(Special Judge). — This suit, which is an action brought by Agnes Fuchs, to recover damages for the death of her husband, Carl E. Fuchs, was filed September 16, 1892, in the circuit court of the city of St. Louis. The defendants in the suit, as originally brought, were the city of St. Louis and the Waters-Pierce Oil Company. The case first came to trial in April, 1893, in which plaintiff was forced to submit to a nonsuit. She appealed to this court, which affirmed the judgment of the lower court as to the Waters-Pierce Oil Company, but reversed the judgment as to the city of St. Louis, and remanded the ease for a new trial. The opinion of the court on that appeal, will be found in Fuchs v. St. Louis, 133 Mo. 168.
The case was finally tried in the circuit court of St. Louis county, to which it had been taken by change of venue, resulting in a verdict and judgment for plaintiff against the remaining defendant, the city of St. Louis. After an unsuccessful motion for new trial, the city appealed to this court.
The petition alleges that her deceased husband, on or about May 26, 1884, became the owner of a lot of ground lying on the east side of Fourth street, about one hundred and thirty-nine feet southwardly of Chouteau avenue, and that in the following year he erected on said lot a building covering the entire width of the lot and extending back about seventy feet. That said building was of brick, three stories high, with a cellarthe cellar and the first floor being designed for the storing of wines and liquors, and the carrying on of a wine and liquor business; and that upon the completion of said building her husband fitted up and furnished said cellar and first floor with bar fixtures, shelving, etc., and thereafter, and until his death, carried on a wine and liquor business in said premises.
The petition further alleges that on July 22, 1892, the Waters-Pierce Oil Company, engaged in the business of buying, storing and selling oils, had on hand in its premises a large stock of oil, in barrels and other packages, and that on said *628day a fire occurred in said company’s premises, “and that said oil company then and there carelessly and negligently did cause, suffer and permit the said oils to escape and to run into the sewer hereinafter mentioned and to fill up said sewer and to generate the gases which caused the explosion in and destruction of said sewer on the twenty-sixth day of July, 1892 the city of St. Louis constructed a main sewer, known as the Mill Creek sewer, leading from the center of the city to the Mississippi river, and designed to drain the surface waters falling within the reach of the sewer, and to carry off into the river waters from private dwelling and certain city buildings. That so far as it was practical to do so, the sewer was constructed beneath the public streets and alleys of the city, and that is thus crossed underneath Fourth street to the east line of said street, where said line intersects the lot of ground purchased by the deceased husband of plaintiff; that under a license from the then owners of said lot the city was permitted to construct said sewer underneath the same, and to carry it eastwardly under said lot towards the river, said sewer being located below the cellar thereafter constructed by said deceased. That when the city obtained said license from the owners it agreed with them and their assigns to keep and maintain the sewer in good order, and to care for the same so that said lot and any improvements which might thereafter be placed thereon would be free from danger of injury on account of said sewer or the use thereof.
The petition then alleges “that said sewer was provided with openings specially designed to carry off any gases which might arise in said sewer and be liable to combustion and explosion, and that said sewer and the openings thereof aforesaid on and prior to the said twenty-sixth day of July, 1892, were in the sole care and control of defendant, the city of St. Louis, its agents and servants, yet the said city, its agents and servants, knowing that said defendant. t]m 'Wra+m”=!-T>ifn'ep Oil Company, had flooded said sewer with oil, neglected to open *629said vents and carelessly and negligently failed to take measures and precautions to prevent gases arising and accumulating in said sewer so as to endanger the same; and that between the said twenty-second and twenty-sixth days of July, 1892, gases did arise and accumulate in said sewer in great and very dangerous quantities, and on the date last named, and within six months ne.xt before the commencement of this suit, ignited and exploded with great force, throwing open said sewer underneath the property of said Carl E. Euchs, shattering his said building,, and also then and there causing the death of said Carl E. Euchs.” That by reason of said wrongful acts of defendants, whereby the death of her said husband was caused, she has been damaged in the sum of five thousand dollars, for which she prayed judgment.
The answer of defendant was a general denial.
At the close of plaintiff’s case, defendant offered an instruction in the nature of a demurrer to her evidence, and at the close of the whole case it offered a peremptory instruction to find for the defendant. Both instructions were overruled by the court.
The evidence shows that “Mill Greek Sewer” as it is called, is one of the leading public sewers of the city of St. Louis, and, considering its length and dimensions, is one of the largest sewers in the world. It was built some thirty-four years before the explosion therein, in a most substantial manner, that left nothing to be desired, of heavy, massive masonry sides and arch. It takes its name from the fact that it follows an old natural creek which was known as “Mill Greek,” and-which formerly constituted the natural drainage of a large portion of the city, emptying into the Mississippi river, at a point between Chouteau avenue and Convent street, about four blocks east of the Euchs premises. At its mouth at the river, the sewer is about twenty-four feet by fourteen feet in dimension, and grows smaller as it reaches its western or beginning point. At the place where it runs under Euehs’s premises its *630dimensions, according to the testimony of Mr. Oolby, the sewer commissioner, is twenty by fifteen feet. The natural creek ran in a diagonal direction over the lot acquired in 1884 by Mr. Euchs.
At the time when he purchased, the sewer had already been constructed and covered with earth or ballast. This he removed in order to construct his cellar, leaving about six inches of earth or cinders between the floor of the cellar and the top of the sewer arch.
