Our Code, section 3755, provides as follows : “The testimony of a single witness is generally sufficient to establish a fact. Exceptions to this rule are made in specified cases; such as to convict of treason or perjury, in any case of felony where the only witness is an accomplice, and • to rebut a responsive statement in an answer in equity. In these cases (except in treason,) corroborating circumstances may dispense with another witness.”
In the case made by this record, which is a felony, tire sole witness connecting the defendants with the transaction is one who admits that he -was one of (he robbers, and there are no circumstances proven by other witnesses corroborating his statement that the prisoners were with him, or that they were connected, in any way, with the commission of the crime. It is contended, however, that as other witnesses do prove there was a robbery, and that the time, place and circumstances *111were such as stated by the witness, this amounts to such corroborating circumstances as fulfill the terms of the Code, leaving the weight of these circumstances to be determined by the jury. A majority of the court are of the opinion that such circumstances, however numerous and detailed,' are not corroborative circumstances in the sense of the Code; that the first to be established is the connection of the prisoners with the crime, and that the “corroborating circumstances” intended by the Code are such as go to show that the prisoners were, in some way, implicated in, or connected with, the robbery. That the witness himself, who comes avowedly before the court as a perpetrator of the crime, should be able to tell the time of night, the manner of the crime, that more than one were engaged, the amount stolen, etc., etc., is corroborative of his own guilt, and is involved. in the very admission that he is an accomplice. But that he tells the truth about such things as these is, as it seems to us, nothing going to show that the prisoners are guilty, and do not at all corroborate his story implicating the defendants. And this is, in our judgement, really the only thing in issue, “the fact to be established.” This section of the Code is peculiar and significant in the language it uses.
It sets out with the statement that ordinarily one witness is sufficient to establish a fact. It then declares there are exceptions to this rule. It mentions treason as an exception, and then perjury, eases of felony where the only witness is an accomplice, and finally where it is undertaken to rebut a responsive statement in an answer to a bill in equity. In these eases, except in treason, corroborating circumstances may dispense with another witness.
It seems to me that it is difficult to read this language without coming to the conclusion that more is intended than that the one witness, the accomplice, shall tell a reasonable story ; that he shall show that he knows a crime was committed, and tell truthfully the'time, place and manner of it, as they are known, perhaps, to half the community, by the story of the person wronged, or by the traces which almost every crime *112leaves behind. As I have said, it seems to me, these words of our law mean something more than this; and that their language can only be met by requiring proof from other witnesses of facts going to show that the persons on trial, whose guilt-is the only fact to be established, are implicated in the crime. Nothing else is any corroboration of the fact in dispute. Nobody denies the guilt of the witness; that fact needs mo confirmation. He goes upon the stand avowing that the whole point of the case is the guilt of the persons on trial, and .nothing which does not tend to confirm his story, as to them, can be fairly spoken of as corroborating the evidence he gives as to the fact to be established.
It will be noticed that the statute does not say that the jury shall not convict on the testimony of- one witness. The provision is that a <l fact” cannot be-established thus. The significant, controlling, vital fact of this issue is not the robbery, not the time, place or manner of it. Had that been the question, this witness would not have been put up at all. The person' on whom the outrage was committed was present and a witness. Not a single new fact of a material character, except his own presence, is stated by the witness; and at last this verdict is founded, the fact of the guilt of these prisoners is taken as established, solely on the statement of the accomplice. It is trifling with the principles of justice to sgy that the circumstances in reference to Sandy Boothe, as that he first suggested the guilt of the witness, that he tried to hire a buggy next morning to go into the country, and said he had the money to pay for it, are facts going to show his guilt. His pointing to the prisoner strikes me as a strong fact in his favor, and suggests a motive for the witness to turn the tables on the one who put the officers after him, and the other facts may be true of one hundred men in Athens that very morning, and are worthy of no attention. Indeed, the whole argument here has been on the assumption that there is nothing to show the guilt of the prisoners but the statement of the accomplice, and the conviction is sought to be sustained by insisting that, under the law, circumstances confirmatory of the witness’ gen*113eral narrative are sufficient, though those circumstances do not at all tend to connect the prisoners with the matter.
