Gaynor, J.:
I cannot bring myself to assent to the proposition that proof that a person fell or an accident happened at a given place is evidence that the conditions there were “ obviously ” dangerous. If the decided cases be examined superficially, and more with a view o'f getting hold of something that was said in them than of what was decided, sentences may be found here and there to support it. But that is nothing; it is not everything that may be said, but only what is actually decided, that may serve as authority. What has to be proven in a given case of alleged negligence is not merely that the place or condition was in fact dangerous, but that it *696was “ obviously ” dangerous. The question always is whether the place or condition was such as persons of the required prudence, foresight and skill would deem dangerous. That an accident happens cannot prove this. Aftersight cannot be thus substituted for foresight on the question of that culpable omission or commission which is necessary to ruáke negligence. Proof of a prior accident may be proof that the place or condition was dangerous in fact, but it cannot be proof that it was obviously dangerous; and upon the latter only can negligence be predicated. It is admissible, however, as proof of notice to the defendant that the place or condition was dangerous, in every case where the defendant had actual notice of the accident, or will be presumed to have had; for it brings home to him knowledge that the place or condition was dangerous in fact, and that makes it obviously dangerous to him. The first case we have on this subject is that of Quinlan v. City of Utica (11 Hun 217; affirmed without opinion, 74 N. Y. 603). While our highest court found that the record there required an affirmance, it took care not to adopt the opinion below; nor do I understand everything said in that opinion is to be accepted for law. The case mainly relied on by the learned judge who wrote it was that of Dougan v. Champlain Transportation Co. (56 N. Y. 7). The action there was for damages for the death of one who' slipped on the icy deck of defendant's 'steamboat, and slid under the guard railing and overboard. That there was so wide a space between such railing and the deck was the negligence alleged. In reversing the judgment for the plaintiff the Court of Appeals said: “ There was no proof tending to show that any one had ever before fallen and gone overboard under the railing, or that any such danger had been apprehended by any one.” After quoting this the learned judge in ■ the Quinlan case strangely enough says: “ This plainly implies that proof of a similar prior accident at that place would have been competent as tending to show that the space was dangerous, and consequently that, the defendant was negligent in allowing it to remain open and exposed.” On the contrary, it “ plainly iniplies ” no such thing, but only implies and says that if such an accident had happened before on the defendant’s boat, that would have been notice to it of what it could not otherwise be deemed to have known, viz., that the space was dangerous. That the accident sued upon proved it dangerous in fact was not enough to permit of a finding of negligence. It was necessary that it should also be *697“ obviously ” dangerous, and a prior similar accident on the defendant’s boat would have been evidence that it was obviously dangerous to the defendant. In the recent case of Lundbeck v. City of Brooklyn (26 App. Div. 595), the decision seems to be that the exclusion of evidence of previous falls at the same place on the sidewalk from the same cause was erroneous for the reason that the fact of such falls would be “ constructive notice ” to the city that the place was dangerous in fact even though not obviously dangerous. This would of course be so if the accidents were so many and so notorious that notice of them could be imputed to the city. It does not seem to me that the decision could have been' meant by the court to go further than this, andhold that apart from the question of notice the evidence was admissible to merely prove that the condition was dangerous in fact; but I am not altogether sure from the opinion. The learned reporter seems from his head note not to have understood the decision to go so far. Unless such evidence be admissible on the head of notice, on what principle can it be admissible? The learned opinion cites the case of District of Columbia v. Armes (107 U. S. 519) as approving the Quinlan case, but it will be found on examination that the accidents proved in the Armes case occurred in the presence of a municipal policeman, and were therefore notice to the corporation. It may approve the actual decision in the Quinlan case, so far as in point, but can it be deemed an approval by the Supreme Court of the United States of everything said in the opinion in that case?
In the present case the question excluded did not call for an answer showing that the previous derailment was at the same place and under the same conditions as the one the plaintiff complained of. It was left to the witness to- say or assume that it was “ at this point in question ”, and the speed was not shown. Proof of such prior derailment was therefore not competent as notice to- the defendant of a defect in the track, i. e., the absence of a flange or guardrail, at the point where the derailment complained of happened, if it be granted that'the so-called curve-, i. e., with a radius of 2000 feet, was one upon which danger of derailment could be based, so that a flange or guardrail was needed. But there was no evidence that a flange or guardrail is necessary or usual on such a slight curve.
The motion for a new trial is denied.