—This is an action brought by the county of Sonoma against Stofen, its ex-treasurer, and against his bondsmen, to recover the sum of seven thousand eight hundred and fifteen dollars and seventy-nine cents, alleged to have been converted by Stofen while he was treasurer. The defendants denied the conversion, and set up as a defense that Stofen had been robbed of the money. Judgment passed for plaintiff and defendants moved for a new trial. Upon denial of their motion they appealed, both from that order and from the judgment.

It is strenuously insisted that the uncontradicted evidence in *34this case establishes a robbery. The account of the affair given by the defendant Stofen is as follows: Upon the morning of the twenty-eighth day of December, a Friday morning, he had gone to his office in the county courthouse in the city of Santa Rosa. He entered his rooms at about the usual hour, at 9 o’clock, or a few minutes thereafter. He opened the outer and inner doors of the vault in his office and proceeded to take out the drawers or money trays for use in his daily business. Turning round with a tray in each hand, he confronted a man with an uplifted dagger who ordered him to drop the trays. “This is the last that I knew. 1 felt no blow, and did not see him strike me, but I presume as I put down the money he struck me on the back of the head, because that is where 1 had a lump and discoloration of the skin. That is the last I know until I came to in the vault, and I was lying on my back with my head right under the opening of the vault door. The first thing that I realized after I came to, I raised up and bumped my forehead again on the metal. That threw my head back again on the floor, and I felt up around, and I felt the safe floor. Then I comprehended for the first time where I was. I remained in the vault until I was taken out by my wife. They said it was about half past 4 in the afternoon of the same day.” He further testified that the money, which was in its proper place at the time he entered the vault in the morning, was taken. From time to time during his imprisonment ho kicked upon the door with all his might, “and that made a great racket.” Mrs. Stofen, wife of the defendant, testified that in the afternoon of that day, visiting the courthouse in search of her husband, she became alarmed at his absence. She procured the janitor to unlock the outer door of her husband’s office, and entered the room with one or two others. She saw her husband’s overcoat and hat on the desk and became greatly excited. She thought she heard a noise in the other room, which was the room containing the vault, and Judge Moore said, “He is in the vault.” She hurried to the vault door and tried the combination, which she knew, but, being excited, she did not succeed the first time in opening the door. Judge Moore stood by her, endeavoring to calm her, and in time she succeeded in opening the outer door. Then she had to unlock the inner sheet-iron door. The key *35of this door, upon the hunch of defendant’s keys, was in the lock. Opening the door, her husband was found dazed and weak. . There was an injury, a bruise upon the hack of his head, a -swelling and discoloration. J. H. Wise and A. P. Moore accompanied Mrs. Stofen when she unlocked the vault. Their testimony corroborated hers, saving that Wise says he could not say whether Mrs. Stofen actually unlocked the inner door or not, or whether the inner door was locked or not. Moore could not testify whether the inner door was locked or not. He states that as they were going from the front room to the inner room Mrs. Stofen, who was very much excited, exclaimed: “Oh, they have locked my husband up!” These two witnesses agreed in their account of the distressed condition of Stofen when released from the vault, but neither observed any injuries to his person.

This account of the robbery stands uncontradicted and unimpeached by any direct evidence in the case. So much must he conceded, and under this concession appellants contend that the decision is unsupported by the evidence. But, while it is true that uncontradicted evidence of a positive fact must generally be controlling, there are exceptional cases in which the jury or the judge as a trier of the facts is held justified in rejecting uncontradicted evidence, even the most positive. In Quock Ting v. United States, 140 H.S. 417, Mr. Justice Field uses this language: ‘‘Undoubtedly, as a general rule, positive testimony as to a particular fact, uncontradicted by anyone, should control the decision of the court; but that rule admits of many exceptions. There may he such an inherent improbability in the statements of a witness as to induce the court or jury to disregard his evidence, even in the absence of any direct conflicting testimony. He may he contradicted by the facts he states as completely as by direct adverse testimony, and there may he so many omissions in his own account of particular transactions, or of his own -conduct, as to discredit his whole story. His manner, too, of testifying may give rise to doubts of his sincerity, and create the impression that he is giving a wrong coloring to material facts. All these things may properly he considered in determining the weight which should be given to his statement, although there he no adverse verbal testimony adduced.” To like effect are the cases of Blankman v. Vallejo, 15 Cal. 639, and People v. Milner, 122 Cal. 171.

