Sutton, J.
The plaintiff brought suit against the defendant for damages on account of injuries alleged to have been sustained from the explosion of a bottle of coca-cola which, in the performance of his duties, he had placed in an ice-box in the restaurant of his employer. He alleged that the bottled drink had been obtained by his employer from the defendant, and that he placed the bottle in the ice-box in the usual and customary manner, and was free from negligence, and that the bottle was negligently carbonated, so that when its temperature was reduced 'by being placed in the ice-box the bottle exploded, and that if the carbonation had been regulated in accordance with the advertisement of the defendant that the beverage was harmless, the explosion would not have occurred. The defendant denied the material allegations of the petition, and charged that the injuries of the plaintiff were the result of 'his own negligence. By amendment- the plaintiff alleged that the bottle used was insufficient and contained thin places, so that it was not capable of withstanding, upon handling in a normal way, the pressure of the carbonating gas, and that if it had been suited to the purpose for which it was used, and had been in proper condition, a proper charge of carbonation would not have exploded the bottle. In the trial of the case the plaintiff relied on the doctrine of res ipsa loquitur. The evidence is not here set out, but the main facts are discussed in the opinion. The jury returned a verdict in favor of the plaintiff. The defendant filed a motion for new trial, and by amendment added a ground in elaboration of the general grounds. The court overruled the motion, and the defendant excepted.
The maxim res ipsa loquitur, as was said by Judge Hill in Cochrell v. Langley Manufacturing Co., 5 Ga. App. 317, 322 (63 S. E. 244), “has been a prolific inspiration to much useless and .wasted juridic erudition,” and it was added in the same case: “Practically, as we said in [Monahan v. National Realty Co.], 4 Ga. App. 680 (62 S. E. 127), the doctrine is simply a rule of evidence, which permits an inference to be drawn from proved facts. It furnishes a working basis for reasonable hypothetical conjecture, and gives scope for legitimate reasoning by the jury. The philosophy of the doctrine is stated in section 5157 of the Civil Code [Code of 1933, § 38-123]. ‘In arriving at a verdict, the jury, from facts proved, and sometimes from the absence of coun*575ter-evidencé, may infer the existence of other facts reasonably and logically consequent on those proved.’” This doctrine has been sanctioned for many years in other jurisdictions, but strangely enough it did not find recognition in any of the reports of this State in personal-injury cases until Chenall v. Palmer Brick Co., 117 Ga. 106 (43 S. E. 443). That case involved a suit against the master by a servant, and notwithstanding that relationship the Supreme Court held that the maxim could be applied. It was there said, with reference to deductions that might be made by the jury, and quoting from the old English case of Scott v. London & St. Katherine Docks Co., 3 Hurl. & C. 596: “There must be reasonable evidence of negligence; but where the thing is shown to be under the management of the defendant or his servants, and the accident is such as in the ordinary course of things does not happen if those who have the management use proper care, it. .affords reasonable evidence, in the absence of explanation by defendant, that the accident arose from want of care.” That deduction that might be made by the jury under this maxim of res ipsa loquitur has been referred to in many jurisdictions as a presumption of law, and in many others it has been designated as an inference of fact; for a full and interesting discussion as to which the reader may consult the elaborate annotations following the report of the ease of Glowacki v. Northwestern Ohio R. &c. Co., in 53 A. L. R. 1486. That it is not a substantive rule of law but only a rule of evidence was pronounced by the Supreme Court of this State when the Chenall case was again before that court in 119 Ga. 837, 842 (47 S. E. 329), where it was said: “The maxim res ipsa loquitur is simply a rule of evidence. The general rule is that negligence is never presumed from the mere fact of injury; yet the manner of the occurrence of the injury complained of, or the attendant circumstances, may sometimes well warrant an inference of negligence. It is sometimes said that it warrants a presumption of negligence; but the presumption referred to is not one of law but of fact. It is, however, more correct and less confusing to refer to it as an inference rather than a presumption; and not an inference which the law draws from the fact, but an inference which the jury are authorized to draw.” These principles were restated and applied to questions of negligence in Atlanta Coca-Cola Bottling Co. v. Danneman, 25 Ga. *576App. 43 (102 S. E. 542); Atlanta Coca-Cola Bottling Co. v. Shipp, 41 Ga. App. 705 (2) (154 S. E. 385); Atlanta Coca-Cola Bottling Co. v. Dean, 43 Ga. App. 682 (160 S. E. 105). As distinguished from a presumption of law, which may be overcome, the inference which the jury is authorized to draw as to negligence, when once drawn, is never overcome. The only inquiry thereafter is: Whose negligence is it? And the defendant may satisfy the jury that it was not his negligence that caused the injury, although the cause of the occurrence, as well said by Judge Russell in Sinkovitz v. Peters Land Co., 5 Ga. App. 788, 795 (64 S. E. 93), “might still be involved in unsolvable mystery.”
The above would seem to make unnecessary any consideration of presumptions of law or of how they may be overcome, and to leave us to examine the present case with reference to the doctrine of res ipsa loquitur authorizing the jury to draw cm inference of negligence. In the Sinkovitz case, supra, which involved a suit brought by one not a servant of the defendant, Judge Russell, now Chief Justice of the Supreme Court of this State, said: “Upon the second appearance of the Chenall case in the Supreme Court (119 Ga. 837), the particular question involved was how far the application of the maxim res ipsa loquitur was affected by the fact that the plaintiff was a servant of the defendant, and thereby had assumed the risks incident to his employment and to the negligence of his fellow-servants. In so far as the opinion in the later case, delivered by Justice Cobb, appears to confine or limit the application of the maxim, it must be borne in mind that the lehrned judge was dealing only with that specific point. But there was no retraction or modification of the original holding (where the relation of master and servant does not exist), that ‘ordinarily, extraordinary and external causes may be treated as the exception, to be established by the defendant Judge Russell then laid down the burden which rests on the plaintiff, quoting from the Chenall case when it first appeared in the Supreme Court, 117 Ga. 109: “All that the plaintiff should be required to do in the first instance is to show that the defendant owned, operated, and maintained, or controlled and was responsible for the management and maintenance of the thing doing the damage; that the accident was of a kind which, in the absence of proof of some external cause, does not ordinarily happen without *577negligence. When he has shown this he has cast the burden upon the defendant, who may then proceed to show that the accident was occasioned by vis major, or by other causes for which' he was not responsible.” Headnote 2(b) is as follows: "If the extraordinary character of the occurrence is sufficient to raise an inference of the negligence alleged, a prima facie case is established, and the burden of disproving negligence, especially in a case where the parties do not sustain to each other the relation of master and servant, is cast upon the defendant, to disprove negligence upon his part; this for the reason that it is more particularly within his power to explain the character and condition of the instrumentality which may have occasioned injury than within the power of the injured party.” Headnote 2(c) is: "In the absence of any satisfactory explanation that the occurrence was accidental or providential, or other sufficient explanation, if something unusual happens in respect to a defendant’s property or to something over which he has control,- whereby the plaintiff is injured, and the natural inference on the evidence is that the unusual occurrence is due to the defendant’s act, the occurrence, being unusual, is said to speak for itself that such act was negligence. Bahr v. Lombard, 53 N. J. L. 233 (21 Atl. 190).” Thus it is seen that the burden upon the plaintiff, when the maxim res ipsa loquitur is applicable, is short of being required to prove actual negligence by direct evidence, and that he may make a prima facie case by proving an unusual occurrence under circumstances referred to in the above-quoted headnotes, and thereby cast upon the defendant the burden of exculpating himself by showing to the jury that the accident (using the word in its loose sense) was occasioned by vis major or by other causes for which he was not responsible. The defendant is not freed from responsibility, therefore, by showing that the negligence might have been that of an external agency. He must show that he did not act negligently, or that the accident was due to causes for which he was not responsible.
