ARANT, Circuit Judge.

Appellee, an Ohio corporation, brought suit upon an insurance policy in the Court of Common Pleas of Hamilton County, Ohio, to recover for alleged windstorm damage to a gasholder it was constructing for the Ford Motor Company, at that company’s River Rouge plant, near Dearborn, Michigan. The case was removed to the United States District Court upon petition of appellant, an English corporation. This appeal challenges the correctness of the judgment for appellee for the full amount of the policy, with interest from date of the loss.

The two principal questions raised are whether there was substantial evidence that a windstorm, within the meaning of the policy, occurred at the gasholder when the conceded loss occurred, and, if so, whether there was substantial evidence that such windstorm proximately caused appel-lee’s loss. In addition, there are exceptions to the Court’s admission and exclusion of evidence.

Appellant introduced evidence that, at several places of varying distance from the gasholder, there was no wind of unusual velocity on the night in question. This evidence included anemometer readings taken at the Detroit City and Wayne County Airports, approximately ten miles northeast and southwest of the gasholder, as well as similar recordings at a moveable coal bridge at the River Rouge plant, which may have been as much as half a mile away; testimony of lay witnesses as to wind conditions at places from one-fourth to four miles distant; and testimony that certain other property in the vicinity of the gasholder was not injured. None of the anemometers recorded a wind velocity in excess of twenty-one miles per hour.

Appellee introduced the testimony of four witnesses who were at the gasholder when the damage occurred. Two were welding inspectors employed by the Ford Motor Company, and the other two were employees of appellee. They testified that a wind of unusual velocity was blowing and described its force and effect in considerable detail. They said that they had to hold the rail when ascending the stairway on the outside of the gasholder to avoid being blown off; that the tool house where they stayed between inspection trips to the interior of the gasholder was rattling as if it would be blown away; that the derrick boom was swaying; that at the height of the windstorm a' terrific noise came from the gasholder, sounding as if the derrick had fallen; that immediately afterward the boom, which had previously stood in an upright position, was leaning to the south and they could see only about two feet of it whereas eighteen or twenty feet had been visible before. They estimated the wind velocity at from forty-five to *704 sixty miles per hour, with gusts up to seventy or seventy-five miles per hour. Theirs was the only direct evidence as to wind conditions when the damage occurred.

Appellant urges that its evidence, which it characterizes as scientific, compels the conclusion that the testimony of appellee’s witnesses is unreliable and, in consequence, that there is no substantial evidence of a windstorm that could have caused the loss.

We are unable to agree with this contention. The anemometer readings, the nearest of which may have been taken half a mile from the gasholder, were from Burton cup-type anemometers, which an expert witness testified are unreliable in recording sudden gusts of wind. The testimony of those who were at the gasholder was both direct and substantial, and the evaluation of their testimony, as well as contradictory testimony, was for the jury.

But appellant contends that, even if there was a windstorm, there was no substantial evidence that the damage was caused thereby. Consideration of this contention requires a somewhat detailed description of the structural aspects of the gasholder.

The structure here involved is a “Klon-ne” type of dry holder, consisting primarily, when completed, of a cylindrical steel shell, steel floor and roof, and a weighted dome-shaped piston capable of ascending and descending as the quantity of gas in the holder increases or decreases.

Appellee first constructed the floor of the holder on a foundation built by the Ford Motor Company, by welding steel plates together to form a flat disc. The frame of the piston was then constructed over a temporary wooden structure erected on this disc; to this frame three-sixteenths-inch steel plates were riveted and welded together, forming an arched surface. A circular twenty-four by fourteen inch steel H-beam, or girder, adjacent to the shell of the holder, forms the principal supporting member of the piston. To it are attached forty-two main, and the same number of secondary, radial curved steel rafters, to which in turn are attached twelve concentric girders, all reenforced or stiffened. A flexible piston seal was later built between the principal girder and the shell to make the holder gas tight. It was injury to this seal, as will appear, that immediately caused the damage.

A second circular H-beam was erected approximately thirty-two feet directly above the lower ring girder, supported by eighty-four vertical five-inch H-beam columns, equally spaced around the circumference of the piston, which were braced by steel trusses and struts, referred to as back-bracing, 'attached to the rafters supporting the piston surface. To guide the piston in its upward and downward movements and keep it in an approximately level position, two series of one hundred and sixty-eight rollers each were affixed to flanges, one attached to the upper, and the other just above the lower, ring girder.

On temporary struts attached to the back-bracing, a circular rail was constructed, sixteen feet inside the. upper ring girder and on a level with it, to carry two eighteen-ton trammel cranes, operated opposite each other and connected by radial arms to the center of the piston, each having a seventy-two-foot boom for lifting materials from outside the holder.

When construction of the piston was completed, erection of the shell began. The piston was floated on compressed air and served as a platform for materials and tools, as well as a scaffold from which the work was done;

The shell was composed of forty-two equally spaced vertical I-beam columns, and horizontal stiffeners spaced approximately five feet one inch apart, to which steel plates were riveted. The shell was further stiffened by outside circular girders, approximately thirty-one feet apart, forming balconies reached by a stairway on the south side. One of the steel plates had been left out on the south side at the third balcony, to make the “door sheet,” through which workmen entered the shell.