*36Are any of these elements of doubt or suspicion present in this case, and, if so, are they sufficient to have warranted, the court in finding against the defense of robbery? In this, as in all such cases where the question has arisen, the offered evidence is the evidence of a person directly interested in the outcome of the litigation. This fact itself is a circumstance of importance in weighing the testimony given. With the manner of a witness upon the stand, and with his deportment under examination, circumstances which have always been recognized as affecting one way or the other the credibility of his testimony, this court can have nothing to do; but when the matter of defendant’s testimony is considered and analyzed, it will certainly be shown to present features of suspicion and doubt. To begin with, it is extraordinary that in the city of Santa Rosa, upon the morning of a business day, one or more robbers could have thus assaulted the treasurer in a public building containing the courts and offices of the county officers, have robbed him, have locked him in the vault, and have escaped with their plunder without leaving the slightest clew to or trace of their identity or whereabouts. Again, the robbery took place but a few days before the expiration of the defendant’s Stofen’s -term of office, and, consequently, but a few days before it was necessary for him to turn over the county funds to his successor. This is but a circumstance, and nothing more, but it certainly gives room for suspicion that this extraordinary robbery should have taken place at so opportune or inopportune a time. Moreover, the vault in which the treasurer was locked was incased upon three sides with sheet iron, the fourth side being the iron door. These facts were well known to Stofen. A blow upon the sheet iron produces a loud, resonant sound, like that of a bass drum, and can be heard for considerable distances throughout the building. Yet Stofen, by his own testimony, although he could hear sounds which he thought were steps in the corridor outside, never struck the sheet iron walls of his prison. He did kick upon the iron door many times, he says, but there is upon this point at least the negative testimony of other occupants of the building that they never heard or noticed any such sound. Again, if it were established that Stofen was found locked in the vault, and that he could not so have locked himself therein, it would tend very *37strongly to corroborate his account of the robbery, for there is not the slightest pretense that there was anyone in Santa Rosa upon that morning who acted as his confederate; but upon this it is shown that the bolts of both doors to the vault may be shot in place by anyone who is inside, and, while the shooting of the bolts does not lock the doors in the sense that the combination must be employed upon the one and the key upon the other to open them, yet it creates the appearance of their being locked. Mrs. Stofen and those who were with her assumed that the outer door of the vault was locked. The first time she essayed the combination she failed, and this, of course, actually set the lock. As to the inner door in which the bunch of keys was found, she testified that she unlocked it. She herself was under great excitement, and may have been mistaken. Certainly, she was not a wholly disinterested witness, and neither of the two who accompanied her testified to the unlocking of this inner door.

It is the law of this state that robbery is a defense to an action for the recovery of public moneys. (Healdsburg v. Mulligan, 113 Cal. 205.) The evidence of one witness entitled to full credit is sufficient to prove the robbery. (Code Civ. Proc., sec. 1844.) But the evidence of a defendant so situated does not as a matter of legal compulsion command this full credit. Robbery is a defense which may be easily simulated, and the temptation to do so must needs be strong upon one who has misappropriated trust funds. In consideration of these facts it is not insisting upon too much to say that the defense must be satisfactorily established. If it be an honest defense, this in the great majority of instances can readily be done. If exceptional cases shall arise where the defense is rightful, yet conviction of its truth is not brought home by the evidence, it must be remembered that the imperfect machinery of the law cannot always measure out perfect justice, and that public policy forbids putting a premium upon peculation by permitting this facile defense to be too lightly established. We think it unnecessary to enlarge further upon the question. What has been said is not designed as expressing this court’s opinion of the evidence, but it is in illustration of the fact that the trial judge, who rejected the evidence of the robbery because of the circumstances of doubt and suspicion attending the account of it, was, under *38the exceptional features of this case, justified in so doing, and that his finding in this regard cannot here be disturbed.

Defendant Stofen testified that while in the vault he kicked many times upon the inner door with the heels of his boots, and that “it made a great racket.” He further testified that he did not kick or strike upon the sheet-iron lining of the sides of the vault. He knew its general construction and. that there was a hollow space several inches in diameter between the sheet iron and the brick outer wall. Evidence was offered and admitted touching experiments made by striking with clinched hands and books of account upon the sheet-iron sides, and also of experiments made by striking or kicking upon the inner door. Witnesses stationed in different parts of the building testified to hearing the sounds of these blows, and that those from the blows upon the sheet-iron sides were much louder than those from the blows upon the door. The admission of this evidence is urged as error, because the atmospheric and climatic conditions are not shown to have been the same upon the two days; the condition of the courthouse was shown to have been different by the removal of one door between the treasurer’s and sheriff’s offices and the replacing of the wooden panels in the upper portions of the other doors with glass; further, that it was not and could not be shown that in any blows struck by the experimenters they used the same amount of force as that employed by Stofen, and finally, as to the blows upon the sheet-iron walls, that the evidence was pertinent to nothing in the case, for Stofen positively testified that he did not strike upon these walls at all.

