DeWITT, J.

This action was brought by De Kalb county against the Tennessee Electric Power Company to recover damages for the destruction of a bridge across Caney Fork river at Holmes creek, which bridge was owned by the county. The bridge was washed away on the night of Saturday, March 23, 1929, by an unprecedented flood in Caney Fork river. The county claimed that this was the proximate result of negligence and improper manipulation of the dam owned and maintained by the Tennessee Electric Power Company at Rock Island, situated 58 miles by river above the bridge; also the proximate result of negligent maintenance of a dam at McMinnville, across Barren fork of Collins river, 25 miles by river above the Rock Island dam. It is claimed that the company improperly and negligently caused an unnatural and excessive volume and pressure of water to go down the Caney Fork river, augmenting the natural volume of water to the extent that it washed away the bridge.

The defendant resisted the action on the grounds that it was not guilty of any negligence or improper use of the dam at Rock Island, or manipulation of its gates; that the gates were operated so that no more water was passed over the dam or through the gates than was coming into the pond above the dam; that the break which occurred in the McMinnville dam had no effect upon the bridge, no part in the destruction of it; and that the damage suffered by the county was due solely to the act of God, as it was caused solely by the natural rise of the river and flow of its waters due to an unprecedented rainfall.

Upon the trial to a jury, at the close of the evidence in behalf of the county, the circuit judge sustained a motion for a directed verdict in defendant's favor on the third count of the declaration, which was based on the alleged negligence in maintaining the McMinnville dam; and at the close of all the evidence the circuit judge sustained a motion for a directed verdict in favor of the defendant on the other two counts of the declaration, holding that the facts were undisputed; that during the great rain and flood the defendant kept the water at the Rock Island dam at a uniform level as long as the same could be regulated by the opening of flood gates; that thereafter the defendant could have no control over the flood, and there was therefore no negligence shown on its part. He thereupon dismissed the suit upon the directed verdict, and his action in so doing is the principal ground relied upon for reversal of the judgment of dismissal.

The question here presented is whether or not there was any material conflict in the evidence; whether reasonable men might differ in their conclusions from the evidence if it is undisputed on all material matters. Defendant company's dam is 75 feet high, running 800 feet across the Caney Fork river just below the mouth of Collins river. This forms a pool or reservoir in the Caney Fork river and its tributaries, comprising something over 2,300 acres. The Caney Fork river and its tributaries drain a total area of about 2,600 square miles, 1,640 square miles being above the dam. The river rises in the Cumberland Mountains, as do its principal tributaries above the dam, Rocky river, Calf Killer river, Collins river, and Cane creek. The river is subject to sudden rises due to the fact that these streams run through rocky gorges or canyons for the most part. About 25 miles above the dam along the course of Collins river and its tributary Barren Fork river, and at McMinnville, was another dam across Barren Fork river, also maintained and operated by the defendant corporation for use in generating electric current. One of the charges of negligence in this cause against the defendant company is that this dam was defective and a part of it washed out during the flood, suddenly releasing a large volume of water.

Between Collins river and Caney Fork river at the main dam is a narrow but high neck of land through which the defendant company constructed a tunnel, which carries the water to the Caney Fork side, where it drops perpendicularly through steel chutes to a power house, and by the operation of turbine engines causes the generation of electric current by other machinery. This power house is situated about three-fourths of a mile in a gorge of the river below the dam. The water so used passes through the power house and through tail-races in the Caney Fork river. The dam is built of concrete and is Taintor gate or flood gate type, and it is one of the standard recognized types of dam. Set in the face of the dam and forming a part of it are eighteen of these Taintor gates, each 14 feet high and 25 feet wide, so arranged that they can be raised or lowered mechanically to any desired height. The top of each gate is set flush with and forms a part of the dam. It appears that this type of power dam is extensively used by power companies, and that the purpose of the gates is to control the level of flood waters in so far as possible by opening the gates and passing along the flood waters which flow in above the dam.

