Opinion,
Mr. Justice Green :In no possible point of view were the declarations which were admitted in evidence under the 13th, 14th, and 15th assignments of error competent. The testimony of Vaughan' as to what Hopkins had told him was not admissible as a contradiction of anything as to which Hopkins had been asked; and as declarations by agents while in the course of their duty, all the declarations of the three persons, Warg, Hopkins, and Smith, were incompetent, because they were not part of the res gestee. The rule of law upon this subject is so perfectly familiar that it is not necessary to refer to the authorities. These three assignments are, therefore, sustained.
The assignments from number 11 to 15, both inclusive, relate to the admissibility of evidence in regard to the condition of the locomotive. It was all received on the trial, but in the charge the learned court below withdrew it all from the consideration of the jury as being entirely insufficient to convict the defendant of any negligence in relation to the engine. After the testimony was closed and before counsel addressed the jury, a motion was made to strike out this testimony, but was refused' by the court, under exception.
It has long been held, and in many cases, that where evidence has been improperly received which tends to prejudice the minds of the jurors, and the court in the charge directs the jury to disregard the evidence and withdraws it from their consideration, this instruction comes too late and does not cure the error of admitting it: Del. & Hud. Canal Co. v. Barnes, *26531 Pa. 193; Penn. R. Co. v. Butler, 57 Pa. 335; Huntingdon, etc. R. Co. v. Decker, 82 Pa. 119. In tlie last case Mercur, J., said, referring to D. & H. Canal Co. v. Barnes, “ Tlie manifest reasoning of tlie court was to hold, that whenever the testimony received was of such a character as to inevitably tend to prejudice the minds of the jurors, the error was not cured by the court telling them after the argument had closed not to consider the testimony.” .And in Penn. R. Co. v. Butler, Sharswood, J., said, “It is in entire accordance with the opinion in that case to hold, as we do here, that if improper evidence is given tending to inflame the damages, and it is not struck out at or before the close of the testimony, so that counsel shall not be allowed to refer to or dwell upon it in their address to the jury, it is altogether too late to cure the mistake by directing the jury to disregard it in the charge.”
These several assignments are therefore sustained, and also that part of the 16th assignment which relates to the testimony of Caskey and Hayes.
The remaining portion of the sixteenth assignment relates to the refusal of the court below to strike out the testimony of the plaintiff’s witnesses Stevenson and Rodman, who gave evidence in regard to the curve in the track and the condition of the road-bed. The court was of opinion that this portion of the testimony might be considered by the jury. In an exceedingly fair and clear manner the judge explained to the jury just liow far they might consider the testimony on this subject and in what event they might find for the plaintiff. He said the railroad company is not responsible to its employees for the manner in which the engineer lays out the curves upon the road; the jury could not be permitted to review the skill and judgment of the engineer in such a mattei', but he said they might review the manner in which the rails were laid. After much careful reference to the testimony of the two witnesses, and they were all who testified on this subject, in regard to an alleged irregularity in the line of the curve as the rails were laid, he said, “If you find that the irregularity existed, and that it existed prior to this accident, then the question is whether such an irregularity was the occasion of danger such as to make this company guilty of negligence.” The evidence in question was limited to that of these two wit*266nesses. The substance of their testimony was that on the 22d of April, which was thirty-three ■ days after the accident, they went to the place of the accident and examined the track and took various measurements to show the degree of the curve, and any variations from uniformity or regularity that might exist. They said they found it to be a 9 degree curve, which varied to as low as 7 degrees 20 minutes in one place, to 10 degrees 20 minutes at the place of the accident; that this variation would make a slight difference in the line of the curve which rvould not be sufficient, however, to be visible to the naked eye, and that it would require a variation of three or four degrees to make it visible.
The witness Stevenson said that there was more or less variation in the curves on all railroads, but that there was more in this instance than he had seen on other roads. The witness Rodman gave no measurements, and in fact said nothing about any variations. He was examined as to other matters. There was no other testimony for the plaintiff in regard to this matter. These witnesses were not on the track on the day of the accident nor until the time of their measurements on April 22d. Of course the testimony would have but little value as to the condition of the road on the 19th of March. They went back in October, 1887, and said they then found the road in good condition, the curves more regular and with little variation. The defendant’s road-master testified that nothing had been done to the track from the time of the accident to the time of the trial, and no witness testified to having done any work on the rails or seen any done in that time. The learned court submitted the evidence as to the variations in the curve to the jury with instructions that they might find negligence on the part of the defendant sufficient to enable the plaintiff to recover.
If any testimony had been given to the effect that the condition of the rails as described by Stevenson and Rodman was the cause or might be the cairse of the engine mounting on the rail, or if they had described how such a result might, or could, or probably would have resulted from such a condition of the rails, there would be something from which a jury might be able to draw an inference of negligence in that respect, but there is not a particle of such evidence in the case. Neither of these witnesses nor any other, testified that the slight variation *267from a regular curve in the line, so slight that it could not be seen, as described by Stevenson and Ilodman, could or probably would causo tlie wheels of the engine to mount the rail. On the contrary the evidence was positive, not at all contradicted, that this was tlie only instance in which such an accident had happened; that this same engine, and other engines were constantly running over this curve for many months before and after, and on tlie day of tlie accident, and not one ever mounted tlie rail. Hence it was only possible to conjecture that tlie mounting of the rail might have been caused by tlio variation from the line of a regular curve in tlie rails, but it would be at best but a conjecture without any evidence to sustain it and opposed to all the probabilities of tbe case as illustrated by tlie whole experience of tbe road. Had tbe fireman been a passenger he would have had the benefit of a presumption of negligence which it would have been tbe duty of the company to rebut. But with an employee there is no such presumption and he must prove affirmatively tlie fact of negligence, and that it is such a kind of negligence as violates the special and limited duty of an employer to an employee. The case is simply and utterly destitute of that kind of evidence, and we cannot sanction the result that was reached without a palpable disregard of all tlie well considered decisions of our own and other courts which so carefully prescribe tlie conditions of liability of an employer to an employee on tlie ground of negligence.
There was some little testimony in regard to tbe ballasting of tbe track, but it proved nothing whatever as to the cause of the accident. It did not show that there was or could be from that source any depression of tlie track at tlie place of tlie accident. Tlie engineer, OTIara, wlio was running tlie engine at tbe time, said that after they had gone about a hundred or a hundred and fifty feet on the curve the engine “ seemed to sag over just as though she struck a soft spot in the track;” that “she came back again, tliat is, sbe recovered from tlie sag, and it did not seem to be over tlie length of a rail before slic struck another such place and she never recovered entirely from that. She seemed to go over and come part way back and then tipped right over.'” Ho adds, “ I don’t know what caused her to roll in that way, whether it was a soft spot in the track or whether she left the track. .... Whether she was on tlie track or off *268from the track, I couldn’t tell.” This is a description of the fact of the accident and the manner of its occurrence, but it altogether fails to show the cause of it. As a matter of fact the ground was frozen and no soft spot was discovered, but even if there had been, there could have been no recovery on that ground by an employee without much further testimony. The cause of this accident was entirely unexplained and it therefore can only be regarded, in an action by an employee, as one of the ordinary risks of the business for which there is no liability. The defendant’s 3d, 8th and 9th points should have been affirmed and the case withdrawn from the jury. These views sustain the 4th, 5th, 6th and 9th assignments, and also that part of the sixteenth which relates to the testimony of Stevenson and Rodman. It is not necessary to consider the remaining assignments.
Judgment reversed.