Fred J. Muhle recovered a judgment for $7,175.00 against the defendant insurance company on an insurance policy covering loss to his farm buildings from windstorm. The policy provided that the defendant should not be liable-

'* * * for loss caused directly or indirectly by (a) frost or cold weather or (b) ice (other than hail), snowstorm, tidal wave, high water or overflow, whether driven by wind or not.'

The defense was that the insured's damage was caused either entirely by flood water or by flood water driven by wind and hence was not covered by the policy. Mr. Muhle died after the trial. His widow was substituted for him in this action.

Plaintiff owned a farm near the Missouri River, three miles south and one-half mile west of Norborne, Carroll County, Missouri. During the month of July, 1951, his farm and the surrounding territory for miles were inundated by an unusually high flood. The water stood some eight feet deep on the Muhle place, forcing them to move to Norborne. Sunday, July 15, Mr. Muhle went to the farm twice in a boat to remove chickens from the loft of the barn. The second trip was at about 7 p.m. All of the insured buildings, although partially inundated, were in place and in good condition at that time with the exception of such damage as water standing in the dwelling house would cause to plaster, floors, and woodwork. The water was backwater from the Missouri River and had little if any current. Such current as one or two witnesses observed was flowing to the northeast. Between 6 and 7 p.m., Monday, July 16, Mr. Muhle was on higher ground about three miles from his farm, when he saw a very dark funnel-shaped cloud come from the northwest. He observed its path and testified that it passed over or near his farm buildings. He could see where the farm was from where he was and could have seen the buildings if some trees had not obstructed them from his view. There was considerable wind and it became very dark for that time of day. He observed the direction of the cloud and at the trial drew a line on a map indicating its course. He could not get to the farm that evening, but the next morning when he did, the walls of the house were blown outwardly, the roof torn loose from the rafter plates, the shingles loosened, the doors and windows blown out, and the building twisted out of shape. It had also shifted from its foundation. It was torn down and rebuilt. The barn, brooder house and chicken house were gone. All that was found of the barn were some articles which had been in the barn loft, which were scattered over a course of about a quarter of a mile to the southeast of where the barns had stood. Doors were blown out of other buildings, the roofs torn loose, rafters, joists and girders torn and ripped loose. The north side of the house was gone. All of these conditions and others were discovered after the water receded. The storm was observed by others, including Mrs. Muhle, who was in Norborne at the time. There was testimony that the windstorm was a tornado and that the effects of tornadoes were freakish and some effects were unexplainable. Testimony on behalf of the plaintiff showed that the twisting motion of the tornadic wind created a comparative vacuum at the vortex of the funnel, and when the vortex passed over or near a building, the vacuum created on the outside of the building caused the air within to 'explode', pushing out the walls and lifting the roof. Plaintiff's expert testified on cross examination that the water inside the buildings would be forced downward by such an explosion, causing it to push out the walls below the water level. The walls of the house showed evidence of having been pushed out from within. Mr. Muhle and other witnesses testified to the similarity of damage caused to buildings on other farms in the path of the storm as fixed by Mr. Muhle, which were likewise flooded, and to the dissimilarity of damage caused to buildings also in the flooded area which were not in the path of the storm. That testimony indicated that buildings which were partially under water like plaintiff's, but which were not in the path of the storm, received slight damage and were repaired at the expense of plastering and painting.

Defendant presents four assignments of error which will be considered in the order they are stated. The first assignment is that the court should have sustained appellant's motion for a directed verdict. It is based on three propositions: first, that there was no substantial evidence that plaintiff's buildings were in the path of a windstorm; second, that there was no substantial evidence that plaintiff's loss or any part thereof resulted from the direct action of wind, as distinguished from damage caused by 'high water or overflow, whether driven by wind or not'; and third, that even if there had been any evidence of damage by the direct action of wind, the evidence did not separate such damage caused by standing water from that caused by moving water and that such separation was essential to plaintiff's recovery under the policy.

Was there substantial evidence that the tornado passed over or near plaintiff's buildings? As heretofore noted, Mr. Muhle observed the cloud from a point about three miles from his buildings. At the trial, over the objection of defendant, he traced a line on a map of Carroll County indicating the course of the cloud through the vicinity of his place. Other witnesses observed the cloud but did not undertake to establish its course with reference to the Muhle buildings. There was expert testimony of the characteristics of damage done by tornadoes. Those characteristics were present on the Muhle place. Testimony of a number of witnesses was to the effect that buildings submerged like the Muhle buildings, and in that neighborhood, but not in the path of the tornado as fixed by Mr. Muhle, received only slight damage and of a different nature from that done to the Muhle buildings. There was also evidence to the effect that other buildings, also submerged, and in the path of the tornado as fixed by Muhle, received damage similar to that done to the Muhle buildings. The eyewitness testimony of Muhle, the evidence of others that there was a storm, the similarity of damage to buildings in the path of the storm, as that path was fixed by Muhle, and the dissimilarity of damage to other buildings in the same neighborhood outside of the path of the storm, constituted substantial evidence that the Muhle buildings were in the path of the storm and that damage to those buildings was caused by the storm.