On the afternoon and night of Friday, March 22, 1929, there was a tremendous and general rain over the entire drainage basin of the Caney Fork river. In the region above the Rock Island dam the rainfall was between 8 and 10 inches. Below the dam it was about 6 inches. The evidence shows without dispute that in the region above the dam this rain produced the heaviest flood in the streams within the memory of living persons, or according to the records which had been kept. This rainfall appeared between the evening of Friday, the 22d, and about 4 a.m. on Saturday, the 23d. The water rose very rapidly in all the streams from Friday night until about noon on Saturday. It was not until 9 o'clock on the night of Saturday that the flood arose so high as to damage the bridge 58 miles below the dam. About that hour one section of the bridge was washed away. Another section was washed away about midnight. The water rose to the height of about 60 feet at the site of this bridge by that time. The rain had virtually ceased very early in the morning of Saturday, so that the flood reached its height from seventeen to twenty hours after the rain had ceased. The evidence shows without dispute that the current of the river below the dam ran during this time at the rate of 8 to 10 miles an hour. The evidence shows without dispute that about 5:30 p.m. on Friday, the operators of the dam discovered that the water was rising above the dam, so the first gate was raised, releasing from the pond above the dam a stream of water 14 feet high and 25 feet wide; that thereafter, as will be shown in detail, the gates continued to be raised at intervals, releasing the water until about 3 a.m. on Saturday; and that during this time the pool level was maintained uniformly at about six-tenths of a foot above the dam, which was according to the rule of the defendant company, until the flood came beyond control. During all of this time the water was flowing both through the gates which had been opened and over the dam. Of course, this shows undisputably that there had been a great augmentation of water coming rapidly down the river and to the dam. If no gates had been opened, this water would have flowed over the dam. If there had been no dam, this water would have flowed on through the gorge where the dam is situated and on down the river. Before the last gate had been opened, it was no longer possible to control the flood by maintaining the pool level at six-tenths of a foot above the dam. The evidence is undisputed that finally the water rose to a height of 15 feet above the dam at Rock Island, and at the same time it was rushing through eighteen open gates.

The work of raising the gates was done by foremen with their assistants in shifts of eight hours each. At 6:30 p.m. on Friday, the second gate was raised. The pool level was more or less standard until 10:15 p.m., when another gate was raised; another was raised at 10:40; another at 11; another at 11:15; another at 11:25; another at 11:45; another at 11:55; another at 11:59; another at 12:10; another at 1:45; then others at 1:55; 2:20; 2:30 (two gates); 2:40; 2:45. At 3:05 the last gate was raised. These gates were thus raised rapidly in succession; water was pouring over the top of the dam; the pond was rising all the time; but the level of about six-tenths of a foot above the dam was maintained until it was no longer possible.

Only one conclusion can be drawn from these facts — that until the last gate was raised no more water was going out than was coming in. Of course, after this time this was also true, but the flood was far greater. The amount of flood and the pressure of the water, after the effect of raising all the gates was over, cannot be considered in any view as involving negligence on the part of the defendant company.

It can only be concluded from the evidence that if no gates had been raised about the same volume of water would have gone over the top of the dam, and that the same water that went through the gates would have gone over the dam. It is insisted that the defendant company was negligent in not opening all the gates in the afternoon of Friday, the 22d; that if this had been done the water which would have gone through at that time would have been water which was merely stored up above the dam and would have flowed on and gotten out of the way before the tremendous rise came. In the first place, there is no evidence that the defendant company could reasonably have anticipated any necessity for so doing, which would be due to the probability of rise and unprecedented flood coming down the river. The rain came suddenly and in enormous quantity. The same water which would have thus gone through the gates went through afterwards accordingly as the gates were raised; and as it came through it did not come all at once in the enormous volume with which it would have come had all these eighteen gates been raised at once. The sole conclusion that can be drawn is that inasmuch as the pool level was kept standard when the gates were raised, an even lesser volume went down the river than would have gone down the river had all these gates been raised at once or within a very short period. The evidence shows that when a gate was raised the water went through it at the rate of 5,000 cubic feet per second. The evidence further shows without dispute that although the water left the dam under heavy pressure from the sudden release, the force of this pressure was broken and became abated in a few miles as this water went over rocks, came to pools, and went around the many curves in the river.

The break in the wall connecting the dam at McMinnville with the railroad abutment occurred about midnight on Friday. There were cracks in this wall and water had been flowing through the cracks. After this break the water rose to a height of 20 feet above that dam. This dam is 25 miles from the Rock Island dam. Of course, the break did release a volume of water; but the undisputed evidence shows that first this water came some 5 miles or more along Barren fork to Collins river, where it struck a tremendous flood going counter to this oncoming water, and that the amount of additional water which thus came suddenly from the break in the McMinnville dam would not raise the pool above the Rock Island dam more than an inconsiderable amount. When we consider that the pool level was maintained uniformly at the Rock Island dam until the flood could no longer be controlled, we must conclude that the additional water from Barren fork due to the break in the dam did not augment materially the volume at the Rock Island dam, unless it arrived there after the pool level at that dam began to rise. However, if this occurred, the additional volume was so small as to be immaterial. The difference of this small amount in a rise of 60 feet at the site of the bridge, 58 miles below the Rock Island dam, did not cause the damage to the bridge. The evidence shows without dispute that it was the last 7 or 8 feet of water at the bridge that caused the damage, and it cannot reasonably be concluded that but for the small amount of additional water the damage would not have been done to the bridge.

