Reversing.

On May 11, 1934, appellee, operating a lunch and soft drink stand in Elizabethtown, was severely injured by the explosion of a beer keg in the manner later described. He filed suit against the Jefferson Woodworking Company and appellant, charging that the woodworking company had negligently manufactured and furnished a keg made of defective material, and that the brewery, knowing the proposed use of the keg, that it would be subjected to heavy pressure from the inside, negligently, without inspection, filled the keg with beer and furnished it to appellee, and further that the brewery made the beer sold to him "in a manner and to an extent * * * that said beer fermented after it came into his hands," and that such fermentation contributed directly to the explosion.

Defendants by answers denied the allegations of the petition and pleaded contributory negligence, completing issues which were after proof, submitted to the jury under instructions which are not subjected to criticism. The jury exonerated the manufacturer of the keg, but returned a verdict for $1,500 against appellant. Appellee sought recovery of $3,000.

The woodworking company admits the manufacture of the keg, and the brewery admitted making the beer which was sold to a distributor and by it sold to appellant. Reversal is sought on the ground that the verdict of the jury is not sustained by the evidence *Page 310 and is contrary to law, and that the evidence was not sufficient to take the case to the jury.

Appellant contends: That as to the keg, appellee's evidence fails to show (a) that the keg was defective at the time it left the hands of the brewery or (b) that any alleged defect could have been detected by the brewery upon inspection. That as to the beer appellee's evidence fails to show (a) that the beer was negligently brewed, or (b) that it generated, or could have generated by fermentation the pressure which caused the blowing out of the head of the keg, and that the doctrine of res ipsa loquitur had no application.

The jury had before it parts of the keg and the apparatus which controlled the flow of the beer from the keg, and some parts of the gas tank, but these exhibits are not presented here, hence it is difficult to accurately describe the combined outfit, more so because of the numerous exemplifying expressions of witnesses, such as "here" and "there," and "at this or that point." However, it appears that there was in use a tank of carbon dioxide gas, which supplied pressure to force the beer into and through the coils to and out the faucet. At the top of this tank was a square head valve turned with a small removable wrench, which controlled the flow of gas from the tank through a short metal pipe to a gauge, thence to a rubber hose, thence into a metal pipe which fitted with the siphon or spigot into the head of the beer keg. The tank supplied gas to two kegs of beer.

At some distance from the gauge the rubber tube connected with two rubber tubes, one to supply each keg of beer; in each of these there was a stopcock, which closed the flow of gas to its particular keg. The spigot, by a double tube arrangement in the head thereof, permitted the gas to enter the keg. It was forced through a bung hole and went to the bottom of the keg, the beer entering the spigot through the holes near the lower end, thence into the coils, its flow being controlled by still another valve.

The accident occurred on May 11, 1934. Appellee had left his place of business the evening before prior to the time his helper, Payne closed the stand. Early on this morning appellee opened up, went first to *Page 311 the beer stand, packed ice around the coils, and turned the valves which permitted the beer to run into the coils. He then looked after his hot-dog stand, and in passing the beer stand first noticed beer spewing or foaming out around the metal tube or spigot. Believing that something was wrong, he reached to turn off the valve in the line which carried the beer from the spigot into the coils. He does not recall whether or not he turned it off, because at this point the head of the keg blew out with such force as to break appellee's arm, otherwise injuring it, and throwing splinters in his face. The force was so great as to throw parts of the beer keg through the ceiling. He says that the valves controlling the flow of gas into the kegs were closed. He does not claim to have looked at the valve on the head of the tank, and on cross-examination says that he only glanced at the others.

Immediately after the explosion, appellee was taken to the office of a near-by doctor who rendered him first aid, and while there he asked Riney to go back and see if the gas was cut off and this witness did go back to the stand where quite a crowd had gathered, and says he found the gas cut off.

Appellee in brief urges that the evidence satisfactorily shows that the explosion was not, nor could it have been, caused by pressure from the gas tank. The jury eliminated from the case the contention that the keg was made of inferior timber, which conclusion obviates necessity of discussion on this point, or on lack of inspection, except incidentally. In speaking of alleged defects in the material, the appellee in brief says:

"The jury decided that question in favor of the defendants, thus holding that the keg was sufficient and that the injury came from some other source,"

contending that the improper manufacture of the beer resulted in fermentation, thus creating gas to such extent as to force the head from the keg in the manner described.