But can the doctrine of res ipsa loquitur be applied in a case like the present one, where the injury is alleged to have been caused by the explosion of a bottle of carbonated beverage which was too highly charged with gas or too weak to stand a normal pressure ? It is true that in other jurisdictions the decided weight of authority is to the effect that the maxim res ipsa loquitur is *578not applicable to the breaking, bursting, or exploding of a container in which a commodity, ordinarily harmless, is sold. See annotations following Grant v. Graham Chero-Cola Bottling Co., 4 A. L. R. 1090, where many cases are discussed, including some from Southern States. But in Georgia, in Payne v. Rome Coca-Cola Bottling Co., 10 Ga. App. 762 (73 S. E. 1087), it was squarely held that “Where an action is brought to recover damages for an injury caused by the explosion of a bottle, the contents of which were manufactured, bottled, and sold by the defendant as a harmless beverage, an inference of negligence on the part of the manufacturer arises, when it is shown that all the persons through whose hands the bottle had passed were free from fault, and that the condition of the bottle and its contents had not been changed since it left the defendant’s possession.” In that case a bottle of coca-cola, manufactured and sold by the defendant, exploded, and fragments of glass flew into the plaintiff’s eye and destroyed the sight. The plaintiff alleged that the water in the bottle had been charged with carbonic acid gas, and that the explosion was due to the fact that the bottle was too highly charged with the gas by the defendant. A nonsuit was awarded, and the plaintiff excepted. This court said: “Granting, for the sake of the argument, that prima facie inferential negligence will be imputed to the person who sold the bottle to plaintiff’s brother, or to the brother himself, the inference is completely rebutted when it affirmatively appears, as it does here, that neither was at fault, that neither handled the bottle improperly or did anything to change the condition from that in which it was when received. Since for every effect there is a cause, where negligence exists some one must have been the responsible author. If he can be found, it is right that he should pay the penalty. The bottle exploded. Inferentially some one was negligent. It was not Cook, the last vendor of the bottle, nor the plaintiff’s brother, nor the plaintiff, nor yet Barnett, because they all stand exonerated by direct or circumstantial evidence of their freedom from fault. But the inference of negligence remains, and some one is prima facie to blame. By a process of elimination we get back to “the manufacturer who set the dangerous agency in motion, and upon whom the blame ought inferentially to be fastened. It is certainly no hardship to require at the manufacturer’s hands an *579explanation of the occurrence, that the jury may say whether it, like the other persons who handled the bottle, has been exonerated. If a manufacturer should sell to a jobber a gun, and, after passing through the hands successively of the wholesaler and retailer, it finally reaches the marksman, and explodes in his hands while being used in the ordinary and usual manner, and injury results, it is plain there was a defect in the gun. Somebody ought to be responsible. Conqéde that inferentially it could be said that the marksman must have done something to the weapon to cause it to explode, if he disproves this, and the retailer, the wholesaler, and the jobber all in turn show that they kept and handled the gun in the usual way, and did nothing to change its condition, the inference of negligence would be shifted back upon the manufacturer, who put the weapon of destruction in circulation with his endorsement that, when used in the ordinary and usual manner, no harm would come to him who used it. In such a case it would be no answer, when the maxim that the thing spoke for itself is invoked, to say that when the injury resulted, the thing was not in the possession, power, or control of the manufacturer.” Although that decision dealt with the question of a nonsuit, it was unequivocally ruled that the doctrine of res ipsa loquitur might be applied in a ease where a bottle of carbonated beverage explodes, and showed what sort of evidence from the plaintiff will make a prima facie case, requiring the defendant to exculpate himself as in other cases where the doctrine has been held. It is true that the court added: “As to whether an inference of negligence would arise against the manufacturers upon mere proof of the explosion, without more, we express no opinion,” but this, of course, had reference to the necessity of affirmative proof that none of the intermediaries had been negligent.
In Atlanta Coca-Cola Bottling Co. v. Danneman, supra, it was held: “But where the event is unusual and extraordinary in its nature, and there is nothing to indicate an external cause, but the peculiar character of the accident is sufficient within itself to indicate that it must have been brought about by negligence on the part of some one, and where the most reasonable and probable inference which can be rationally drawn from the happening of such an event is that it would not and could not have taken place had not the person charged with furnishing or maintaining the instru*580mentality causing the accident been guilty of the particular acts or omissions set forth by the plaintiff as constituting the actual cause, then the jury is authorized to apply the rule of evidence known as the doctrine of res ipsa loquitur, in determining whether or not the accident must’ have been thus occasioned.” See also Atlanta Coca-Cola Bottling Co. v. Shipp, supra. In the Monahan case, supra, the plaintiff was injured by the fall of a heavy window in a new building, and it was said by the court that neither the trial judge nor the appellate court could say that the jury were not authorized to apply the maxim res ipsa loquitur in that case against the landlord; and on the question of a prima facie case being made, it was said: “In the present case there was a heavy window in a new building, which fell, certainly without any fault of the plaintiff, and according to some of the testimony, without any apparent cause. In determining whether it was an accident for which no one would be liable, • or whether it was due to negligence on the part of the landlord in the original construction of the building, it is to be borne in mind that where injury is shown, and it is also shown that the party injured is free from fault, and there is no proof of any external cause, a prima facie case of negligence may be raised, because the party managing or controlling the agency which created the injury is naturally the one responsible, and the burden is thereby placed upon the defendant to show proper original construction. . . Whether the explanation that the chains used in the windows in the defendant company’s building were improved appliances, similar to those used 'in two other modern buildings in the City of Savannah, was sufficient to rebut the- inference of negligence in construction, due to the unaccountable fall of the window in question, which was reinforced by evidence of continual recurrence of similar accidents all over the building and the fact that there was never any test of the tensile strength of the window chains, was a question for the jury.” If, as we hold, the maxim res ipsa loquitur is applicable under the facts of the present case, it follows that whether the defendant’s explanation that the machinery used in bottling coca-cola and introducing gas into-the bottles was improved, and modern and equal to that used by the best equipped bottlers, was sufficient to rebut the inference of negligence due to the unaccountable explosion, was a question for the jury.
*581The judge properly charged the jury that they might apply the maxim res ipsa loquitur in the present ease, and that their verdict should be based on a preponderance of the evidence. It can not be said that the jury did not apply the doctrine in returning a verdict for the plaintiff. We now proceed to ascertain if that verdict was authorized. The evidence showed that the plaintiff was injured by a fragment of glass which flew from a bottle of coca-cola which exploded'as he was removing his hand from an ice-box into which he had placed several bottles of the beverage. They were taken from a case near the ice-box, and had been purchased directly from the defendant. It was shown that on the morning of the injury the temperature was such that a bottle of the beverage taken from the case, the bottle being properly constructed and carbonated, would not explode merely because of the cooling process when introduced into the ice-box. It was testified by an expert witness of the defendant that if such a bottle, at the temperature that existed at the time, were placed in ice-water or in an ice-box on top of ice, without anything between the bottle and some other object, and if an explosion occurred upon the bottle being subjected to the cooling process, the explosion would be due to a defect in the bottle or to excess earbonation. For six years the plaintiff had been handling bottles of the beverage from the case to the ice-box; and he testified that on the occasion of the injury he handled them in a very careful manner, that he did not strike'any bottle against another or against any piece of ice, and that the explosion did not' occur until he had leaned over to close the lid of the ice-box. One of the defendant’s employees testified that he made deliveries to the plaintiff’s employer, and that there was no opportunity for any tampering with the bottles from the time they left the defendant’s premises until delivered to the customer. Therefore it can not be said that the jury was not authorized to find that the bottles had not been negligently handled by any intermediaries, and that the plaintiff was himself free from fault. That being true, a prima facie case was made against the defendant. The jury would next consider whether the defendant had exculpated itself by showing freedom from negligence, or that the injury was the result of vis major or causes for which it was not responsible. Evidently the jury did not find that the defendant had satisfactorily explained the occurrence, *582and in our opinion it was justified in so finding. Notwithstanding the testimony of the defendant’s witnesses as to the operation of its plant, its modern machinery, and its usual care in bottling its product, it was testified by its general manager that “The water is carbonated before it goes into the bottle, and is put in by an automatic measuring machine, measuring five ounces for each bottle, making six ounces as the total contents of the bottle. There is a possibility, but slight, that this machine may get clogged up and out of gear.” If the device on the machine, intended to prevent excess earbonation by dissipating into the air any excess of gas, becomes clogged, the only reasonable inference to be drawn from a clogged condition is that under such circumstances excess earbonation would result. Obviously, the very purpose of placing the device upon the machines was to prevent just that thing. It is not shown that the defendant, as was its burden to do, established any vis major or external causes for which it was not responsible.
