This is an action to recover damages for the washing away of a part of plaintiff’s land by the waters that flowed down the Lexington Wash, near the town of El Monte, in Los Angeles County, during the flood of *715 February 20, 1914. In August, 1913, the Bailway Company had reinforced its bridge across the wash by placing sway-braces on the bents. It is claimed by plaintiff that in thus placing sway-braces on the bents of the bridge, defendant was guilty of culpable negligence, and that this negligence was the proximate cause, or concurring proximate cause, of the diversion of the flood waters on to and through his land. The case was tried before a jury, a verdict was rendered for defendant, and plaintiff appeals from the judgment.

The following summary of the principal features of the case is sufficient for an understanding of appellant’s claims: Out of the steep mountains lying a few miles to the north of the area immediately surrounding plaintiff’s land, the San Gabriel Biver debouches. As the waters of the river emerge from the mouth of the canyon at the foot of the mountains, they follow, generally, and for some distance, the course known as the San Gabriel Biver. A little below where the Santa Fe bridge crosses the stream, the waters of the river, during some of the floods occurring over a period of years commencing at a time prior to the reinforcement of defendant’s bridge to the time of the flood of February 20, 1914, have divided, a part of the waters flowing in a southeasterly direction in the old course of the river, known as the Old San Gabriel or East Biver, and a part flowing southwesterly into and through Lexington Wash, which is now a broad, irregular sand and gravel wash, extending generally in a northerly and southerly direction. As is frequently the case with the streams of Southern California, there is, at the mouth of the canyon, a broad debris cone. Prior to 1891 Lexington Wash was a small wash extending generally in a northerly and southerly direction, and situated west of the town of El Monte, receiving its water from four or five small canyons lying to the northwest. Prior to 1891 this wash, which was then quite small, received no water from the San Gabriel Canyon, excepting once, in the year 1870. In 1891, by breaking through the debris cone at the mouth of the canyon, a considerable quantity of the waters of the San Gabriel Biver flowed westerly over into the Lexington Wash, widening it to some extent. This condition did not change materially until W10, when, on January 1st, during a severe rainstorm, a large part of the San Gabriel Biver cut through the debris cone below the canyon’s mouth, swung westerly *716 into the wash, widened it considerably, and, racing down the wash, followed its course down to and through the Bio Hondo into the Los Angeles Biver and thence to the sea. This condition was repeated in 1911, during which year two floods occurred. In the floods of that year Lexington Wash received more water from the San Gabriel Biver than it had the previous year. As we have said, the damage to plaintiff's land occurred during the flood of February 20,1914.

Defendant’s bridge, as originally constructed, was built some years prior to 1913. It crosses Lexington Wash about eight hundred feet north of and above plaintiff’s land. The bridge runs approximately due east and west. North of defendant’s bridge, and substantially parallel with it, are two other bridges—the county road bridge, which is about one thousand feet north of defendant’s bridge, and the Southern Pacific Bridge, which is about two thousand four hundred feet north of defendant’s bridge. As originally constructed, defendant’s bridge was about six hundred feet long. After the flood of 1910 it was lengthened to about one thousand three hundred feet. Prior to August, 1913, there was no sway-bracing on the bridge, but in that month defendant strengthened and reinforced its bridge by placing sway-braces upon and across the bents. It is these sway-braces that plaintiff claims are responsible for the damage done to his land. In his complaint he alleges that on or about August 15, 1913, defendant “negligently, carelessly, and wrongfully . . . added to . . . reinforced . . . braced . . . and changed said bridge. . . . That such negligent, careless and wrongful acts were calculated to and did interfere with, impede, delay, and modify the natural, usual, and ordinary course, direction, and flow of said river, and was calculated to and did cause said river to be diverted and changed from it old, original, and accustomed bed and channel. That on or about February 20, 1914, waters gathered in and flowed down the course and channel of said river, as they were accustomed to, until they reached and arrived at the bridge 'of said defendant, where, because of the negligent, careless, and wrongful acts of the defendant, . . . they were impeded . . . and modified from their usual and ordinary course . . . and changed in such a manner . . . that they formed a new and independent course and channel . . . almost due south ... so that they came into and upon the lands of plaintiff, . . . and cut ... in *717 . . . the land of plaintiff a deep, wide, new, and independent hed or channel.” One of plaintiff’s expert witnesses, Wilkie Woodard, a civil engineer, testified that, in his opinion, the bridge, prior to the addition of the sway-braces, was sufficient to carry all the waters of the river during ordinary high water, including the waters of the flood of February, 1914, without injury to plaintiff’s land; but that “after the placing of those sway-braces upon those bents ... in my opinion the bridge was not in such condition as that it would be sufficient to carry off the ordinary high water of the river, and especially the flood of February, 1914, without doing injury.” By its answer defendant denied the allegations of negligence in its reinforcement of the bridge by the addition of the sway-braces, and further denied that the bridge was the cause of the damage. Defendant also, by way of affirmative defense, alleged that the flood of February 20, 1914, was the result of an unusual, unprecedented, and extraordinary storm, and that it was of such volume that it could not have been reasonably anticipated. The flood washed away two bents in defendant’s bridge, and otherwise damaged it. The ■witnesses on both sides described in detail this storm of February, 1914, and its effect upon the water flowing down and through Lexington Wash.

