1' performing13 act in ordinary way. On the 13th day of June, 1916, the defendant Cedar Rapids Oil Company was engaged in the .business of wholesaling and retailing gasoline. Its place of business was located at Cedar Rapids. The defendant Barth was its local agent at the town ^ paye|ocjc jn Pocahontas County. On that day, the plaintiff ordered, through Barth, a barrel of gasoline, to be delivered at his garage and implement store in the town. Barth, for the company, accepted the order, and undertook to deliver the gasoline; did deliver it in a metal barrel. Plaintiff’s and Barth’s places of business were on the same street, and about 240 feet apart. Barth delivered the gasoline by rolling the barrel from his place to plaintiff’s. When he reached the place of delivery, he undertook to empty the barrel for the plaintiff, into an underground tank provided by the plaintiff for its reception. While in the act of emptying the barrel, for some reason it became ignited, and the fire extended to and burned plaintiff’s buildings. It is to recover the value of the buildings so destroyed that plaintiff brings this action.

The action is bottomed on negligence, and the negligence charged consists:

(1) In that the gasoline was kept in a metal storehouse.

*555(2) In that Barth rolled the barrel on the street from the storehouse to plaintiff’s place of business on an unusually hot day.

(3) In rolling the barrel into the implement house and there removing the bung and releasing the gas, while the usual and proper way would have been to have carried it from the store in buckets.

(4) In that the barrel was too full.

(5) That the defendant Oil Company had never instructed Barth how to handle high test gas so as to avoid an explosion.

The cause was tried to a jury. At the conclusion of the evidence, the court instructed the jury to return a verdict for the defendants. A verdict was returned, and judgment entered upon the verdict. Plaintiff appeals, and contends that the cause should have gone to the jury; that the court erred in instructing a verdict for the defendants.

To sustain plaintiff’s claim, two things must appear: (1) That the acts charged constitute negligence; (2) that this negligence was the proximate cause of the fire which destroyed plaintiff’s buildings.

Considering the grounds of negligence relied upon in the order in which they are presented, we have to say that there is absolutely no evidence that the defendant is guilty of any negligence in storing its gasoline in a metal storehouse. The evidence rather tends to show that it is the usual, ordinary, and safe method of storing it.

There is no evidence that this gas was high test. The evidence is that it was not high test gas. The charge in the fifth division is that Barth was never instructed how to handle high test gas so as to avoid an explosion. It is not claimed that any instructions are necessary where it is not high test gas.

*5562‘ n^uSusative connection. *555The second and third may be considered together, to wit, the negligence of the defendant, predicated on the *556manner in which, the gasoline was conveyed to plaintiff’s storehouse, and the maimer in which Barth undertook to empty it into the underground tank placed to receive it. The contention of the plaintiff is that the rolling of the barrel 240 feet on a very hot day had a tendency to increase the pressure by causing the formation of gas in the barrel, which, upon opening the barrel and exposing the gas to the air, produced spontaneous combustion; that this was especially true when the weather was hot; that the acts of defendant done in this way produced a gaseous formation in the barrel, which escaped when Barth undertook to empty the barrel in the manner in which he did, and that this gas, meeting with the air, produced instantaneous combustion, from which, as a natural sequence, the results complained of followed. We may assume, for the purposes of this case, that the rolling of this barrel on this day had a tendency to, and did, in fact, increase the pressure, and cause the gaseous formation within the barrel which escaped immediately upon the opening of the barrel for the purpose of emptying it into the tank.

That the barrel was opened in the usual and ordinary way and in the proper way, is not controverted. The question is: Was the defendant negligent in opening it and attempting to empty it into the tank after the conditions had been produced and existed in the barrel as the natural result of the rolling? And next, Were the consequences that followed due to the condition of this oil in the barrel, and did the fire follow as- a proximate result of such condition upon the opening of the barrel ?

