The petitioner, having been convicted of the crime of rape upon the person of one Ellen Warner, brought his petition to the Superior Court for a new trial for newly discovered evidence; and the case was reserved for the advice of this court.
The newly discovered evidence is all contained in two affidavits annexed to the petition, which were taken before the petition was served, without any notice or opportunity for cross-examination in behalf of the state.
Assuming for the present for purposes of discussion that the ex¶arte affidavits are competent evidence, do they disclose sufficient ground for granting a'new trial ?
The evidence relied upon consists of two items; — 1st. That Mrs. Warner, the victim of the crime, before her testimony was given in court, made an admission to a Mrs. Collins, *267■which is claimed to be inconsistent with her testimony as given at the trial. 2d. That after the conviction of the petitioner, upon the suggestion of a fact by Mrs. Collins, Mrs. Warner believes she made a mistake in charging Shields with the crime.
1. The admission to Mrs. Collins, according to her affidavit, was as follows:—
“I know Mrs. Ellen Warner of New Haven; have known her the past four or five years; took care of her the day after this outrage. I asked her if she knew these parties, or any of them, and she said, ‘No; but they were a rough crowd of fifth-warders.’ ”
It seems to us that such testimony is too uncertain to affect the case. It appears that several persons were concerned in the assault on Mrs. Warner, all of whom were entire strangers except Shields. The case, as stated in the petition and in the affidavit of Mrs. Warner, proceeds on the idea, that she knew him, and had just seen him with this rough crowd in the saloon, at the door of which she was seized by several persons when she was going out. Now, when asked in'that general way “if she knew these'parties or any of them,” no particulars being given, may she not have referred to these strangers ?
But however this may be, the most that could be claimed for it is, that it was a statement out of court somewhat inconsistent with her testimony at the trial, and as such would be one mode of impeaching the credit of the witness. 1 Greenl. Ev., § 462. And.as it is well settled, as laid down in Parsons v. Platt, 37 Conn., 564, and other cases, that the new evidence must not be offered to impeach a witness, why should not the rule exclude such evidence' as this? If a direct impeachment, that tends utterly to destroy the credibility of the witness, is excluded, it would be strange to admit an indirect impeachment, which, as in this case, could only affect the witness in some slight degree.
.In Commonwealth v. Drew, 4 Mass., 399, the defendant, after he had been convicted of the crime of murder, ascertained that a material witness for the government had declared *268before the trial that he would hang the prisoner by his testimony if he could,-and for this newly discovered fact applied to the court for stay of sentence and relief; but the court held that it was insufficient to avail the prisoner on any legal principles.
2. In regard to the other item of new evidence, that Mrs. Warner made a mistake in her identification of Shields, we think the fact of a mistake is not sufficiently shown.
Wheré a mistake is made to appear plainly by extrinsic facts and circumstances, as for instance where persons are convicted of theft and the alleged stolen property is after-wards found by the owner not to have left his own possession; or where, after conviction for murder, the alleged victim is found alive, justice as well as public policy would demand the correction of the mistake and the speedy relief of the parties affected by it.
But in the case at bar the mistake is not otherwise evidenced than by the altered opinion of the witness.
The affidavit of Mrs. Warner is as follows: — “I was the complainant in the above entitled case and testified that Shields was one of those who made the attack upon me. I then believed that I was telling the truth. In a conversation this day with Mrs. Hanna Collins I was reminded of a circumstance which showed me that I was wrong in saying that Shields was one of them. I was excited at the time; it was dark, and I had seen Shields in the saloon before the attack; and when I testified I felt certain that he was one of those who attacked me, but now I feel sure he was not. I make this statement because I now believe Shields to be innocent, • and wish to undo the wrong which,, my mistake' has brought upon him.
“ July 31st, 1877. '
her Ellen Warner.”
The following is the affidavit of Mrs. Collins, so far as it relates to this point: — “I had been told by somebody that when the crowd were going out of the door after Mrs. Warner, some one of them told Shields to get out of the way, for Mrs. Warner knew him, and that he went away. On this, the 31st day of July, 1877,1 saw Mrs. Warner and asked her *269if she did not remember that as she was going out of the door some one told Shields to get out of the way, as she knew him ? Mrs. Warner said that she had not thought of it before I spoke of it, but now she remembered it, and she must have made a big mistake in her testimony at the time of the trial in saying that Shields was the one who took hold of her by the throat, that she was very sorry for having made such a mistake and would be willing to do anything to correct the mistake. I told her then that she ought to go to Judge Robinson and see if he could do anything about it, and she came here with me. Dated at New Haven, Conn., July Blst, 1877.
her Hanna + Collins.”
. It will be seen from the above that Mrs. Warner’s change of opinion, after the conviction of Shields, originated in the suggestion of a fact by her friend, and a process of reasoning; thereon. And it is to be noticed also that the fact referred to had only a very remote connection, if any, with the question of identification. If Shields was told by somebody to get out of the way, it implied that he was in the way, but it cannot be implied that he obeyed the request; and without this last mentioned fact, (which is not found,) the process of reasoning has no basis to rest upon.