Great stress is laid by the defendant on the fact that the bottles were purchased from an efficient and reputable manufacturer; but the best of bottles will naturally explode from super earbonation. To prevent that eventuality, proper machinery is necessary, and likewise adequate inspection is necessary; and yet there is nothing in the record which shows any inspection at any time of the device upon which so much depended. In the absence of any information on that subject, and from the admission of the defendant’s general manager, the jury could reasonably have inferred that the plaintiff had been injured by the explosion of a bottle which had been filled and carbonated at a time when the automatic device, so called, was clogged, and when an excess of gas was introduced with water into the bottle, or that the bottle was weak and insufficient to withstand a normal and proper pressure. Because of the above facts, as shown by the record, we are of the opinion that the jury was authorized to apply the doctrine of res ipsa loquitur, to find that the plaintiff was free from fault, that no external cause of the operation was shown, that no intermediaries had handled the bottle from the time it left the possession of the defendant, and that the defendant did not show that it was not its negligence that caused the explosion, or that it was due to vis major or causes for which it was not responsible. Accordingly, the verdict will not be disturbed.
*583
Judgment affirmed.
Stephens, P. J., concurs. Felton, J., dissents.
Felton, J.,
dissenting. Eobert L. Crane sued the Macon Coca-Cola Bottling Company for damages, alleging substantially that it was engaged in bottling coca-cola, buying syrup, bottling the drink and furnishing the same to the general public, claiming it to be a harmless beverage; that on or about August 7, 1935, defendant sold a number of cases of bottled coca-cola to Ben F. Lyles, the operator of a restaurant in Macon, Ga.'; that plaintiff was an employee of Lyles in said restaurant; that the coca-cola supplied to Lyles was not cooled, and it was necessary to take the bottles from the cases and place them in an ice-box; that as plaintiff, in pursuance of his duties as an employee, was taking bottles of coca-cola from the crates in which they were delivered to the restaurant and putting them into an ice-box, a bottle which he was handling in a careful manner exploded and injured him; that the defendants’ process of bottling was to place a quantity of syrup and water in each bottle and charge the mixture with carbonic acid gas for the purpose of producing an effervescence; that the bottle which exploded was prepared and furnished by the defendant to the restaurant, and was so highly charged with the gas by the defendant as to cause the explosion; that if the bottle had been properly charged the explosion would not have occurred; that plaintiff did nothing whatever to cause the explosion; that plaintiff is not in position to know or charge just how the carbonic acid gas was placed in the bottle or the quantity of gas used, but that if said bottle had been properly charged so as to render the bottle and its contents a refreshing and harmless beverage, the explosion would not have occurred; and that the bottle was one belonging to the defendant and designed in the exclusive right of the defendant. By amendment the plaintiff added to his petition the following paragraphs: “ The bottle hereinbefore referred to, which was by the defendant filled with coca-cola syrup and carbonated, was insufficient and contained thin places, so that it was not capable of withstanding, upon handling in a normal way, the pressure of the carbonating gas. Petitioner alleges that the defendant was negligent in using and distributing to its customers, to be by said customers dispensed to consumers, a bottle which was not sufficient to withstand proper and expected handling without being *584exploded by the pressure of the carbonation.” “Petitioner charges that if said bottle had been suited to the purpose for which it was used and had been in proper condition, a proper charge of carbonation would not have exploded said bottle.” The defendant’s answer denied the allegations of the petition, and alleged that it exercised all the care and diligence required of it, that it was not negligent as alleged in the petition or in any manner, and that the plaintiff could have avoided the injury by the exercise of ordinary care. The jury found for the plaintiff, to which finding the defendant excepted on the general grounds. The case is here on exceptions to the refusal of a new trial.
The portion of the plaintiff’s testimony material for this discussion was as follows: “On August 7, 1935, I was employed by Ben 3?. Lyles’ restaurant in Macon. My duties were to put coca-colas or any kind of cold drinks in this tub. . . On the date I mentioned I had put the bottles on ice, as was my duty, at about nine o’clock in the morning. I should say the cases were about two or three feet from the ice-box. . . The case of coca-cola . . was put in there the day before this accident occurred, Tuesday or Monday, I don’t know exactly which. In transferring these bottles of coca-cola from the case to the ice-box I was taking them by hand . . out of the crate, sit them over on the table, and put them down in the water. The tub contained ice and water. A bottle exploded while I was doing that. In transferring that particular bottle . . to the ice-box I did not knock that bottle against any piece of ice or any other bottle. . . I put the bottle in the ice-box, -laid it down, put my hand back on the lid to close the lid, and it said ‘bow’ and hit me in the eye. The bottle exploded and bursted all to pieces. . . That case of coca-cola came from the Macon Coca-Cola Bottling Co. . . There -was nothing to indicate to me that there was anything the matter with this bottle or why it should explode. I didn’t look for it. I did nothing out of the ordinary to cause this bottle to explode. I just put them in the ice-box just like I always did, just as easy as I could. I reached down and got a handful of bottles and put them into the box. I did pick up three or four bottles with one hand. I did not have any in my hands when the explosion occurred. I had finished putting them in the box, and was fixing to' close the lid. I did not save the pieces of that bottle. *585. . I don’t know what the temperature was that day, but we had our electric fans running since six o’clock when I opened up. The ice-box is . . just a few feet away from the kitchen. These cases of coca-cola were sitting under the counter. The fan was blowing on these particular bottles from above. After the explosion I saw it was a coca-cola bottle.” Several disinterested witnesses for the plaintiff testified that the bottle burst, and some that it was a coca-cola bottle.
O. O. Beasley testified for the defendant, in brief: “I am employed by the Macon Coca-Cola Bottling Company, and have been so employed for twelve years in the capacity of salesman. . . This truck is loaded with my colored help at the plant. There is no opportunity for any one to tamper with these bottles in any manner from the time they leave the plant until the time they are delivered to the customer. There was no unusual handling on my truck during August of last year that would cause the bottles to be affected in any way unduly. In the restaurant . . Mr. Lyles keeps the cases of coca-cola on the right side under the counter as you go in,' near the kitchen. ’ I would say the ice-box is about six or seven feet from the partition which divides the ■restaurant proper from the kitchen end. My best estimate is it is just six or seven feet from the ice-box to the partition.”
