The judgment must be reversed for several reasons.
The_ trial court erred in changing the finding of the jury on the subject of damages, — nearly doubling the same. That is so clear as to be unexplainable, seemingly, except upon the theory that, it was thought the evidence of one interested witness on the subject of damages, not contradicted from the mouth of any other, is, necessarily, to be taken as true. Not so by any means. True, the nneontradicted, positive, consistent evidence of one witness, who personally knows whereof he testifies, even though a party, is controlling, if not affected by any inherent weakness materially impairing its credibility, and not purely opinion evidence in the field of common knowledge, or the special field in which the judge, presumably, has the special knowledge of an expert. That leaves a broad field where the court and jury are not bound by the evidence of a party, even though entirely uncontradicted by any other witness.
The mere words in which the evidence of a witness may be given, is one thing, the probative effect is another. In the .former it may be very positive, direct, and consistent, yet in the latter be of very little weight, or of no weight whatever. It may be so very unreasonable, or contrary to matters of common knowledge, as to be self-destructive, or so based on matter of hearsay as to greatly impair its credibility, if not leave the matter involved not established to the reasonable certainty required to warrant a finding in favor of the party on whom the burden of proof rests. These suggestions are so elementary that the mere statement thereof is sufficient for this case.
*496The application of what has been said will be apparent from the following: The only evidence on the subject of damages was given by the one mainly interested in securing a judgment, and as large a one as practicable. He, doubtless, testified to the truth, so far as he knew it. The infirmity was, in that he had very little, and no precise, personal knowledge of the extent of the damage. Many years elapsed between the occurrence complained of and the trial. The witness depended, in the main, upon an eighteen-year-old record made by himself from written data furnished him by many persons who were not witnesses and whose work was entirely unverified. The original writings, even, were not produced. The witness had, practically, no personal knowledge of the subject of his testimony, except that there was a large amount of tin stock in the basement; that the basement was flooded to some extent, and that the necessary effect of such an occurrence on such property was to damage it. What the precise effect' was in the given case, the amount of property in the basement, the extent to which it was flooded with water, and all the particulars of the damages, were outside his personal knowledge. All those matters were given to him at his desk by verbal explanations and written data from many persons. From that he made up the record presented to the court. He knew his statement to be correct according to the second-hand information which came to his desk, gathered by those who were working in his interest, and then compiled with his estimates and preserved on file, not in the regular course of business, but for the very purpose of making up a claim against defendant. Whether the information which came to him was true, he, manifestly, did not know.
The data furnished the witness as aforesaid, were not, as suggested, in the nature of entries, or reports for entry, in the regular course of business. The happenings were not before the particular use made of them was contemplated so as to make the record, or evidence depending thereon, admissible in evidence under an exception to the rule as to hearsay evi-*497deuce illustrated in F. Dohmen Co. v. Niagara F. Ins. Co. 96 Wis. 38, 71 N. W. 69. On tbe contrary, all tbe writings, as before indicated, were made in tbe process of laying a basis for efficiently making a claim against tbe city. Instead of tbe situation being sucb as to create a strong probability of tbe work being correct, it was one well calculated to produce unsatisfactory results. None of tbe reasons characterized tbe matter rendering entries of transactions in tbe regular course of business in which many persons are concerned, evidence for what they are worth, but tbe one circumstance that there were many actors in tbe matter rendering it difficult, if not impossible, to produce them in court to verify their work several years after tbe transactions. There was not even tbe merit of an attempt to produce those who gathered the information tbe witness used in making bis statement. Moreover,' no effort was made to have tbe work done so as to support tbe correctness of it by verification of those who did it. Even tbe written data used by tbe witness were destroyed. Men, in number and of a character, were used in gathering tbe information, in utter disregard of tbe necessities of tbe future as regards judicially establishing a claim. Tbe dilemma which left tbe case to rest, largely, on tbe unverified stater ments, was created by respondent at tbe very time when great care should have been exercised to gather and preserve evidence, competent in a court of justice to establish tbe facts.
When it came to giving a money measure of tbe diminished value of tbe property, tbe witness gave opinion evidence only, and not even that, as we have seen, on personal knowledge of tbe damage, or facts in that regard verified upon tbe trial, or at all. Moreover, not even opinion evidence was given of tbe value constituting tbe legal basis for a. recovery. He failed to testify at all to tbe diminution of the market value of tbe stock caused by tbe flood. His evidence related to what tbe property cost and what it was worth to respondent after tbe flood.