For myself, I cannot so read the Code. Its language is, to my imind, plainly to the contrary. Its suggestion, that these circumstances may dispense with another witness; its classification of the case with cases of treason, perjury, and rebutting an answer in equity, impress me overwhelmingly with the idea that there must be more evidence of the guilt of the defendants than the prisoner’s statement; that the corroborating circumstances must be cumulative, that they must add to the testimony of the accomplice. In other'words, I think our law — this section of the Code — requires more evidence of the guilt of the prisoners than the oath of an accomplice.
The facts of this case, as detailed by the accomplice, are simply the robbery, and its details; and he says that the prisoners were there. The other witnesses say nothing implicating the prisoners at all, even the most remotely. A majority of this court think this is no corroboration at all of the fact in issue, to-wit: the guilt of the defendants. In England, the law upon this subject is in a very uncertain condition. It is laid, down in all the books that even in felonies it is competent for the jury to convict on the uncorroborated evidence of an accomplice, and there are several cases where a new trial has been refused when this was the cáse. On the other hand, it is a well recognized rule for the court, on the trial, to tell the jury that they ought not to convict unless the witness is corroborated. In other words, there is no rule.of law on the subject. The courts advise the jury what is right and proper, but at last, whether a conviction shall be had or not, depends on the opinion of the jury, and the judges let the verdict stand or not, accordingly as they think upon the whole the defendant is guilty or not. But even under this very uncertain rule, it is our conviction that the great burden of authority is, that the judges advise the jury not to convict unless the accomplice be corroborated, not generally, but in that part of his story which connects the prisoner with the offense.
"Whilst we are not able to see what this practice of the Eng-*1143ish judges, as to their advice to the jury as to the weight of evidence under a rule of law which authorizes the jury to convict without any eorroboratjon, has to do with the construction of our statute which declares, as a rule of law, that in cases of felony the evidence of an accomplice is not sufficient, still, as much stress was, in the argument, laid on this practice, I propose briefly to discuss it.
As we have said, it is competent in England for a jury to convict even of felony, on the uncorroborated evidence of an accomplice: Per Lord Denman in Hastings’ case, 7 C. & P., 152. But it has long been the practice of judges at nisi prius, to advise the jury that they ought not to convict if the accomplice be uncorroborated. There is, however, some uncertainty as to the precise terms in which it has been the custom to give this advice. The truth is, the terms have not always been the same, and that, for the simple reason, as it is mere advice, and not obligatory either on the judge to give or the jury to adopt it, and not the subject of review, (see Regina vs. Stubbs, 33 E. L. & E. R., 551), it has always depended on the temper and tendency of the mind of the particular judge presiding at the trial, and has not been, and cannot be, precisely settled, because no reviewing court can define and precisely mark it out. Still I think, as I have said, that the large balance of the English eases state the practice to be to advise the jury not to convict unless the accomplice be corroborated in his statement connecting the prisoner with the crime. The practice is so stated by Roscoe, 155 to 157, by Philips, 1 Phil. Ev. 30, by Wharton, American Criminal Law, sec. 789, and is elaborately defended by Amos & Philips in their note to the 8th Ed. of Philips. See that note in full in Coven’s Notes to Philips, Part 1, 20 to 23. In these text books the authorities are quoted, that is, the cases in which the advice was given, and I refer to them simply, without taking the trouble to set them out in detail. Against this is the essay of Baron Joy, late chief baron of his majesty’s court of exchequer, in Ireland.