The premises of the Waters-Pieree Oil Company were situated on Gratoit and Thirteenth streets and in their yards were erected a number of sheet iron tanks and stored a number of barrels containing oils. On July 22, 1892, a fire broke out in these premises, during which a number of tanks and barrels caught fire. Some were left intact, the oil in some exploded and was consumed in the explosion, while in others it was displaced by the throwing of water into the tanks by the fire engines. The water, mixed with burning oil, ran through the premises and down into what is known as the railroad valley. There being danger that some of the standing cars might catch fire, drains or chutes were constructed by the railway people and by the firemen, and it appears that this running water and oil was led between the tracks to a drain or inlet in the valley, the surface of which was covered with cinders. Some of the oil and water ran into a drain or inlet in the yard of the oil company’s premises.
At the time of the fire the Mississippi river was at a very high stage, and the mouth of the sewer was submerged by high water. The explosion which caused Mr. Euch’s death occurred in the afternoon of July 26, 1892, or four days after the fire. It appears that the Peters Fish & Oyster Company, a concern doing business at the French Market, which was somewhat south of Mr. Fuchs’s place, used in connection with their business a basement under house No. 1026 South Fourth Street, about five stores south of Mr. Fuchs’s place. An em*631ployee named Hnmpert was at work in the cellar storing melons, and happening to pass over a drain in the cellar floor with a lighted candle, there was a sudden flash of flame, and almost simultaneously the explosion in the Mill Greek sewer occurred. The drain or sewer thus referred to connected with a sewer in the rear of the premises, which in turn connected with the Mill Creek sewer, as one of its numerous connections. Only that portion of the sewer was blown up which was immediately under the cellar. In the backyard, where-the earth had not been removed, the sewer remained intact.. According to the testimony, the break in the sewer at Mr„ Euchs’s place, indicated clearly the outlines of the former cellar floor. About two blocks east of the Euehs place, and nearer the river, where the slope of the ground is downward with but little surface ballast to cover it, there was an open break of three hundred feet or more in length.
These are the general facts of the case as shown by the voluminous record in this case. Other specific facts, bearing more immediately on the principles of law involved, will be noticed and mentioned hereafter.
Counsel for appellant, in his brief and in his printed and oral argument, insisted that the court erred in not sustaining its demurrer to plaintiff’s evidence at the close of her case. When that demurrer was overruled, defendant waived its objection to Hie action of the court, by afterwards introducing evidence-in its behalf. But having offered a peremptory instruction to find for the defendant at the close of the case and having duly preserved the point, we are required to review the evidence taken as a whole [Hilz v. Railroad, 101 Mo. 36; Weber v. Railroad, 100 Mo. 194; Hite v. Railroad, 130 Mo. 132. l. c. 141], and when this is done, then, in the light of the principles of law applicable thereto, there can be but one conclusion, and that is that the plaintiff can not recover in this action.
At the outset, it may be stated that there is not the slight*632est testimony in the record tending to prove the allegation in plaintiff’s petition that when the city constructed the sewer through and underneath the lot-, subsequently acquired by the deceased, it agreed with the then owners and their assigns to keep and maintain said sewer in good order and to care for the same, so that said lot and any improvements which might thereafter be placed thereon would be free from danger or injury on account of said sewer or the use thereof.
This allegation, of necessity, implies a contractual relation existing between the deceased and the city. There being no evidence to support it, it follows that any idea of such contractual relation between the parties, must be eliminated from this inquiry and it- further follows that whatever claim the plaintiff may have against the city must be predicated upon the non-performance of a legal duty due and owing by it to the deceased.
Such duty and its violation, upon which the plaintiff bases her claim for damages, must be found in and determined by the allegations of the petition.
These allegations are: That oils escaped from theWatersPierce Oil Company at the time of the fire hereinbefore referred to; that they ran into the Mill Creek sewer; that such oils generated gases, and that such gases caused the explosion in and destruction of the sewer on July 26, 1892, by which explosion deceased lost his life; that the sewer was provided with openings specially designed to carry off any gases which might arise in said sewer and be liable to combustion and explosion ; that, nevertheless, the city knowing of the presentee of oil in the sewer neglected to open said vents and carelessly and negligently failed to take measures and precautions to prevent gases arising and accumulating in said sewer so as to endanger the same.
It will thus- appear from these allegations, that plaintiff charges that the gases which were generated in the sewer were so generated from the oils which escaped and ran into the sewer *633from tbe Waters-Pierce Oil Company at the time of the fire on its premises which caused the explosion. It will also appear that there are two charges of negligence, namely, one a specific charge of negligence in that the city knowing of the existence of oils in the sewer did not open the vents specially designed for the purposes of, presumably, permitting the gases thus generated to escape, and, in addition, another charge of negligence to the effect that it carelessly and negligently failed to take measures and precautions to prevent gases arising and accumulating in the sewer. In these allegations of non-action on the part of the city are contained the charges against it of legal duties not performed.
I. It- is conceded that oil escaped from the Waters-Pierce Oil Company at the time of the fire, and it may fairly be inferred from the evidence that some of the escaping oil ran into the Mill Creek sewer; but a careful reading and re-reading of the record herein fails to disclose any evidence from which even an inference could be drawn, that sufficient oil flowed into this large sewer from which were generated sufficient gases to cause the explosion complained of. Indeed, there is a total lack of evidence to prove that the explosion was, in fact, caused by gases generated from the escaping oils as charged. On the other hand, the evidence showed that there were gases other than those arising from oil present in the sewer at the time of the explosion, gases which are always present in sewers, and which form as the result of decaying organic and vegetable matter and which are both inflammable and explosive.