Mr. Goranflo, a plumber and steamfitter, said that he installed a gas apparatus for appellee some time in 1933, and supposed it was the same one in use at the time of the explosion. He set the regulator at 8 *Page 312 pounds' pressure; no more gas could go through the line, "unless the regulator was out of condition." "If the pressure is needed you turn the regulator to the right and to reduce you turn to the left. It is easily turned." He never saw it after it was installed.

Witness Payne gave a detailed description of the outfit used in supplying gas pressure, and just how the appliance was rigged. He says the gauge was set at 8 pounds, and, as far as he knew it had not been changed since installation by Goranflo. Payne tapped this keg of beer at 3 o'clock in the afternoon of May 10th. It was known as a "half-barrel" and contained 15 1/2 gallons. He says when he gave this keg gas he would turn it off down at the gas tank, and also at the stopcock. He had never noticed any leak in the valves up to the time of the explosion; he put on gas twice during the afternoon and closed the stand at 12 o'clock that night. When he came down immediately after the explosion, he mopped up the floor and wrung beer out of the mop several times; mopped up as much as 2 or 3 gallons. Witness had never used apparatus of this kind prior to working for appellee. He says when he quit work the gas was off; that he always turned the gas off because of fear of a leak in the gas line. Goranflo had advised him to do so. As to the amount of beer that was in the keg when he closed, Payne says he had only sold out 2 or 3 gallons; appellee says that there were only 2 or 3 gallons left in the keg after the day's sales. Riney says that there was about 8 inches of beer, or from 2 1/2 to 3 gallons left in the keg. Appellee and Payne said there appeared to be nothing wrong with the beer; that it looked clear, had the proper taste, and there was no complaint from the customers; it drew out the usual way. It is shown that the same outfit was used some time after the explosion, but was finally supplanted by a device that pumped air into the keg.

Several witnesses stated that after the explosion there was a white or blue fog or haze appearing in the keg. None of the beer from the keg was saved nor was any examination made to ascertain whether or not it was sour or had set up fermentation.

On the part of appellant (and its codefendant below) it was shown by witnesses experienced in making *Page 313 beer kegs that the material from which this particular keg head was made was of good oak; that the manufacturers tested all their kegs before sale, subjecting them to 40 pounds' direct pressure. After a keg is furnished the brewery, it is inspected; this keg had been in use some time prior to May, 1934. After use they were inspected by coopers before refilling. In this case inspection was made by a cooper who had been in the business for many years, examining the heads, staves, and hoops for defects. There was also an inner inspection for cleanliness. While no one of the inspectors could identify this particular keg, it was shown that their duties required them to, and they did inspect every keg bought and used by the brewery before, and more particularly after, each use.

Testimony showed that beer kegs are strongly made of extra heavy material so that refrigeration would be maintained; weight and bulk are required because of rough handling of the kegs. The average life of a beer keg was said to be about 10 years; coopers had known some to last for as long as 40 years. As to the quality of the beer, it will be noticed from appellee's proof that no witness says there was any fermentation of the beer; no circumstance save the foaming before, and a blue or white fog noticeable in the keg after explosion, is put forward. There is a lack of evidence that fermentation of beer in any stage will produce a spewing or foaming, which would differ from that caused by gas pressure. On the other hand, it is shown that if there was fermentation the beer would bubble rather than spew or foam, and no vapor would be created.