It is insisted that since Muhle was not an expert on tornadoes, and was three miles from his farm when he saw the cloud and observed its path, his testimony was not of probative value, should not have been admitted and must be disregarded. His testimony is likened to that of expert medical testimony in Kimmie v. Terminal R.R. Ass'n, 334 Mo. 596, 66 S.W.2d 561, where it was held that expert testimony that a cancerous bone tumor could have resulted from a back injury from a fall was insufficient to furnish more than a basis for speculation for a finding that the cancer did result from the fall. We do not find the situation in this case analogous to that in the Kimmie case. In the latter case the cancerous tumor was not present after the fall and did not develop until long afterwards. The facts and circumstances were such that it was as probable that the tumor was caused from causes other than the fall as from it. Naturally, under those circumstances the court held that the evidence failed to establish the fall as the cause of the cancer, although the evidence was admissible to establish the existence of the cancerous tumor. Here not only the occurrence of the storm was shown, but facts observable immediately thereafter, and later after the high water had gone down, were sufficient to justify the inference that the tornado caused the major part of the damage to the Muhle buildings.

Did the evidence justify a finding that the major part of the damage to the Muhle buildings was caused by the windstorm as distinguished from damage caused by a combination of both wind and water? The responsibility rested on plaintiff to produce evidence which justified the factual inference that wind alone caused the damage for which he sought recovery. If the evidence he offered was as consistent with the inference that wind and 'water driven by wind' caused that damage, as with the inference that wind alone caused it, the evidence established neither, and he could not recover. Wilson v. Borden, 61 App.D.C. 327, 62 F.2d 866; State ex rel. Trading Post Co. v. Shain, 342 Mo. 588, 116 S.W.2d 99. As the trial court construed the insurance contract in submitting the case to the jury, and as we construe it, if the damages sued for were caused by wind, and the damage sued for would have resulted from the wind alone without the presence of the water, recovery was proper. The defendant cites four cases construing contracts containing similar exclusion clauses as being windstorm policies only and if the damage is caused by a combination of wind and other elements there can be no recovery. Newark Trust Co. v. Agriculture Insurance Co., 3 Cir., 237 F. 788; Palatine Insurance Co. v. Petrovich, Tex. Civ. App., 235 S.W. 929; National Union Fire Ins. Co. v. Crutchfield, 160 Ky. 802, 170 S.W. 187, L.R.A. 1915B, 1094; Hartford Fire Ins. Co. v. Nelson, 64 Kan. 115, 67 P. 440. That is the proper construction of this contract.

Was the jury warranted in concluding that the damages sued for resulted from the tornado and not from a combination of wind and water? Defendant lays great stress upon the testimony of one of plaintiff's expert witnesses to the effect that the tornado caused the air inside the buildings, particularly the house, suddenly to expand with an explosive effect, that this caused a pressure on the water standing in the building and the water in turn pressed outwardly against the walls, causing or at least directly contributing to the damage involved. From that premise it is argued that the high water was an active instrumentality in causing the damage. The difficulty with defendant's theory that this established the fact that water was an active contributing element is that there is nothing in the record that would prevent the finding that substantially the same thing would have happened to the walls had the water not been there. The evidence justified the conclusion that a tornado struck these buildings; that the effect of a tornado was to cause a vacuum in the vortex of the tornado; that the air in enclosed buildings suddenly expanded with an explosive effect. These factual premises being established, it was not an unreasonable inference that the explosive effect of the air inside the building would have pushed out the walls if the water had not been there. That being true, plaintiff should not be denied recovery because the buildings were standing in water instead of in air. In the light of the evidence we cannot say that the conclusion of the jury, inherent in the verdict in the light of the instructions, that damage would have occurred regardless of the presence of the water, was not supported by substantial evidence. The argument is advanced that waves always result from wind and hence they must have contributed to the damage done. There was evidence that a tornado is usually preceded by a period of calm. There was evidence that it was calm before this storm struck. Waves do not ordinarily precede a wind. The evidence permitted a finding that there was substantial damage from wind alone before there could have been any effect from waves.