The record discloses that at no time on this day did the temperature exceed 76 degrees Fahrenheit. The plaintiff’s testimony tends to show that gasoline vaporizes readily; that it combines with the air at practically any temperature readily; that, if the temperature is raised, it *557vaporizes more readily; that liquid gasoline will not explode. It must first be broken up into vapor, and that vapor, escaping and combining with the oxygen in the air, causes' an explosion. It is due to chemical action — the union of the oxygen from the air with the carbon in the gasoline. A glowing or red hot coal will not ignite the gas. An open flame will. A spark is a flame. It is possible for vapor to explode without coming in contact with either a flame or a spark. The conditions are similar to spontaneous combustion. The generating of an excessive quantity of gas from the gasoline, if given a chance for it to unite with the air, would mean a chemical compound or combining chemically of the gas and the oxygen in the air, and that would result in an explosion. Generally, however, sparks and flames are the two essential causes of ignition of the gas that forms from gasoline. It is exceptional for gasoline to ignite from any other causes than sparks or flame. This witness, however, testified that a temperature of 76 degrees Fahrenheit would not be sufficient temperature to suggest any unusual precaution in handling gasoline. A witness for defendant testified that it would be impossible, with a temperature of 76 degrees Fahrenheit, for gasoline or gas vapor passing through the venthole or bunghole of a barrel to be raised to a temperature of 900 degrees; that it takes a temperature of 900 degrees Fahrenheit to ignite gas; that gasoline vapor will not explode without a spark or flame.

Another witness testified that there could never be a flame from gas escaping from a crack or vent in one of these drums when the pressure would not be sufficient to burst the drum; that there could not be a flame without external fire or flame.

In fact, all the testimony offered both by the plaintiff and the defendant shows that spontaneous combustion could not follow and produce flame and fire under the con*558ditions and at the temperature of the air at the time this barrel was opened.

The evidence is undisputed that, at the time Barth undertook to empty this gasoline, there was no fire anywhere about the place. The question then arises: Was the fire that destroyed these buildings traceable to the manner in which the oil was handled, or to the manner in which Barth attempted to empty it into the tank, as its proximate cause?

There is evidence that one Murray, an employee of the plaintiff’s, was present at the time the gasoline was .being turned into the tank; that he had a cigarette in his hand; that he was under the influence of liquor; and that, about the time the gasoline started fairly running into the tank, someone struck a match. Negativing this, however, there is testimony that no match was struck, or, at least, that no one saw a match struck. So, upon this point, there might well have been a question for the jury. But we have to say that the evidence is uncontroverted that spontaneous combustion could not take place under the conditions that attended the emptying of this barrel. Plaintiff’s whole contention rests on the theory that the handling of the barrel produced a condition in the barrel which, in attempting to empty the barrel, produced spontaneous combustion, from which the fire followed that produced the injury. It positively appearing that spontaneous combustion could not follow, the question arises: What caused the fire? As there was no fire or light or flame in the building when defendant' Barth took the barrel in, something must have happened which brought about the condition which produced the fire and destroyed the building. What it was, if no match was struck, does not appear. It must, then, have been caused by the intervention of some independent agency hot known to or controlled by these defendants; an agency intervening which they could not and had no reason *559to* anticipate. If it could not occur without a spark or a flame, as all of the witnesses agree, under the conditions there existing, the spark or flame must have been produced by someone. We are inclined to think that it is traceable to the action of Murray, who is not a witness in this suit. It is certainly not traceable to any conduct of these defendants for which they are culpable. Negligence must be proven. Some act must be shown to have been committed by these defendants, or either of them, which produced the condition which caused the injury. The condition which they did produce was not the proximate cause of the injury. It could not come without the intervention of an independent agency. They are in no way shown to be responsible for the intervention of the independent agency. Therefore, the acts of this defendant are not shown to be the proximate cause of the injury, and they cannot be holden for the injury.

Some question is made about the introduction of evidence, but we have examined these points, and find no ground for reversal there.

Upon the whole case, we find that the court was right in directing the jury to return a verdict for the defendant because of a failure of proof of the issue that the fire was due and traceable directly to the acts of the defendant as its proximate cause. The case is, therefore, — Affirmed.