There is testimony of the brewmaster who made the beer, fortified by that of other experienced brewers, as to its treatment in manufacture. Of course, no one person could be produced who could identify the beer which was in this keg, but appellant's master brewer, who had been brewing beer for 20 years, and who had been making beer for appellant since March, 1933, described in detail the process, and from his testimony we gather that this brewery makes 500 barrels at one time. Starch grain, the basis, is converted into sugar during the mashing process, then the liquid is separated from the mash, cooled, and put into a blending *Page 314 vat, and at this point yeast is added and the process of fermentation is begun, being completed in about 12 days. The mass then goes into storage, remaining from 60 to 90 days, then taken from storage, carbonated, filtered, and put into finishing tanks, from whence it is put into bottles, kegs, and in some cases tin containers. The advance and completion of fermentation is indicated by the constant use of a saccharometer which definitely determines when fermentation has ended, and no beer is put into containers until the fermentation is completed. The process of filtration removes all traces of yeast. It is proven by the brewmaster that each unit into which the 500 barrels is divided for market is of precisely the same grade and texture. There would be no difference in quality in any one of the 500 barrels conditioned at one time. It was shown that during the month of May, 1934, there was no complaint from any customer of any beer sold by appellant. It was also shown that at the time the keg left the brewery the added inside pressure on a keg was 8 to 10 pounds, which would increase slightly if the weather was warm. The pressure would increase if the keg were placed where the heat would materially affect it, but not sufficiently to explode the keg; that, as the beer was withdrawn the pressure reduced, and if only three or four gallons remained in the keg there would be little, if any pressure from the beer and the gas contained in it; perhaps 2 or 3 pounds.

The brewmaster also testified that beer which was normal, drew, tasted, and looked all right at 4 o'clock could not possibly go into secondary fermentation by 7 o'clock the next morning. It was shown that, if the beer had not completed fermentation when leaving the brewery, or if fermentation should set up, it would be unpalatable; would have a bitter taste easily detected by customers.

On cross-examination one witness testified that beer in a normal condition would not under any circumstances or conditions ferment so as to create more than "one pound or so" of pressure in a period of 7 1/2 hours; at the end of 12 hours the added pressure would be about 2 pounds, and would not be sufficient to explode a beer keg. The brewery experimented with beer, permitting it to ferment in a warm room for 2 *Page 315 or 3 weeks; it would turn sour and develop some small extra pressure.

As to the apparatus used by appellee for gas pressure, one witness who had been engaged in installation of beer equipment for 35 years visited appellee's stand some time after the accident and examined the tank and its connections. He found the screw which regulates the flow of gas into the gauge turned all the way in, which he said would indicate that one would not know what pressure he was getting. The drum was shut off, as was the stopcock. This witness then turned the drum on, leaving the stopcock off, screwed the regulator all the way out, and the "gauge hand went deliberately all the way round, showing that the gauge was defective." He also noticed that the valves used on the gas line were not carbonic gas valves, but valves for use on illuminating gas fixtures, not suitable for carbon gas, because of the vast difference in flow and pressure.

Another witness, for 38 years a dealer in beer equipment, examining an exhibited valve, or what he calls a "gas-cock", said it was made for use in connection with illuminating gas; he described it as an "old-timer." He inspected the one shown him and said it was leaky; that its use was "a very unreliable procedure; it's dangerous." He demonstrated that it was loose and said it "lacked bearing; not enough to stop leaking of carbon gas."

Counsel insists that the doctrine of res ipsa loquitur is applicable here, though further arguing that appellee is not necessarily relegated to reliance upon the doctrine above mentioned, because "the evidence satisfactorily showed that the beer did ferment," and we may as well dispose of that contention now, which we do by expressing the opinion that there was not sufficient proof that the beer was in fermentation process at all. The only proof on this point is the spewing around the socket into which the spigot was fitted into the keg, plus the blue or white haze. There is no proof that fermentation caused either of these conditions. The spewing could easily have occurred from a very slight pressure in the keg, if there had been a slight leak around the head of the spigot. Since the *Page 316 beer comes out of the faucet in a foaming condition, it would come out of a leak with the same appearance. It is not shown that fermentation would create a bubbling condition rather than foam, nor yet that, if there was fermentation, it was sufficient to cause the explosion.

So we approach discussion of the point to which appellee almost necessarily brings his case, and that is, Does the doctrine of res ipsa loquitur have such application here as would authorize the court to submit the case to the jury on the question as to whether fermentation of the beer was the proximate cause of the explosion?

