dissenting: The plaintiffs allege that their property was destroyed by fire which negligently escaped from the engine of the defendant. and the rule uniformly applied in cases of this character since the case of Ellis v. R. R., 24 N. C., 138, is that a presumption of negligence arises from proof of the origin of the fire, and that when this is shown the burden is “on the defendant to rebut the presumption of negligence arising from proof connecting it with the origin of the fire, by evidence which would satisfy the jury that the engine was properly equipped, that competent men were in charge of it, and that it was prudently operated.” Currie v. R. R., 156 N. C., 423.
*319It follows necessarily, as.the burden of proof is on the defendant, if there is evidence of the origin of the fire, that we cannot consider the evidence of the defendant that the engines were properly equipped and skillfully operated, on a motion for judgment of nonsuit, and, as said in McRainey v. R. R., 168 N. C., 571, “The. only question presented by this appeal is whether there is any evidence that the fire of which the plaintiff complains originated from defendant’s engine and passed to his land, causing him damage.”
Is there evidence- of the origin of the fire?
The fact may be established by circumstantial evidence, and it is not necessary for a witness to testify that he saw sparks coming from the engine. Recoveries were sustained in McMillan v. R. R., 126 N. C., 725; Williams v. R. R., 140 N. C., 623; and in Deppe v. R. R., 152 N. C., 82, on circumstantial evidence, and when no witness testified to seeing sparks.
The rule laid down by Shearman and Redfield on Negligence, see. 58, as to what amounts to evidence, is that “The plaintiff is not bound to prove more than enough to raise a fair presumption of negligence on the part of the defendant and of resulting injury to himself. Having done this, he is ■ entitled' to recover unless the defendant produces evidence to rebut the presumption. It has sometimes been held not sufficient for the plaintiff to establish a probability of the defendant’s default; but this is going too far. If the facts proved render it probable that the defendant violated its duty, it is for the jury to decide whether it did so or not. To hold otherwise would be to deny the value of circumstantial evidence.”
In Fitzgerald v. R. R., 141 N. C., 535, the rule- is stated to be that the case cannot be withdrawn from the jury “if the facts proved established the more reasonable probability that the defendant has been guilty of actionable negligence,” and this was approved in McRainey v. R. R., supra.
In Henderson v. R. R., 159 N. C., 583 : “But if the more reasonable probability is in favor of the plaintiff’s contention the question ought to be submitted to the jury”; and in Kelly v. Power Co., 160 N. C., 285: “If the evidence for the plaintiff renders it probable that the defendant neglected its duty, it is for the jury, not for the court, to- decide whether it did so or not.”
If, therefore, upon proof of the origin of the fire, there arises a presumption of negligence, and the burden of proof is then on the defendant, and if the degree of proof required of the plaintiffs is that they furnish evidence that establishes the “more reasonable probability” or “renders it probable” that the fire originated from the defendant, the *320only inquiry with us is whether the plaintiffs have furnished evidence which makes it reasonably probable that the defendant set out the fire; and in determining this question the unvarying rule is that the evidence must be construed favorably for the plaintiffs, and that every fact which the evidence tends to prove must be accepted as established.
What, then, does the evidence tend to prove?
The fire was discovered about 1 o’clock at night, and Frank McArthur, who was on the premises at the time, testified “that the fire in the boiler was only pulling the dry-kiln, and there was no other fire around the place.” The witness McAllister also testified that “when he got there the fire was burning from the railroad, coming to the mill.” All the evidence is to the effect that the boiler was 250 or 260 feet from the place where the fire started, and that the wind was blowing from •this place to the boiler and not from the boiler to the origin of the fire. The wind was also blowing from the railroad to the place where the fire started.
We have it, then, established for the purpose of a motion for judgment of nonsuit that there was no fire on the premises before the engines of the defendant passed except in the boiler, and’that the wind was blowing so it could not carry the sparks from the boiler to the place where the fire was first seen. Two engines of the> defendant passed, and within thirty minutes the fire was discovered burning on the side of the shed next to the track, and within 60 or 70 feet from the track, and the wind was blowing from the track to the shed. If there was fire in the engines of the defendant and in the boiler of the plaintiffs, and no other fire on or about the premises, the fire which caused the damage must have originated from the engines or the boiler. Which is the “more reasonable probability” — that it was from the engines, which passed within 60 or 70 feet, with the wind blowing- to the place of the fire, or from the boiler, distant 250 or 260 feet, with the wind blowing in a different direction? The evidence also shows that the fire caught on the side next to the railroad, and that there was a large shed between the boiler and the place of the fire.
The Court evidently felt the probative force of this evidence, and thought it necessary to furnish some explanation for the fire.
It is, therefore, suggested that the fire might have started from a match carelessly dropped during the day, or from a spark from a workman’s pipe, or that the lumber might have been greatly overheated in the dry-kiln and developed a fire in that way, when there is no evidence that any one about the jiremises had a match or dropped one, or that any workman smoked, or that the lumber was overheated in the dry-kiln. It is cetainly more reasonable to conclude that the engines, which *321passed with fire in them, set out the fire, than that the fire was caused by a match or a workman’s pipe, which have no existence so far as this record discloses, except in the imagination, or by the overheating of lumber in a kiln, which had been out of the kiln in the open air, in March, at least six or seven hours.
