlie effect of pouring kerosene on a piece of charcoal that had been heated but had cooled down to a point where he could take it in his hand. He lighted a match, but the flame did not touch the charcoal. The vapors became ignited. This proved indisputably that kerosene when heated above 115 degrees, or thereabouts, will throw off an inflammable vapor. The testimony of these witnesses is neither contradicted, impeached, nor discredited by any other testimony. It is well settled that such testimony must be accepted as true and the jury cannot reject it. If it is not contradicted, impeached, or discredited, it cannot be permitted to be discredited or disregarded arbitrarily or capriciously. Frank v. Wright, 140 Tenn., 535, 205 S. W., 434; Biggs v. Johnson, 1 Shan. Cas., 622, 623; De Kalb County v. Tennessee Electric Power Co., 17 Tenn. App., 343, 350, 67 S. W. (2d), 555. If such testimony is discredited in any of the modes recognized by law, the fact is not established as a matter of law for purpose of a motion for peremptory instructions. Welch v. Young, 11 Tenn. App., 431. The testimony of these witnesses who used this liquid is clear, uncontradicted, and is not discredited. There is no evidence that in using this liquid they poured it on live coals. As aforesaid, some of them testified that they had poured kerosene on live coals of fire and it flared up and exploded. In no view of it can this testimony of these witnesses be regarded as unreasonable. We find no other evidence in the record of facts or circumstances which easts suspicion upon the truthfulness of this testimony. Now while the jury are warranted in drawing fair and reasonable inferences from the facts and conditions shown, it is only from those shown, and not from those imagined or inferred, that such inferences could rightfully be drawn. To prove a theory by circumstantial evidence so that it may be accepted as a fact proved, the known facts relied on as a basis for the theory must be of such nature and so related to one another that the only reasonable conclusion that may be drawn therefrom is the theory sought to be accepted. Chicago, R. I. & P. Railway Co. v. Rhoades, 64 Kan., 553, 68 P., 58. 'Every case of this kind must be determined upon its own facts. In our opinion the verdict in this ease is based upon mere guess or conjecture. The facts relied upon do not tend reasonably to exclude any other cause, but, on the other hand, it is equally probable that there was nothing but kerosene in the can, that the explosion occurred in the stove from pouring it on live coals, and that the can burst from vaporization of kerosene in it. We are therefore compelled to apply the rule that the verdict based upon one theory of the cause, when under the facts another theory of cause was equally probable for which the defendant was not liable, cannot be allowed to stand. Nashville Railway & Light Co. v. Harrison, 5 Tenn. App., 22, and eases there cited. *675 Tbe motion for directed verdict should have been sustained. It results that the judgment must be reversed, the verdict set aside, and the action dismissed at the cost of the administrator. Faw, P. J., concurs. CROWNOYER, J., dissents, holding that there was sufficient evidence for the ease to be submitted to the jury, and that its verdict should be binding on this court.