Plaintiff’s witness, Frederick Egner, a gas engineer, testified that from his general reading, he knew there was a constant generation of inflammable gas in sewers from vegetable and animal matters, which is called sewer gas, and is a mixture of sulphuretted hydrogen and marsh gas, the latter of which emanates from decaying vegetable matter. He had heard of gases which would ignite and burn spontaneously on the sur*634face of water in a stream or lake when stirred up. He had read of such cases and believed such gas was common in sewers because the conditions were favorable. Such gas was generated from human excrement and rotting vegetables and animal •matter. He believed that the explosion of sulphuretted hydrogen and marsh gas would be of the same violence as the explosion of petroleum vapor.
So also testified plaintiff’s expert, Huniclce. He states that a gas generates from decomposing vegetable matters which is called methane or mársh gas, and which is' explosive. Another gas, according to witness, which generates from human excrement, being carbonic acid, more or less, and sulphuretted hydrogen; the latter gas smells like rotten eggs and is quite explosive when mixed with air. These are the only witnesses who refer to the subject on behalf of plaintiff.
That explosive gases form in sewers and are ever present therein was also shown by experts introduced by defendant.
Since, then, the gases which form naturally in sewers are as highly inflammable and explosive as the vapors thrown off by kerosene or other oils, and since there was no evidence furnished by plaintiff from which the jury could draw the inference that the explosion was of the latter rather than of the former character of gas, it can not be said that the allegations of the petition were established, or that the verdict should be permitted to stand. In fact the theory that the explosion was caused by the usual sewer gases is the more likely one, since all the witnesses agreed that the oils which escaped during the fire at the Waters-Pierce Oil Company’s premises will throw off vapors in comparative abundance only when subjected to heat. Upon this subject Egner stated that it could not be told how much oil it would take to generate enough gas in Mill Creek sewer under ordinary temperature to cause such an explosion as that which occurred. He did say, however, “I should think it would require a good deal of it in a large sewer like that. Assuming the temperature of the sewer to have *635been 60 or 70 degrees,’ or even 75 degrees, the quantity of vapor generated from crude petroleum would be comparatively small.” Witness further stated that the temperature in a sewer would likely be lower than the temperature in the exposed atmosphere.
Mr. Oholler, also a gas engineer, who had made a special study of the subject, testified that even if petroleum or crude oil got into the sewer it would not have cast off gasoline or coal oil under a temperature below 95 degrees. Therefore, if at the fire, crude oil found its way into the sewer, that would not account for the presence of either coal oil or gasoline, because the application of heat would be necessary to produce either. As for coal oil, or kerosene, they are combustible but not explosive.
It was for plaintiff to account for the explosion and the causes of it, and there being present at least two independent-causes, and no proof as to which of the two was responsible for the accident, there can be no recovery. [Breen v. St. Louis Cooperage Co., 50 Mo. App. 202; Searles v. Railroad, 101 N. Y. 661; Dobbins v. Brown, 119 N. Y. 188.]
II. There is no evidence in the record to show that there were openings into the sewer “especially designed to carry off any gases which might arise in said sewer and be liable to combustion and explosion.” The evidence, what there was of it on this subject, tends to show that there were certain manholes and certain inlets. The former were intended as a means of ingress and egress for making inspections of and repairs in the sewer, whilst the latter were intended to permit the surface water from the streets to flow into it. Moreover, the evidence tends to show that from the river to Sixth street, one block west of Broadway, there were, besides the many private sewer connections, altogether six manholes, which, with the exception of perhaps one, had perforated covers, twenty-seven open inlets and two open vaults. Two of these, with open, grated iron covers, were situated in the alley in the rear of the Fuchs *636place: an open inlet on the southwest corner of Broadway and Chouteau avenue opposite theretój and one manhole somewhat south thereof near the meat market which had a solid cover. But as to this, the removal of the cover would have been useless, because at the request of the people of the neighborhood, the city had put a “goose-neck” device into that manhole to prevent the odors and gases of the sewer from emanating therefrom. It was so constructed that the gas could not issue out of the sewer, cover or no cover. Notwithstanding all these openings, it is apparent that whatever gases there were in the sewer did not rise and escape therefrom. But suppose we concede, as is claimed by respondent, that there were certain closed covers to some of the sewer openings which might have been removed; it does not appear that their removal would have done any practical good. To say so is to indulge in mere conjecture.
III. Assuming that the allegations of respondents’ petition, with respect to the matters hereinbefore referred to have been proved, what was, as a matter of law, the appellant’s duty with respect to the premises?