The case most assuredly does not come within that class of cases in which the doctrine has often been applied, where one who furnished the instrumentality was in control thereof, though used by another; nor to that class in which there existed a direct contractual relation; nor yet to that class of cases in which the seller (by himself or through another) supplies a chattel which he knows to be inherently dangerous, or likely to become so in its ordinary or intended use. The case here clearly falls within the class of cases where the elements above set out have not been present, wherein we have held the rule to be that the doctrine is applicable only where the existence of negligence is reasonably deductible from all the circumstances, and should not prevail where on mere proof of the occurrence the causation is speculative or conjectural.

We do not agree that appellee has completely removed the possibility that the gas tank was not a factor, since, while the testimony with regard to all the valves permitting the gas to flow into the lines were closed, there was still evidence of the leaky stopcock, the proof as to faulty gauge, and some proof that the persons in charge were not operating the fixtures in a correct manner, together with the testimony of appellee to the effect that he did not look at the main valve, and only glanced at the others.

Counsel in brief says:

"The explosion was caused by a great pressure from the inside of the keg, which under the evidence could have been caused only from the gas *Page 317 from the tank or gas generated by the beer fermenting," with which conclusion we agree.

Res ipsa loquitur is only evidentiary; it is but "a general way of saying that the circumstances attendant upon an accident are of themselves of such a character as to justify a jury in inferring negligence as the cause of that accident." Wright v. Elkhorn Consolidation Coal Coke Co., 182 Ky. 423,206 S.W. 634, 638. The doctrine is only to be applied when the nature of the accident itself not only clearly supports the inference of negligence, but excludes all others, or such as might have been due to one of several causes, of or for which the defendant is not responsible. Lucid v. E. I. Du Pont de Nemours Powder Co. (C.C.A.) 199 F. 377, L.R.A. 1917 E, 182. The general doctrine and its applicability is clearly laid down in Stephens v. Kitchen Lumber Co., 222 Ky. 736, 2 S.W.2d 374, which quotes with approval Louisville N. R. Co. v. Comley, 173 Ky. 469, 191 S.W. 96, L.R.A. 1917 C, 978, and Louisville N. R. Co. v. Mink, 168 Ky. 394, 182 S.W. 188.

Appellee relies mainly on what is referred to as an exception to the general rule or rules, which we have stated above, which exception is set out in Payton's Adm'r v. Childers' Electric Co., 228 Ky. 44, 14 S.W.2d 208, 209, in which we said:

" 'An act of negligence of a manufacturer or vendor [of an article] which is imminently dangerous to the life or health of mankind, and which is committed in the preparation or sale of an article intended to preserve, destroy, or affect human life, is actionable by third parties who suffer from the negligence.' [Huset v. J. I. Case Threshing Machine Co. (C.C.A.) 120 F. 865, 61 L.R.A. 303.]

"The early cases limited this exception to things in their nature destructive, such as poison, explosives, and deadly weapons, but the tendency recently of the great majority of the courts has been to extend this exception to include any article imminently dangerous, whether inherently so or not, and we think the exception as extended is sound in principle. * * * 'If the nature of a thing *Page 318 is such that it is reasonably certain to place life and limb in peril when negligently made, it is then a thing of danger. Its nature gives warning of the consequences to be expected. If to the element of danger there is added knowledge that the thing will be used by persons other than the purchaser, and used without new tests, then, irrespective of contract, the manufacturer of this thing of danger is under a duty to make it carefully.' "

MacPherson v. Buick Motor Co., 217 N.Y. 382 [111 N.E. 1050,1051, L.R.A. 1916F, 696, Ann. Cas. 1916C, 440].

Inherently there is no danger in a keg of beer. The fact is both parties admit, and the brewers say that in their long experience, they have never heard of an injury from the explosion of a keg of beer, or a beer keg, nor have we after considerable search been able to find a case where such an injury occurred, except in the reported case of Greeney v. Haberle-Crystal Spring Brewing Co., 190 A.D. 785,180 N.Y.S. 648, wherein the only question was whether or not Greeney, a servant, was acting in the line of duty when he failed to observe his master's direction to wash out an empty beer keg in a certain way, and instead, attached a rubber hose to it, forcing steam into the keg, exploding it and killing him. The case throws no light on the question here.

The Payton Case, supra, and citations therein would be applicable if it were shown that the beer had been knowingly improperly made, but it must be remembered that it is sought to bring the case within the exception to the rule on the mere circumstances of the foaming of the beer and the presence of the mist or haze, which do not create an inference that the beer was in process of fermentation.