W. T. Williford, testified for the defendant, in brief: “I am general manager of the Macon Coca-Cola Bottling Company,- and am secretary and treasurer in charge of the plant. Our bottles are purchased from the Chattanooga Glass. Company, Chattanooga, Tennessee. The concern is about thirty or forty years old, and is a reputable concern. I have nothing to do with designing the bottles. I think they were designed fifteen years ago. It is the same bottles used by all bottlers of coca-cola, with the exception of one other plant in Greenville, S. C. With that exception the specifications for all coca-cola bottles are the same throughout the United States. The specifications are as follows: it is a six and one half ounce bottle, height seven and three quarter inches, and the diameter in minimum is two and one eighth inches and maximum a two and three thirty-seconds, I believe, diameter; it contains fourteen ormces of glass, that is the weight of the bottle for standard specifications, and are the specifications by which we ordered bottles from the Chattanooga Glass Company. These bot-*586ties are built to withstand a pressure of a minimum of seven hundred and fifty pounds up to a thousand pounds. The gauge on the machine registers 1000 pounds, and some bottles stand more than that; but the minimum it has got to stand is 750 pounds, to be shipped. I have no knowledge of any other bottles of coca-cola put up by the Macon plant bursting on or about August 7, 1935, or prior thereto. We have not had any trouble from breakage of bottles purchased from the Chattanooga Glass Company. We put out approximately ten million bottles of coca-cola last year. I have been with the plant since November 1, 1929, and prior to that time was connected with the Chattanooga Glass Company as secretary and treasurer, at which time they manufactured bottles for many coca-cola dealers. I have had occasion then and have had occasion since to visit other coca-cola bottling plants in this section. When I was with the glass company I visited practically all of them in four or five States, and in the past few years visited other plants. I have had occasion to observe machinery used in other plants, and to observe how they inspect the bottles and inspect the product after it is bottled, and to observe those plants with reference to cleanliness, and to find out from whom they purchased their machinery. Recently at a convention of coca-cola bottlers I have had occasion to observe the newest and most modern machinery exhibited there. Taking all this into consideration, the Macon plant compares 100 per centum. The Crown Cork and Seal Company, Baltimore, Maryland, is the manufacturer of the carbonating machine or saturator used in our plant. They are a reputable concern, and have been in business possibly longer than thirty years. In inspecting bottles in our plant as they come in and are unloaded off the trucks we try to give as good inspection as we can. When they are put into the soaker they are given another inspection; that is when they are washed; and they stay there about eighteen minutes. As they come out empty and clean they get another inspection. Then they travel around with the coca-cola syrup and gas, and the cap put on after it is filled, and there is another inspection. Of course, the persons that take them off and put them in cases like we load, they make another inspection.
“The empty bottle . . is an odd shape. The bottle is patented, and must be made by specifications in Washington by some *587commerce, or whoever handles it. The glass companies have a rule that bottles be made according to the Government in "Washington. In Macon we put one ounce of coca-cola and five ounces of carbonated water, the bottle holding six and one half ounces. Dimensions of that bottle by United States Department of Commerce, the Bureau of Standards, are the figures I am quoting you the glass factories make them by. They have got to make the bottle by that. . . The design is blown into the bottle and is an indentation in a mold. This Chattanooga Glass Company makes 210,000 (gross) a year. Last year I would say I used about two cars or about six hundred gross of bottles, considering my revolving stock I had in there. My stock on storage varies. About four million bottles went through my plant with the others and the used ones put in, what I threw away and what I put in. This bottle of the same grade, size, make, and manufacture has been in use by us for thirteen years. Some of our bottles have been in use six or seven years, the length of time depending upon the amount of business done in that territory, and the fact that some bottles are held out longer than others before being turned back to us. We expect our drink to be chilled before it is used by the consumer. We try to get dealers to chill coca-cola to about forty degrees; over half of them do not serve it that low. Some serve it at thirty-five degrees. . . I don’t know of any inspection of new bottles after being delivered before taking them olf the trucks, and do not know of any explosion of a bottle of coca-cola in my territory. There has never been any explosions reported to me. I supervise the inspection of bottles in the plant here. The Chattanooga Glass Company is a reputable and responsible company, and is a solvent company. When I know as much as I do about the making of these bottles, I think I have a right to depend on what I call a guaranty, either implied or expressed. If I had a claim against them, they are able to respond to it. I am familiar with the methods of the Chattanooga Glass Company, having been secretary and treasurer of the company, and have had occasion to visit their plant since leaving them in 1929. They use modern methods in manufacturing their bottles and modern methods in inspecting their bottles. From time to time, I don’t know how many times during the day, they put what is called a pressure test on their bottles, whatever it will stand; *588and if iliey find any bottle standing less than seven hundred fifty pounds they go looking for trouble. Also they give them hot and cold water test. As bottles come out, it takes around 2600 degrees Fahrenheit to mold that bottle. It is cooled down where it will stand up, and when impression is put in and it goes to “annerly’ beyond, and that is in sections, it takes two and a half or three hours to go in, and it slowly cools down with a register in every compartment showing what the heat test is. When it comes out of the oven it comes out in the open, from the end of the table to this. Those bottles are not cooled by hand. Every bottle is inspected from the Teer’ before it goes into cases or storage-bins; they get an order for may be eight or nine hundred bottles they won’t order out; they won’t order out but one car of bottles at one time. The rest goes into storage-room, and when they are loaded on the car I would say there is another inspection. In the last five years I have been to the Chattanooga bottle-manufacturing place about twenty-five times. Since I left them in 1929 I would say I have been there about thirty-five to forty times. I generally go through the plant when I am there. May be once out of five times I have not been through the plant. As nearly as possible, we put in one ounce of coca-cola in each bottle, measured by our machines and in other ways. The machines are 98 per cent, accurate, and were obtained from the Liquid Carbonic Company, a 'perfectly reliable and responsible company. The water is carbonated before it goes into the bottle, and is put in by an automatic measuring machine, measuring five ounces for each bottle, making six ounces as the total contents of the bottle. There is a possibility, but slight, that the machine may get clogged up and out of gear. Ninety-eight per cent, of the coca-cola bottles are uniform height, so far as contents are concerned. I will say that our machines are not infallible; but I will say this, that there is nothing a hundred per cent, perfect, and, while there may be a little variation, they are as perfect as they can be made.”
George Clark testified for the defendant, in substance as follows: “I am in charge of the weather bureau in Macon. On August 7, 1935, from 5 a.m. to 11 a.m. the hourly temperature readings are 76, 75, 78, 80, 85, 88.10, 91, 93 at noon. The machine recording this temperature is a sheltered house on the roof of the post-office building, and is in the shade. The temperature *589on the street is usually a little different, but in the majority of cases I think it would be about the same. In the sun it seems to be warmer, due to the fact that the body of any object does not radiate the heat as fast as it receives it. The actual temperature is very little different.” W. T. Williford, recalled, testified: “When I stated that I did not inspect new bottles, I meant that we did not inspect them as we loaded them from the car where we put them in stock. But they get the same inspection just like any old bottle that goes through the machine, and they are washed and get the same inspection. It is customary in the bottling trade to use bottles over and over again.”
R. L. Rankin, testified for the defendant, in substance: “I am employed as superintendent of the Macon Coca-Cola Bottling Company, where I have been employed in this capacity for seven or eight years. Prior to that time I was employed by the Macon Coca-Cola Bottling Company. Altogether I have been with them twenty-three years. As superintendent I am in charge of supervising the operations connected with the bottling of coca-cola. When the bottles are received they are taken from the train . . and laid on a table, . . where the empty bottles are inspected before they even go into the washing-machine. They go into number one tank after leaving there; with anywhere from a hundred and ten to a hundred and twenty degree temperature. This tank contains a solution of caustic soda with water; and they go from there to number two tank with about four per cent, caustic soda with a hundred and forty to a hundred and forty-five degree temperature, and out they go into number three tank which is plain water at a hundred and ten degrees. After passing out of number three tank they go on to the rinsing spindle which rinses bottles with water on the outside. Then they go on a set of spray tubes which spray water on the inside of the bottles and then go on to twenty-four spindles with brushes, where they are brushed on the inside with water. After leaving the brushes they go on to two more tubes of spraying water that shoots up into the bottle, and then it is discharged under two sixty-watt bulbs to be inspected empty as they are discharged. If a bottle had a crack or defect in the glass it would break on being immersed in these hot solutions of caustic soda and being brushed out with various brushes and washed out with these various tubes. If it did not *590break during the process, when the cold water hit it under the filler it would not fill, it would just spew out, and it would just be gone to nothing. When the bottle leaves the washing-machine it travels on this conveyor to the syrup-machine which has sixteen spouts, and those bottles are set on the machine. It passes on to another machine which is the carbonated water and gas, and goes on and discharges off; and when it discharges off this filler-machine, they travel about four feet not sealed at all, no cap; the mouth of the bottle is open, and passes on to the crown and cap-machine, and on from the crowner, down the conveyor to a machine which takes the bottles sixteen at a time automatically to a neon inspecting light, and from there right on to the accumulating table where they are picked up and cased. There is an inspector at the neon light, whose duties it is to inspect the bottle for everything that does not look right; if a bottle is not properly filled, he takes it out. This inspector has no other duties. At the time the bottles leave the washing-machine, there is an inspection made by a man whose only duty is to inspect the bottles. These inspectors are relieved from time to time.
