The defendant’s only assignment of error is that the court erred in refusing to grant him a new trial on the ground of newly discovered evidence; hence we will confine ourselves to that assignment.

The action is one for damages caused to the plaintiff by the bite of a vicious dog, which the defendant harbored. The main issue was the defendant’s previous *80knowledge of the vicious propensities of the dog. On that subject Kors, the defendant’s clerk, as a witness for plaintiff, testified that he had informed the defendant prior to the accident that the dog had bitten the plaintiff on a previous occasion, and had bitten one other person, and snapped at several. The recovery-under our decision in Staetter v. McArthur, 33 Mo. App. 218, hinged mainly on the testimony of this witness.

The newly discovered evidence, on the strength of which a new trial is sought, is contained in the affidavit of this witness filed in support of the motion. In this affidavit the witness states “that he was mistaken as to time of giving Mr. Crohn (the defendant) knowledge of fact that the dog had shown signs of being vicious or of having attempted to bite any one, except that he snapped at a negro, and that such information was imparted after the plaintiff Dennehy had been bitten by the dog as charged in the complaint filed in the case.”

The evidence of a witness, who has been examined. at the trial touching the very circumstances contained in his affidavit filed in support of a motion for new trial, is in no sense newly discovered evidence. That term means evidence which the party seeking relief neither knew nor could have procured by reasonable diligence for use at the trial. Waiving, however, this technical defect in the assignment of errors, yet it must fail for want of merit. Suppose the witness would testify upon the new trial as stated in his affidavit, what would prevent the plaintiff (if defendant’s position is correct) from obtaining a new trial again by procuring an affidavit from the same witness that his testimony on the first trial was correct, and that on the second trial was a mistake, and so ad infinitum. Is the number of new trials to be limited by the conscience of a witness! *81Is the court bound to give credence to the affidavit of a witness who impeaches his own veracity! These questions, it would seem, admit of but one answer. Such a practice offers a premium for tampering with witnesses of the adverse party after the trial.

We need only add to the foregoing that the witness was very fully examined both on his direct and cross-examination as to the date of the alleged communication, and was very positive that it was before the accident to plaintiff. The court, which saw and heard him, is in a better position to judge than we are whether the testimony thus given was true or not.

All the judges concurring, the j udgment is affirmed.