A sewer is defined to be “a drain or passage to carry off water and filth under ground; a subterraneous channel, particularly in cities.” [Webster’s Dictionary.] Sewers are constructed, as sanitary measures, for the public good, to carry off all sewage, consisting of human excrements and refuse animal and vegetable matter which, as the testimony shows, and as everybody knows, constantly and continuously generates gases, noxious and dangerous, as the result of the constant and continuous process of nature. It is intended and is the object of sewers'to carry off and to guard the community against these gases as much as it is to carry off the substances from which they spring. Sewers are supposed to be covered and to be so constructed as to prevent the escape of gases generated in them. It is not intended that they be permitted to disseminate and breed disease or to cause injury to personal or property rights. If this is not so, then there is no need of sewers, and *637whilst it is riecessary in order that the city may, in the exercise of its ministerial function, properly repair and maintain its sewers, that there should be certain openings or manholes, to permit ingress thereto and egress therefrom, and, hence, it becomes necessary to occasionally remove the covers, the city would be remiss in its duty were it to deliberately remove the covers for the purpose of permitting the escape of all the vile, noxious and dangerous gases which, through nature’s laws, are constantly produced therein.
Thus in City of Atlanta v. Warnock, 91 Ga. 210, Mrs. Warnock brought her petition for injunction against the city of Atlanta, alleging that the city had opened two manholes, two feet in diameter, with perforated tops, into the large sewer extending along Wheat street both within a few feet of petitioner’s property, which fronts two hundred and ninety feet on Wheat street, and one hundred and fifty feet on Oourtland street, where she and her family had resided for twelve years, one of the manholes being on the comer of the two street's and the other on Wheat street, with no trap or obstruction to prevent the foul sewer gas from coming up through the same; that this gas was exceedingly offensive and dangerous to the health and lives of petitioner and other occupants of her premises. It came up in great volumes, especially after a short dry spell and was distressingly troublesome and annoying in warm weather, driving petitioner and her family and friends from the veranda of her house, compelling her to shut the windows in the warmest weather and making life unbearable; that she had made frequent applications to the various officers of the city, all with no effect, etc.; that the way to abate the nuisance was to place a solid instead of a perforated top on the manhole, and she prayed “that the State’s writ of injunction do issue, restraining the said defendant from keeping said mouths of said manholes open, and from continuing said nuisance in front of petitioner’s premises.” The court below, after hearing, ordered “that the defendant be enjoined from *638continuing said manholes in such condition as to allow the escape of noxious gases,” and this action of the lower court was affirmed by the Supreme Court.
Hardy v. City of Brooklyn, 90 N. Y. 435, was an action against the city for damages resulting from a nuisance alleged to have been caused by the negligent construction of a sewer; it appeared that by the plan of the sewer,, as adopted and filed by the board of water commissioners, it ran past plaintiff’s premises to a point where it would find a proper discharge. The sewer was constructed to a point a short distance above the plaintiff’s premises and there a wooden-trough or shute was built to carry off its contents, in consequence of which, noxious and deadly gases were emitted, injuriously affecting plaintiff’s premises. The plaintiff recovered damages and the Court of Appeals affirmed the judgment.
The plaintiff in this case, charges a dereliction of duty on the part of the defendant in not opening its vents or manholes to permit the generated gases to escape. In the Georgia case referred to, the city was enjoined from opening the manholes permitting the gases to escape. In the New York case, the city was mulcted in damages becauses it permitted gases to escape.
We have thus presented to us the anomaly, that if plaintiff’s theory be correct, the city is liable on the one hand for not opening the manholes and for not permitting the 'gases to escape, and on the other hand to be enjoined or mulcted in damages for doing so. As Judge Sherwood says, in his dissenting opinion on the former decision of this case (133 Mo. l. c. 200) : “Again, if the city is to be held responsible for failing to keep open the vents to the sewers within its jurisdiction, is it tp be held liable also if some person passing while the vents are open casts a lighted match into one of them, or the gas from it rises and catches fire from a street lamp, thereby causing an explosion ? Is it possible that the city be held thus responsible whether it does or does not open vents ? *639And yet if the position taken by plaintiff as ground for recovery in this action be correct, that the city is responsible for the gases which breed in its sewers, then the spectacle will soon be presented of actions for damages against the city, because: first, it does not open its sewers and thereby allow the gases therefrom to escape, thereby causing an explosion; because, second, it does open its sewers and thereby an explosion is caused; because, third, it opens its sewers to allow the gases to escape and thereby becomes liable for disease and death, scattered by reason of the escape of such gases.”
That the manholes or vents should not be kept open to permit the escape of dangerous gases, is demonstrated by the facts of this case. If the drain or sewer in the Peters cellar, four or five stores south of Euchs’s place, into which gas was evidently forced from the Mill Creek sewer through connecting sewers, had not been open, permitting gas to escape into the cellar and to come into contact with Humbert’s lighted can-file, the catastrophe would never have happened. The cause of the unfortunate death of Mr. Euchs was because there was an open inlei, through which the gas escaped' which, coming into contact with the flame of the candle, caused the explosion.
We conclude, therefore, that it was not the duty of the city to open its manholes in order to permit the escape of gases that had accumulated in the sewer.
IV. The other allégation of negligence, namely, that the defendant “carelessly and negligently failed to take measures and precautions to prevent gases arising and accumulating in said sewer,” is easily disposed of. There was no evidence to sustain this allegation. As we have stated, the testimony shows that there is a constant generation of inflammable gases in sewers and there was not the slightest suggestion on the part of any one of the witnesses as to any known method of arresting what counsel for appellant aptly calls “one of the never-ceasing processes of nature.” On the contrary, the testimony was all the other way.