There is evidence that at the site of the bridge the water rose steadily until about noon on Saturday and then remained stationary for a short time, and began to rise again. At this time, noon, the water had not done damage to the bridge. If we infer that by noon of Saturday the height of the water at the bridge was due to the 6 inches of rain which had fallen in the region below the dam, plus the natural flow from the region above the dam, plus the stored-up water which had come through the gates, it is impossible nevertheless to conclude that this water which had come through the gates was still there at 9 o'clock on Saturday night when the flood reached 60 feet. We must bear in mind that the last gate had been opened at 3 a.m. on Saturday; that the water was flowing down the river below the dam at the rate of at least 8 miles an hour. A large part of this stored-up water, which had been released through the gates, had been released long before 3 a.m. on Saturday. Now as this water flowed down the river at the rate of at least 8 miles an hour, it would reach the site of the bridge in seven hours, or a little more, going 58 miles. Therefore, the stored-up water which was released must have passed the site of the bridge by noon on Saturday. Thereafter the water which came to that place was water which in no view of it was in any way augmented by any manipulation of the gates or maintenance of the dam. It was simply the natural flow in unprecedented volume which had gotten beyond all possible control upon the part of the defendant company. It is perfectly clear from the undisputed evidence that the water which caused the enormous rise that washed away the bridge was the natural volume of water due to the flood, and that no more water had been going beyond the dam than was coming through and over it.

There was nothing in this case to be submitted to the jury for determination. His honor, the trial judge, was correct in so interpreting the evidence. Testimony of witnesses who were not impeached or discredited, must be accepted as true, so that any argument that the testimony of experts as to these physical facts should have been submitted to the jury along with the rest in order that the jury might have the opportunity to refuse to believe such testimony, is not sustainable. Frank v. Wright, 140 Tenn. 535, 205 S.W. 434. Neither did that testimony invade the province of the jury, for it was of such character as merely might be of aid to the jury in reaching a conclusion upon the ultimate issue. McCravy v. State, 133 Tenn. 358, 181 S.W. 165.

To allow a jury upon this record to determine the issue would have been to leave to them mere matters of conjecture. A verdict for the plaintiff would have had to be based upon mere conjecture or upon inferences, which could not be drawn reasonably from the evidence. It is a familiar rule that the law requires an open, visible connection between the principal and evidentiary facts and the deduction from them, and does not permit a decision to be made on remote inferences. U.S. v. Ross, 92 U.S. 281, 23 L.Ed., 707. In Noble v. C. Crane Co. (C.C.A.), 169 F., 55, 59, the rules, aptly and soundly stated and approved by this court in certain reported cases (Fitch v. American Trust Co., 4 Tenn. App. 87; Louisville N. Railroad Co. v. Jackson, 3 Tenn. App. 463), are given as follows:

"It is the duty of the court to direct a verdict unless the conflict is positive. The conflict must be real, not merely apparent. A mere dogmatic assertion, which does not appeal to the reason of the court, which does not have substance and relevant consequence, which does not have fitness to induce conviction, is not proof, even if uncontradicted, and does not interfere with the duty of the court to direct a verdict."

It is true that even where the facts are undisputed, if intelligent minds may draw different conclusions as to the matter at issue, questions of negligence are ordinarily for the jury. Carey Roofing Manufacturing Co. v. Black, 129 Tenn. 30, 164 S.W. 1183, 51 L.R.A. (N.S.), 340. But this rule does not overcome the effect of the rule that a verdict cannot be based upon mere guess or conjecture if intelligent minds may not draw therefrom different conclusions upon the question of negligence. The exercise of guess or conjecture is not the operation of such reasoning as justifies a verdict upon an issue, for an issue is a matter of great seriousness, the rights and interests of parties are deeply involved, and it is the policy of the law not to allow them to be thus passed upon.