In the Payton Case, supra, the only question before the court was the sufficiency of an amended petition, which in substance charged that certain machinery was so improperly constructed as to be a source of imminent danger by reason of the presence of a deadly and dangerous electric current in use, and that it was known by defendant that the device was imminently and inherently dangerous. In the Olds Motor Case, *Page 319 145 Ky. 616, 140 S.W. 1047, 1049, 37 L.R.A. (N.S.) 560, Ann. Cas. 1913B, 689, cited in the Payton Case, supra, it was said:

"There is convincing evidence that the box upon which the rear seat rested was not only insufficient in itself, but was insecurely fastened, * * * and that the manner in which it was fastened was unsafe and dangerous in the extreme."

The Threshing Machine Case, also cited in the Payton Case, was one testing the quality of "a pleading, the court saying:

"But upon the facts alleged in this complaint, the act of delivering it to the purchaser with a knowledge and a concealment of its dangerous condition was flagrant disregard of the rule."

In the case at bar there is no charge of knowledge or concealment.

In the Buick Case, cited in Payton, supra, the court not only found proof of a defective wheel, but evidence "that its defects could have been discovered by reasonable inspection, and that inspection was omitted." Here it is shown that the keg of beer in question was taken from the brewery a day or two before the accident; it was delivered to appellee on the same day. The proof is conclusive that the beer was in a normal condition up to midnight before the morning of the explosion.

Treating the partly filled keg as standing free from connection with the gas tank brings it in the class of containers dealt with in the bottle explosion cases, of which the last in our jurisdiction is Loebig's Guardian v. Coca-Cola Bottling Co., 259 Ky. 124, 81 S.W.2d 910, 911, and where a boy had lost the sight of his eye by the explosion of a Coca-Cola bottle on a warm day in September. The court below found for defendant, and judgment was affirmed by this court, distinguishing Coca-Cola Bottling Works v. Shelton, 214 Ky. 118, 282 S.W. 778, because in that case it was shown that numerous bottles of Coca-Cola had exploded the same day, by reason of which fact we said negligence was reasonably inferred. The court found the Loebig *Page 320 Case to be analogous to Stone v. Van Noy R. News Co., 153 Ky. 240, 154 S.W. 1092, 1094. In discussing the res ipsa loquitur doctrine in the Loebig Case we said:

"Since the principle applies only to cases where the existence of negligence is a more reasonable deduction from the circumstances, it should not be allowed to prevail where, on proof of the occurrence, without more, the matter still rests in conjecture alone. * * * In the instant case, we are still left to conjecture as to the cause of the defect in the bottle and its contents or whether it was such a defect as might have been discovered by a more thorough inspection."

In the case of Stone v. Van Noy R. News Co., supra, we said:

"There is no evidence in this case tending to show that an ordinary bottle of pop is intrinsically or inherently dangerous to health, life or limb. On the contrary, millions of these bottles are sold every year without such an accident as occurred in this case. Manifestly, such bottles do not belong in the category of poisons or dangerous drugs that are labeled as containing innocent or harmless ingredients. That being true, can this case be said to fall under exception 2, as above stated? There is no proof in this case that the bottle itself was too heavily charged. Nor is it shown that the Merchants' Company's bottles frequently exploded. * * * When the question is one of negligence or no negligence, it is the well-settled law that where the evidence is equally consistent with either view — the existence or non-existence of the negligence — the court should not submit the case to the jury, for the party affirming the negligence has failed to prove it. * * * Having failed to prove negligence, the trial court properly directed a verdict in favor of the defendants."

We think the line of cases embodying the principles laid down in the Loebig Case apply here.

In so far as it is claimed appellee's injury was attributed to the explosion of the keg from pressure by fermentation, we conclude that the proof adduced was *Page 321 not sufficient to justify the court in submitting the question to a jury. Under appellee's proof it would be left to speculate on the causation. Having thus determined we are of the opinion that the judgment should be reversed, with directions to award appellant a new trial consistent with views expressed herein.

Judgment reversed.

Whole court sitting.