[1] A railroad company in constructing a bridge across a stream is bound to exercise ordinary care to avoid injury to neighboring lands. Ordinary care is such care as is exercised by ordinarily prudent persons under the same or similar circumstances. In the exercise of ordinary care, it is the duty of a railroad company to guard against such floods or freshets as men of ordinary prudence can foresee, but not against such extraordinary floods and accidental casualties as cannot reasonably be anticipated. It not only should employ engineers of at least ordinary ability, who will bring to bear such engineering skill as is ordinarily applied to works of that kind, but it is its duty to exercise a reasonable degree of care and prudence in the construction of the bridge, taking into account the laws of hydraulics, the natural formation of the country, and the character of the stream, its habits and history, to the extent of learning its probable behavior under conditions which experience has shown are likely to recur, so as to guard against injuries which may reasonably be anticipated, not only from the usual state of the stream, but

*718 also from such extraordinary floods and freshets as may reasonably be anticipated, in view of the stream’s past history and known behavior, even though such floods or freshets have occurred but infrequently. For the bridge should be so constructed as not to be subject to the risks and perils arising from rainfalls known by experience to be incident to that section of the country, though rarely occurring, or which ordinarily competent and skilled engineers should reasonably anticipate. [2] But there is no liability for floods which would surprise ordinary caution. The company is not bound to provide against such extraordinary floods as have never been known to occur, and which competent and skilled engineers could not reasonably anticipate. And if the company has employed engineers of at least ordinary ability and skill, and has used all ordinary precautions in the construction to have the .work done properly, it will not be liable to others for injury to land caused by an extraordinary and unprecedented storm and flood, unknown to common experience, and which could not reasonably have been anticipated.

Over the objection of appellant, the court permitted several witnesses for respondent, owners, some of them, of lands a mile or so upstream above appellant’s land, and others, owners of lands some distance downstream, from appellant’s property, to testify respecting the actions and behavior of the waters in the wash opposite their respective land holdings during the flood of February 20, 1914—the flood that resulted in the injury to appellant’s land. It is the admission of this testimony of which appellant principally complains. His contention is that it is prejudicial error to admit testimony as to injury done to other lands along the course of the river during a flood, unless a proper foundation be first laid by showing that the conditions at such other points on the stream were the same as at plaintiff’s land. [3] Undoubtedly it is the general rule that, in an action against a railroad company to recover damages occasioned by an overflow of plaintiff’s land, resulting from the railroad’s negligence in the construction of its bridge, evidence that other lands upon the stream were inundated by the same flood is inadmissible to support the defense that the flood was extraordinary and unprecedented, unless it be shown that there is a similarity of location, surroundings, and conditions as between the respective lands; for, unless such similarity be shown, it may *719 be that these other lands were so situated with reference to the stream that they were subject to overflow, to the extent shown, even upon the occasion of ordinary floods, or of floods that reasonable prudence could have foreseen and anticipated. (Bell v. Missouri K. & T. Ry. Co., 36 Tex. Civ. App. 569, [82 S. W. 1073]; Missouri K. & T. Ry. Co. v. Bell (Tex. Civ. App.), 93 S. W. 198; Farnham on Water and Water Rights, sec. 595; 40 Cyc. 586.) It may be conceded, therefore, that the general rule is as contended for by appellant, where such evidence is offered solely for the purpose of showing that the flood was of unprecedented dimensions; but the difficulty that confronts him is that in this case the evidence was admissible for an entirely different purpose.