It is insisted that the court erred in permitting the witnesses Crouch, J.C. Cooper, W.M. Cooper, and Steuhser to testify as to floods occurring in the Cumberland Mountains and in the Emory river and Tennessee river basins; that is, being on the other side of Cumberland Mountains and in a different drainage basin. This testimony merely went to show that this heavy rain was of a general character east or northeast of the Caney Fork basin, causing there an unprecedented flood. We are unable to conclude that it would have materially affected the verdict of the jury. If it had been submitted to the jury, it would not have been reversible error.

It is further insisted that the court erred in sustaining objections to the testimony of many witnesses that had lived on or about Caney Fork river all their lives, had observed the river for many years, were familiar with the rises admitted to be caused by the operation of the dam, and gave as their opinion that the second rise in the river, which carried away the bridge in question, was caused by what is known along the river as a "dam" rise, and by the opening of the gates of the dam. There was no error in sustaining the objections to this testimony, for the opinions sought to be expressed were determinative of the ultimate issue which would have been submitted to the jury. Cumberland T. T. Co. v. Mill Co., 129 Tenn. 374, 164 S.W. 1145, L.R.A., 1915A, 1045. But beside all these questions as to the admissibility of evidence, it is clear that the other evidence in the cause was of such character as alone to justify the action of the circuit judge in granting peremptory instructions in favor of the defendant company. That evidence presents, without material dispute, the determinative facts in the cause. This statement also applies to the assignment challenging the introduction at the trial of a purported miniature of the dam and the demonstration of the operation of the gates thereof upon the trial of this cause. The trial judge did not base his action upon this miniature or the demonstration thereof. The facts which were shown without dispute, and which we have carefully endeavored to set forth in the preceding part of this opinion, were in themselves determinative of the question whether or not the cause should have been submitted to the jury.

In the case of Tennessee Electric Power Co. v. Robinson, 8 Tenn. App. 396, the determinative fact was that over two billion gallons of water was drawn out of the pool above the Rock Island dam in addition to the amount that was flowing in above the dam; in other words, this was the amount of water that was drawn from the pool; and there was an increase of volume and acceleration of flow beyond the natural volume and flow due to the negligent opening of the gates in the dam. We do not have in this case such a state of facts. There is no evidence that the water which went below the dam included a volume that had been stored up and that would not have gone over the dam if the gates had not been raised.

This action could only be maintained upon the theory of negligence on the part of the power company in manipulating the gates in the dam, for the company had the right to maintain the dam and to permit flood waters to pass through or over it in such quantities as flowed into it. Crawford v. Cobbs Mitchell Co., 121 Or. 628, 253 P. 3, 257 P. 16; Tennessee Electric Power Co. v. Robinson, supra. Furthermore, even if the defendant was guilty of negligence, the plaintiff could not recover unless such negligence was the proximate cause of the injuries complained of.

There is no evidence that the destruction of the bridge was proximately due to any other cause than the natural flow of the water in unprecedented volume. This court is not at all unmindful of the terrible destruction and consequent loss of property which was suffered by the county of De Kalb, and by many landowners from this great flood in the valley of the Caney Fork river. If there were evidence tending to show that this was due proximately even in part to wrongs committed by the defendant company, this court would not hesitate to hold that the issues in suits arising therefrom should have been submitted to the juries for their determination. It is the duty of the trial judge when the material, determinative evidence is undisputed and only one inference therefrom can reasonably be made, and that in favor of the defendant, to treat the whole issue as an issue of law and to try it by directing the jury to return a verdict in favor of the defendant. In such a case there is no denial of the right of trial by jury. That right accrues only when there are such material conflicts in the evidence, or such diverse inferences can reasonably be made therefrom, that a jury instead of the judge has the power and duty to arrive at conclusions upon the issues. It results that there is no error in the judgment of the circuit court upon the directed verdict, dismissing this suit, and therefore that judgment is affirmed. The costs of the appeal will be adjudged against the county of De Kalb and the sureties on its appeal bond.

ON PETITION TO REHEAR.

De Kalb county has presented to this court a petition to rehear this cause, vacate the judgment formerly rendered, affirming the judgment of the circuit court, and to find some additional facts. This cause was very carefully investigated and considered prior to the decision already rendered. The petition presents no new facts or phases of this controversy and this court is satisfied with its former decision.

