Plaintiff was the owner of a building which was nearing completion. It was located on a lot in San Francisco. For several days prior to February 13, 1909, and on that day defendant had been installing gas pipes in the basement of plaintiff’s building. The said building extended from Valencia Street to San Jose Avenue and from the latter thoroughfare defendant had laid a three-inch gas main which passed under the sidewalk into the basement and was carried a distance of thirty-five feet along joists to the floor of a, light-well, where it turned and was conducted near the wall of the light-well to a height of fifteen feet. The light-well, which was about ten feet square, was almost in the center of the building and extended the whole distance to the roof, but was itself open to the sky. Two of the store rooms of the building had been thrown into one. This large apartment was completed and had been occupied by the United States government as a branch post-office for about two weeks before February 13, 1909. No other rooms in the building had been plastered and none was occupied. On the morning of February 13, 1909, about 2 o’clock the building caught fire and was almost entirely destroyed, the four walls and a portion of the room occupied by the post-office being all that was left. Nearly two years later plaintiff brought this suit for damages, alleging in its complaint that defendant had negligently and carelessly omitted to fasten and construct the pipes in said building in a substantial manner or to inspect said pipes; that the main pipe fell from its support on the morning of Feb ruary .13, 1909, thereby breaking the coupling and allowing large quantities of gas to escape into the plaintiff’s building; and that the fire which destroyed the building was caused by the ignition of this escaping gas. The amount demanded by way of damages was eighteen thousand dollars. After hearing all of the evidence for the plaintiff the court granted a nonsuit. Plaintiff appeals from the judgment of nonsuit and from the order denying a motion for a new trial.

The appellant’s counsel insists that a prima facie case was established. The court found that the defendant negligently failed to strap its gas supply pipe and rested said pipe on two wooden arms, one by three inches, nailed to the wooden uprights of the building. There was some evidence on the part of the firemen that this was a “gas fire”—that is that when they arrived and found the building in flames they recognized the fact that the gas was escaping and burning. From these circumstances appellant’s counsel reasons that, the existence of a “gas fire” being shown, the burden is on the defendant to prove that the escaping and burning gas was not the cause but the result of the fire which consumed the building. In this behalf counsel cites such cases as Judson v. Giant Powder Co., 107 Cal. 549, 552 , [ 48 Am. St. Rep. 146 , 29 L. R. A. 718, 40 Pac. 1020 ], and Linforth v. San Francisco Gas etc. Co., 156 Cal. 60 , [19 Ann. Cas. 1230, 103 Pac. 320 ]. In the first case it was held that proof of the explosion and consequent damage to plaintiff’s property shifted the burden to defendant to prove its freedom from blame, because dynamite factories do not blow up when they are properly operated. In that ease there was no doubt of the explosion itself. In the Linforth case there was proof, partly furnished by the defendant’s witnesses, that the explosion which wrecked the property of Mr. Linforth, was caused by gas. The explosion being proven and its origin having been established as escaping gas, the burden shifted to defendant to show that something other than its careless installation and imperfect maintenance of the pipes caused the leak which produced the explosion. So in this ease, if the origin of the fire had been established as escaping gas and its ignition, the same rule would apply. To complete that part of its ease it would be necessary for plaintiff to show, either directly or circumstantially: 1. That the pipe was not properly secured; 2. That its weight caused it to break away from its supports and to fall; 3. That in falling it broke from its connections, thus permitting the gas to escape into the building; 4. That the escaping gas was so diffused that it passed through small spaces in a door leading into the post-office; 5. That it was there ignited, because the only burning gas jets in the building were in the post-office; and 6th, that it was so diffused and ignited within seventy, seconds, because it was proven that the jets in the post-office would continue to burn that length of time after the shutting off of the supply at the main.

Mr. Clute, president of the plaintiff corporation, testified that when he examined the ruins some days after the fire he went into the basement and found the pipe broken off. The whole theory of plaintiff’s case was based upon the supposed fact that the break in the pipe preceded the starting of the fire. The testimony tended to show that the gas was burning at the jets in the post-office while the building was in flames. One of the firemen, George H. Thomas, said, while he was on the witness-stand, that he could not be positive, but his recollection was that the gas at the jets was burning when he reached the building. At that time the fire had “a big start” and was “burning rapidly.” Three of plaintiff’s witnesses described flames around the connections of the meter through which the gas supply for the post-office passed. This was located on the wall of the light-well about thirty-five feet from the place where the pipe was broken. The fact that such flames were seen is absolutely at variance with plaintiff’s theory that the liberated gas from the broken pipe caused the fire. Mir. Anderson, an expert witness produced by plaintiff, testified, with reference to the lights around the meter: “If the light was burning for any length of time and kept burning persistently, of course the main could not be broken. However, there could be gas burning there daring the course of the fire, because the gas pipe would be broken and the gas would exude from the gas pipes.”

Part of plaintiff’s theory was that the gas escaped from the broken main and entered the post-office through openings in a door. The proof was that the glass part of the door had been broken but the opening had been boarded up with tongue and groove lumber which left a space an inch and a half wide on the side of the boards. At the bottom of the door was an aperture through which mail was sometimes thrust. Through these openings the gas must have passed if the fire was caused as plaintiff contends. It must have traveled very rapidly to have caused an ignition before the lights would disappear because of lack of gas in the pipes due to the cutting off of the supply. In order that there might have been an ignition without a serious explosion it would have been necessary that the amount of gas in the room should have greatly exceeded the amount of air: The expert so testified. He also gave this answer to the question which is quoted below:

“Q. And if that window was boarded up—had boards nailed across it—and the only aperture was the space through which letters could be stuck that was underneath this door, •you would not expect the gas to get into that room in such quantities and with such rapidity that there would be only a small explosion, would you ? A. Under the conditions as you have mentioned the window being pretty thoroughly boarded up and no other accsss to that flame, why the answer certainly is that there would not be.” Yet there was no evidence of any explosion and the testimony showed that the post-office was less damaged than any other part of the building.

It will thus be seen that there was an utter failure to establish a causal connection between the only facts that may be taken as proved,—namely the insecure fastening of the pipe and the burning of the building. All the rest of plaintiff’s case rested on inference piled upon inference. The building may have been set on fire by an incendiary, or a lighted cigarette dropped by a workman may have ignited debris which smouldered for hours before bursting into flame. These theories are pure conjectures, yet they are no more unsubstantial than the attempted proof that the defendant’s negligence was the proximate cause of the fire. The fact that the pipe was insecurely fastened, coupled with the other facts that it was found broken after the fire and that gas was burning in the building during the conflagration, does not establish plaintiff’s case. The learned judge of the trial court said: “As there is no evidence in this record from which I can say whether it was the burning of the building that broke the pipe, or the breaking of the pipe that caused the building to be burned, the plaintiff has not made out a ease.” And, to quote Puckhaber v. Southern Pacific Co., 132 Gal. 364, [ 64 Pac. 480 ]: “By proving defendant’s negligence, without in some way fastening that negligence to the injury, a case is not made out.” The connecting links between the negligence and the catastrophe are wholly lacking in this case. The judgment and order are affirmed.

Henshaw, J., and Lorigan, J., concurred.