delivered the opinion of the court.

On February 19, 1908, Charles S. Waller, a bralceman in the employment of the plaintiff in error, while endeavoring to bring a flat car, loaded with green hickory timber, down grade from a switch known as a “Y” to the main track, by means of hand brakes, lost control of the car, and, as was supposed, to avoid the result of collision with a box car on the track below, either jumped or fell from the car and was killed.

There was a verdict and judgment for the plaintiff, to which judgment this writ of error was awarded.

The case has been twice tried, and this is its second appearance in this court. On the first writ of error it was held that “Where evidence shows that the method adopted by the defendant railroad company of handling its cars on a siding was a reasonably safe one, and had been used for many years without injury to any one, and was in common use on railroads, it is error to submit to the jury, in an action for personal injury to a servant, the question whether or not such method was a negligent one.” Ill Va. 152, 68 S. E. 404. For that error the judgment was reversed, and the case remanded for a new trial.

The grounds relied on to justify the present recovery are twofold: 1st, That the brakes on the flat car were defective; and, 2d, that the car was overloaded.

There are also but two main assignments of error. The first involves the ruling of the court on the admission of evidence, *335and the second the overruling of the defendant’s motion to set aside the verdict as contrary to the law and the evidence. Of these assignments, the first alone demands our attention.

The court, over the objection of the defendant, admitted the stenographic report of part of the examination of W. H. Richards, general manager of the defendant company, on the former trial, in respect to the condition of the brakes on car No. 12, in March, 1909, more than a year after the accident. The theory on which the court admitted this statement was that it constituted an admission by an official of the company that the brakes on the car were in the same condition when examined by Richards as when the accident occurred. Upon that assumption, the court permitted the plaintiff to introduce the testimony of several of the jurors at the first trial as to the condition of the brakes at that time.

The statement attributed to Richards was as follows: “There is the same lever—the same thing—to the very best of my knowledge and belief and inquiry, as there was when the accident occurred, and is to-day; the same chain is upon it.” Q. “That dead lever is, you say, the same to-day as when the accident occurred?” Answer: “Yes, and that pin that is there, it is the same pin.”

It is quite apparent that, while the witness identifies the various parts of the brake as being the same as when the accident occurred, he does not say that they were in the same condition.

Richards, in his testimony at the second trial, says: “I meant to say that it was the same pin and the same lever; that no other had been substituted in place of it. I did not mean that it was in the same condition as it was, because there was wear. I mean to state that the same lever and same pin were there.” Q. “Did you or not mean to say by that answer that the brakes on No. 12 were in the same condition on the day of the accident as they were on the day you were testifying?” A. “No, sir.” Witness then proceeded to state that he had no personal knowledge as to what the condition of the brakes was when the accident occurred.

But, independently of Richards’ explanation, it is clear that the language attributed to him at the other trial was not susceptible of the construction placed upon it. If, therefore, the state*336ment was admissible for any purpose, Richards being present in court and subject to examination, it was unambiguous, and should have been interpreted by the court. The condition of the brakes at the time of the accident was a vital issue in the case, upon which the evidence was conflicting; and the effect of the court’s ruling was to cast in the scale against the defendant the inadmissible evidence of jurors at the former trial.

The court correctly held that evidence of the condition of the brakes at the date of the first trial was too far removed in point of time (about thirteen months) to be admissible, in the absence of evidence that their condition was unchanged. The general doctrine on that subject is that evidence of the condition of machinery, etc., a reasonable time after an accident (determinable by the character of the appliance) is admissible, in the absence of evidence of a change of condition in the meantime. But when considerable time has elapsed, the burden of proof shifts, and such evidence is not admissible, unless it is made to appear that the condition has not changed. 1 Wigmore on Evidence, sec. 437; 37 Cont. Dig., “Negligence,” see. 251; 29 Cyc. 614; Wash., Alex. & Mt. V. Ry. Co. v. Vaughan, 111 Va. 785, 69 S. E. 1035.

The situation was not relieved by the court’s instruction that if the jury believed from the evidence that Richards did not intend to say that the brakes were in the same condition at the last trial as at the date of the accident, they must disregard the evidence of the jurors on that subject. Richards made no such statement, and the only remedy for the original error in admitting the evidence based upon the false premise was to have unqualifiedly instructed the jury to disregard it.

For these reasons the verdict of the jury must be set aside, the judgment reversed, and the case remanded for further proceedings.

In these circumstances it is unnecessary, if not improper, to consider the second assignment of error.

Reversed.