“In making carbonated water, . . plain water . . comes . . down to the carbonator downstairs, where it is mixed with the gas and water and the gas goes into it, and this water goes into the machine from a spray and gets the mixture with the gas and water. The gas comes from the gas drums into this machine. We do not manufacture this gas, but buy it from the Crystal Carbonic Company in Atlanta. . . When the filler takes charge of carbonated water and gas it automatically cuts off from the mixing machine to the filler until the water level is dropped down to a certain height where it will automatically take on another charge. Coca-cola is bottled at about 40 degrees Fahrenheit. The pressure of the carbonator is approximately around 48 pounds. The machine that puts the syrup in has nothing to do with the machine that puts in the water. The syrup tank is this machine with sixteen valves on it. . . At the time the carbonated water is being put into the bottles it is under pressure, which is taken off after the carbonated water is put in and before it is capped. This is a low-pressure machine. From the time the carbonated water is put in until the time the bottle is capped, it is under pressure in the filling machine. For the distance of *591four feet the bottle travels to the capping-machine there is no pressure on it. The carbonating machine has a gauge showing the amount of pressure put into the bottle, the amount of gas-pressure being mixed with the water. The maximum pressure registered on the carbonator gauge is approximately 48 to 50; that is, it is regulated to that pressure, although the figure on it is as high as 300 pounds. The gauge is controlled either way it is set, or it is controlled by some other gauge. There are two gauges. On the filling gauge which will run approximately 33, there is a pressure against that of about 34 or 35. If you put more than 50 pounds of pressure on it, it would blow up. If it blew out, it would blow out gaskets, and everything would shoot out everywhere and would not go into the bottle. The carbonated gas would simply be dissipated in the air. In capping the bottles it is necessary to use a pressure by that machine of not more than 50 or 53 pounds, to where the crimp in the crown will hold, so it can not be turned by hand. After coca-cola is bottled, we make a test to determine the pressure on the bottle of coca-cola, once a day; and records are made of these tests. We made tests every day we bottled coca-cola-. These dates on this record are August 3, 1935, August 3, August 5, August 7, August 9, 1935, and are the original records, made by me. On August 3, 1935, carbonator pressure 53; filler pressure filling-machine 33; and carbonator temperature 41; filling-machine temperature was 43. August 3, 1935, carbonator pressure 53; filling-machine 33; carbonator temperature 41; filling-machine 43. August 5, 1935, carbonator pressure 53, filling-machine 33. August 7, 1935, carbonator pressure 60, filling-machine pressure 33; temperature carbonator 44; temperature filling-machine 46. . . After they are delivered by our salesmen . . the company has no control over these bottles. Prior to August 7, 1935, we had no trouble either at our plant or in our trade territory with broken bottles. The pressure just described is checked by a regular pressure gauge, made on each individual bottle. I take two bottles just being filled right on the bottling machine and two bottles out of what we -call order stock, say filled a day may be. Eecently I made an experiment by cutting down on my water supply, to see what would happen. The result was that I would not have as much pressure on the bottle as I would ordinarily for filling use. I cut the water off *592entirely, and there was nothing but syrup and gas in the bottle and no pressure at all registered. The normal pressure at which we desire coca-cola put on the market- runs anywhere between 30 to 35 pounds. Up to a few years ago all our bottles were agitated by hand by turning the bottle upside down. The bottles are agitated and inspected at the same time. During the last several years I have had occasion to visit other coca-cola plants and inspect their machinery, and to observe their method of inspection. We have just as modern a plant as any I have ever been in. The Chattanooga Bottle Company is a reputable concern. Our carbonating machine is manufactured by the Crown Cork and Seal Company, Baltimore. The mixer is manufactured by the Miller Mixon over at Bainbridge, Georgia, a reputable concern.
“The water in the bottle is carbonated with an ounce of syrup, and this mixing machine turns the bottle upside down and mixes it with the water to get the syrup and water mixed together. . . The pressure on the gas-drum varies -a point or two; but if the pressure on the gas and water is about 50 pounds when it goes into the filling-machine, it will blow out gaskets, and the stuff will shoot out. I have seen this happen. Sometimes the regulators have got safety-devices on them, and we have got a safety-valve on top in case the machines get overcharged; instead of building up pressure, it would be decreasing on account of this valve blowing off on top. That regulator regulates the gas in the gas-drum. I have never seen it blow out. It is the mixing of that water and the gas together that will vary just a little bit. Just the agitation it gets in there, and that creates a greater pressure than just the gas itself; . . just running gas into water will have more pressure than the gas being agitated. It is the gas that causes the pressure. . . When these bottles are filled and capped they are then inspected by being placed under a neon light, which is on a conveyor; this neon inspection has a hood over here with those neon lights across here. The inspector stands behind the lights. The bottles move at the rate of about ninety-four bottles a minute. One purpose of the inspection is to see whether the bottles are properly filled. There are forty valves on the filling-machine; and if the valve gets a little out of whack, it will allow a bottle to go through without being filled, or to fill too much. When this happens we take these bottles out. These two bottles *593are not uniformly filled. This one is overfilled, but not enough to be taken out. This inspection will show any default in a bottle filled with coca-cola. I don’t know about a thin place, but a crack or anything like that will show up. When we find a bottle that is . . not properly filled, then it is taken out. If there is a crack in the bottle, it will show up under the neon light. Some bottles are sometimes filled up more than others, due to the valve on the machine. These valves do not frequently fail, but are often removed and repaired or replaced. The valve will get a little off, just like the valve in an automobile or any other piece of machinery. Any excess gas that might get into the bottle would go off into the atmosphere in passing from one machine to the other. The plant is the same now as it was on August 7, 1935, and a week or ten days prior thereto. We have the same machinery, same lights, same equipment. I put some bearings in down underneath the filler; otherwise little minor repairs just like upon an automobile or some little average minor repair. There has been no change at all in the place. I put some new bearings in ten days ago. I do not know whether or not this case had been set at that time. The bearings went into the filler-machine, which is one of the principal machines involved.”
Two witnesses for the plaintiff testified that each heard a coca-cola bottle explode without having been hit against anything.
