Shortly after one of defendant’s passenger engines had passed over the road, a building belonging to the plaintiff Chandler, ninety-three feet from the center of the rail, was found on fire. The fire started in the roof, and there was no other apparent cause than the sparks from the defendant’s engine. That building burned and another building, and still another about thirty feet therefrom, and again another building one hundred and fifty-eight feet from the last, all upon plaintiff’s premises. Plaintiff collected an insurance, and this action is brought both by the plaintiff and the insurance companies against the railroad company to recover damages suffered from the burning of said buildings.
First. It is contended that there is an improper joinder of the parties'plaintiff; but the facts appear upon the face of the complaint, and by not demurring this objection has been waived.
Second. Defendant contends that there was no proof of its negligence. The defendant by three of its employees showed an examination of the spark arrester before and after the fire, by which examination it was found to be in perfect order. There is no question that the proper style of spark arrester was used by the *70defendant upon its engines. The plaintiff’s expert swears that if the spark arrester had been in order, a live spark could not have been thrown more than thirty feet. This evidence was corroborated by the expert for the defendant. This engine was shown to have thrown large sparks at other places upon this trip. The fire occurred in the middle of the day. All other causes for the fire were eliminated. The season was very dry, and the jury has found that 'this engine caused this fire and its spark arrester was defective. If the spark were thrown from this engine to this building, in view of the evidence both of the plaintiff and defendant that it could not have happened if the spark arrester had been in good condition, I think there was enough evidence to go to the jury upon the question of whether in fact the spark arrester was intact.
Third. In making its proof the plaintiff was allowed to show that other engines upon defendant’s road threw large sparks more than ninety-three feet. This evidence was admitted over the defendant’s objection and to the ruling exception was duly taken. In addition to this the plaintiff’s counsel requested the court to charge “ that in arriving at their verdict they may take into consideration the evidence of the various witnesses of other engines upon the defendant’s road at other times throwing out cindei-s as stated in the evidence.” This the court charged, and the defendant’s counsel duly excepted. This evidence was I think improperly admitted and the plaintiff’s request to charge improperly granted.'
In this discussion I will assume as matter of élementary law that negligence in a particular instance cannot be inferred merely from the fact of negligence at other times and places. If one charged with negligence were called upon to justify every act of his life claimed by the plaintiff also to be negligent, his burden would be intolerable. If one or more of defendant’s engines, other than the one which caused this fire, had been so defective as to throw out cinders a considerable distance, that fact of itself could not be deemed proof that the identified engine which caused this fire was also defectively constructed, or, if so, that its defective construction caused this fire. Again : A distinction must be remembered between a fire caused by a spark arrester in good repair, but of improper construction, and a fire caused by a spark arrester of proper coni *71struotion, but out of repair. In the first case it may properly be presumed that different engines upon the defendant’s road were equipped with similar spark arresters, and the fact that one of such spark arresters allowed large cinders to escape through its meshes would be some evidence that another spark arrester presumed to be of similar construction and in similar condition would also permit large cinders to pass through its meshes. This was the case presented in Sheldon v. Hudson River Railroad Co. (14 N. Y. 218). And of the rule there stated as applied to the facts of that case I have no complaint. In Wigmore on Evidence (Vol. 1, § 455, p. 547) the right to the admission of evidence of sparks thrown from other locomotives is based upon the presumed or proven similarity of construction. The text in part reads : “ Rot all engines are of the same construction or in the same condition with reference to a capacity to emit sparks, and, therefore, the instancing of such a capacity in some of them does not evidence such a capacity in another unless they belong to the same class with reference to construction, etc.; in short, unless the principle of substantial similarity of conditions is fulfilled. * * It seems wiser, where the other engines belong to the same owner or run over the same lime, to assume this similarity.”
In the case at bar the spark arrester used by the defendant upon its engines was of approved style and construction. The plaintiff’s expert himself says that if in proper repair it could not emit a spark over three-sixteenths of an inch in size. A spark of this size could not have caused this fire. The defendant’s liability must be based, therefore, upon the defective condition of this spark arrester in being out of repair. It will be seen at a glance that this presents an entirely different question than was presented in the case of Sheldon v. Hudson River Railroad Co. In the first place there can be no presumption of similarity as to the extent of the defect. Because one engine is so defective as to emit a spark the size of a chestnut, it cannot for a moment be urged that another engine is presumed to be equally defective so as to emit a spark equally large. The presumption of similarity, which arises when the question is simply one of construction, entirely vanishes when the question is not one of construction, but of want of repair. It was clearly competent to show that this particular engine within a reasonable *72time before or after this fire emitted sparks which could not have passed if the spark arrester had been in good repair. I can conceive, however, of no possible inference that the spark arrester upon this engine was defective from the fact that spark arresters upon other engines were shown to be out of repair. And still less right would the court or jury have to draw the inference from such comparisons as to the extent of the defects in the different spark arresters.
It is speciously argued, however, that this evidence is competent to show that a defective engine is capable of throwing a spark ninety feet. ISTo one in the world would question this proposition if the opening in the spark arrester be of sufficient size. Cinders from burning buildings are blown many hundred feet and fires are caused thereby. If the opening in the spark arrester is allowed to become so large as to pass a large enough cinder, it is of common knowledge that the fire could have been caused thereby. I need not rely upon the common-sense presumption for this, for the respondent’s counsel in .his brief states the same proposition. He . says: “ If there is a defective screen large sparks can be carried by the wind alive and set fires over one hundred feet from the track. This - is not disputed by the defendant.” If there were in the case any proof or presumption that the defects in the different engines were of equal extent, the evidence might be justified, but the fact that other engines had defects in their screens of sufficient size to emit cinders large enough to fly 100 feet while burning cannot in the nature of things be any proof that the defect, if any, in the engine in the case at bar was of equal, extent. The admission of the evidence, therefore, is not justified to show a fact of common knowledge, and not anywhere questioned in the case, that if the opening in the defective spark arrester be large enough a cinder can be thrown therefrom sufficient to set fire to this building.
. I am not unmindful that there are some cases in the courts of this State which seem to have gone further in the admission of this class of evidence than would be justified by the rule which I am declaring. Those cases, however, without exception, refer to the Sheldon case as authority, and are apparently based thereupon. (See Jacobs v. N. Y. C. & H. R. R. R. Co., 107 App. Div. 134.) If the distinction which I have attempted to draw between the facts *73in the Sheldon case and the facts in the case at bar be sound, it will be seen that the Sheldon case is no authority for such a ruling under the circumstances as are here presented.
These views lead to the reversal of this order and judgment and the granting of .a new trial. It is not necessary to this decision, therefore, to pass upon the question of damages here presented. The question is a complex one, not entirely free from doubt. If the case be retried the trial judge can ask a special finding upon the value of the several buildings, which would enable this court or the Court of Appeals, upon appeal, if such should be taken, to adjust the judgment in accordance with its opinion of the extent of the defendant’s liability.'
I recommend that the judgment and order be reversed and a new trial granted, with costs to appellant to abide the event.
All concurred ; Kellogg, J., in result.
Judgment and order reversed and new trial granted, with costs to appellant to abide event.