Identification from -.the nature of the case is not susceptible of demonstration; it rests mainly on impressions derived through the senses, and does not ordinarily depend on any conscious process of reasoning. As we remarked in giving the opinion in Sydleman v. Beckwith, 43 Conn., 13: “A witness well acquainted with another usually identifies him without conscious mental effort in the way of comparison or inference. In the absence of striking peculiarities of form or feature the identification may be, and often is, by the mere expression of countenance, which cannot be described. And the witness may be correct in his opinion, and yet be unable to give a single feature, or the color of the hair, or of the eyes, or any particulars as to the dress. In such cases the distinction between opinion and fact is so very nice that it might perhaps have been as well to consider such identification as a fact, like any other direct perception of the senses.”
*270To allow a case depending on opinion or identification to be re-opened on such grounds as are here disclosed, would open a wide, convenient, and tempting door to defeat the ends of justice. After the trial is over and the accused stands convicted, with the heavy penalty of the law impending and just ready to fall upon him, how easy by artful or even honest suggestion to awaken a sympathy even in the heart of the victim, who was the main, perhaps only witness against the accused, and who naturally feels responsible for the conviction ; and how easy for such witness by a process of speculation, colored by feeling, to feel and express a doubt about the correctness of the opinion entertained at the time of the 'transaction.
While we would gladly grant relief for a manifest mistake, yet in such a case as this the maxim, “Interest reipublicoe ut sit finis litium,” applies in full force.
We know of no adjudicated case precisely like this, but there is a well settled general rule, sustained by numerous authorities, which applies; this is, that a new trial will not be granted on the mere after-recollection of a former witness. Duignan v. Wyatt and another, 3 Blackford, 385; Johnson v. Blanchard, 5 R. Island, 24; Bond v. Cutler, 7 Mass., 205; King v. Gray, 17 Texas, 62; Archer v. Herdt, 55 Geo., 200; Ainsworth v. Sessions, 1 Root, 175.
There are exceptions to the above rule, as in the case last cited, but there is nothing in the case at -bar to make it an exception.
3. The only remaining question is, whether the existence and character of the new evidence can be shown by ex parte affidavits. The disposal of the present case does not require an answer to this question; but if we now avoid the question doubts will be raised about the proper rule of practice, which will embarrass' the profession in bringing future petitions.
Until this case we had supposed there was no doubt that by our practice the allegations relative to the new evidence should be sustained by testimony, either ore tenus or by deposition, as in other cases at law. In other states and in England where the chancery practice is different, affidavits have been *271received. The argument in behalf of the petitioner is founded upon the practice of other states; and also upon the p.1 aim that when the General Assembly of this state in 17(32 committed to certain courts the power “ to grant new trials according to the common and usual rules and methods,” (which expression is still retained in our statutes,) the rules of the common law as evidenced by the decisions in England must have been referred to and adopted.
We cannot accept this statement as correct. The rules and methods of our own state were intended. Judge Swift says in his introduction to the second volume of his Digest, p. 17, that “ as our courts of law succeeded to the General Assembly as courts of equity they adopted their mode of proceeding instead of the English, especially in regard to taking-testimony by vivd voce examination of witnesses.” See also 1 Swift’s Digest (Rev. ed.,) top p. 271. Affidavits that fill» so many volumes in the chancery records of England and of some of our sister states are almost unknown in our practice.
This court at an early day, in Jerome v. Jerome, 5 Conn., 352, decided that “ an affidavit, by our practice, is not required to be annexed to a bill in chancery in any case.” In the earliest reported cases in this state, as well as in the later decisions, there is no reference to affidavits, but in many cases the expressions used indicate that the witnesses were personally present. For example, in the case of Noyce v. Huntington, Kirby, 282, decided in 1787, it is said “that a witness named in the petition was offered ly the petitioner but the court refused to allow him to he sworn, because due diligence had not been used to procure his testimony at the former trial. The petitioner then offered a witness to be sworn, who was not named, and the court for this reason refused to allow him to be sworn. Some members of the court however thought he might be admitted after witnesses had testified who were named. No after witnesses were offered and the petition was withdrawn.” If the practice had been to prove the case by affidavits alone such expressions as are contained in the above would not have been used.
But it is said that the application for a new trial is always *272to be determined, not by the merits of the cause in -which the application is made, but by the merits of the application itself, and therefore affidavits are the only proper testimony. We do not see how the conclusion follows from the premises. To determine the merits of. the application the court must put ■itself, so to speak, in the place of the court before whom the original cause was heard, and carefully weigh all the new evidence with that previously offered, in order to decide whether a re-trial would probably change the result.- The new evidence, it will be conceded, must be under the sanction of an oath, and must be thoroughly sifted and weighed by the court; why then should it not for the same reason have applied to it the test of cross-examination, justly described as “one of the principal and one of the most efficacious tests which the law has devised for the discovery of the truth.” 1 Greenl. Ev., § 446. To allow the rights of parties to be affected by eayparte affidavits without the protection of a cross-examination, is almost as unjust as to allow the statement of a witness not under oath. Such a practice would hot only affect injuriously the rights of parties, but would open a wide door for imposition upon both witnesses and court.
The case at bar well illustrates the danger of such evidence. The entire evidence consists of two affidavits signed only by a mark. In all probability the affiants could. neither read nor write. The contents of the affidavits might not have been made known, or may have been imperfectly understood; and how, and under, what .circumstances and inducements, they were obtained, could only be known through the sifting process of a cross-examination. In this case it is found that the witness might have been produced in court without trouble or difficulty.
We think therefore a new trial should be refused on both the grounds before mentioned, and we advise that the petition be dismissed.
In this opinion the other judges concurred.