[4] The burden rested upon plaintiff to show, by a preponderance of the evidence, not only that defendant was guilty of negligence as alleged in the complaint, but that such negligence was the proximate, or a concurring proximate, cause of the injury. For if, notwithstanding any negligence in reinforcing and sway-bracing the bridge, the injury to plaintiff's land would have occurred, nevertheless, even had there been no such structure, then its existence could not have caused the damages and defendant would not be liable. (San Antonio & A. P. Ry. Co. v. Kiersey, 98 Tex. 590, [86 S. W. 744] ; Vyse v. Chicago, B. & Q. Ry. Co., 126 Iowa, 90, [101 N. W. 736].) [5] Also, if the direction of the flow of the flood waters, during these excessive rainfalls, constantly changes, so that no engineer, however familiar with the known habits of the stream, could foresee at what angle the waters of any flood might strike the bents and sway-braces, it could not be claimed that defendant was negligent simply because it did not place the bents or sway-braces at some particular angle with the longitudinal direction of the bridge or with the ever-changing “axis” of the flood waters, if such waters can be said to have any particular axis of flow.

[6] To sustain his contention that the bridge was a concurrent proximate cause of the damage to his land, appellant, both by the allegations of his complaint and evidence introduced in his behalf, presented the theory that during the flood of February 20, 1914, the river had a definite “axis” at the place where the bridge crosses the wash. By evidence produced in his behalf, appellant presented the theory that immediately above respondent’s bridge, i. e., between re *720 spondent’s bridge and the county road bridge about one thousand feet to the north, the river had such definite “axis”; that in this section of its course, the axis of the stream was at an angle of about forty-five degrees with the sway-braces that had been placed across the bents for the purpose of reinforcing the bridge; that thus the flood waters, impinging upon these sway-braces at an angle of forty-five degrees, were deflected in a southerly direction toward and through appellant’s land, instead of passing on under the bridge as, in the opinion of appellant’s expert witness, Woodard, they would have done had there been no sway-braces. Appellant’s land, where it was cut through by the flood waters of February 20, 1914, lies to the south of the bridge. It is altogether probable that, between a point in respondent’s bridge where one of the bents went out and the middle of the county road bridge that crosses the wash about one thousand feet to the north, the stream as it flows in the wash after the subsidence of a flood, does flow at an angle of about forty-five degrees with the sway-braces. But it is a question whether the stream has any such well-defined axis during those periods when the waters of great floods are tearing through the broad gravel wash, carrying before them and turning over and over large uprooted trees, boulders, and other debris. This was one of the principal questions in the case—a question as to which there was a sharp conflict in the evidence. If the river did not have any well-defined axis during flood periods, and if, theretofore, regardless of the utmost care with which defendant may have sought so to construct the sway-bracing that the waters would not strike at an angle that would deflect them toward the plaintiff’s land, the waters, nevertheless, by a sudden switch of the stream and change in direction of flow, might strike the structure at any angle whatever, or if, in the absence of any bridge, plaintiff’s land, by reason of the switching of the river and rebounding of the stream, would have been cut into by the waters, then a state of facts would be presented that would afford an occasion for defendant to contend before the jury, either that no negligence could be predicated of defendant’s acts or that plaintiff’s land would have been damaged to the extent that it was even if there had been no bridge. In other words, if defendant could show that the waters in the wash, during these great floods, have no eon *721 stant and definite axis of flow, it could well contend that its alleged negligence was not the cause, or contributing cause, of the injury.