*640V. At the trial of this case, the plaintiff, against defendant’s objection, was allowed to show by her expert witness, Prof. H. A. Hunicke, that if a manhole cover had been removed, in his opinion, the use of a fan or blower of seven horsepower would have eliminated the gases so as to- render them non-explosive, by blowing air into or sucking the gases out of the sewer.
The petition in the case alleged two specific acts of negligence against the city: a failure to prevent the formation and accumulation of gases, and a failure to ventilate the sewer by opening the vents specially designed to carry off gases. There was no allegation of neglect of duty on the part of the city in not using a fan or blower or any other contrivance or design to carry off or destroy the gases. It is settled that where specific acts of negligence are charged, evidence of other acts is inadmissible. [Atchison v. Railroad, 80 Mo. 213.]
The testimony should therefore have been excluded.
There are other reasons why the testimony of this witness should have been disregarded, and the jury should have been instructed so to do as requested by ¿ppellant.
His testimony, so far as it is necessary to refer to it, is substantially as follows:
“I am a consulting chemist and mining engineer, and have been for sixteen years. I was for four years professor of applied chemistry at Washington University, St. Louis, from whiehs university I am a graduate. I have had practical experience in mines in New Mexico, Colorado, and Illinois, and am familiar with coal oils and other substances. Petroleum is the term applied to a mineral oil occurring in nature. Petroleum means oil from the rock and is equivalent to crude oil as it is pumped from the earth. Petroleum when released will evaporate at all temperatures. It will give up the lighter constituents until the air about it is saturated. It is the same as with water. The air around it will take up so much water as will saturate it. The evaporation naturally increases as the *641temperature increases; that is, the higher the temperature, the more rapid the evaporation.' The’less volatile and more solid portion of the oil will remain. All the vapors of petroleum are combustible, and will explode when mixed with the proper amount of air or oxygen. Explosion is a rapid burning. In order to produce an explosion there must be an ignition. . . . None of these oils or products that I have mentioned appear in mines, but gases do. Gases accumulate more in coal mines than in metal mines. I had some experience in taking care of mines and removing gases from them. I have heard the manholes on Fourth and Fifth streets described, but don’t know that I have ever seen them.
“Q. Now, suppose these manhole covers were removed, or a monhole cover was removed from one of the manholes, what scientific appliance was there, well known previous to 1892, for creating a draft through this manhole ?
“To which question counsel for defendant objected on the ground that the witness had not qualified or undertaken to. qualify to answer the question, he not having pretended that he ever made a study of the science or method of removing gases from sewers.
“The Court: From the experience you have had, practically and scientifically, can you answer that question? A. To know what to do ?
“The Court: Yes. A. I might say at the outset that it is an engineering, not a scientific problem, -and would depend very largely on the individual who has charge of it.
“Q. As an engineer, can you answer us ? A. Well, things that might suggest themselves to me as a mining engineer might not present themselves to others.
“Q. Professor, I will ask you the question this way: Was there previous to 1892, any well-known appliance or means for ventilating, by means of which a sewer could be ventilated — a manhole cover being removed ?
*642“To which question defendant, by its counsel, again objected, on the ground that the witness had not qualified as an expert on the ventilating of sewers, and had further already stated that he is a mining engineer, and what would suggest itself to his mind might not suggest itself to an engineer in charge of a sewer.
“The Court: If you know of any one you may answer it.
“To which action and ruling of the court, defendant by counsel then and there duly excepted.
“A. So it refers only to. one hole ?
“Q. Yes.”
Counsel for defendant again objected on the ground that this question of the witness and answer of counsel made the question still more incompetent.
The witness proceeded: “The means would have been possible, but under the circumstances it would require considerable time to eliminate the gases by simply opening a manhole. It would be necessary to apply means, and nothing short of a blower would have done in that case. A blower of say seven horsepower would have removed the air, or replaced the air sufficiently to be non-explosive within thirty-six hours. When the crude oil got into the sewer, it would tend to evaporate and rise, like sugar in coffee, although the sugar is heavier than the coffee. This vapor rising up in the sewer, would continue through the air until the air was saturated and could take up no more, and will tend to creep out through any overhead holes. The most natural point for placing the blower would be at the opening closest to the river. Either a pressure blower or suction fan might have been applied at a point further up. Fans will either blow or draw air, ,so the point may differ. It would be preferable to blow the air in instead ■of drawing it out.”
On cross-examination he testified:
“I never saw any practical attempt made, either in the city of St. Louis or in this country, or anywhere in the world, *643to extract gas out of a sewer by .«means of a fan or blower. Neither did I ever know of any sewer commissioner anywhere in the United States attempting to blow air into a sewer filled with gas, or suck air out of a sewer filled with gas, by means of a fan. I can’t give any positive information touching that, and I never saw it done. I know that experiments have been made in London, but I have no knowledge or observation of its ever having been put to practical use. I don’t know whether in order to suck air out of a sewer it would be necessary to close all of the apertures except the one at which the fan was at work in order to create a suction. That would depend upon what the purpose of the suction was. If you were trying to suck gas out of this courtroom and the windows were all up, and you put your fan at one of the windows, it would not work. In order to make the fan work and create a suction, I would want an opening at which to put the fan, and the balance of the openings closed. I would want a certain ratio of inlet and outlet. I don’t know how many drains, inlets, openings or connections there were to the Mill Creek sewer within a circle of five hundred feet around the manhole where I stood and would station the blower or fan, but I say that without knowing that and without myself having experimented with this sewer, or having seen it attempted anywhere in the world, my plan would have prevented an explosion. I say that because the connections can be opened and closed, and if the gas 'had been sucked out the fresh air would have rushed in to supply its place to such an extent that a flame held at the opening would not have ignited it. I have never attempted to use a blower myself in connection with a sewer, nor have I ever seen one used, and, as I have stated, I have no experience or observation in that direction. I know that the Mill Creek sewer extends to a mile west of the river, and that other sewers of various sizes and from all directions lead into it, and I know that some of them have open or perforated covers. But notwithstanding all these connections and openings into the main *644sewer, I should not hesitate to apply the blower on my plan. I know that a blower would have been effective without ever having experimented. I do not recollect whether or not at the last trial, I testified that the thought of' using the fan or blower had occurred to me as a scientific possibility, but that without a practical experiment I could not say whether it would accomplish any result; since I testified last time I have made no experiments to ascertain whether the plan was a practicable or feasible one.”
