In this case the plaintiff offered evidence tending to show that she sustained an injury by slipping on a formation of ice, which had remained about four days on a side-walk in the city of Hartford, extending entirely across the walk somewhat diagonally, from one and a half to two feet wide at the outer edge of the walk and some three feet wide in the center, about four and a half inches thick in the thickest part and sloping to a thin edge, and that it was dangerous to cross.
The defendants offered evidence tending to prove that there was no such body of ice, and that, if there was, it did not remain so long and was not dangerous.
To sustain their claims the defendants offered evidence to show that a number of persons during the whole time claimed repeatedly passed along and over the side-walk in question, without slipping thereon or experiencing any inconvenience whatever. The court admitted this evidence to prove that there was no such body of ice, and that, if there was, it remained there for a much shorter time than the plaintiff claimed, but not to prove that, if there, it was not dangerous.
We think the court erred in not allowing this evidence to go to the jury for what it might be worth. We do not see how this evidence could consistently be admitted as tending to show that there was no ice, and not as tending to show that the ice, if there was any, was not dangerous by the use of ordinary care.
One of the most satisfactory kinds of evidence is that which is founded on the natural relation of cause and effect as ascertained by observation and experience. The plaintiff founded her claim upon the result of the common experience of mankind, that a glade of ice usually causes any person passing *59over it to slip. If she could succeed in satisfying a jury that there was a formation of ice of the description claimed, she would make out a prima facié case, that the side-walk was dangerous, not from the single fact that she slipped, but from the fact established by the common experience of mankind, that any person crossing such a formation in its primary condition would be hable to slip. But the same observation and experience teach us that such a formation may in a few days, by the effect of frequent passing, become soiled and worn, and by the action of the sun and other causes become rough and cease to be slippery, and consequently, in that respect, will cease to be dangerous. One important question in this case was, whether, if the ice was there, it was or was not in a slippery and dangerous condition.
If the plaintiff had offered evidence to show that a number of persons had actually slipped upon - it-, it would have been strong proof that it was in a slippery and dangerous condition. Men always act on such evidence in deciding whether they will risk their limbs or not. Why then should not proof that a number of persons passed over it and did not slip, be admitted as tending to show that it was not in a slippery condition ?
We have been asked' whether the fact that a number of persons passed in safety a bridge in which it was claimed there was a dangerous hole, would be admissible to prove that there was no such hole, or, if there was, that it was not dangerous. It is unnecessary to answer that question. Many persons might cross the bridge without passing over the hole. If the claim was that a plank reaching the whole width of the bridge was gone, it would bear a stronger resemblance to the present case. Here, all who passed over that side-walk must have either stept upon the ice, or have taken an unusual stride to step over it. Their attention would necessarily have been drawn to it.
We think the rejection of this testimony would be to some extent inconsistent with the decision of this court in House v. Metcalf, 27 Conn., 631. In that case it was necessary for the plaintiff to prove that a water wheel, revolving in an *60exposed condition, near a highway, was dangerous, and that the plaintiff’s horse was frightened by it. To establish the fact that it was dangerous it was held to be competent to prove that other horses were frightened by the same wheel. The objection was pressed in that case, as it has been here, that it would ^aise a collateral issue. But Sanford, J., answers that-objection by remarking that every piece of evidence, in every case, might be exposed to the same objection. It is plain that a collateral issue may be made at any time by an attempt to impeach a witness. Still the jury ought to have the benefit of all testimony from which a reasonable. inference may be drawn of the existence or non-existence of any material fact which is in issue in a case.
The court below may have been led to the decision in question by several cases which have been decided in Massachusetts. Whether we should he disposed to follow those decisions to the full extent it is not necessary to determine. A plain distinction can be drawn between those cases and the present. In the first of those cases, Collins v. Dorchester, 6 Cush., 396, the plaintiff claimed that there was a line of posts on the outside of a highway which passed through a swamp, to protect the traveled path, and that the plaintiff’s carriage ran against one of these posts,, in consequence of the absence of some others on the line which should have been near it. The plaintiff offered evidence to show that the carriage of another person had previously run against a post under similar circumstances, at or near the same place, from the same alleged defect. The supreme court of Massachusetts held this evidence inadmissible, chiefly on the ground that it would raise a new issue. There the alleged defect was entirely outside of the traveled path, and would not affect any person who kept on it. The case of Aldrich v. Pelham, 1 Gray, 510, turned in part on the question whether the traveled path was wide enough to allow two carriages to pass conveniently. Evidence that other persons had passed without collision or accident was rejected. The decision seems to have been governed by that in Collins v. Dorchester, with a,- suggestion of the court that the proper evidence in such a case would be *61an actual measurement of the width, as that could easily have been made.
The case of Kidder v. Dunstable, 11 Gray, 342, was merely a reiteration of the doctrine of the former cases, and an application of it to general evidence that other teams passed and re-passed the place where the plaintiff claimed to have been injured, without accident or difficulty. None of these cases presented a state of things where, as in this case, the experience of all who passed must have been substantially the same.
We advise a new trial.
In this opinion the other judges concurred; except Hinman, C. J., who dissented.