There is but little, if indeed my, material difference between the evidence as shown in the record now before us and that disclosed on the former appeal (Jones v. The State, 4 Texas Ct. App. 436), when the case was reversed solely because in our opinion the evidence was insufficient to support the verdict and judgment. In this case, as in that, the question is, Has the testimony of the accomplice who turned State’s evidence been corroborated in material matters tending to connect appellant with the commission of the homicide? It is a rule well settled, that in a case where corroborating evidence is required, it is always permissible to strengthen a witness’s testimony by connected incidents showing its consistency and reasonableness. Burton v. The State, 21 Texas, 337.
This has been attempted in this case by showing that many incidents and facts related by the accomplice have been fully established by indubitable testimony. And we are free to confess that in ordinary cases the corroboration of such and so many incidental facts would go a great way to strengthen the testimony. But this is no ordinary case. Appellant is notcharged, either by indictment or by the accomplice, as having been present at and an active participant in the homicide ; he is only held to be an accomplice himself, — one who was not present, but whose guilty complicity consisted in his aiding,-encouraging, and inciting the actual perpetrators to the commission of the deed, before its accomplishment. These incidental facts and circumstances which are supported relate mainly to the actions of the active participants in the crime, and to the conduct and actions of the State’s witness himself. All these.collateral facts which are substantiated so fully may and can be true, and yet be reasonably and fully accounted for upon the hypothesis of the guilt of others and the perfect innocence of defendant.
There are six separate and distinct interviews specified and named by the witness, in which this, accused and some one or more of the other parties charged with the homicide *459are said to have planned, discussed, and arranged the details of this murder ; and in all but one he is directly and positively contradicted by the witnesses for the defence, some of whom had been tried for and acquitted of this murder, and others of- whom had had the case dismissed by the prosecution after they were indicted. It will not do to say that the State’s witness was entitled to more credit than they ; on the contrary, they were freed by the law from all imputation of crime, whilst the witness was by his own admission an accomplice in the deed. The interview not positively denied was one which the witness said took place between himself, defendant Jones, and one William Blocker, about six weeks before the killing. This testimony was denied by Blocker on the record in the former appeal; and the "same record shows that though Blocker had also been indicted, the prosecution was dismissed as to him. There is not one single witness who corroborates the statement as to either of these interviews and conversations.
We come, then, to the only point in which there is the slightest tendency to corroboration in a material matter connecting defendant with the homicide. Expert testimony was introduced to show, and the experts did testify; from a comparison of writings admitted to be genuine with the writing found pinned to the lapel of the murdered man’s coat, that in their opinion defendant wrote the latter.
Heretofore, in the cases of Nourse v. The State, 2 Texas Ct. App. 304, and Jones v. The State, 3 Texas Ct. App; 575, the rule established in California is adopted, to the following effect: “The corroborating evidence may be slight, and entitled to little consideration ; nevertheless the requirements of the statute are fulfilled if there be any corroborating evidence which of itself tends to connect the accused with the commission of the offence.” The People v. Melvane, 31 Cal. 614. When either of these cases (Nourse’s, Jones’s, and Melvane’s) is examined, it will be seen that in fact the corroborating evidence was positive *460and satisfactory, not only as to material matters, but matters which could not be explained consistently with the innocence of the accused.
Our statute provides that it is competent in every case to give evidence of handwriting, by comparison made by experts or by the jury.” Rev. Code Cr. Proc., art. 754. The fact, however, that our statute permits such evidence does not change the well-established rules as to the value of such testimony. Such evidence has always been considered feeble, and in some States unsafe to act upon. Burman v. Plunkett, 2 McCord, 518, cited in Hanley v. Gundy, 28 Texas, 211. In Adams v. Field, 21 Vt. 265, it was said: “But those having much experience in the trials of questions depending upon the genuineness of handwriting will not require to be reminded that there is nothing in-the whole range of the law of evidence more unreliable, or where courts and juries are more liable to be imposed upon.” 1 Greenl. on Ev., sect. 580, note 2.
It makes no difference, however, in our view of the testimony before us, whether this character of evidence be weak or not. We are of opinion that if it were admitted that the writing was defendant’s, still the evidence would not necessarily be conclusive of guilt, or tend so strongly to establish his guilt, even in connection with the testimony of the accomplice, as to warrant us to allow the conviction to stand as a precedent upon which a human being’s life should be taken at the demands of the law. The writing may create a suspicion, a strong suspicion it may be, — it may create a probability, a strong probability if you will, that his guilty complicity in the crime is established as charged. But the law, in its wise and hümane demands for the life of a human being, demands more than strong suspicion, more than strong probabilities, — it demands that the evidence should lead, to a moral certainty, to the conclusion of -guilt beyond every other reasonable hypothesis. Less than this will not suffice.
*461The evidence in this case does not fill the full measure of the standard which the law requires, and the judgment is therefore reversed and the cause remanded for a new trial.1
Reversed and remanded.
Clark, J, having been of counsel in the court below, did not sit in this case.