The holder, upon completion, was to have a height of approximately three hundred and four feet, a base area of thirty-eight thousand square feet, or seven-eighths of an acre, and a capacity of ten million cubic feet.

On the night of September 30, 1936, when the damage occurred, the piston, which was ultimately to weigh approximately fourteen • hundred tons, weighed eight hundred and thirty tons. One hundred and eighteen feet of the shell had been completed, and the fourth balcony affixed at a height of one hundred and twenty-four feet. The piston was floating with the upper girder at the third balcony and the lower girder slightly below the second balcony, or fifty-eight feet above ground. At the same time, about forty-four lineal *705 or seventy-three square feet of each crane's boom was exposed to the wind above the completed shell. As a consequence of damage to the piston seal, the cause of which was the real issue in this case, the compressed air escaped and the piston fell in a tilted position, one side resting on the bottom of the holder, and the opposite side against the shell at a height of seventeen feet.

Appellee claims that the force of the wind against its exposed boom toppled the south crane off the rail into the back-bracing, causing the piston to tilt and the top rollers on the south side and the piston seal on the north side to give way, thereby permitting the supporting air to escape. Appellant, on the other hand, contends that there is no substantial evidence that either derrick was derailed before the piston began to fall, 'maintaining that the piston, never in an absolutely level position, fell because of the accumulation on the south side of many tons of water from heavy rains, which were not carried off with sufficient rapidity by the temporary drainage system. To appellee’s recovery it was essential that the jury find that the wind first derailed the south crane and that its derailment caused the piston to fall. We are of the opinion that there was substantial evidence to support the jury’s finding of these essential facts.

Bertram, one of the welding inspectors, testified that he was in the tool house near the gasholder shortly before the damage occurred; that the wind was blowing very rapidly; that a noise developed in the tank that sounded like a cable or something slapping against the inside of the shell; that he and the other three men immediately went out of the tool house and heard a terrible noise in the holder that sounded "“like a great thud of something that might have fell down or been bumped hardly against the tank”; that the wind was then blowing at least fifty miles an hour; that he immediately turned his flashlight up toward the top of the shell; that only about two feet of the south crane’s boom were visible and “appeared to be laying to the right if you were standing facing the door sheet,” whereas it was straight up earlier.in the evening; and that a “terrible whistling and screeching noise started immediately after the thud.”

Curlee, the other welding inspector, and Lawson, an employee of appellee, testified to the same effect.

Rowland, appellee’s night supervisor, testified that he heard something like a cable slapping against the shell of the holder when he was in the tool house with the other men; that he went outside but couldn’t see anything and, the noise quieting down, he returned to the tool house; that, in a couple of minutes, “a gust of wind came that seemed it was going to take the tool room off, and this loud crash came from the holder”; that he jumped up, went outside, looked up at the gauge and found that the air pressure was normal; that he went up the stairway to the door sheet and found the south crane, which had been on a level with it, off the track down in the back-bracing, its boom lashing back and forth; that the piston was then five or six feet below its former level, quivering and rocking sideways with the crane; that, in a few minutes the piston settled “to the bottom, with thumping and screeching noises that “sounded something like a train might be going around a real sharp curve with the flanges biting on the rails.”

Appellee also introduced expert testimony that the pressure exerted upon the exposed portion of the boom by a fifty-mile wind could cause such a failure of the sill, or axle, of the crane as would result in its overturning.

Appellant has included in its brief numerous charts and drawings, based upon record data, and has made calculations thereon for the purpose of showing that the testimony of appellee’s witnesses as to how the collapse occurred could not be true, in view of certain physical facts, particularly markings on the inside of the shell, claimed to have been made as the piston fell; that appellee’s evidence is therefore without probative value; and that the verdict, in consequence, is unsupported by substantial evidence. But no special findings were requested or made, and we cannot conclude from the data properly before us, nor may we assume, that the particular markings relied upon to discredit appellee’s evidence were made as claimed by appellant. The inference as to what caused the piston’s fall was for the jury. Viewing the evidence in the light most favorable to appellee, as we must after verdict, we are of the opinion that appellant’s claim that there was no substantial evidence that appellee’s damage was caused by windstorm is without merit. Accordingly, we find no error in the District Court’s denial of appellant’s motions for a *706 directed verdict, for judgment notwithstanding the verdict, and for a new trial.

Appellant further contends that the weight of such water as appellee’s own witnesses testified had accumulated, necessarily contributed to the piston’s fall, and that this fact relieves it from liability because of the following provision of its policy: “This company shall not be liable for any loss or damage caused by water or rain, whether driven by wind or not.” However, the policy expressly insured against “all direct loss and damage by windstorm” and we are of the opinion that this coverage extends to losses where windstorm is a contributing cause. Cf. Jordan v. Iowa Mutual Tornado Ins. Co., 151 Iowa 73, 130 N.W. 177, Ann.Cas. 1913A, 266.

There was no exception to the Court’s charge. Nineteen of appellant’s twenty-one assignments of error relate to the admission or exclusion of evidence. 'It would unreasonably lengthen this opinion to discuss each of these assignments. It must suffice to say that we have 'carefully considered them and find no prejudicial or reversible error.

The judgment of the District Court is affirmed.