MALTBIE, C. J.

The plaintiff brought this action against the defendant town to recover for injuries due to a fall upon an icy sidewalk, claiming to recover under the statute on the ground that it was defective. General Statutes, § 1420. She also brought an action against Talcott, the owner of the property abutting upon the sidewalk, claiming that the ice formed as the result of water dripping from the roof of the building standing upon the property. The two actions were tried together. At the conclusion of the plaintiff's case each defendant moved for a judgment of nonsuit, which in this case the trial court granted. The present appeal is taken from the denial of the plaintiff's motion to set that nonsuit aside. After it was granted, the case against Talcott was fully tried and verdict returned for the plaintiff; an appeal from the judgment was taken to this court; and we found error and remanded the case for a new trial. Young v. Talcott, 114 Conn. 675, 159 A. 881.

The basis upon which we must approach the present appeal is stated in Cook v. Morris, 66 Conn. 196, 208, 33 A. 994. When the issue depends upon the sufficiency of the evidence offered, a nonsuit is properly granted only "when the facts testified to are so clearly without that logical relation to the facts in issue, legally essential to any probative force, as in point of law to constitute no substantial evidence of the facts in issue." Fields v. Fields, 93 Conn. 96, 98, 105 A. 347. In determining whether a nonsuit should be granted, the evidence is to be taken in the light most favorable to the plaintiff and every inference favorable to him is to be drawn, and he has the same right to submit a weak case to the jury as he has a strong one. Fritz v. Gaudet, 101 Conn. 52, 124 A. 841.

The sidewalk was concrete and at the point where the plaintiff fell one of its slabs had been somewhat depressed so that near the inside edge of the walk it was seven tenths of an inch lower than the next flag, the difference in elevation decreasing until at the outside of the walk the flags were substantially at a level. Whether or not this inequality would constitute a defect is not now a material consideration; there was no evidence at all that any inequality existed in the surface upon which the plaintiff walked; on the other hand, one of the witnesses testified that the walk was unusually smooth that day because of the ice upon it having filled the depression. The defect which caused the plaintiff's fall consisted in the ice upon the sidewalk. As to the extent of that ice at the time she fell she presented little evidence. Her own testimony was that it ran diagonally across the walk and in giving it she indicated what she meant upon a map in evidence, in a way not disclosed in the record before us. One witness testified that the ice extended practically across the sidewalk. No one testified as to the extent of walk covered longitudinally. But the plaintiff's son testified that, in the afternoon of the same day, when the ice had melted, there was water in the depression covering about three fourths of the slab; and if this evidence is compared with that of the witness who testified that the walk was unusually smooth that day, a reasonable inference might be drawn that the patch of ice upon which plaintiff fell was at least as large. The slab in question was about five feet square. There was evidence that the ice was smooth and no sand or ashes had been sprinkled over it. While it had begun to snow a short time before the accident and there was a slight skim of snow over the ice, no contention is made that this produced any material change in the situation. Frechette v. New Haven, 104 Conn. 83, 91, 132 A. 467; Keating v. New London, 104 Conn. 528, 531, 133 A. 586. Whether a patch of ice of the size indicated would constitute a defect is a close question, but we cannot say that there was not some substantial evidence to go to the jury upon that issue. Schroeder v. Hartford, 104 Conn. 334, 337, 132 A. 901.

Upon the issue of imputed notice to the defendant of the defect there was no direct evidence of the time the ice had been upon the walk. But the weather reports in evidence showed that upon December 27th, three days before the accident, there had been a storm of rain, sleet and snow, ending at four-twenty-five p.m. and resulting in many sidewalks, before being cleaned, having a covering of soft slush one to two inches deep; that from that time until after the accident the temperature hovered around freezing except for four or five hours on the afternoon of the 29th, when a maximum temperature of thirty-nine degrees was reached, and that this rise in temperature caused even snow to melt so little that it decreased that day but one tenth of an inch. From these facts the jury might have reasonably inferred that, had the slush not been promptly cleaned from the walk, the ice which formed would not have been smooth and that the ice upon which the plaintiff fell had been upon the walk from the afternoon or evening of the 27th. They might also have found from the evidence that this place upon the walk was one where ice often formed, a fact as to which there was both direct testimony and a basis for inference because the walk sloped to it from one direction and ran almost level in the other, and upon the side away from the street was a bank of earth and upon the other a ridge formed when snow had been shoveled from the sidewalk before the accident. The particular danger in such a situation might be regarded as increasing the obligation of the defendant to exercise reasonable care to see that the walk was maintained in a reasonably safe condition. Again we cannot say that there was not substantial evidence to submit to the jury upon the issue whether the defect had existed for such a length of time that the defendant in the exercise of reasonable watchfulness should have discovered it in time to take steps to guard passersby from injury. Frechette v. New Haven, supra, p. 94; Schroeder v. Hartford, supra; Keating v. New London, 104 Conn. 528, 530, 133 A. 586.