The function of this court in jury cases is not to make findings of fact, but merely to recite evidence relied upon, either to require the submission of the issues to the jury or to support a verdict which a jury may have rendered. Anderson v. Stribling, 15 Tenn. App. 267. Section 12 of chapter 100 of Public Acts of 1925, requiring the chancellor and the Court of Appeals to file written findings of fact, has no application to a case tried by a jury demanded by the parties or one of them. An assignment of error in this court that the circuit judge sustained a motion for a directed verdict presents a question of law and not a question of fact, and it is not the function of this court to find the facts, but it is our duty to ascertain whether there were any material conflicts in the evidence, or reasonable minds might reach different conclusions from the undisputed evidence. By "material evidence" is meant the evidence material to the question in controversy, which must necessarily enter into the consideration of the controversy and by itself, or in connection with the other evidence, by determinative of the case. But a conflict of the evidence upon an additional or separate feature or fact, even though it is material, should not of itself prevent the giving of peremptory instructions. Facts are frequently material in themselves, but become immaterial when taken in connection with other facts; so that the disputed fact must not only be material but in itself or in connection with other facts it must be determinative of the real issue and the merits of the case. Knoxville Traction Co. v. Brown, 115 Tenn. 323, 89 S.W. 319.

There is evidence that Caney Fork river in the vicinity of Holmes creek, about 58 miles below the dam at Rock Island, was 7 to 9 feet higher in 1929 than in 1902; that the rain pour below the dam in 1929 was less than it was in 1902; and that certain creeks which were tributaries to Caney Fork river below the Rock Island dam were lower in 1929 than in 1902. But from these facts no inference could reasonably be drawn that the greater height of the flood in 1929 was due to improper operation of the gates of the Rock Island dam, in the face of the evidence that the rainfall above the Rock Island dam in 1929 was greater than it was in 1902 and was unprecedented in volume.

There is evidence that Barren Fork river at the place where the McMinnville dam was washed out was higher in 1902 than it was in 1929; but as pointed out in our former opinion, the only conclusion that can be drawn is that the additional volume of water suddenly produced by the break in that dam had no part in the destruction of the bridge; and therefore this fact is immaterial.

The petitioner is mistaken in the averment that there is evidence that the raising or opening of the gates caused the rise of some 25 or 30 feet. The evidence cited in support of this statement does not bear out the statement. The rises mentioned in the testimony cited were natural rises, designated by defendant's witnesses as the extent of the rises at the various times when the gates were opened — not rises caused by the opening of the gates. Furthermore, we have shown that it could only be concluded from the evidence that the water which was let loose through the gates at successive times had passed this bridge at the time the damage was done late in the night of Saturday.

The question of the average stream flow of the Caney Fork river is immaterial. The material fact was the rate at which the river was flowing during this heavy flood. This is recited in our former opinion. Nor is the difference in the way Caney Fork river rises and falls after rains since the dam was put in at the Rock Island site, and the way it rose and fell theretofore, of material or determinative quality. It is not a question of the habit of the river in rising and falling during ordinary rains. The material fact is that the pool was maintained at a virtually uniform level above the dam by the operation of the gates until the flood was beyond control. We have heretofore held that the undisputed evidence showed that on the afternoon and evening of Friday the rainfall was not of such magnitude as to cause the power company reasonably to expect that the flood would be as great as it became; and we have stated the conclusion from the undisputed evidence that had all the gates been opened on that afternoon the result would not have been any better than it was. In other words, the same water which would have thus gone through the gates went through afterwards accordingly as the gates were raised, and as it came through it did not come all at once with the enormous volume with which it would have come had all these eighteen gates been raised at once. In our former opinion we recited that there was evidence that by noon on Saturday the stored-up water, which was released by on Saturday, and then remained stationary for a short time and began to rise again. At this time, at noon the water had not done damage to the bridge. We showed that it could only be concluded that by noon on Saturday the stored-up water, which was released by opening the gates, must have passed the site of the bridge. Counsel seem very confidently to ask the question in their petition: "What caused this second rise if not by the opening of the gates of the dam?" Of course, as we have pointed out, it can only be concluded from these facts that what caused the second rise was the enormous natural flood which had come down the Caney Fork river and from its upper tributaries, and which rose to a height of 15 feet above the dam with all the gates open. If there was any artificial flood that was let loose by the opening of the gates, it had passed this bridge by noon on Saturday.

The purpose of the flood gates in the dam at Rock Island is to maintain the pond level so as to protect not only the bridges and property of the power company above the dam, but also the property of the riparian owners above the dam; and the further purpose of them is so to regulate the flow of the water as to maintain a uniform pool level at the dam so that as long as the flood can be controlled no more water will go over or through the dam than comes into it. There is no evidence that the purposes of the flood gates are without regard for the rights of the riparian owners below the dam.

The petition to vacate the former judgment of this Court is denied.

Faw, P.J., and Crownover, J., concur.