The defendant has been convicted of the crime of arson in the third degree. She is a woman, forty-three years of age, and at the time of the alleged crime was living with her husband on a farm a short distance from the scene of the fire, which occurred on the farm of James Kintz, at Meyers, N. Y. It is assumed that the fire was of incendiary origin. The prosecution, after showing the facts indicating an incendiary origin of the fire, produced evidence tending to show that tracks were found in a plowed field near the scene of the fire; that a pair of rubber overshoes found at the defendant’s home fitted into these tracks, and that the defendant had been in the habit of wearing these overshoes; that the tracks indicated shorter strides going to the point where the fire originated than those going in the opposite direction, and that there were other indications that the person was running on leaving the fire; that these tracks proceeded across the plowed land, into a clump of sweet clover, and were finally lost in a beaten path; that sometime soon after the discovery of the fire the defendant was seen by the roadside, between her home and the place of the fire; that a beer bottle which had contained kerosene oil was found near the place where the fire originated, and that similar bottles were found in the defendant’s house, though upon this point the evidence does not disclose with any degree of detail the points of alleged similarity; that pieces of newspapers were found at greater or less distances from the point where the fire originated, which corresponded with newspapers shown to have been taken by the defendant.
*484It is not to be doubted that the prosecuting attorney had made a case which entitled him to go to the jury, though it is by no means certain that the defendant was so intimately connected with the circumstances that the jury would be' bound to find her guilty.' It may be read between the lines that the defendant had been associated in the public mind with various other crimes of a like nature, and there was direct evidence that she .had made threatening suggestions against the family of Kintz, but there was a decidedly weak link in the chain of evidence between the time that the tracks merged into the beaten path and the discovery of the defendant in the highway some distance away, and the mere fact that a neighboring woman was discovered, between her home and the scene of the fire, a considerable time after the fire was under- way, was not persuasive evidence that she had set the fire; any number of innocent women may have been attracted by the blaze and been present to view the spectacle.
It is unnecessary to consider what would have been the attitude of this court if the evidence had closed at this point, for we do not know what the jury would have done. There is no means of knowing what a jury, confronted with circumstantial evidence, would conclude, and the judgment of this court may not be invoked upon a' mere hypothesis. The difficulty with the case, as it is presented here, is that the prosecuting attorney introduced in evidence the conduct of a so-called German police dog, which was put upon the trail of the tracks across the plowed ground some five days after the fire, and the testimony goes to show that this dog followed the trail to the point where the visible tracks were lost -in the beaten path, thence along the pathway and the highway to the point where the defendant was shown to have stood on the occasion of the fire, and with this fink in the evidence thus strengthened the case was submitted to the jury, resulting in the conviction of the defendant, The question thus presented is whether the evidence of this dog’s conduct was, under the circumstances of this case, properly admitted.
It may be assumed that under proper conditions such evidence is admissible, without deciding the question. The weight of authority appears to be in that direction, though *485it is conceded that no case has been decided by the appellate courts of this or of any controlling jurisdiction upon the exact point. Whatever we might determine upon this point, which has in it the suggestion of.romance and dramatic art, we are of the opinion that under the rules laid down by the courts in those jurisdictions in which such evidence is recognized as having probative force, it was error to permit the evidence to go to the jury upon the highly vital point to which it was necessarily addressed.
In State v. Dickerson (77 Ohio St. 34) the court say that there may be deduced a rule which until shown untrustworthy may be followed in cases where this class of evidence is offered. It is apparent that before the act and conduct of the dog can be shown a proof of preliminary foundation must be laid, and, to establish such foundation, it must be shown that the particular dog used was trained and tested in tracking human beings and by experience had been found rehable in such cases; and the dog so trained was laid on the trail, whether visible or invisible, at a point where the circumstances tended clearly to show that the guilty party had been. In addition to this, the reliability of the dog must be proved by persons having a personal knowledge thereof. This foundation may be strengthened by proof of pedigree, purity of blood, or the exalted standing of his breed and performance of such peculiar work. In the case now under consideration, the owner of the dog was called, and testified that the dog was what was known as a German police dog; that he had owned him for about three years, and since he was three months old; that he had a registered pedigree, which was given, and then the witness was asked to detail what he had seen the dog do. He testified that at the Litchfield horse show he had taken a handkerchief of his own from which the dog could get the scent, and had then made a detour of a quarter of a mile, making several turns to the right and left and going through some pieces of woods and hid the handkerchief; that the dog was held by the judges in position where he could not see the direction taken by the witness; that on his return the dog was let loose and told to go and get the article; that the dog immediately started over the course, following the trail closely, and returned with the handkerchief. The witness said he had seen substan*486tially this same feat performed many times; that the dog immediately after being let loose followed the trail and never failed; and then added: “ Of course that is a trail taken up at once. I have never tried him on an older trail, but I am sure it oan be done.”