It will appear from the foregoing that the witness is a mining engineer and a consulting chemist; that he never, as a scientist, investigated the ventilation of sewers and never gave the subject of sewers a particular or special study; that he did not claim that his proposed remedy had ever been practically applied or even made the subject of experiment by himself; he says that he “never saw any practical attempt made in St. Louis, or in this country, or anywhere in the world,” to extract gases out of a sewer by means of a fan or blower, nor did he ever know of any sewer commissioner anywhere within the United States attempting to blow air into a sewer filled with gas, or suck air out of a sewer filled with gas, by means of a fan. He could not give any positive information about that and he never saw it done. Experiments were made in London, but he had no knowledge or observation of its ever having been put to practical use. He never had attempted to use a blower in connection with a sewer, nor had he ever seen one used, nor had he ever had any experience or observation in that direction. He had testified in the case twice before, but had made no experiments with respect to the matter since that time. So far as the record shows, Prof. Hunicke’s proposed remedy is a mere theory of his own. Practical engineers who had devoted years to the study and management of sewers testified that Mr. Hunicke’ scheme is entirely impracticable. Not one of them had ever heard of it. Not one of them knew of any known practical or feasible method by which the gases *645could have been extracted from the sewer. It would be monstrous to predicate a charge of negligence against the city on such testimony as that.
It should never have been admitted; having been admitted, it should have been entirely disregarded and the jury instructed to do so as requested by appellant in its instruction numbered 2 refused by the court. [Graney v. Railroad, 157 Mo. 666.]
VI. Eliminating from consideration the testimony of witness Hunieke, and from what has been heretofore stated, it will appear that the respondent had nothing on which to go to the jury save the bare fact that the sewer exploded. But the mere fact of an accident and the consequent injury resulting therefrom does not, as a rule, make out a prima facie case. There are cases where the doctrine of res ipsa loquitur applies, but we can find no adjudicated ease where that doctrine was applied to a case like the one before us. Except in cases relating to common carriers of goods and passengers or arising out of other contractual relations, the mere fact of an explosion, without affirmative proof of negligence, does not raise a prima facie presumption of negligence on the part of defendant. [Huff v. Austin, 46 Ohio St. 386; Cosulich v. S. O. Co., 122 N. Y. 123; Walker v. Railroad, 71 Iowa 658; Losee v. Buchanan, 51 N. Y. 476.] The plaintiff, therefore, made out no case to be submitted to the jury and she is not aided by any testimony introduced by defendant.
VII. Irrespective of what has been said before, there is a vital and fatal objection to respondent’s right of recovery in this action.
In the case of the American Brewing Ass’n v. Talbot, 141 Mo. l. c. 683, the court says: “Numerous authorities hold that it is not negligence not to take precautionary measures to prevent an injury which, if taken, would have prevented it, when the injury could not reasonably have been anticipated and would not, unless under exceptional circumstances, have *646happened.” And in support of that doctrine, the court cites, with approval, in addition to a large, number of cases, Ray on Negligence of Imposed Duties, pp. 133, 134, as follows:
“Mischief which could by no reasonable possibility have been foreseen, and which no reasonable person would have anticipated, can not be taken into account as a basis upon which to predicate a wrong. A reasonable man does not consult his imagination, but can be guided only by a reasonable estimate of probabilities. The reasonable man, then, to whose ideal behavior we are to look as the standard of duty, will neither neglect what his reason and experience will enable him to forecast as probable, nor conduct, on a basis of bare chances, a business whose success is dependent upon his accuracy in forecasting the future. He will order his precaution by the measure of what appears likely in the usual course of things. The proper inquiry is not whether the accident might have been avoided if the one charged with negligence had anticipated its occurrence, but whether, taking the circumstances as they then existed, he was negligent in failing to anticipate and provide against the occurrence. The duty imposed does not require the use of every possible precaution to avoid injury to individuals, nor of any particular means which it may appear, after the accident, would have avoided it. The requirement is only to use such reasonable precautions to prevent accidents as would have been adopted by prudent persons prior to the accident. The prudence and propriety of men’s doings are not judged by the event, but by the circumstances under which they act. If they act with reasonable prudence and good judgment, they are not to be made responsible because the event from causes which could not be forseen nor reasonably anticipated had disappointed their expectations.”
