The only witness who actually saw the burglar when the crime was committed was unable to identify the accused as the criminal, and testified only that the burglar was a negro man of the same size, color, and general description as the accused. Even conceding that the evidence showed that on the night of the crime and after its perpetration a track-dog was placed upon the trail of the perpetrator and followed the trail to the house occupied by the accused, the evidence disclosed that the dog was not permitted to enter the'house and did not follow the trail to the bed then occupied by the accused, and also that in the same house there was another man, who may have been the person actually trailed by the dog. Neither does it appear from the evidence that such circumstances as that the accused, when found in bed, was moist, as if from a long walk or run, and wore socks of a coarse weave, which would make an impression resembling the tracks made on the soil by the burglar, and that the foot of the accused was of about the same size as the foot of the criminal as indicated by his track, would not apply equally to the other man living in the house to which the trail pursued by the dog led. Though the facts may be otherwise, the other man occupying the house with the accused was, so far as this court is apprised, fully as able as he to commit the offense, may have been of the same color, size, and general description, may have wore the same kind of socks, may have had a foot of the same size, and may have made the trail followed by the dog from the scene of the crime. There was nothing else developed which even remotely pointed to the accused as the perpetrator of the crime or tended to arouse suspicion against him, except the statement, which the testimony discloses he made just after his arrest, to the effect that he had “not been prowling around any white folk’s house.” If this statement was made spontaneously, before he had *722been informed by the officers as to the nature of the crime they suspected he had perpetrated, and not in response to an inquiry whether he had been in or near the burglarized house, it would naturally excite suspicion against him, by betraying knowledge as to the particular crime for which he was wanted, which a wholly innocent man would probably not have possessed. But if the remark was made in response to an inquiry as to whether he had been in or near the burglarized house, it would have no special significance as tending to indicate criminal knowledge on his part concerning the commission of the crime. The record does not disclose how or for what reason he made this statement after his arrest, or whether it was voluntarily offered by way of denial before he had been charged with the commission of any particular offense, or in response to a direct inquiry, or after he had been apprised of the specific charge against him. It can not, therefore, be said that the evidence, which was purely circumstantial in its nature, was sufficient to exclude every other reasonable hypothesis than that of the guilt of the accused. It is somewhat doubtful that there was sufficiently clear and comprehensive proof that the dog used to follow the trail left by the actual criminal was of a breed or stock “characterized by acuteness of scent and power of discrimination, and had been trained or tested in the exercise of these qualities in the tracking of human beings” (Fite v. State, 16 Ga. App. 22 (4), 84 S. E. 485). The testimony as to the breed of the dog appears rather to have been a conclusion on the part of the witness than a statement of facts within his knowledge. We do not, however, pass u]3on this question, since the proof may be different on the next trial, and the rules touching such evidence have been clearly laid down by this court in Fite v. State, supra, as well as in Aiken v. State, 16 Ga. App. 848 (86 S. E. 1076).
The exceptions other than those relating to the verdict are either without substantial merit or relate to alleged errors which can hardly recur on another trial. In view of the fact that a jury of the vicinage has three times found the defendant guilty, this court is loath to set aside the conviction now before it for consideration; but more important even than the possible escape from punishment of one who may by chance be guilty is the maintenance of the orderly and systematic administration of the law, since the rule applied-in one case must be followed in all those of *723a similar character, and, whether guilty or innocent, every man accused of crime is entitled to an acquittal unless his guilt is made legally manifest. A wise provision of our law declares that, “to warrant a conviction on circumstantial evidence, the proved facts must not only be consistent with the hypothesis of guilt, but must exclude every other reasonable hypothesis save that of the guilt of the accused.” Penal Code, § 1010. It can not be said that in this case the proved facts excluded every reasonable hypothesis save that of the guilt of the defendant; and since his conviction was not authorized under the plain terms of the statute, it must be set aside and a new trial ordered; upon which trial the proof offered may come up to the requirements of the law, and legally establish the guilt of the accused, or may completely exonerate him from all connection with the crime.
Judgment reversed.
Russell, O. J., absent.