Cliff Bowden, testified for the defendant, in brief: “I am employed by the Macon Coca-Cola Bottling Company, and have been in the bottling business 37 years the 22nd of last month. . . I have been in my present position over 18 years, and it is my duty to help in the bottling of coca-cola. . . The bottles are brought into the plant on trucks. . . These bottles are loaded on a conveyor chain holding 144 bottles. It is continuously being set at all times, because machinery moves up a space and then stops. On this conveyor chain these bottles are placed in rotation, twelve bottles in a row on the chain run by machinery. This carriage chain moves a space, and raises those bottles up; it is continually going by; and then when these bottles raise a certain height on a machine in the carriage chain, it comes in contact with water. That water shoots in all twelve of the bottles from a reservoir, and when the chain moves into that certain space that reservoir shoots into them. As the chain moves from the reservoir they go *594over to a certain distance and go down in a hot solution that is water and caustic soda, and the temperature is 150 degrees. Then this carriage moves on on the same chain, and this tank is 140 degrees in heat. I spoke to you of the first tank of that solution; that solution consists of from two and a half to three and a half per cent, in strength. The second tank carries 140 degrees in heat, and consists of four and a half to five per cent, in strength of caustic soda; and then these bottles move on continually, and then get to another tank of water; a spray of water, goes over them and down and up like that. When it goes into the last tank of water it is plain water and let out every afternoon after a day’s bottling, to prevent any kind of slime that may settle in the water during the day. The tank is 110 degrees. The bottles move from that tank and proceed oír 12 little rods. As they pass, those little rods are mounted with rubber tips and the reason is to prevent any breaking of the bottle’s mouth; then they go between two long brushes. Those brushes are run by machinery and revolve; the rod sending the bottles revolves at the same time, and at the same time there is a spray of water on the outside of the bottles, -and also there is a brush that is stationary that brushes the bottom of the bottle. Then they pass from that in rotation, twelve bottles in rotation, and when they pass it catches rinsing water on a brushing machine; it is done automatically. Then when it passes from there in rotation it passes to another; it gets the same rinsing there it got here. Then it passes from there to a spindle. The brush goes up through the mouth of the bottle, and has a spout on it, and goes up in the mouth of the bottle like that. At the same time a spray of water is going up through the spindle; it is making a very fast revolution; . . it comes down with the spindle turning, and goes back up and releases itself. From there it passes to another, and there is the same work done over again; then it passes from there and into another; in fact it has seven it goes through; the last one is like this, it goes up into the bottle. . . Then after that is done the bottles leave the brush machine slowly in traveling coming down to the inspection light. It takes one bottle eighteen minutes to come from the rear end of the machine to the end; one bottle gets as much washing as the other bottle; when the bottle comes to inspection lights, there are two lights, and a white gentleman sitting there *595by the name of Mr. Ed Bussell, who has plenty of time in operating, to inspect these bottles.”
Dr. Henry Bogers,. testified for the defendant: '“I am . . professor of physics at Mercer University. . . At Mr. Popper’s [defendant’s counsel] request I have had occasion within the last week or so to inspect the plant of the Macon Coca-Cola Bottling Company, and have familiarized myself with the machinery there and with the formulas for making glass. I am somewhat familiar with the glass manufacturing business, from a theoretical study from text-books, and am familiar with the effects of pressure on glass bottles, brought about from filling bottles with carbonated water, from laboratory experience. I made a particular examination of the machinery of the Macon Coca-Cola plant that is used in making carbonated. water and the other machinery that puts the carbonated water into the bottle. Glass does not become weak from age, unless it is subjected to some sort of force or unless given chemical treatment; normally it does not. A glass bottle ten years old will be just as good as a new bottle, provided it had not been struck or cracked in any way. I have tested the thickness, height, and effect of temperature on coca-cola bottles in the laboratory. A bottle of that character which is filled with coca-cola syrup, carbonated water at 35 and 40 degrees Fahrenheit ’under a pressure of 35 pounds, is not in any way dangerous.- In my opinion if there was a crack or defect in the bottle under the type of manufacture and glass, it would break either when it was first filled or when it was capped and the capping process was put on. . . At the plant . . they use carbonic acid gas under very high pressure, which is admitted by valve into a machine which is called a saturator or mixer. At the same time the water is put in under high pressure, and the gauge will register up to exactly 300 pounds as the maximum limit on that gauge. They run in the plant around'50 pounds or under pressure at the gas, and 50 pounds or under at the mixer, where the gas is mixed with the water shooting in under pressure. You have then, instead of carbonated gas and water, carbonic acid or H2C03. . . The filling-machine has a gauge on it, which I examined. The purpose of that gauge is to check the pressure of the admitted carbonic acid you get into the bottle, so it will not enter at high pressure, and you will fill your bottle uniformly, and they have a rub*596ber diaphragm gauge with a maximum of 50 pounds on it to limit and advance and regulate the filling of the bottle. If the pressure got higher than that, the rubber piece in the gauge would blow out. It would allow any pressure over 50 pounds to diffuse in the atmosphere. This is a low-pressure machine; after the bottle is filled it passes out into the open air before it is capped. Assuming that more than 35 or 40 pounds were put into the bottle, as soon as it left the filler all but 15 pounds per square inch would dissipate itself into the atmosphere, because there is nothing there to hold it but air pressure. The vapor pressure of the carbon dioxide from gas in the bottle gives a pressure of 15 pounds to a bottle of coca-cola after it has been capped. . .
“I made a pressure test with a pressure gauge on a bottle of coca-cola, and the pressure shown was 35 pounds, which is sufficient to break the bottle if it had a crack in it, immediately after the crack occurred. The bottle is supposed to average in thickness throughout a quarter of an inch, and is an average diameter at the bottom of two and one half inches. It does not vary more than one sixty-fourth of an inch at any point, and averages a quarter of an inch from top to bottom. . . A sudden change on the temperature of glass is due to the fact that by each degree rise in temperature you have eight and one third pounds pressure due to the expansion in the glass. . . I picked up five bottles.' Only one of the five bottles broke at the temperature of 90 degrees, and the highest was 105. . . The bottle broke at a point 90 degrees above 32 degrees Fahrenheit, which would be 122 degrees. . . The tensile strength of this type of glass is 3000 pounds per square inch. The better type is 4000 pounds. . . If you raise that bottle to 85 degrees and then immerse it in ice water 32 degrees, you would submit that bottle to a change in temperature of 53 degrees. Multiplying that by eight and one third pounds, you would have a forced pressure on that bottle of 442 pounds. If the pressure inside, 45 pounds, adding the two would give you 487 pounds pressure on that bottle at that time, and the bottle will stand 750 pounds. . . A sudden change in temperature on glass has the effect of a sudden change by submitting the glass to eight and one third pounds per square inch for each degree rising temperature or falling. One would -be an expansion force, and one would be a contraction force. All bottles are not built *597to withstand the same pressure and changes in temperature. You can make glass to withstand practically any temperature you want to. In my opinion this glass is made to withstand a certain pressure, and has a proper margin of safety for that pressure; for a bottle submitted to around 50 pounds pressure, 750 pounds per square inch is a pretty good margin of safety. The thickness of the glass does not increase its resistance. The thinner the glass the more it will resist the change in temperature. The force on a glass due to change in temperature depends on the speed at which it can distribute the temperature throughout the bottle. . . If the temperature is the same inside and outside, the bottle will not break. It is a matter of common knowledge that glass heated to high temperature and suddenly immersed in ice cold water is likely to break. . . A bottle filled with carbonated water carried from hot water to cold water would surely crack and fall apart. All the pressure does is to throw glass in proportion to its force. The bottle is supposed to stand one fourth the pressure of its thickness, if it had a thin part in it that was only one fourth of the average thickness of the bottle, which would be one fourth of 750 pounds; approximately 187 pounds. This is .certainly higher than the pressure inside.