Also Webb’s Pollock on Torts (Enlarged Am. Ed.), pages 45 and 46, as follows:
“Now a reasonable man can be guided only by a reasonable estimate of probabilities. If men went about to guard *647themselves against every risk to themselves or others which might by ingenious conjecture be conceived as possible, human affairs could not be carried on at all. The reasonable man, then, to whose ideal behavior we are to look as the standard of duty, will neither neglect what he can forecast as probable, nor waste his anxiety on events that are barely possible. He will order his precaution by the measure of what appears likely in the'known course of things. This being the standard, it follows that if in a particular case (not being within certain special and more stringent rules) the harm complained of is not such as a reasonable man in the defendant’s place should have foreseen as likely to happen, there is no wrong and no liability. And the statement proposed, though not positively laid down, in Greenland v. Chaplin, 5 Ex. 248, namely, ‘that a person'is expected to anticipate and guard against all reasonable consequences, but that he is not, by the law of England, expected to anticipate and guard against that which no reasonable man would expect to occur,’ appears to contain the only rule tenable on principle where the liability is founded solely on negligence. ‘Mischief which could by no possibility have been foreseen, and which no reasonable person would have anticipated’ may be the ground of legal compensation under some rule of exceptional severity, and such rules, for various' reasons, exist; but under an ordinary rule of due care and caution it can not be taken into account.”
The same principle is recognized and applied in Sullivan v. Railroad, 133 Mo. l, and also in the large number of cases cited in the dissenting opinion in Fuchs v. St. Louis, 133 Mo. 168. A reference to a few other cases, very analogous to the case before us, is not inappropriate.
In Sofield v. Sommers, 9 Ben. 526, the facts were these:
The fumes of crude petroleum, carried in a tank on a lighter used in the oil trade, escaped into a locker, which locker — there being no watchman on board, when the lighter lay one night with other vessels at a pier in Jersey City — was *648forced open during the night by a thief, who, exploring the locker with a lighted match, set fire to. the gas and caused an explosion and a fire, whereby the lighter and the libellant’s lighter that lay alongside were destroyed: Held, that the escape of the gas into the locker was an accident, and the presence of a lighted match in the locker not the natural result of the absence of a watchman; that between the act of omission charged upon the defendant and the explosion charged, there intervened an independent human agency, the presence of which had no natural relation to any act of defendant and which therefore entailed no responsibility upon the defendant for the explosion.
In the case of Stone v. Railroad, 171 Mass. 536, decided by the Supreme Judicial Court of Massachusetts, in July, 1898, the facts were that the railroad company maintained a passenger station, freight house and freight yard, all adjoining a public street or highway, in the village of Spencer. On one. side of the freight house was a wooden platform about eight feet wide and set on posts in the ground about four feet high. The under side of the platform was left open and exposed. It seems that the railroad company was in the habit of storing oil in barrels on and under this platform until called for and taken away by the consignees, and that in consequence thereof the platform had become saturated with oil, much of which dripped to the ground underneath, saturating the rubbish and straw which had collected there. On September 13, 1893, one Casserly, a teamster, brought a load of goods to this platform to be shipped by the railway company. Lighting his pipe, he threw down the match, which fell to the ground, underneath the platform, and in consequence of the saturation of the rubbish and the platform with oil a fire resulted which spread with great rapidity, exploding several barrels of oil, and soon communicating with plaintiff’s buildings fronting upon the street, and about seventy-five feet distant from the platform, which buildings it destroyed. The testimony of plaintiff *649showed that some twenty-five to thirty barrels of oil had been left upon the platform for more than forty-eight hours prior to the accident; that the platform had been thoroughly saturated with oil for at least eight years. The action was to recover from the railroad company the value of plaintiff’s buildings thus destroyed. The lower court directed a verdict for defendant, upon which judgment was entered, and this action of the lower court is approved and affirmed by the Supreme Court. After considering the evidence somewhat in detail, the court reaches the conclusion that the railroad company was guilty of negligence in the keeping of the oil and in 'the maintenance of the platform saturated as it was with oil. Having reached this conclusion the court proceeds to inquire whether this state of facts was the proximate cause of the accident, so as to result in the liability of the company. Discussing this feature of the case, the court say:-
“Nevertheless the question remains, and in our view this becomes the important and decisive question of the case, whether, assuming that the defendant was thus in fault, the plaintiff introduced any evidence which would warrant any finding by the jury that the damage to his property was a consequence for which the defendant is responsible; or, in other words, whether the act of Casserly in starting the fire was such a consequence of the defendant’s original wrong in allowing the oil to remain upon the platform that the defendant is responsible to the plaintiff for it.....The rule is very often stated that in law the proximate and not the remote cause is to be regarded; and in applying this rule it is sometimes said that the law will not look back from the injurious consequences beyond the last sufficient cause, and especially that where an intelligent and responsible human being has intervened between the original cause and the resulting damage, the law will not look back beyond him. This ground of exonerating an ■original wrongdoer may be found discussed or suggested in the following decisions and text-books, among others.”