Defendant contends that even if damage from windstorm alone was shown, the evidence did not separate that damage from the damage caused by high water. Plaintiff asserts that damages caused by high water or high water driven by wind were an exception to liability which it was incumbent upon defendant to show. We need not be concerned with the latter theory. It has no application here. The policy was a fire and lightning policy to which had been added the extended coverage endorsement now involved, insuring against windstorm but not covering wind and water. The authorities dealing with exceptions to general coverage are not applicable. Plaintiff was entitled to recover damages caused by windstorm. There could be no recovery for damages caused by high water or high water driven by wind. It was incumbent upon plaintiff to prove the amount of damages caused by wind alone. Plaintiff's evidence showed damage from both wind and water. The evidence showed a total damage to all buildings of from $10,550 to $11,150. The amount of insurance carried was $7,175. That was the verdict. Under the circumstances of this case the separation of the damage from the two elements is unimportant except insofar as that was necessary in order to show the amount of loss from the element insured against.

The trial court properly and explicitly charged the jury that plaintiff could recover only for the damage caused solely as a result of windstorm, and further instructed that:

'If you should find and believe from the evidence that the damage described here was the result of the combined force of the water and wind, or the water driven by the wind, then it would be your duty to find for the defendant because that would not be direct loss or destruction as a result of the windstorm.'

And again in the summarization of the issue:

'If you find it was the result of water, a flood, then it is your duty to find for the defendant. If you find it was the result of a combination of a windstorm and water, the waves, the current whipped up by the windstorm, then it would be your duty to find for the defendant.'

Did plaintiff prove that damage from windstorm was as much as the verdict? It is suggested by plaintiff that since damage to other houses in the flood but not in the path of the tornado was described, the necessary repairs therefor enumerated, and the exact extent of the water damage to plaintiff's house and the cost of repair of such damage could not be ascertained because of the havoc wrought by the wind, it would be a matter of common knowledge that the water damage would not amount to the difference between the total damage and the amount of the verdict. The necessary implication is that the jury was warranted in so assuming, without any evidence whatever as to the damage in dollars from either element. The water was at least five or six feet deep inside the house. It could not be assumed that damages to the house would have been 'minimal', as plaintiff's counsel expresses it. Although it is not any more necessary that the amount of the damage from windstorm be fixed by the evidence in this case at an exact amount than it is in eminent domain proceedings and other cases where the judgment of witnesses as to values and damages may differ, it is necessary that there be an evidentiary basis for the verdict and that the verdict be based on evidence, not on mere speculation. Plaintiff is entitled to recover for damages caused by windstorm, but there must be evidence which will furnish reasonable support for a verdict therefore. That evidence is now lacking. As stated in Chicago, B.&Q.R. Co. v. Gelvin, 8 Cir., 238 F. 14, 22, L.R.A. 1917C, 983:

'The record showing different conditions, occurring subsequent to the fire, naturally affecting the gain in weight of these cattle, for which it is conceded the defendant could not be held responsible, without any proof of how much of the damage resulted from these conditions as distinguished from the damage resulting from the alleged injury, with nothing in the way of testimony of any witness pretending to even estimate the proportion of the damage resulting from either of the causes, is far short of that reasonable certainty required by law, and upon such a record a jury cannot arbitrarily apportion a part, or all, of the proven damages to the cause for which the defendant is responsible (Knowlton v. Chicago & N.W. Ry. Co., 115 Minn. 71, 131 N.W. 858), and the verdict of the jury upon the first cause of action, predicated upon a consideration of this evidence, is erroneous.'

Counsel for plaintiff state that Mr. Muhle's testimony fixing the amount of damage done to the buildings excluded the damage done from high water. His testimony is not fairly susceptible to that interpretation.

There was objection to the evidence of the nature of the damage done to other buildings in the path of the storm as that path was fixed by Mr. Muhle. The objection was that the evidence was immaterial. It was admitted as being material on the question of whether there was a storm. It was material for that purpose and in connection with the testimony of Mr. Muhle as evidence of the course of the storm and the nature of damage done by this storm, i.e., whether it was a tornado.

Objection was also made to the testimony of witnesses that the damage to plaintiff's buildings was similar to the damage to other buildings in the alleged path of the storm, on the ground that such testimony was an improper conclusion and opinion on the part of these witnesses. This evidence was admitted by the court for such value as it might have on the question of whether there was a windstorm and the direction it took. As heretofore indicated, the evidence was admissible. But in several instances the answers were in the nature of conclusions, unnecessary under the circumstances. Sometimes the nature of the descriptive facts is such that they are incapable of exact description, such as that a person appeared to be suffering, was weak and helpless, appeared sick, looked pale, or paler than usual, or was declining in health. In those instances a non-expert is allowed to give testimony which is in the form of an opinion. Partello v. Missouri Pac. R. Co., 217 Mo. 645, 117 S.W. 1138. The subject is discussed in Pillsbury Flour Mills Co. v. Miller, 8 Cir., 121 F.2d 297. For the proper application of this rule of evidence we need only direct attention to the Pillsbury case and the cases cited therein.

Because of the failure of the evidence to show loss from windstorm equalling or exceeding the amount of the verdict, the cause must be remanded for a new trial.