There is this further infirmity in the evidence as regards *498furnishing any certain basis for an assessment of damages: There was a second flood and a second infliction of damages, as was claimed, and statement of loss made up including both occurrences. The jury found, there was no right of recovery shown for the second occurrence, and that was not disturbed. An attempt was made to separate the claim so it would be definitely seen what amount was chargeable to the first flood and what to the second. In the end, the precise amount charged by the witness to each flood was left quite uncertain. If removed from the field of conjecture, it was not from the field calling for jury interference, if, indeed, it had sufficient probative power to locate it there at all.
On the whole, we are not prepared to say, there was sufficient competent evidence of the damages to warrant the jury in finding in respect thereto to a reasonable certainty, or even that the evidence depended on to prove the claim was competent at all. It would not be well to venture upon another trial without being prepared to furnish some substantial verification of the data upon which the witness relied in making up the claim for damages, not only as regards the amount and kinds of property damaged, but the precise condition thereof after the flood. In attempting to make a finding as to damages, under the circumstances, that the jury placed the amount much less than the total claimed, is most natural. It was competent for them to do so, — conceding for the point that the evidence of damages was competent to establish some amount with reasonable certainty, — upon the theory that no more than the $2,500 found was established, satisfactorily. At the best for respondent, conceding as indicated, the court grievously erred in changing that finding by greatly increasing the amount.
Another serious error was committed by submitting questions to the jury, especially in connection with the charge, as to whether the rainfall was so excessive as not to be reasonably anticipated by men using ordinary care, — the thought being *499that answers thereto would settle the question as to municipal liability. That idea pervades tbe whole verdict. The error, in prejudicial character, was intensified by refusal of the court to submit to the jury, whether the rainfall was extraordinarily heavy. Evidently the learned trial court- entertained the idea that, in maintaining a system for carrying off surface water, or such and sewage, the city should provide for any accumulation reasonably to be expected. That idea was, unmistakably, embodied in the verdict as explained by the instructions. That it was erroneous seems plain from the following:
The leading case in this state on the subject under discussion is Allen v. Chippewa Falls, 52 Wis. 430, 9 N. W. 284. There rainfalls are plainly differentiated or divided into ordinary, extraordinary, and unprecedented. The latter class, is not expressly spoken of, but is inferentially. The plain logic of the case is that there are ordinary rain storms,— such as frequently and ordinarily occur, — extraordinary rain, storms, — such as occur sometimes, but are not unprecedented, hence may be reasonably anticipated will occur once in a while, — and the unprecedented, — such as have never, so far as known, definitely, occurred before, and so are not reasonably to be expected even at long intervals, though not impossible. '
It is only the first class mentioned that the municipality, in. maintaining its sewer system, must use ordinary care to provide for. What the court meant in Allen v. Chippewa Falls, supra, by “extraordinary” is plain from the fact that the-word was not used as synonymous with “unprecedented” but with unusual, — that which is rare or uncommon; happens, sometimes but not so often as to be regarded a common occurrence.
The conclusion stated will be appreciated by a study of the-opinion of Denio, C. J., in Mills v. Brooklyn, 32 N. Y. 489,, *500495, which was followed in the Allen Gase, this language being quoted:
“If the city provides drains and gutters of sufficient size to carry off in safety the ordinary rainfall, or the ordinary flow of surface water, occasioned by the storms which are liable to occur in this climate and country, it is all the law should require.”
Note the words “storms which are liable to occur.” The court in the Allen Gase correctly interpreted the thought of the New York court as not to be of all storms which are liable to occur, but storms which usually occur. That is manifest by the phrasing of the matter by this court thus:
“If the city did provide gutters which were sufficient to conduct away in safety the rainfall of ordinary storms,— such freshets as usually occur in this climate, — it would seem to be all the law should require.”
What has been said is confirmed by Hophins v. Rush River, 70 Wis. 10, 34 N. W. 909, 35 N. W. 939, where the court declared that in the Allen Gase it was held that the limit of municipal duty was to provide for such storms as are “usually liable to occur.” Note the significance of the words “usually liable.” That plainly excluded storms which are liable to occur and so are within reasonable anticipation that they may or will occur, but only at long intervals.