“Prior to ten days ago I knew nothing about this machinery. The pressure I spoke of was the pressure they used on the days I was there. . . With an equal amount of carbonation, which would have the higher pressure, a bottle that is filled like that one or v bottle filled like this one, would depend on the glass you have. If the bottles had the same pressure of gas it would make no difference, because as the pressure becomes greater and greater the carbonic acid is forced back into the liquid. The weaknesses in a bottle that would cause explosion is overheating and possibly a crack in the glass, or it was worn down there; its tensile strength would decrease in proportion as the thickness of the glass is diminished. The thinnest part of the bottle would not be the same oar any bottle; for no bottle would be exactly like the other. . . Caustic soda is a chemical if mixed with a solution. . . The more it is used, the more it affects the bottle. If you take a bottle of coca-cola at 89 degrees under a 35 to 50 pounds pressure on the inside and insert it in either ice water or place it in an ice-box on top of ice, without anything between that bottle and *598some other object, and it explodes, if no contact or physical force were present, there would be only two forces that could affect the bottle; one would be the pressure inside the bottle and the other the change it went through. If it broke, it was thinner and could not stand 150 pounds, or there was more than 150 pounds inside the bottle. Picking up three or four coca-cola bottles at the same time would be likely to increase the likelihood of breakage to the extent that force is present. I might say if-you have a bottle that will stand 400 pounds, and you already have it loaded to 400 pounds, just a camel’s hair, and it is gone. One pound would break the bottle if it is loaded up to 400 pounds. A solution of caustic soda of a strength of five and one half per cent, would affect glass only over a long period of years, by daily immersion. If you wash a dish daily with caustic soda, it will last you quite a long time. If there is a coating inside of glass and you wash it with caustic soda so as to remove this coating, it would affect the coating, and it would etch on the glass. Bottles that go into the tank with five and one half per cent, solution of caustic soda for three or four minutes would not become impaired.”
Robert Crane, the plaintiff, recalled, testified: “I testified this morning that the ice-box was four or five feet from the partition that separates the restaurant proper from the kitchen. I measured with a yardstick, and it is about eleven feet. It is about fourteen feet from that partition to the range.”
The question to be decided is whether the rule of res ipsa loquitor would be applicable as a final proposition, in view of the uncontradicted evidence of the defendant negativing negligence on its part or showing that the injury could reasonably have occurred without its negligence. While I seriously doubt the applicability of the rule in this case, even to the extent of supporting a prima facie case, for the reason that it is easily conceivable that the explosions of bottles generally in isolated cases are not due to the negligence of either the manufacturer of the bottle or the dealer handling the bottle in the manufacture and sale of drinks in which the bottle is used, a discussion of its applicability as a final basis for the jury to base a verdict oil will suffice for the purposes of this case. It will be well at the outset to determine just what the act of negligence is for which recovery is sought in the case. In the original petition the plaintiff alleged: *599“The bottle hereinbefore referred to, which was prepared and furnished by the defendant to. the restaurant at which plaintiff worked, was so highly charged with the said gas by the defendant as to cause the explosion described. Petitioner alleges that if said bottle had been properly charged the said explosion would not have occurred.” By amendment the plaintiff added the following paragraph: “The bottle hereinafter referred to, which was by defendant filled with coca-cola syrup and carbonated, was insufficient and contained thin places, so that it was not capable of withstanding, upon handling in a normal way, the pressure of the carbonating gas. Petitioner alleges that the defendant was negligent in using and distributing to its customers, to be by said customers dispensed to consumers, a bottle which was not sufficient to withstand proper and expected handling without being exploded by the pressure of the carbonation.” The plaintiff so amended paragraph 23 of the petition as to read as follows: “Petitioner was in the exercise of all ordinary care and diligence in and about the matter referred to, and the explosion of the bottle and the consequent injury to petitioner was caused solely by the negligence of the defendant in charging said bottle excessively and in using a bottle which was insufficient as hereinbefore charged, so that upon the handling of the same it would and did explode, causing the injury to the plaintiff.” The construction of the petition as amended, against the plaintiff, confines the action to one for the negligent use of a weak bottle. Otherwise it would be problematical whether' the action were one for the excessive charge of a good bottle or the normal charge of a bad bottlq, which normal charge would be excessive by reason of the insufficiency of the bottle.
The propriety of the prima facie application of the rule of res ipsa loquitur must be determined by the specific act or acts of negligence alleged in the petition. The same is true as to its final application. Palmer Brick Co. v. Chenall, 119 Ga. 847 (supra). The authority for the application of the rule in Georgia is in the Code, § 38-123, which is: “In arriving at a verdict, the jury from facts proved, and sometimes from the absence of counter-evidence, may infer the existence of other facts reasonably and logically consequent on those proved.” There seems to me to be a distinction between a prima facie and a final application of the *600rule. It may be entirely reasonable, legal, and proper to shift to a defendant the burden of proceeding with evidence upon proof of an injury which reasonably would not have happened but for his negligence, and at the same time be entirely unreasonable and unjust to apply the rule after the defendant has shown by uncontradicted and positive testimony that the injury was not due to his negligence, or presenting another reasonable hypothesis. The reason for this is that the rule is based on two propositions: one is the proof of the injury which serves as tentative evidence to take the case to the jury, and the other is that in the absence of testimony which explains the occurrence by showing how it happened without negligence, or which shows that the defendant was not negligent, the jury may finally apply the rule and find against the defendant because of the failure to produce the proof required. 4 Wigmore on Evidence, § 2509. After the defendant explains the injury and thereby excludes his negligence, or presents one or more reasonable hypotheses any one of which excludes his negligence, the rule no longer applies, and if the plaintiff would recover he must rely on evidence, and not the arbitrary and artificial rule. When this is done, the question of the propriety of a verdict for the plaintiff is determined by a comparison of the evidence for one side with the evidence for the other, and not a comparison of the rule of res ipsa loquitur on the plaintiff’s side and the evidence on the defendant’s side. The fact that the rule may be prima facie applicable does not mean that the jury must have the right finally to apply it; Chenall v. Palmer Brick Co., 117 Ga. 106, 110 (supra), where Judge Lamar said: “Of course, the defendant could have made out a case which would warrant the judge in holding that the presumption arising from the fall of the arch had been rebutted, and under such circumstances he could have directed a verdict in favor of the defendant.” So, when a plaintiff relies solely on the rule, if the defendant makes out such a case, it is a matter of law, and subjects the case to a determination on such basis either by the trial court in the direction of a verdict or by an appellate court in ruling on the insufficiency of plaintiff’s evidence. See Langley Bus Co. v. Messer, 222 Ala. 533 (133 So. 287); Manuel v. Pacific Gas & Electric Co., 134 Cal. App. 512 (25 Pac. (2d) 509); Crooks v. White, 107 Cal. App. 304 (290 Pac. 497, 92 A. L. R. 655); Edwards v. *601Cumberland County &c. Co., 128 Me. 207 (146 Atl. 700); Black Mountain Corporation v. Partin, 243 Ky. 791 (49 S. W. (2d) 1014, 92 A. L. R. 657); Potomac Edison Co. v. Johnson, 160 Md. 33 (152 Atl. 633); Plumb v. Richmond Light &c. Co., 223 N. Y. 285 (135 N. E. 504, 25 A. L. R. 685).
