(concurring). I concur in the reversal of this case on the ground set forth in the majority *854opinion; and it is manifest that the harm resulting from the admission of this unlawful confession could, not, on the facts in this case, he cured by instructing the jury to disregard such confession. With the confession, eliminated there was nothing left to secure a conviction, except circumstantial evidence, and this was not, in my opinion, strong enough to exclude every reasonable hypothesis except guilt. The rule governing the sufficiency of circumstantial evidence is well stated in Algheri v. State, 25 Miss. 584, 1 Morris’ State Cases, 658, as follows:
“In the application of circumstantial evidence to the-determination of a case, the utmost caution and vigilance should be used. It is always insufficient where, assuming all to be proved which the evidence tends • to-prove, some other hypothesis may still be true; for it is the actual exclusion of every other hypothesis which invests mere circumstances with the force of truth. Where the evidence leaves it indifferent which of' several hypotheses is true, or establishes only some finite probability in favor of one hypothesis, such-evidence cannot amount to proof, however great the-probability -may be.”
Applying this rule to the facts in this case, the proof is wholly insufficient to establish either: First, the corpus delicti; or, second, the defendant’s connection, with the burning.
In my opinion -the corpus delicti is not sufficiently established. Pitts v. State, 43 Miss. 472, 2 Morris’ State Cases, 1655; Stringfellow v. State, 26 Miss. 157, 59 Am. Dec. 247; Spears v. State, 92 Miss. 613, 46 So. 166, 16 L. R. A. (N. S.) 185. It appears that the barn-was burning and the fire was under good headway when it was first discovered by the owner and his employees.. No person was seen near the barn, nor were any such Circumstances in evidence as would show beyond reasonable doubt that the barn was set on fire by human agency. It appears from the evidence of numerous *855witnesses in the ease that there was a good deal of' lightning and thnnder on the night of the burning, and that several houses in the community had been destroyed by being set on fire by lightning within recent times to the knowledge of the witnesses. Under these circumstances it would appear probable that the barn was set. on fire by lightning. There was certainly a reasonable hypothesis that it could have been fired by this agency; and the evidence should establish the corpus delicti to-the exclusion of every reasonable hypothesis where there is no direct evidence of human agency in.setting fire to the barn.
Again, I do not think the evidence to connect the* defendant with the burning is sufficient to exclude every reasonable hypothesis if we consider, for the sake of' argument, that the barn was burned by human agency. John v. State, 24 Miss. 569, 1 Morris ’ State Cases, 608 ; Calebb v. State, 39 Miss. 721, 2 Morris’ State Cases, 1490; Algheri v. State, 25 Miss. 584, 1 Morris, State Cases, 658. After the fire was discovered the parties present made every effort to keep other persons away from around the barn until bloodhounds could be obtained. "When the party owning the bloodhounds appeared upon the scene he circled the barn with the dogs, and they picked up the trail at some little distance from the barn, and pursued the trail to the house-of one Dillard, a negro living upon the premises of the owner of the barn. The dogs trailed the tracks onto the-gallery and to the door, and stopped and when the-door opened and' the dogs went on the inside, smelled around the house on the inside, and, to use the language of one of the witnesses, “appeared to be satisfied.'”' Thereupon the parties in charge of the dogs questioned a little negro, being a son of Jim Dillard, and he stated that in the night he heard some party come upon the gallery and go off the gallery at a certain corner. The dogs were then taken outside of the house, and circled the house, and took a track at a corner of the *856gallery about where the little boy said the party went ■off of the gallery. There was no effort to have Jim Dillard brought in contact with the dogs to see whether they would recognize him, nor was his track examined, ■or his shoe measured, or anything done to negative any inference that he may have made the track that the dogs trailed from near the barn to his house. The dogs, when leaving Dillard’s house, pursued or trailed until they reached the house of Ward. They trailed to the .house, and into his house, and Walked up to Ward, and smelled of him, and walked away, seeming, to use. the language of the witness, “to be satisfied.” It does ■not appear that the dogs were trained in a manner to make any demonstration on finding the person they were trailing. The owner said they were trained so that they would usually go to the party and smell of him and appear satisfied, but would not bark at him, or make any effort to take hold of him.
It is exceedingly unsatisfactory to me to convict a man upon evidence that a dog trailed a track to his house, smelled of him, and looked satisfied without making any other demonstration. The testimony does not negative the idea that other persons than Ward could have committed the crime, if it was a crime under the proof; that is to say if there was any human agency in burning the barn. It does not appear conclusively to my mind that the dogs ran the same track all the way from the barn to the appellant’s house. It would look to me like a dog well trained would have continued the track upon and across the gallery and into the yard and on wherever it may have gone. Just why the dogs would stop and leave the trail and go to the door and into the house (when, concededly, under the testimony, the party who crossed the gallery never entered the room at all), instead of keeping the track, is not apparent. Just why the dogs would “look satisfied” when reaching Dillard’s house, and why that would not be accepted as evidence as well as when the lpoked *857satisfied at Ward’s house, is a little mysterious. It rather appears to my mind that the parties did not want. Dillard. He probably was a valuable plantation hand,, and it would be inconvenient to send him to the penitentiary, and for that reason no examination was. made of him, his tracks, or of his premises on the occasion in question.
Unless the state can materially strengthen its evidence on these points on a new trial, I think it would be a good opportunity, for the state to show magnanimity and enter a nolle prosequi. Ward is shown to-have been a person of good reputation in his community and he made all the defenses that a man ordinarily can make — that is, by the evidence of a man’s wife that he was at home on the occasion; the wife testifying positively that she waked up while it was thundering, and that her husband, was then in bed. The time of the thundering shown by other witnesses would make it impossible for Ward to burn the barn and reach home and be in bed at that time. After a consideration of the-facts in the record, three of the judges of this court-are of opinion that the facts proven are insufficient to-prove guilt, which, under the rules of circumstantial evidence, ought to be enough to show that there is a reasonable hypothesis other than guilt.