In the Hophins Gase, as counsel for appellant contend, the judgment was reversed because the trial court, the same as here, submitted the case as if all storms within reasonable anticipation are to be provided for. The question submitted was, “Whether the freshet in question was one reasonably to be anticipated ?” This court commenting on that said:
“In this latitude and climate, unusual and extraordinary rainfalls and freshets may reasonably be anticipated by any' one at long intervals and uncertain times. Hence the jury may have found that the freshet in question was an unusual or extraordinary one, and yet that it might reasonably be anticipated by the town authorities to occur some time.”
*501Because it is for the latter duty only a municipality in maintaining a sewer system is to provide. The judgment was reversed.
The idea expressed in the Sopldns Case was in mind in Schroeder v. Baraboo, 93 Wis. 95, 100, 67 N. W. 27. There Allen v. Chippewa Falls was referred to as correctly holding that “negligence on the part of the municipality cannot he predicated on mere failure to construct gutters or sewers of a sufficient capacity to carry off surface water in case of an extraordinary storm, such that a person of ordinary prudence would not ordinarily anticipate and provide against.” The word “anticipate” was added to what was said in the Allen Case. But it is evident there was no thought of adding a new element. “Anticipate” was used in connection with “ordinary” with the idea that what ordinarily, — commonly,—occurs, must necessarily he within ordinary anticipation. Obviously there is a wide difference between the things ordinarily anticipated because of their common occurrence, and things within reasonable anticipation, which extends to everything which may reasonably happen, — things which are within the realms of possibility, but extraordinary in that they seldom occur.
There was much evidence tending to prove that the basement was flooded before the sewer broke and that, had the break not occurred, the damage would have happened just the same. The evidence strongly tends to show that the breaking of the sewer, at the worst, merely added to a flood already existing; the water from the sewer uniting with an existing flow in the street in which that from the sewer lost its identity. So it seems quite clear that the court should have submitted questions on both subjects, as requested. If the damages, in the whole, would have occurred regardless of the breaking of the sewer, then such breaking was not the proximate cause of the injury to respondent’s property. If the water from the sewer united with, and lost its identity in, *502an independent flow of water which, as an entirety subsequently reached respondent’s basement, then the breaking of the sewer was not the proximate cause of the respondent’s damage except in so far as the damage would not have occurred had the water from the same not been added to that in the street. The rules governing these phases of the case are fully discussed in Cook v. M., St. P. & S. S. M. R. Co. 98 Wis. 624, 14 N. W. 561. All such should have been submitted to the jury, if the case had to be submitted at all.
The point is made that the evidence conclusively shows that the rainfall and flood were extraordinary and, therefore, the court should have taken the case from the jury in appellant’s favor. That matter is by no means free from difficulty.
The evidence very strongly tends to prove that the rainfall was out of the ordinary. It is not unlikely that the trial court would have so held, as matter of law, had the rule been understood to be as we have indicated. The court might well have thought the rainfall was extraordinary, — unusual and yet not beyond reasonable anticipation, — something between an unusual occurrence and the unprecedented. Had it been supposed that such medium line was fatal to respondent, the result, perhaps, would have been different. There was undisputed evidence .that before the sewer broke basements in the vicinity of respondent’s were flooded, water was over sidewalks in some places to a considerable depth, and in one instance the sidewalk actually floated off. The evidence on the part of appellant made out a very strong case of an extraordinary rainfall and freshet, — such as occurs, if at all, only at long and very uncertain intervals, — indeed one, the like of which, no evidence was produced. Opposed to that there was only the weather bureau record produced by the person in charge, and his evidence in connection therewith respecting rain precipitation. At one point he was inclined to call the rainfall in question excessive, but not extraordinary, and at another excessive though not to such extent but that it was *503within reasonable expectation. He said a precipitation of one inch per hour would be excessive. The person in charge of the forecasting department of the local station of the weather bureau, showed by his record that the precipitation on the day in question for the hour about the time the sewer broke was considerably more than one inch, had not been exceeded but some three times in about seventeen years, and had not really been exceeded at all under the same circumstances. He called it extraordinary. Neither of these witnesses pretended to know anything about the particular freshet, except as indicated by their records.'