I recognize the distinction between the presumption of negligence against a railroad company arising on proof of certain injuries, and the rule of res ipsa loquitur. Were it not for the presumption in many cases the rule of res ipsa loquitur would not apply, because the nature of the accident and the attendant circumstances do not exclude every reasonable hypothesis other than the negligence of the defendant; for, after all, the rule of res ipsa loquitur is but a rule of circumstantial evidence. As Judge Cobb said, in Palmer Brick Co. v. Chenall, 119 Ga. 837, the rule is of slight value in any case, and it should be applied with great caution. Part of the weakness in the presumption against railroad companies inheres in the res ipsa loquitur rule, and the same reasoning which led to the overthrow of the former except as a prima facie substitute for evidence (W. & A. Railroad v. Henderson, 279 U. S. 639, 49 Sup. Ct. 445, 73 L. ed. 884), prevents the final application of the rule of res ipsa loquitur in certain cases. In other words, the law says that the proof of an unusual occurrence in certain instances is sufficient to prove that the mere accident under the circumstances amounts to proof that there is no reasonable explanation except the negligence of the defendant. If the defendant then shows by positive, uncontradicted evidence that he was not negligent, or explains by the same kind of evidence that the injury could reasonably have occurred in one or more ways than having been caused by his negligence, and the plaintiff introduces no more evidence (which would render the rule inapplicable in itself), still to permit the rule to be applied would run into the teeth of the logic, reason, and compelling authority which has declared that _ such an application is arbitrary and unreasonable. In this case the defendant showed (1) that it exercised ordinary care in purchasing bottles suitable for the purpose for which they were intended, (2) from a reputable bottle manufacturer; (3) that it exercised ordinary care in testing and inspecting its new and old bottles, and threw out the ones found to be defective; (4) that it exercised ordinary care in so bottling *602its drinks as not to charge into a bottle more gas than it was designed to hold; and (5) that the plaintiff might have caused the last damage to the bottle by striking it against the bottles he had in his hand with it, which could have caused it to burst. The testimony of Dr. Eogers was to the effect that knocks, jars, etc., weakened the bottles even when it did not crack them. Being unable to see the weak points in a bottle because of such weakness, or to see a weakness caused by the long use of caustic soda, the defendant in this case seems to have done all it reasonably could to test its bottles in the ways shown by the testimony. If the bottle came from the manufacturer in bad condition, weak but not cracked, it could have come without negligence on the part of the manufacturer, and its weakness could have escaped the defendant without its having been negligent. To expect 100 per cent, perfection in such mass manufacture of bottles is to expect a miracle, and to expect a bottler to detect a weakness in a bottle caused by the rough treatment and other weakening processes, by methods outside and beyond what was shown to have been done in this case, is to require an unreasonable and maybe a prohibitive task. The defendant is not an insurer of the bottle.
The testimony of Dr. Eogers seems to be emphasized by counsel for the defendant in error and the majority opinion. Dr. Eogers swore that if the bottle broke under circumstances embraced in a hypothetical questioú which covered the conditions in this case, it was due to a weak bottle or an overcharge of gas. It was evidently overlooked that he also testified that bottles were weakened by knocks, etc., and that the plaintiff might have done the remaining damage to the bottle needed to weáken it to the breaking point by striking it against the bottles the plaintiff picked up with one hand together with the bottle which burst. Even if that is not so, a weak bottle does not necessarily mean that the defendant was negligent. All it has to do is show its ordinary care in trying to find the weak bottle and throw it out, either when it was new or after it had been used. Then there is the testimony of Mr. Williford, which is stressed in the majority opinion. He swore that the gas-charging machine was set to diffuse all surplus gas into the air, but on cross-examination he swore that there was a slight possibility that the machine might *603become clogged. That is not evidence that the machine did get clogged, or, if it did, that the clogging was dne to defendant’s negligence. Conceding for the sake of argument that it was such evidence, then in that event the plaintiff is compelled to rely upon his evidence of negligence, and not on' the rule. This evidence alone is insufficient to show that the bottle was overcharged, because it does not appear what the result of the clogging of the machine would be. If the petition is treated as charging negligence in using a weak bottle and overcharging, the result in this case would be the same. There is nothing in Payne v. Rome Coca-Cola Bottling Co., 10 Ga. App. 762 (supra), contrary to what is here' held. The case of Atlanta Coca-Cola Bottling Co. v. Danneman, 25 Ga. App. 43 (supra), is not contrary authority for two reasons at least; the defendant did not buy the bottle from a reputable manufacturer, and it was negligence for him blindly to use a stray bottle he did not know anything about; and because the rule did not apply in that case, because there was evidence in addition that the bottle was defective. The foreign-substance cases are so manifestly different from this and similar cases that I shall not dwell upon them. Suffice it to say that their applicability was due to insufficiency of the explanations and the lack of reasonable probability that negligence of one other than the defendant caused the injury, or that the injury might have occurred without the negligence of any one. “A theory can not be said to be established by circumstantial evidence, even in a civil action, unless the facts relied upon are of such a nature, and are so related to each other, that it is the only conclusion that can fairly or reasonably be drawn from them. It is not sufficient that they be consistent, merely, with that theory; for that may be true, and yet they may have no tendency to prove the theory. . . If the facts are consistent with either of two opposing theories, they prove neither.” U. S. F. & G. Co. v. Des Moines National Bank, 145 Fed. 273; Taylor v. State, 44 Ga. App. 387, 417 (161 S. E. 793); Federal Reserve Bank of Atlanta v. Haynie, 46 Ga. App. 522 (168 S. E. 112).
My conclusion in this case is consistent with the weight of authority on the question; and without discussing the cases in detail, I shall simply cite some of them. Guinea v. Campbell, 22 Quebec Law R. 257; Glaser v. Seitz, 35 Misc. 341 (71 N. Y. Supp. *604942); Smith v. Peerless Glass Co., 259 N. Y. 292 (181 N. E. 576); McPherson v. Buick Motor Co., 217 N. Y. 382 (111 N. E. 1050, L. R. A. 1916E, 696, Ann. Cas. 1916C, 440); Loebig v. Coca-Cola Bottling Co., 259 Ky. 124 (81 S. W. (2d) 910); Stone v. Van Noy Railway News Co., 153 Ky. 240 (154 S. W. 1092); Hughes v. Atlantic City &c. R. Co., 85 N. J. L. 212 (89 Atl. 769, L. R. A. 1916A, 927); Burnham v. Lincoln, 225 Mass. 408 (114 N. E. 715); Wheeler v. Laurel Bottling Works, 111 Miss. 442 (71 So. 743, L. R. A. 1916E, 1074); Gausman v. R. T. Pearson Co., 284 Pa. 348.(131 Atl. 247); Standard Oil Co. v. Murray, 119 Fed. 572; O’Neill v. James, 138 Mich. 567 (101 N. W. 828, 68 L. R. A. 342, 110 Am. St. R. 321, 4 Ann. Cas. 177); Bates v. Batey, 3 K. B. (Eng.) 351, 29 Times L. E. 616. Citation of authority for the proposition that a statutory presumption based on the value of proved facts as circumstantial evidence is entirely overcome as a matter of law by uncontradicted, positive, and unimpeached testimony to the effect that the fact presumed is not true, seems to be superfluous. The establishment of another reasonable hypothesis by the same kind of evidence would seem to me to be sufficient as a matter of law to preclude the final application of the rule of res ipsa loquitur. The main point of difference between the majority of the court and me is just what is the defendant’s burden to preclude the final application of the rule. The majority seems to be of the opinion that after a prima facie case is made, the burden is on the defendant to prove that he was not negligent, or to prove that the accident was " occasioned by vis major or by other causes for which he was not responsible.” I do not think, that is the law. I am of the opinion that the defendant in such a case may exculpate himself by showing positively that he was not negligent, or by showing that the accident could just as reasonably have occurred in a way which was consistent with his diligence. 3 Cooley on Torts, 378, and cit. I do not think the defendant has to prove just exactly what caused the accident, or just exactly whose negligence besides his caused it. If the defendant shows a theory consistent with due care on his part which is as reasonable as the theory that he was negligent, he has met the requirements of the law, especially because the reason for the final application is based on the defendant’s failure satisfactorily to explain. To illustrate further the point I *605am trying to make, if the plaintiff had procured and introduced all the evidence, and the defendant none, he could not recover, because he would- not have excluded every reasonable hypothesis save that of the defendant’s negligence. I think the same thing is true when the defendant produces the evidence, such as in this ease, when there is no question of the conflict of evidence of credibility of witnesses to make the questions those exclusively for a jury. Furthermore, it is obviously impossible in a case like this one for the defendant to prove his diligence with reference to one bottle out of ten million. It would be unreasonable to require him to show the course of the burst bottle from its cradle to its grave, and to show that on each step of its journey due care had been exercised with reference to it. I am of the opinion that the evidence was insufficient to support the verdict, and that a new trial should have been granted.