Plaintiff assigns error:
1. That it was error for tbe court to sustain defendant’s objection when plaintiff offered to show by the witness J. R. Bowden that the day before the fire complained of the defendant’s locomotive on the through freight, going north, put out fire on the witness’s lands near the plaintiff’s mill.
2. That the court also erred in allowing the defendant’s motion for judgment as of nonsuit.
The first exception cannot be sustained. Ice Co. v. R. R., 126 N. C., 800; R. R. v. Smith, 55 Southern, 871; Kerner v. R. R. 170 N. C., 94.
In the last case it is said: “It is conceded that where a fatal fire has been set out from a designated or known engine, it is admissible to introduce evidence of other fires previously set out by the same engine, for the purpose of showing its defective condition, but-the rule has never been extended iso as to permit evidence of sparks emitted by some other engine at some other time and place.”
The motion to nonsuit was likewise properly allowed.
If the evidence is taken in the light most favorable to. the plaintiff, as it must be on a motion for nonsuit, the following facts are shown: That the mill plant was located on the east side of the railroad, 70 feet from it; that the wind was blowing across the railroad track towards the mill, but there was not much wind blowing. The fire caught on the side of the mill next to the railroad, about 1 o’clock in the morn*313ing, and two trains bad passed along the railroad track within thirty minutes before the fire was discovered; that the mill was located on a grade which started about 200 yards below or south of the mill. There was a fire in the boiler of the plant, which was. about 280 feet from where the fire started. No other fires were kept going at night except that.
When first discovered, about 1 o’clock at night, the fire was next to the railroad and the side of the planing mill next to the railroad was burning at the north end. The lumber shed and mills are all together. When discovered, the fire had not reached quite to the top of the lumber shed.
There is no evidence tending to show' that either of the engines which passed the mill within thirty minutes before the fire were throwing any sparks, nor were any burned cinders discovered between the track and the mill. There was evidence that there were rotten lumber, trash, and shavings in the mill shed. There was no evidence that the fire was communicated from the right of way or that the right of way was in a foul condition, or had any inflammable matter on it.
The defendant’s, evidence established the fact, if true, that the engines were properly equipped with spark arresters .and skillfully operated.
There is no difference of opinion as to the láw applicable to this case. It is settled that' if the plaintiff has introduced evidence sufficient in probative force to justify a jury in finding that the fire was caused by a spark from defendant’s engine, the issue should have been submitted, the weight of the evidence being a matter for the jury. In such case' the defendant is called upon to prove that its engine was properly equipped and operated. If so equipped and operated, there is no negligence or liability upon the part of defendant. Williams v. R. R., 140 N. C., 624; Aman v. Lumber Co., 160 N. C., 371; McRainey v. R. R., 168 N.C., 571.
It is undoubtedly true that the fact in controversy here, as to the origin of the fie, may be established by circumstancial evidence, but the circumstances proven must have sufficient probative force to justify a jury in finding that the fire originated from a spark from defendant’s engine before the issues can be submitted to them. What is sufficient evidence to justify the court in submitting a controverted issue to the jury has been much debated.
This Court has used various forms of expression in commenting on the subject. In Lewis v. Steamship Co., 132 N. C., 904, the Court holds that if the evidence is “conjectural or speculative, it should not be submitted to the jury.” In S. v. Satterfield, 121 N. C., 558, the Court *314bolds tbat it is for tbe court to find “tbat tbe evidence is sucb as would satisfy a jury in proceeding to a verdict, sucb as will satisfy an impartial mind.”
In Young v. R. R., 116 N. C., 932, we find tbis language: “Judges are no longer required to submit a case to tbe jury merely because some evidence has been introduced by tbe parties having tbe burden of proof, unless tbe evidence be of sucb character as tbat it would warrant a jury to proceed in finding any verdict in favor of tbe party introducing sucb evidence.”
In Cobb v. Fogalman, 23 N. C., 440, tbe Court says tbat tbe evidence must amount to more than tbat which raises “a possibility or conjecture of a fact.”
In Wittkowsky v. Wasson, 71 N. C., 451, tbe Court says: “There must be evidence from which they might reasonably and properly conclude tbat there was negligence.”
Tbe subject was considered by tbe House of Lords in Bridges v. Ry. Co., Law Journal 1874, 7 H. L. 213, all tbe judges rendering opinions. Tbe rule is thus stated by Mr. Justice Brett: “Are there facts in evidence which, if unanswered, would justify men of ordinary reason and fairness in affirming tbe question which tbe plaintiff is bound to maintain?” This is regarded by Professor Wigmore as tbe best and fairest statement of the most satisfactory test tbat can be adopted. 4 Wig., 2495.
“Are tbe circumstances in evidence adequate to convince a reasonable man?” asks tbe Supreme Court of Kansas in R. R. v. Matthews, 58 Kan., 447.