On the state of the case indicated the majority of the justices are constrained to hold that there may be, in some permissible view of the evidence, room for holding that the rainfall was not extraordinary. As we have indicated, neither trial judge nor jury passed on the matter from a proper point of view as to the law. Therefore, it is thought best to leave it so, upon another trial, such court can exercise original judgment in respect thereto.
• In view of the foregoing, a few general observations respecting other matters which seem within the exceptions, will suffice to guard against a recurrence of the errors found in the record, and others that might be cognate thereto.
The issue of whether the sewer was sufficient in size for the service put upon it was submitted, found in the negative, and that such- defect was a proximate cause of the injury. In this, the court seems to have overlooked the settled doctrine that, if a municipality, acting reasonably, adopts a plan of sewage, including the size, form, and character generally of the conduits, and executes the same, it is immune for consequences inherent in the plan itself, including the proposed method of construction. Schroeder v. Bamboo, 93 Wis. 95, 67 N. W. 27; Hart v. Neillsville, 125 Wis. 546, 104 N. W. 699. Mistakes of the character of those mentioned are referable to mere errors of judgment made by a body clothed by *504law with discretionary authority, and have the same status as regards liability for negligence as mistakes of any person or body exercising quasi-judicial authority.
True, if upon executing a plan of sewage and the system being put in operation it proves defective, endangering private rights, the duty arises to exercise ordinary care to remedy the matter, and failure to perform such duty may he actionable. Hart v. Neillsville, supra; Hart v. Neillsville, 141 Wis. 3, 123 N. W. 125. Whether the municipality had such knowledge or not in this case, was submitted to the jury and found in respondent’s favor; but, inasmuch as the record is barren of evidence that the sewer at any time before the instance in question had proven insufficient, it seems likely that the jury might have so answered upon the theory that the plan itself was insufficient, and the municipal officers should have known it. Of course the rule which renders the municipality immune for mistakes in the plan will render it immune from any charge of constructive knowledge of insufficiency from mere implied knowledge of the plan. Knowledge, actual or constructive, in such a case, must come from performance of the system or something other than knowledge of the plan itself. The municipality does not need to search for defects till something occurs independently of the plan to suggest the existence thereof. The submission of the case in this respect was not free from criticism. The appellant having in the exercise of its discretionary authority adopted the plan, it had a right to presume that such plan was sufficient till, by performance, the contrary was brought home to it, actually or constructively. It did not have to suspect the existence of defects and search for them.
This must be remembered: (1) A municipality is not responsible for mistakes in a duly adopted plan of sewage. (2) A municipality is responsible for a defective original construction of a system of sewage, as said in the authorities already cited, but that means negligent execution of the plan, *505not defective original construction inhering in tbe plan itself. Proper construction according to tbe adopted plan can never be a defective original construction, witbin tbe rule mentioned. (3) If a duly adopted and executed plan of sewage does not prove defective in operation while in a proper state of repair, but becomes out of repair to tbe knowledge, actual or constructive, of tbe municipality, tbe duty devolves upon it to remedy tbe matter, and it is liable for failure to exercise ordinary care in respect thereto.
One thing more. We observe that tbe fprm of tbe questions on tbe subject of proximate cause, was such that, as to three distinct matters, tbe jury found that each was a proximate cause of the injwry. There was no finding that any one, or all, was tbe proximate cause. Since one of tbe questions might well have been interpreted so tbe answer would not render appellant liable, and there were matters, particularly tbe great amount of water which gathered in tbe street independently of tbe sewer, indicating that there were other causes and may be some one of them was tbe proximate cause of tbe injury, it is difficult to see bow tbe verdict could, in any event, support a judgment in respondent’s favor.
There may be more than one proximate cause of a personal injury, all concurring to produce it. But neither can, by itself, be said to be tbe efficient or proximate cause, unless, bad tbe others not operated, it would have produced tbe injury. .So in case of there being two causes, one which is of an actionable nature, and tbe other not, it is not sufficient to find that tbe former was a proximate cause, because it may or may not have been an efficient cause; for that, it must have been an element added to tbe other producing tbe result, which, without it, would not have occurred. Cook v. M., St. P. & S. S. M. R. Co. 98 Wis. 624, 638, 74 N. W. 561.
By the Court. — Tbe judgment is reversed as to tbe first flood complained of, and action remanded for a new trial as to such first flood.