The preface to this essay states the object of the book to be *115to show that certain opinions of judges, to-wit: two cases lately decided, in which the terms of the advice to be given to juries are stated as requiring confirmation, as to the connection of the prisoner with the crime, do not state the law correctly. Those two cases are Rex vs. Addis, 6 Car, & P., 388, and Rex vs. Webb, Ibid., 595.
The text of the essay contends elaborately: First, that there is no rule of law on the subject; and second, that the practice of the judges has been to give the advice simply to the effect that there must be confirmation, without limiting the corroboration to any particular part of the testimony. It will be found, however, that whatever may be the rule as deduced from-the cases referred to by Baron Joy, the English judges have.continued to give to the juries the advice given by Patteson, Judge, in 6 Carr. & P., 388, and in 6 Carr. & P., 595, by Justice Williams, which it was the object of the essay to show was not law. And a careful reading of the essay will, I think, show that Baron Joy has, in the cases he referred to as supporting his view, laid too much stress upon those cases which hold that there is no rule of law requiring any corroboration, and has not given due weight to the practice of the judges in their charges to the juries, which, in truth, is the only force the rule has in England at all. In Wilkes’ case, 7 Carr & P., 272, after the publication of Baron Joy’s essay, Aldebson, B., in summing up, said, “ The confirmation of the accomplice, as to the commission of the felony, was no confirmation at all. And the confirmation which he always advised juries to require, was in some fact which went to fix the guilt on the particular persons charged.” And in Moore’s case, 7 C. & P., 270, the same judge told the jury that where a thief and a receiver were jointly charged, a confirmation, as to the thief, did not advance the case against the receiver. In 8 Carr & P., 106, Lord Abinger said he was clearly of the opinion that the corroboration of the accomplice must be as to the particular prisoner. In 8 C. & P., 261, Guerney, B., held that in a majority of recent cases it was laid down that the confirmation must be of some matter which went to con*116nect the prisoner with the transaction. In Regina vs. Stubbs, 33 Eng. L. & E. R., a case decided in 1855, twenty years after the publication of Baron Joy’s essay, in a reserved case before the court, the point was whether a conviction was legal, on the evidence of an accomplice, entii’ely uncorroborated. The court held that it was. Jarvis, C. J., said they regretted the conviction because it was contrary to the ordinary practice. The judge had charged that if the accomplice was corroborated as to one prisoner, that was sufficient to justify the conviction of the others, and the jury had convicted. And he added : “ It is not a rule of law that the judge must-tell the jury corroboration was necessary. But it is the practice to advise the jury they ought not to convict without it. And I think it is proper for the judge to tell the jury it is not safe to act on his testimony, as to any prisoner, with regard to whom he is not confirmed.” Park, B., said: “My practice has always been to tell the jury not to convict the prisoner ixnless the evidence of the accomplice is corrobox’ated, not only as to the circuixxstances of the offense, but also as to the. person of the prisoner.” Creswell, Judge, said: “ I have always acted upon the view of the subject px-esented by my brother Park.” Wigi-itman, J., said: “The practice had not been uniform,” and Welles, J. said : “This is not a question of law but of practice, and questions of law only can be reserved for our opinion.” That there should be some diversity of px’actice in a matter which could not be reviewed, is not strange. But here is the opinion of a majority of the judges of the courts of queen’s bench, the highest criminal court in England, as to what the rule under discussion is.
As I have said, it is not clear how a matter of advice only to the jury can be used to aid in the construction of a positive rule of law, as laid down in our Code. Still, as I have shown, even this rule of advice conforms to the view we have taken.
It may be these men are bad, guilty men, but if they are convicted, it ought to be under, the rules of law. To justify this verdict would be, in our judgment, to make the best men *117in the land subject to the danger of conviction by any guilty scoundrel who might attack them. His guilt is easily corroborated. The rule is, and ought to be, that some fads must be shown by other witnesses, tending to show guilt in the person on trial.
Judgment reversed.
Trippe, Judge, concurred, but furnished no opinion.