This appeal is from a judgment in favor of the owner of a building covered by a policy insuring against all direct loss and damage by windstorm, cyclone and tornado.
Suit was filed on the policy following the collapse of a portion of the walls and roof of the building on June 21, 1932. The first paragraph of the bill of complaint alleged that a windstorm occurred in the city of Gary, Indiana, on June 5, 1932, as a result of which the building was greatly damaged in that the walls and roof were twisted and weakened and the roof joints pulled from their places, causing the walls to fall; the second paragraph alleged that a windstorm occurred on June 18, 1932, with the same result; the third alleged that a windstorm occurred on June 5, and/or June 18, resulting in the darhage claimed.
The building in question was a brick and stone structure 125 feet long, 90 feet wide, and 29 feet high as originally built in 1910, for use as a skating rink. The roof was entirely supported by six wooden trusses, resting on the walls and pilasters 20 feet above the ground. There were no columns or other supports between the walls. In 1911 the building was remodeled for use as a theatre, and in 1913, a superstructure 28 to 30 feet in height was added over the building at one end, to be used as a stage loft, for scenery. On top of this there was erected a second superstructure 12 feet high for a ventilator. The building was again remodeled in 1930 for use as an indoor golf course, and had not been used for any purposes from 1931 until its collapse in 1932. It was the roof over, and the part of the wall toward the end where the superstructures were erecte’d which fell.
To prove the fact of the first storm, that of June 5, appellee introduced the evidence of two witnesses. The first, qualified by fifty-five yéars of experience with sailing boats to testify as to weather conditions, testified that on June 5, 1932, while he was about 5% or 6 miles east of Gary in his boat on the lake, he encountered a squall during the course of which the wind attained a velocity of about 60 miles an hour. He used no instruments to measure this velocity but based his opinion on his experience. The other witness who testified as to the storm was the secretary-treasurer of the appellee corporation, the owner of the building. He stated that on June 5, he encountered the storm while he was out at a country club 5 or 6 miles south of Gary. He estimated the velocity of the wind at 50 miles an. hour, but had access to no instruments to verify his estimate which he felt qualified to make by reason of his experience as an air pilot. The only testimony as to the alleged storm of June 18, was that of a pharmacist who lived about four blocks from the building and *777who stated that about 3 o’clock in the morning of June 18, a violent thunder and windstorm occurred, during which the wind blew at a velocity of 50 miles an hour, and the next morning there were branches and twigs blown off in the park and he believed “it was possible that some of the trees were bent over.” His competence to testify as to the wind velocity was based on the fact that he had “made observations as to weather and wind conditions and kept a record more or less of that, and from traveling around.” He also stated that he usually read the charts in the paper each morning, but that he had made no study of wind or wind conditions in Gary and its immediate vicinity with the exception of keeping the records. He had not been outside his house during the storm.
The testimony of these three witnesses was the only evidence relied upon to prove the fact of two storms of such severity that one or the other or both of them moved the superstructures on top of the building, causing one of the trusses to pull out of. the wall. None of the witnesses knew of any other damage resulting from the storm, and no one had observed any change in the condition of the building until its collapse. Appellee also called as witnesses the building contractor who remodeled it in 1930, his general foreman, a builder who examined it'after the collapse, and a general contractor who helped build it. These witnesses were relied upon to prove that the building was in sound condition prior to the windstorm; that it was substantially erected; that the trusses were properly constructed; and that the moving of the superstructures, supposedly by the storm, could have caused the fall.
It was appellee’s theory of the case that all it needed to prove was the fact of the storm or storms, and that the building had been sound. We quote from its brief: “There is but one presumption in this case, one inference, and that is that the windstorms moved the superstructures on top of the building. There is no direct evidence that this is what happened; no witness was there during either of the storms and saw it happen and no witness knew at the time that it had so happened. * * * From these facts the jury might reasonably infer that these storms, one of them or both of them together, moved the superstructures and caused Truss Number 1 to pull out of the south wall. This was the only presumption and1 the only infer-ence. If there had been no positive testimony of the windstorms, then the facts in this case might well fall within the rule argued in the cases cited by appellant.” We cannot agree with appellee’s contention that it needed no direct proof that the storms moved the structures in order to maintain its case. Had the storm and the collapse occurred simultaneously there might be some reason for inferring that the one caused the other. But here, sixteen days elapsed between the first alleged storm and the collapse, and three days between the second and the collapse. No direct effect of the storm or storms was ever observed; appellee and its witnesses simply inferred that because the building was sound, according to their observations, the storm or storms ’must have caused its fall, and this in face of the fact that this record contains undisputed evidence that some of the members of the truss which broke and fell and was pulled from its moorings in the wall, were affected more or less with dry rot.
The policy insures against “all Direct Loss and Damage by Windstorm, Cyclone and Tornado. * * *” We construe this to cover only such losses as are directly attributable to such storms. Here, we think there was a complete absence of proof that the loss resulted from the alleged storm or storms, hence the verdict should have been directed in favor of appellant upon its motion to that effect.
We have refrained from commenting on appellant’s evidence as to the absence of any storm of sufficient severity to be recorded in the official records for the locality in which the damage occurred, or to have impressed itself upon the memory of numbers of witnesses who lived or worked near the building in question; and its evidence as to the condition of the building after the collapse and the likelihood of other causes for such collapse. It is not our province to weigh the evidence where there is conflict as to material facts. Here, however, there was failure of proof of the most material fact, namely, that the storm or storms moved the superstructures, pulling out the truss, thereby causing the damage insured against. Under these circumstances there could be no recovery under the policy sued upon, and the judgment in appellee’s favor must be, and it is hereby reversed, and the cause remanded to the District Court for new trial.