The evidence tended to show that if the fire was set by the appellant’s locomotive, it was by the one in the freight train which passed the station, near which the respondent’s lumber was piled, at 10:15 o’clock in the forenoon, or by the one in the passenger train which passed the same place at 11:50 thereafter. The fire was discovered ¡soon after the passenger train passed that point, and there was no evidence whatever that this particular fire was set by any other engine on the road. The circuit court admitted evidence, against the objection of the appellant, of fires in the vicinity on this same road, both before and after this fire, occurring after the passage of other locomotives. This was clearly erroneous. Such evidence would open the ■door for a wide issue of great importance,— whether such other locomotives caused such fires or not,— and could not ¡affect the issue in the cause, even if it had been proved that other locomotives caused other fires in the vicinity. The rule has never been extended further than to allow proof qf •other fires caused by the same machinery. If it had been *338proved in this case, beyond a doubt, that one of these locomotives — either that of the freight or passenger train passing soon or immediately before the fire occurred — caused the fire, it could not add to the defendant’s liability by showing its habitual carelessness in respect to other locomotives ; and if it had been proved that other locomotives on the same road caused other fires, at other times and places, it would not be even presumptive evidence that the locomotives in question were insufficient in any respect, or that they caused this particular fire.
The evidence in this case was circumstantial, and it should not be extended to circumstances which could not have any logical bearing upon the issue. The syllogism that because other locomotives on this road caused other fires at other times in th§ vicinity, therefore these two locomotives, or one of them, which passed the place at this time, caused this particular fire, would be no more logical than that locomotives on some railway in another state, a thousand miles -distant, caused fires in the vicinity of the railway, on account of insufficient manufacture or repair, or other negligence. Evidence, to be admissible in such cases, should have some relative bearing upon the issues, either directly or remotely, and it is not perceptible how evidence that other locomotives, at other times and places, even on the same road and in the vicinity of the fire in question, could have any such bearing. Evidence that other fires had been set by the fire machinery managed by the company, before or after or about the time of the fire in question, is confined to the same machinery which caused such fire, and which may bear upon the question of the sufficiency of such machinery or its management; and such are the authorities cited by the learned counsel of the respondent. There may be cases which hold that the locomotives of the company on the same road, at different times before and after the fire in question, were so constructed as to scatter fire along its *339.track, was competent evidence to show a possibility, and therefore a probability, that this particular fire was set,.by some one or more of su.ch locomotives. But, so far as I have examined such cases, they are those in which the locomo.-.tives which actually caused the burning were, not. identified.
The late and leading case in which this question is.disr cussed and decided is that of the Grand Trunk R. R. Co. v. Richardson, 91 U. S., 454. In that case, both in the brief of the learned counsel and in the opinion of Mr. Justice SteoNG, the language is very carelessly used, that evidence that the locomotives of the company, at other times and places on the same road, were so constructed as to scatter fire along the track, might tend “ to prove a possibility, and a consequent probability, that some locomotive [of the company] caused the fire, and show a negligent habit of the officers and agents of the railroad company.” . But in that case.it is said in the opinion, “the particular engines [which caused the fire] were not identified.'” In such a case, such evidence might tend to prove the possibility and consequent probability that some locomotive of the.company caused the fire. This wonderfully loose logic may be satisfactory to a judicial mind in cases where there was no proof that any particular and identified locomotive caused the fire in question, if any locomotive of the company did. But in. due deference to the learned judge who wrote the opinion, and .the other judges who have used this language, it is submitted that a possibility can never establish a probability of a fact required to be proved in order to make a railroad company or any party liable in any action whatever, and the proposition is no sounder in logic than in.law. It would be monstrous doctrine that when a party is sued in .tort for a personal injury to another, occasioned by his negligence in not furnishing proper appliances, or otherwise,, his common carelessness, or carelessness in other cases, tend to prove the “possibility,” and therefore “probability,”'that *340the act charged was the result of his negligence, without proof even that he committed it.
In cases where it is shown, either by positive or circumstantial evidence, that some locomotive of the company caused the fire, without the identification' of any particular one, such evidence might have weight in showing the negligence of the company. There may be cases which have gone further than this in the admission of such evidence, but they do not appear to us authority in reason. In Ross v. B. & W. R. R. Co., 6 Allen, 87, it was held competent to show that the engine in question emitted burning sparks a fortnight previous to the fire in question, and that other similarly constructed engines had emitted sparks which set fires. Where there is no proof of what particular engine set the- fire, and the circumstantial evidence is such that there is a strong probability that some engine on the road did set the fire, then it may be proper to show that the engines on that road generally emitted sparks, or that some one or more of them did so at other times and places. Sheldon v. H. R. R. R. Co., 14 N. Y., 221; Field v. N. Y. C. R. R. Co., 32 N. Y., 339; St. Joseph & D. C. R. R. Co. v. Chase, 11 Kan., 47; Huyett v. P. & R. R. R. Co., 23 Pa. St., 373; 1 Thomp. on Neg., notes, 160. Testimony showing that some of the company’s locomotives had previously or-subsequently scattered fire is not admissible unless it is also shown that the locomotive which caused the fire was one of them, or was similar in construction, state of repair, or management. Boyce v. Cheshire R. R. Co., 42 N. H., 97; Phelps v. Conant, 30 Vt., 277; Malton v. Nesbit, 1 Car. & P., 70; Hubbard v. A. & K. R. R. Co., 39 Me., 506; Standish v. Washburn, 21 Pick., 237; Collins v. Dorchester, 6 Cush., 396; Robinson v. F. & W. R. R. Co., 7 Gray, 92; Jordan v. Osgood, 109 Mass., 457; Sheldon v. H. R. R. R. Co., supra; Smith v. H. & St. J. R. R. Co., 37 Mo., 287; Lackawanna & B. R. R. Co. v. Doak, 52 Pa. St., 379.