Appellant’s engineering expert, Woodard, in the course of his testimony, stated appellant’s theory as to the diversion of the river toward appellant’s land as follows: “Because the axis of the river from the highway bridge down to the Pacific Electric bridge and beyond was at an angle of forty-five degrees with the bridge, and at the time that the piles were all in there, the debris had collected and it was hanging on each and every pile and on the sway-braces and was forming considerable of a dam, and there was about thirty feet from the extreme north end to the extreme south end of each bent, and staying in that direction and being dammed up and having the sway-braces on there, they would form a baffle-board and the water striking them would be deflected and the current changed, and I think the change would unquestionably persist for a distance of two thousand feet, unless it got started the other way.” In order to establish his theory of a definite direction of the flow or current, or definite “axis,” as his witness Woodard.called it, appellant introduced evidence respecting the flow and habits of the river in that part of its course which extends from the Southern Pacific bridge to and through the county highway bridge, on southward through the bridge of defendant, and then southward beyond the northerly line of appellant’s land —a stretch of more than three thousand two hundred feet along the course of the stream. To counteract this theory of a definite “axis,” respondent introduced the evidence of which appellant now complains. For example, one of respondent’s witnesses, James Cleminson, testified as follows: “I think there was more water in the Lexington Wash than I ever saw before, more than in 1884. It was a waving, rolling flood. ... I saw the part of my land that was washed away in the daytime, after it went out. The flood was not coming in a straight line, it switched from side to side. It would switch from bank to bank. As fast as it cuts off the lower part of the bank it drives to the other part of the bank and keeps working from side to side.” Another witness for defendant, John B. Snoddy, testified: “I know the Lexington Wash. I have lived on it during my entire life. . . . From 1884 it continued to grow every flood year in some point. *722 There was some points it has never touched any of the old soil since 1884, but it would switch across the channel and take out ten or fifteen acres and the sand. The current would carry the sand, and the sand created a levee, and in high water it would throw over and in low water it would melt away. . . . The river in the February flood was rolling pretty high. . . . The river struck the bank. It had struck the bank south and pretty near reached us through some of Mr. Pierson’s property, and flowed directly west. Mr. Pier-son’s property is right down below the Santa Anita Wash, right across the river from the wash. The water struck the bank of Mr. Pierson’s property and then came across and struck our bank. Then after it struck our property it took out the Cleminson’s property. He lost about ten or twelve acres. From the Cleminson side the river switched over to the Michener side, and took out part of the Baldwin lease. ... In each flood the Lexington Wash makes a separate channel. ... In running of the same flood, if it lasted more than a day, the river would not keep to a true course. The floods do not come down in well-defined channels; they wind about a good deal. The course of the channel in that part of it largely depends on the hummocks or rises in the river-bed. I do not know what makes the course of the river switch; it is beyond my knowledge. It will switch where there is comparatively nothing to be seen. The river does not act like it does in a confined channel. I don’t know as there are any rocks in the channel to amount to anything. There is no certainty where the current of the water will flow in high water in any portion of the river above us. To-day it may be in that portion of the river up there, where it is half or three-quarters of a mile wide, and to-morrow it may move westerly and be against the westerly bank, and in the afternoon against the east bank; then it may switch back to the west bank at the same point. I have always been told that that was caused by the quicksand, of which there is a great deal at that point. That has been the condition of the stream for years. Because of the switching from day to day and from hour to hour almost, it has cut out in that portion an immense territory of land. ” J. F. Killian, a witness for defendant, testified: “I don’t think that the axis of the stream between the Southern Pacific bridge and the P. E. bridge has been constant during the times I have observed it. . . . *723 While the banks of that stream may be definite this year, you don’t know where they will be next year, and this particular bank you speak of between those two bridges has washed out on both sides more or less within that space of time, from 1884 down to the present time. ’ ’ The witness Bradway testified: “I saw the flood of February, 1914. It was very much larger than any of the other floods I have ever seen here. . . . In these floods the river did not have any fixed or definite axis of flow. It changed from side to side, and made some changes during the same flood; that is, it would change in a few hours or over night at times, striking first on one side and then the other.” It is true these witnesses, owners of lands above and below appellant’s property, were addressing themselves, in the main, to conditions immediately opposite to or in the vicinity of their respective holdings but their testimony, in the aggregate, tends to show the habits of the flood waters in the wash during severe storms and heavy rainfalls, and to establish respondent’s claim that, during these great floods, the turbulent waters, that churn and leap, writhe, and tumble, rush down the wash without any definite axis of flow, and, commencing in the upper reaches of the river, the current lashes and switches, from side to side, striking first one bank of the wash, eating out and devouring there its tribute of soil, and then, rebounding to the other side, works there the same ravaging destruction, only to switch back again to the other side, and so on throughout its whole devastating course.