Experimental evidence in corroboration or disproof depends for its value upon the fact that the experiment has been made when the conditions affecting the result are as near as may be identical with those existing at the time of and operating to produce the particular effect. An absolute identity is, of course, impossible, but a substantial identity must exist to give the evidence value. But this identity need not extend to nor be shown to exist as to conditions which could have had no causal operation upon the result. Thus, if a witness testified that at a given distance he had heard a gunshot, before experimental evidence in disproof could be admissible it might properly be required to *39he shown that the force and direction of the wind were the same upon -both days. If the question were upon the stopping of a car within a specified distance by the use of a given brake, experimental evidence would he valueless, unless it were shown that the 'experiment was made under like conditions as to the car, the brake, the amount of force used to set it, and the state of the track. If a witness should testify that from a particular spot, he saw a certain act, evidence that be could not have seen from that spot because of an intervening building would be worthless unless it were also shown that the building was there at the time of the occurrence. So indefinitely might illustrations he multiplied, hut, upon the other hand, it would he entirely immaterial in the instance last given whether the wind was blowing three or thirty miles an hour in one direction or another, while these would constitute most pregnant circumstances in determining how far a gunshot could he heard. So, in experimenting with the car-brake, it might be of no consequence whether it was done by day or by night, but to test the vision of a witness obviously similar conditions of light would be most material. The experiments here made went to the volume of sound and the distance of its audibility. It is not apparent that different atmospheric conditions or the structural changes in the building would have materially affected the result, and as to the force used the experimenter testified that he employed moderate force, while Stofen testified that he kicked with all his might.

The witnesses stationed about the building agreed substantially in their accounts, the differences being such as would naturally be expected from their varying distances from the origin of the sound. The blows upon the door could he heard distinctly, but were not such as would have been likely to attract attention. The blows upon the walls were loud and resonant like the heating of a bass drum, or the pounding of an iron boiler. The evidence as to the blows upon the door was admissible as experimental evidence. It was of little value, assuredly, since in the one ease the witnesses stood with attentive ears to catch the expected sounds, while in the other case there was nothing to attract their attention. The admissibility of the evidence of the blows upon the wall rests upon a somewhat different ground. It was a matter of fair argument that if his *40story were true Stofen would have used every means to escape from confinement, and would have given an alarm; that with his knowledge of the vault and of its construction he would have pounded upon that part of it which naturally would have given out the greatest noise. It was permissible to support this argument by proof of the fact that, if he had done so, the alarm would certainly have been heard and his prompt release would have been assured. That he did not do so certainly gives basis to an argument against the probability of his story, for it is to be remembered that according to that story he was incarcerated for nearly eight hours, and that he never struck upon the sheet-iron walls, and yet within his prison he could occasionally hear the footsteps of those passing outside.

In support of their application for a new trial upon the ground of newly-discovered evidence, defendants presented the affidavit of one George E. Perry. Perry’s affidavit is to the effect that upon the day of the alleged robbery he was in charge of the steam-engine and healing apparatus in the Santa Rosa courthouse, during the temporary absence of his brother, who was the regular engineer in charge. In accordance with his practice he made a tour of inspection of the offices in the building to note their temperature and observe whether or not the heater was performing its work. Upon this particular morning, at a few minutes after 9 o’clock, as he approached the door of the treasurer’s office, the door opened, and a man.stepped out carrying a leather valise in his right hand. His -appearance surprised the affiant, who thought that the treasurer was out of town. He stopped in front of the man and asked if the treasurer was in his office, to which the man replied yes, and hurried on. He tried one of the doors of the office and found it locked. He passed on to the other door and found that also locked. He then concluded that the man had misunderstood his question, and had thought that he had inquired if that were the treasurer’s office, and so attached no significance to the matter, until the evening of that day, when he learned that the treasurer’s office had been robbed. He further deposed that in July or October, 1896, he related these facts to the defendant Stofen, but also stated to him that in attempting to open the side door he noticed upon it a placard stating that the treasurer was absent *41or out of town, and that at the time he so stated he believed that such was the fact, but “I may be mistaken, and I now believe that I am.” After he had told Stofen about the placard, Stofen seemed to take less interest in his narration. Aside from the rigorous scrutiny, amounting to suspicion, with which newly-discovered evidence such as this should be regarded, the court was justified in refusing to grant a new trial, because in fact it was not newly-discovered evidence. Affiant’s statement is, that he informed Stofen of these facts while the action was pending. It was evidence of the highest importance to the defense to show that a man with a valise in his hand was seen coming out of the treasurer’s office upon the morning of the day of the alleged robbery, and at the time of day which Stofen fixes as the hour of its commission, regardless of the question as to whether or not there was a placard upon one of the doors stating that the treasurer was out of town. Aside from the dubious circumstance that the affiant declares that his present belief is that he was mistaken as to the placard, even if he had adhered to his first story in this regard, the evidence would still have been of vital moment. It is conceivable that the robber himself might have placed such a placard upon the door to avert inquiry. But, whatever may be the truth as to all these matters, the fact is, that knowledge of this evidence was in the possession of the defendant while his case was pending, and he did not see fit to avail himself of it.

The judgment and order appealed from are, therefore, affirmed.

McFarland, J., and Temple, J., concurred.

Hearing in Bank denied.