Objection was made to the last portion quoted above and a motion was made to strike it out as a conclusion, but the court denied the motion, saying: “ It may stand for what it is worth.” The defendant excepted. The witness then testified that he had seen the dog work out these tests four or five dozen times in the past three years, and that he never failed. But this does not attempt to show that this particular dog had ever had any training in tracking human beings generally; it simply shows that he had been able to follow the fresh trail of his own master and to find a handkerchief which had been hidden a quarter of a mile away, a clever piece of work, perhaps, but hardly up to the standard required by the rule quoted. Many dogs are able to trace out their own masters, but to place him upon the trail of a stranger, and to have him follow up this trail, is quite another matter; and the witness admitted, on cross-examination, that there was a difference in following a trail by odor and following an article the dog had been allowed to smell of and then hidden away some place. The only other person whom it was claimed this particular dog had ever trailed was the wife of the witness, and his testimony as to this was that at one time he sent the dog to look for her; that she was nearly a mile away; that the dog did not know where she was; that he told the dog to go and find her, and that the dog turned up where the wife was, and that she brought him back with her. How the wife reached the point where she was found by the dog — whether by automobile or otherwise — does not appear, and the most that can be said of this is that the dog went to his mistress when sent. How he reached her, whether by trailing her, or by merely going to a familiar point, is not disclosed, and the incident does not tend to establish any reliability on the part of this dog in trailing strangers. The dog was evidently a clever animal, but the witness does not claim that it was trained for any practical work; he was an exhibition dog at the shows, and the mere fact that he did *487some surprising things in following the members of the witness’ own family is not a proper foundation for the admission of. the evidence complained of by the defendant. There is absolutely no evidence of the reliability of this dog in tracing strangers; no evidence that he had ever taken an old trail and followed it, and the mere conclusion of the witness that the dog could do this is worthless in establishing a foundation for the introduction of evidence as to the conduct of this dog when placed upon a trail five days o d. It appears from the evidence that there had been two rain storms between the time of the fire and the date of the starting of the dog upon the trail; it is clear from the testimony in the case that many persons had been along the trail from the fire to the beaten path, where the tracks were lost, and probably along the highway where it is claimed the defendant was seen on the occasion of the fire, and while we will ¿not suggest that it is impossible that this dog could have followed this one individual trail it is certainly highly speculative under the circumstances here disclosed. It is quite as possible that he may have been following the trail left by the investigators, who had undoubtedly been over substantially the course which it is claimed was followed by the defendant in reaching the point where she was seen. The rule requires that the dog must have been trained and tested in tracking human beings; not that he was clever in finding things which his master hid away from him a few minutes before he was released, or that he had upon one occasion been out and found his mistress, under unknown conditions. Mere private exhibition tests, not designed to prepare the dog for the work of searching out strangers, does not meet up with the requirements of the rule, and without the presence of this evidence in the case, it is by no means certain that a jury would have found the defendant guilty of this crime. There was not time enough after the starting of the fire and the time when she was seen on the highway for her to have gone back to her home and changed her shoes, and no one pretends to say that she had the rubber overshoes on when she was seen, and the evidence is far from convincing that this woman actually set fire to the buildings. The known facts are suspicious, but proof beyond a reasonable doubt is necessary, and if the acts of this *488dog had not fitted into the theory and the known facts, it is difficult to understand how the jury could have been fully convinced by the other evidence. Under such circumstances, if this questionable type of evidence is to be received at all it should certainly comply with all the requirements of the rule, and we are of the opinion that there was an entire failure in this particular. The People did not show the necessary facts to justify the admission of the testimony, and it was error to permit the jury to convict upon the basis of this evidence.
The judgment of conviction should be reversed and a new trial ordered.
All concurred.
Judgment of conviction reversed and new trial ordered."