*341In Pennsylvania R. R. Co. v. Stranahan, 79 Pa. St., 405, the court said: “This was not a case where a certain engine had thrown out the sparks which set fire to the plaintiff’s barn, but it was where the engine was unhiown. Yet the cause of the fire was clearly traced to the railroad track, and left the belief that some one of .the engines of the defendants had emitted the coals which; set the barn on fire. . . Hence, it was necessary to permit the party to show that the emitting of coals and sparks in unusual quantities was frequent, and was permitted to be done by a number of engines.” This is the true-rule and its limitation, approved by the weight of authority as well as by reason. 4 South. Law Rev. (N. S.), 110; Henry v. S. P. R. R. Co., 50 Cal., 176; Woodson v. M. & St. P. R'y Co., 21 Minn., 60.
The case of Brusberg v. M., L. S. & W. R'y Co., 55 Wis., 106, does not militate against this principle, but rather strengthens it. Other fires in the vicinity were allowed to be shown as tending to prove that the particular train in question caused them, which involved the question whether that engine was properly constructed or managed to prevent sparks from being emitted along the track. In this case, as said before, the fire was caused by the engine in the freight or passenger train which passed soon or immediately before the fire. The plaintiff was on the ground, and discovered no fire before one or the other of these engines passed the place of the fire. If there was no evidence that one or the other of those engines caused the fire, there was no evidence that any engine of the company caused it; so that, within the authorities, proof that other engines caused other fires at other times and places,, would not show a “possibility,” and consequently not a “ probability,” that these particular engines caused the fire. The admission of that evidence in this case was, therefore, erroneous. I could write fifty pages of opinion in examining and criticising authorities on this question, on account of their looseness of expression, when ■the question is really in a “ nut-shell,” and we are ready to *342■disapprove of any and all of them which assert a different 'principle.
-The only other question deemed to be material on this appeal is whether the leaving of refuse or other combustible material along the track at this place, by which fires could be 'communicated with the property of others in the vicinity, was per se the negligence of the company. This was a depot for the shipment almost or quite exclusively of lumber, ties, etc. Bark and other material liable to take fire with great facility would necessarily accumulate on and near the track, and, under such circumstances, whether it was the duty of the railway company to keep their track and grounds always free from such material is a very important question, which we do not feel at liberty to decide in this case. And in. such a place, and under such circumstances, whether it may or may not be contributory negligence on the part of those bringing lumber to such a point for shipment, to place it on the grounds of the company amidst such refuse and combustible material, which almost necessarily accumulates at such a place, we need not decide in this case. But we do hold unhesitatingly that it is a question of fact for the jury whether, in any particular place, it was negligence to so leave such material on or. near the track on the grounds of the company, liable to be ignited by the sparks emitted by engines; and such we regard the weight of authority, as well as the effect of the decisions of this court. It is not per se, or as a question of law, negligence, but a question of fact to be determined by the jury in any given case. Pierce on Railroads, 434; Shearm. & R. on Neg., 404; Ohio & M. R. R. Co. v. Shanefelt, 47 Ill., 497; 1 Redf. on Railways, 477, and notes; Smith v. L. & S. W. R’y Co., L. R., 5 C. P., 98; Karsen v. M. & St. R. R’y Co., 29 Minn., 12; Toledo, W. & W. R’y Co. v. Wand, 48 Ind., 476; B. & M. R. R. Co. v. Westover, 4 Neb., 268; P., C. & St. L. R. R. Co. v. Nelson, 51 Ind., 150; Kansas Pac. R'y Co. v. Butts, 7 Kan., 308.
-We need not have gone outside of the decisions of this *343•court for an authoritative decision of this question, but unfortunately it has been customary to cite other authorities, in order to show that this court decided the question correctly. In Kellogg v. C. & N. W. R'y Co., 26 Wis., 223, it is expressly decided that it was not negligence per se for the plaintiff to leave grass, weeds, and stubble on his own land, exposed to fire, in accordance with the same rule that it was .not negligence per se for the defendant to leave such burning material on its right of way. In respect to this principle, •or rather in violation of this principle, the circuit court instructed the jury as follows: “ You see the company has not done all that it might to prevent fires, because it might have kept its right of way free from this combustible material.” Without considering other points made, we regard these two errors as material and prejudicial.
By the Court. — 'The judgment of the circuit court is reversed, and the cause remanded for a new trial.