We think not, in tbis case, in tbe absence of any evidence tbat tbe engines were emitting sparks at tbe time they ^passed. The plaintiff seeks to bolster up its case by attempting to exclude every other possible cause. If tbis were sufficient, then plaintiff has failed. Tbe only evidence as to tbat is tbat tbe sole fire at tbe mill kept up by tbe plaintiff that night was in tbe boiler room. It is likely tbe fire started in tbe machine-room shed, where tbe burned lumber was. Tbe witness Spencer says tbe burned lumber was in a part of tbe machine-room shed, and before tbat in tbe dry-kiln. The room was afire when the witness McArthur first saw it. Tbe fire might have been started by a match carelessly dropped in tbe shed during tbe day or from a spark from a workman’s pipe as be left after bis day’s work, or tbe lumber might have been greatly overheated in tbe dry-kiln and developed a fire in tbat way. A fire just started, as is well known, will sometimes slumber and smolder for hours -and then burst into a blaze.
*315There are many unaccountable ways by which sawmills catch fire, for they are notoriously very bad fire risks. If this fire originated from a spark from the defendant’s engine, it is remarkable that it should have reached such proportions in the space of thirty minutes. It is also remarkable that if the engines were throwing sparks the witness McArthur should not have observed it and testified to it, and further that no cinders were found about the right of way and mill grounds. The cases cited by the learned counsel for plaintiff do not support their contention and are easily distinguished.
In McRainey’s case, supra, the fire stated on right of way at the place where the track crossed the swamp and set fire to combustible matter on right of way. In Deppe’s case 152 N. C., 80, an engine was switching for three-quarters of an hour near the mill; the Ventilator on top of the mill was open; the fire started there; the mill and ventilator were entirely enveloped in dense smoke from the engine; it was a bright summer day and sparks could not have been seen in the smoke. In Ice Co. case, 126 N. C., 798, the evidence showed affirmatively that the engine passed the ice factory emitting sparks, and that soon thereafter the fire broke out. In Kerner’s case, 170 N. C., 94, there was evidence of sparks from the engine, but the jury found that they were not the cause of the fire.
In Hardy v. Lumber Co., 160 N. C., 116, also in Kornegay’s case, 154 N. C., 389, there was abundant evidence that sparks were emitted by the engines. In Williams’ case, 140 N. C., 624, the fire broke out in foul matter on the right of way shortly after the engine passed. In Aman’s case, 160 N. C., 371, there is evidence that the engine threw sparks and live coals. In Maguire’s case, 154 N. C., 384, the fire occurred off the right of way, and, as in this case, there was no evidence that the engine threw out sparks. There was a verdict for plaintiff, but this Court unanimously sustained the motion to nonsuit, saying: “Where plaintiff alleges that he has been ihjured by fire originating from sparks issued from defendant’s locomotive he must not only prove that the fire might have proceeded from defendant’s locomotive, but must show by reasonable affirmative evidence that it did so originate,” quoting from Ice Co. case, supra.
In Kemp’s case, 169 N. C., 731, the fire started in the defendant’s depot, which set fire to a box car and communicated the same to property belonging to the plaintiff. The Court held that where the evidence only tended to show that the defendant’s depot caught fire during the night, supposedly from a passing engine, .and this fire was communicated to the plaintiff’s property, a judgment of nonsuit was properly allowed.
*316In R. R. v. Gossard, 14 Ind. App., 244, it is held that proof that a fire started a few minutes after an engine passed, and that at a point a quarter of a mile distant sparks were escaping from the engine, is insufficient to show negligence on part of the railroad company, especially in .a case where the evidence shows that engine was properly equipped with a spark arrester. In a concurring opinion the decision is sustained on the ground that there is no evidence that the engine on that or any other occasion threw live sparks a distance of 66 feet.
In Musselwhite v. R. R., 4 Hughes, 166, the evidence showed that soon after a freight train passed a large shop that stood near the track was on fire in the upper story near a broken window. It was alleged that the fire was started by a spark from defendant’s engine, proved to be in good condition. The evidence was held to be too uncertain and conjectural to justify a recovery.
We have found cases holding that where engines had passed through open fields covered with dry grass and straw, with high winds prevailing, and a fire broke out shortly after their passage, the coincidence was a fact sufficient to be submitted to the jury as some evidence of the origin of the fire. But a diligent .search has failed to find a case wherein it has been held that the fact that an engine passed a building situated 70 feet from the track and off the right of way, a half-hour before a fire broke out in it, is sufficient evidence to warrant a finding that such fire was caused by the engine.
There are hundreds of lumber mills situated very near railroad tracks in this State, and to hold passing engines responsible for every unexplained fire that breaks out in them, without other evidence, would impose too great a liability upon the common carriers who are compelled to serve them.
In our opinion, the evidence raises nothing more than a conjecture and is not strong enough in probative force to justify the submission of the issue to the jury.
The judgment is
Affirmed.