*650Here follows a host of references bearing upon the question. Thereupon the court continues its discussion:
“It can not, however, be considered that in all cases the-intervention even of a responsible and intelligent human being will absolutely exonerate a preceding wrongdoer. Many instances to the contrary have occurred, and these are usually eases where it has been found that it was the duty of the original wrongdoer to anticipate and provide against such intervention, because such intervention was a thing lilcely to happen in the ordinary course of events.....Was the starting of the-fire by Oasserly the natural and probable consequence of the defendant’s negligent act in leaving the oil upon the platform ?' According to the usual experience of mankind, ought this result to have been apprehended ? The question is not whether it was a possible consequence, but whether it was probable-, that is, likely to occur, according to the usual experience of mankind. That this is the true test of responsibility applicable to-a case like this has been held in very many cases, according to which a wrongdoer is not responsible for a consequence which is merely possible, according to occasional experience, but only for a consequence which is probable, according to ordinary and usual experience. One is bound to- anticipate and provide against what usually happens, and what is likely to happen; but it- would impose too heavy a responsibility to hold him bound in like manner to guard against what is unusual and unlikely to happen or what, as it is sometimes said, is only remotely and slightly probable. A high degree of caution might, and perhaps would, guard against injurious consequences-which are merely possible; but it is not negligence, in a legal sense, to omit to do so.....Tried by this test, the defendant. is not- responsible for Casserly’s act. There was no close connection between it and the defendant’s negligence. There was-nothing to show that such a consequence had ever happened before, during the eight years covered by the plaintiff’s testimony, or that there were any existing circumstances which-*651made it probable that it would happen. It was of course possible that some careless person might come along and throw down a lighted match where a fire would be started by it. But it was not according to the usual and ordinary course of events. In failing to anticipate and guard against such an occurrence or accident, the defendant violated no legal duty which it owed to the plaintiff.”
It being contended, however, that the negligence of the railroad company concurred with that of Casserly to produce the injury, the court met that contention as follows:
“The plaintiff further contends' that the negligence of the defendant in keeping the oil upon the platform was concurrent with the careless act of Casserly, and that therefore it was a case where two wrongdoers, acting at the same time, contributed to the injurious result. But this is not a just view of the matter. The negligence of the defendant preceded that of Casserly, and was an existing fact when he intervened.”
The application of the principles announced in these cases and text-books, to the case before us is obvious.
To warrant the inference of negligence on the part of the city, it is necessary to conclude:
1. That oil works would burn.
2. That by reason thereof oil would flow into’ the sewer.
3. That oil would reach the sewer in sufficient quantities to generate gas sufficient in quantity to explode it.
4. That during the period whilst the oil was in the sewer, some one with a lighted candle would cross a drain leading into a private sewer, leading in turn into another sewer, which, in turn, led into the Mill Creek sewer; and
5. That before the happening of this latter occurrence, the city could have removed the oil or the gases, although as the testimony shows, no method for doing so existed unless we adopt the individual theory of the witness Hunicke which, as has already been stated, should be entirely eliminated from consideration in this case.
*652In short, this ease presents the instance of an injury •which could not have reasonably been anticipated; it would not, unless in exceptional cases, have happened; whilst the explosion was possible, it was not probable, according to ordinary and usual experience; although the sewer had been in existence for some thirty-four years, no such explosion had ever happened before, and in all likelihood, will never happen again. Had it not been for Humpert’s legitimate act, the oil and the gases might have remained in the sewer without injury to any one until the subsidence of the high water in the river and without any foundation for fear of hurtful consequences.
The facts of the case simply present a case of inevitable accident for which the city can not be held responsible because it was not negligent.
VIII. Counsel for respondent strenuously urges upon the court the plea of res adjudicada, relying, in its support, upon the opinion and judgment of this court in the former appeal between these parties.
When the ease was here before, it came up on plaintiff’s appeal from an involuntary nonsuit which she was compelled to suffer. The only question which was then presented to the court was whether upon the evidence then submitted to the jury and preserved in the record, she made out a prima facie case entitling her to go to the jury. The court held that she did make out such prima facie case and reversed the judgment and remanded the case for a new trial. Upon the retrial, her case was sent to the jury, countervailing evidence was offered by the defendant, instructions were given, others refused, rulings were had on the competency and admissibility of testimony and the appellant now appeals from the judgment rendered against it in favor of the respondent. Appellant is, of course, entitled to be heard on the matters of which it complains and to have the case, as it appears from the present record, reviewed by the court. Moreover, a careful comparison of the facts as they appear from the record now here, *653with the facts as stated by Judge Barclay in his opinion based upon the record then before the court, shows that they are not the same, but differ in many essential particulars. Under such circumstances, the doctrine of res adjudicata does not apply. A reference to a few cases on this subject decided by our own appellate courts will suffice. [Steinhauser v. Spraul, 114 Mo. 551; Kelly v. Thuey, 143 Mo. l. c. 437; May v. Crawford, 150 Mo. l. c. 525; Grattan & Knight Mfg. Co. v. Troll, 77 Mo. App. l. c. 344.]
The view we have taken of this case makes it unnecessary to review the instructions given and-refused
This suit was instituted in September, 1892. It was tried three times in the St. Louis Circuit Court, once in the county of St. Louis, and it was argued in this court at least four times. As in our opinion, the defendant can not be held liable in this action, it will serve no useful purpose to remand the case for a new trial.
In so far as the opinion in Euchs v. St. Louis, 133 Mo. 168, conflicts with this opinion, it is overruled and the judgment of the court below is, with the concurrence of the other judges, reversed.
Burgess, C. J., Sherwood, Robinson, Valliant, and Gantt, JJ., concur; Brace, J., concurs except as to paragraph one and five of the opinion; Marshall, J., not sitting.