That the testimony of respondent’s witnesses respecting this switching and rebounding tendency of the river at places above and below appellant’s land was admissible as indirect evidence of a similar tendency in that particular stretch of the wash where the seething waters flow between the county bridge and respondent’s structure—the stretch concerning which appellant’s witness Woodard testified—is shown by the testimony of defendant’s witness J. A. Bell, a civil engineer, who testified as follows: “As to what are the causes of the axis changing, there are a number of causes. ... A stream has a tendency, these torrential streams, once it has deflected at one point to reverse its action. If it strikes a bank and is deflected it naturally flows toward the other bank, and this action is aggravated when each succeeding flood tends to increase instead of diminish. That is what is called *724 switching.” With these excerpts from the testimony of the witnesses, we leave this branch of the case with the declaration that, in our opinion, the evidence of which appellant complains was admissible as tending, in the most practicable, if not the only available manner, to rebut appellant’s theory of a definite axis of the flood waters in the vicinity of respondent’s bridge.

[7] In one of its instructions the court charged the jury as follows: “There is no charge in this case-that defendant was negligent in respect to the condition of the channel of the wash during the years succeeding the building of this bridge, nor that the bridge as built originally was not of proper length, type, construction, or location. You must, therefore, assume conclusively, in considering this case, that the bridge was not as originally built of improper length, type, construction, or location. The charge is made that the said bridge several years after its original construction was negligently reconstructed, retied, rebraced, and reinforced. In determining whether the defendant was negligent in this matter, you should be guided ... in the light of the circumstances and surroundings ... at the time of such reconstruction. ” Appellant complains of this instruction and assigns the giving of it as error. He contends that, as originally built, the bridge was negligently constructed, for the reason that the bents in the bridge, as originally constructed and as maintained up to the time of the injury to his land, were placed at right angles to the length of the bridge, and hence at an angle of forty-five degrees to the “axis” of the stream; whereas these bents should, he says, have been constructed along the axis of the stream, instead of at so great an angle therewith. Upon these premises counsel argue that “when defendant came to reconstruct, reinforce, retie, and rebrace said bridge, it was incumbent upon it to take into consideration the fact that the bridge was not properly constructed in the first instance.” Of course, when defendant came to reinforce the bridge by placing sway-braces on the bent, it was “incumbent upon it to take into consideration” all details of the then existing bridge—its character and also the relative position of the bents and of the whole structure with the axis of the stream at times of flood, if the stream had any well-defined axis at such times, and also the known history and habits of the waters in the wash during flood *725 conditions. But there is nothing in the instruction that told the jury that- defendant, when reinforcing the bridge, was not required to take all these conditions into consideration. Appellant’s complaint did not charge defendant with negligence in the original construction of the bridge, but only with negligence in reinforcing and rebracing it. The instruction was proper, and we fail to see how appellant could have been prejudiced by it.

Judgment affirmed.

Sloane, J., and Thomas, J., concurred.

A petition to have the cause heard in the supreme court, after judgment in the district court of appeal, was denied by the supreme court on October 13, 1919.

All the Justices concurred.