Nor does it appear that the fire attained unusual proportions in thirty minutes. J. A. Powell testified that he - was asleep- when the whistle blew; that he got up and dressed and went to the mill, and “the fire was moving in toward the main body of the mill and' was about 10 feet square.” McArthur saw no spark, because he was in the boiler room, and he says: “The door toward the railroad was shut.”
The evidence as to the origin of the fire is as convincing as in McMillan v. R. R., 126 N. C., 725; Williams v. R. R., 140 N. C., 623; Deppe v. R. R., 152 N. C., 82; Aman v. Lumber Co., 160 N. C., 370; McRainey v. R. R., 168 N. C., 571; and in all of these cases it was held sufficient to be submitted to a jury.
The Court says in the McMillan case: “While it is not negligence for a railroad to run its trains over its roads well managed and well equipped, as it seems the defendant’s train was, yet we know that no spark arrester can be so constructed as to entirely prevent the emission of sparks, without destroying the efficiency of the. engine; and while it was not negligence in the defendant to run such a train over its road, the fact that it had recently passed over the road, and fire was found there, was some evidence tending to show that it emitted sparks that set the grass on fire.”
This is quoted and approved in the Williams case, and in the latter case the Court further says: “No one testified that he saw the sparks fall from the engine upon the right of way. . . . But here the fire was seen on the right of way; it burnt along the track between the ditch and the ends of the ties, and thence had gone into the woods. The wind was blowing from the northwest across the track, the fire being on the south side. Two witnesses testified that they first saw the smoke about thirty minutes after the defendant’s engine passed. How long before that the fire began no one knew, but there was no fire before the engine passed. These were matters for the jury.”
It is true that in each of these cases the fire from the engine ignited combustible matter on the right of way, but, as said in the Deppe case, “In considering the origin of the fire, it is immaterial whether the fire caught on or off the right of way.”
These cases establish the principle that when it is shown that there was no fire before the train passed, and it breaks out shortly there*322after at a point to which the wind is blowing from the train, that this is circumstantial evidence that the fir© came from the train.
In the Deppe case the fire started in a dry-kiln off the right of way, in which there were heated pipes, and it could not have come from the defendant’s train unless sparks from the engine entered a ventilator near the top of the kiln, 414 by 8 feet and opening back 6 or 7 feet. No witness saw any sparks. The plaintiff offered evidence tending to prove that the fire was not near the heated pipes, and that there was no other fire on the premises nearer than 156 feet, and that the wind was not blowing from this place towards the kiln, but in an opposite direction; that an engine of the defendant was shifting within about 60 feet of the kiln three-quarters of an hour or one hour and three-quarters before the fire was discovered, and that the smoke from the engines enveloped the kiln.
The only circumstance in favor of a recovery present in the Deppe case that is not in the case before us is as to the smoke, and this proves nothing except that there was smoke, and that the wind was carrying it to the kiln, while, on the other hand, there were the circumstances in favor of the defendant of the improbability of a spark entering the ventilator and passing back 6 or 7 feet into the kiln, and the strong probability that the lumber in the kiln became ignited by the heat from the pipes. The Deppe case has been approved eighteen times, as appears from the annotations; 164 N. 0., 710, and four times since the annotations were printed.
In Aman’s case there was evidence that sparks were seen to come from the engine on the day before the fire, but that on the day of the fire “no one saw .any sparks emitting from it,” and the circumstances relied on were that there was no fire before the engine passed, that the engine passed and the fire broke out soon thereafter. ■
Jmtice Walicer says, when considering this evidence on the refusal of a motion for judgment of non-suit: “The familiar rule is that the evidence, upon such a motion, should be considered in its most favorable light to the plaintiff, and every fact which it proves or can prove should be taken as established. With this guide before us, we are led unhesitatingly to the conclusion that the ruling of the Court was correct”; and again: “It can make no difference whether the sparks lighted on or off the right of way, if they kindled the fire and destroyed plaintiffs’ trees, there was a sufficient case of prima facie evidence for submission to the jury, upon the whole evidence; to find the ultimate fact of negligence. This Court has been most pronounced in its opinion *323upon this subject, and has adhered steadily and strictly, without the shadow of turning, to the just rules which have heretofore been promulgated.”
In Kemp's case, 169 N. C., 732, no recovery was allowed because, as stated in the opinion, “It was admitted by the plaintiff that he could not prove the origin of the fire,” and there was no evidence that any engine of the defendant ever passed the point where the fire originated, unless this might be inferred from the fact that the defendant was maintaining a depot building and a railroad track.
I have thought it necessary to discuss the principal cases bearing upon the question involved in this case, because, as I see it, a precedent is being established which will prevent just recoveries hereafter in cases which have heretofore been sanctioned by a long line of decisions.
Hoke, J., concurs in this opinion.