Opinion by The only question which it is j>roposed to consider is, did the court err in oveiruling the application for a new trial? The other errors assigned by appellant are not regarded as .showing any ground for reversal, or as presenting any questions requiring discussion. The result of appellant’s efforts immediately after the trial shows that the testimony which he then obtained as to the whereabouts of Mahan in the fall of 1844 might have been discovered and obtained before the trial. Clearly the appellant was inexcusable for failing to produce this testimony on the trial, if the issue as to where Mahan actually was in September, 1844, was one *198on which he was bound to be prepared on the trial. By the affidavit of Mrs. Mahan impeaching the genuineness of the deed of September 8, 1844, from Mahan to Connough, appellant was notified of the necessity of proving its execution; and by the cross-interrogatory propounded to the witnesses, whose depositions he took to prove the handwriting of Thompson, the witness to the instrument, and the officer before whom it was acknowledged, of Mahan, the maker, and that the body of the instrument was in the handwriting of Connough, the grantee, he was notified that Mrs. Mahan was endeavoring to establish that Mahan had .removed from Houston before September, 1844.

As these witnesses testified fully that in 1844 Mahan continued to live in Houston, it is not perceived that the appellant was in any way notified that evidence would be offered to show, not that Mahan, in September, 1844, had removed from Houston, but that he, during that month, was in a distant State. There being nothing to notify appellant in advance that such an issue would be made, we think that he is excusable for having failed to anticipate it. Having proved up tie instrument by four respectable witnesses who knew Mahan, Thompson and Connough for some years prior to 1844, who testify clearly and minutely as to the genuineness of the signatures of witness and grantee, and also that the body of the instrument is in the handwriting of the grantee, Connough, and who testify also as to the residence of all those in Houston during 1844, we think appellant was not chargeable with negligence if he failed to go further and prepare to meet evidence that in September, 1844, Mahan was in Philadelphia and Trenton, H. J. The instrument impeached purports to be over tbirty-five years old, and he could hardly anticipate that an effort would be made to disprove its genuineness by showing that Malian was somewhere else at a date so remote.

Under all the circumstances of the case, appellant might well be surprised when testimony was offered, not that Mahan, in 1844, had removed from Houston, hut that at different times about that date, he was in Trenton and Philadelphia, and therefore could not have been in Houston. True, such evidence tended to disprove the genuineness of the deed impeached, but it was evidence which, under the circumstances exhibited by the record, appellant could not have anticipated, and therefore, in the exercise of ordinary diligence, was not hound to he prepared to meet. The case of Sar*199gent vs. Deniston (5 Cowen, 106-122) is one similar in principle. (Graham and Waterman on New Trials, pp. 952-962 and 1055.)

But it is said that appellant adduced evidence on the trial that Mahan, throughout 1844, still resided in Houston, and that the evidence offered to show that throughout the fall of 1844 he was actually in Houston, is merely cumulative. We do not so regard it. The main issue was the genuineness of the deed, and on the the trial, by witnesses produced on the stand, the subordinate issue was made that at the date of the deed Mahan w'as many hundred miles away from Houston. Appellant, on the trial, adduced no direct evidence on this point. Indirectly his evidence, as to residence had a bearing on this issue, but direct evidence that Mahan was in Houston throughout the fall of 1844 was not cumulative of this indirect evidence. Evidence is not cumulative merely because former evidence may have indirectly tended to establish the same fact. If so, after any evidence tending to show the execution of the deed, evidence of the whereabouts of Mahan at its date would have been cumulative, because whatever tended to show the genuineness of the deed tended also to show that Mahan was in Houston at the date of its execution.

The reason of the rule forbidding a new trial, for the purpose of admitting cumulative testimony, does not apply where the party has had no fair opportunity to procure and adduce evidence. oh an issue raised by his adversary, for the first time during the trial, by the introduction of evidence which could not be anticipated. The reason of the rule is, that public policy, looking to the fidelity of trials, requires that parties be held to diligence in preparing their cases, and that they shall not be allowed a second trial because they mistook the amount of testimony requisite. (Powell vs. Jones, 42 Barb., 30.) But the policy which seeks to limit continued litigation does not apply where a party has had no fair opportunity to present his side of the case — no real day in court.

In discussing this subject, Chief Justice Wheeler said: “It is desirable that there should be an end of litigation with as little delay and expense as possible, consistent with the great end of litigation — a correct decision of causes according to their real merits; but it should always be sought in subordination to the great end to be obtained.” (Mitchell vs. Bass, 26 Texas, 377.)

In the same opinion, he also said: “Where there can be any doubt of the justice of the verdict, to refuse a new trial, when the *200party has really discovered new evidence of a conclusive tendency, would be against justice and precedent.”

In the conflict of the evidence, the justice of the verdict in this case is at least doubtful, and, as the additional evidence sought to be adduced is evidence which the appellant is excusable for not having jiroduced on the former trial, and is evidence tending directly to a material issue, and which might probably lead to a different verdict, we think that the new trial should have been granted. (Railway Company vs. Forsyth, 49 Texas, 178.)

Because the court erred in refusing a new trial, the judgment is reversed and the cause remanded.