The appellee seeks to recover damages against his employer, the appellant, for injuries alleged to have resulted to him from the negligence of the employer in failing to furnish him with safe appliances for use in the performance of the duty required of him hy the service in which he was employed. The injury resulted from the collision of the car upon which the appellee was performing the duties of a brakeman, with another part of the same train, which had been detached for the purpose of making what is commonly called “ a running switch.” The car upon which the appellee was a brakeman was a platform car, laden with large and heavy blocks of stone, and the appellee was at the front end of the car endeavoring to check it by using the brake. Discovering that he was unable to do so, and that a collision was inevitable, he attempted to make his way to the rear of the car, but his feet were caught between two heavy stones and crushed. In the first paragraph of the complaint it is alleged that the accident was caused by the negligence of the appellant in failing to repair a cylinder-cock of the engine, which had been blown out some time before the *245accident, and that the failure to replace the cylinder-cock rendered it impossible for the engineer to get that part of the train which the car on which the appellee was standing was following, out of the way, and this brought on the collision. The second paragraph of the complaint charges that the brake on the car was defective, and substantially repeats the allegations of the first as to appellant’s negligence in failing to replace or repair the cylinder-cock of the engine. The third paragraph is based upon the negligence of the appellant in regard to the. brake, but it also alleges that there was some defect in the engine which was unknown to the appellee. As no question is made upon the complaint, we have given only a general outline of its allegations, which are full and explicit.
The question to which the appellant’s counsel devote the principal part of their argument arises on the ruling of the trial court in permitting the appellee to give in evidence the declarations of the engineer in chai’ge of the locomotive which was drawing the train on which the appellee was acting as a brakeman. The appellee’s counsel argue with earnestness that, even if there was error in admitting the evidence, it was harmless. This contention makes it necessary to dispose of the question as to the efiect of the evidence before considering its competency; for if it was harmless, the judgment can not be reversed for admitting it, although it was incompetent. We are satisfied that if the evidence be conceded to be incompetent, the error in admitting it was not harmless.
The appellee’s counsel assume that the error was a harmless one, even if the incompetency of the evidence be conceded, for the reason that the declarations of the engineer were proved by witnesses called to prove that he had made statements out of court contradicting those made by him on the witness stand. This position is untenable. The witnesses by whom the engineer was contradicted were im*246peaching witnesses, and their testimony went to his credibility, but it did not prove, nor tend to prove, the principal fact. Impeaching testimony goes only to the credibility of a witness, and it can not be given any force as evidence in proof or disproof of a disputed fact, except in so far as it bears upon the credibility of the witness it tends to impeach: In Seller v. Jenkins, 97 Ind. 480 (436), it was said of impeaching evidence that, “ Such evidence does not tend to establish the truth of the matters embraced in the contradictory evidence; it simply goes to the credibility of the witness.” Other cases assert a similar doctrine. Davis v. Hardy, 76 Ind. 272; Hicks v. Stone, 13 Minn. 434.
The position assumed by appellee’s counsel that, as the facts which the declarations of the engineer tended to prove were established by other testimony, the ruling in admitting evidence of such declarations, even if erroneous, was harmless, can not be maintained. There may be cases where the facts are so fully and conclusively proved by other testimony that the appellate tribunal will not reverse the judgment because incompetent evidence to the same facts is admitted; but this is not such a case, for here the evidence was as to a material point, and it can not be justly said that the facts which the declarations tended to prove were established by uncontradicted evidence.
We can not, it is evident from what we have said, avoid a decision of the principal question upon the ground that if the .evidence was incompetent it was not prejudicial. We are required to decide whether the evidence was competent, because its material character creates the presumption that it was probably prejudicial. The rule is well settled that where evidence of an influential character is erroneously allowed to go to the jury it will be presumed to have prejudiced the objecting party, and unless this presumption is rebutted, the judgment must be reversed. See authorities cited in Elliott’s Appellate Procedure, section 594, note 2.
*247It is an elementary rule that the declarations of an agent are not admissible against the principal, unless they were made while the agent was conducting some transaction for the principal, or in a matter where the agent’s act is a part of the res gestee. If the declarations of the appellant’s engineer were not part of the res gestee, there was prejudicial error in permitting them to be given in evidence.
It can hardly he affirmed that there is a general rule which will fit all cases, for each case is dependent, in a great degree, upon particular facts. It is, perhaps, safe to declare that where the declarations of the agent are made to the person whose interests are directly involved, at the place where the transaction or occurrence happened, so near the occurrence or transaction in point of time as to be justly and reasonably regarded as a part of it, refer directly to the transaction or occurrence, and are not narratives of the past, they are ordinarily to be regarded as a part of the res gestee. If the declarations are made at a different place, or are separated from the occurrence or transaction by such an interval of time as requires the inference or conclusion that they were not part of the act, transaction, or occurrence, their, under all the well reasoned cases, they are not a part of the res gestee, and can not be given in evidence against the principal. There is wide diversity of opinion and stubborn conflict as to how great an interval of time must elapse between the occurrence and the declarations, in order to deprive a party of the right to give them in evidence. But we think our general statement is supported by the weight of authority. The difficulty, as we have indicated, is not so much in formulating general statements as in determining under what phase or branch of a general rule the particular case falls. That is here the difficulty; for, while we are satisfied that our general statement is correct, we have found it no easy task to determine under what branch or phase of it this case belongs.
*248The question as to the competency of the declarations of the engineer has two branches; for there is one branch founded on specific objections interposed to the testimony, and another upon a motion to strike out part of the testimony. It will conduce to clearness to consider each branch separately, although both depend upon the effect and application of the rule relating to the competency of evidence as part of the res gestee.
The appellee testified as a witness in his own behalf; and, after giving an account of the collision and the manner in which he was injured, he said: ’ “In the meantime I was getting up. I went to walk, I went to step, and when I stepped on this foot I felt that was the first time I knew I was hurt. I reached down in the dark and felt that my foot was all cut. I crawled over to 'the car and sat down. About that time Mr. Bramley, the engineer, came to me with his torch. I was going' on like, I suppose, anybody would when he was hurt. He says: ‘ That is too bad, Bill.’ I said: ‘Yes.’ He said: ‘What was the matter, Bill? Didn’t you understand the signal, or couldn’t you get out of the road?’” At this point an objection was stated, and then followed questions and answers. Some of the questions, as indicated by the stenographer’s report, were interposed by Mr. McMullen, counsel for the appellants. The statements elicited by the questions addressed to the witness are, in substance, these: From the time the collision occurred until the engineer came to the appellee was. “not over a minute or two.” The engineer left his engine and walked back to the car where the appellee was; the engine had “gone down the track between two and three hundred feet” ahead of the car on which the appellee was at work, but when the engineer left the engine it was about a car’s length from the car on which the appellee was injured; “the engine was stopped,” and, as the witness expressed it, “the collision was all over” when the engineer reached the appellee, who had at that time *249crawled from the end of the car to the center, and was holding his foot and moaning.
It appears from this evidence that the direct collision was over and the injury done at the time the engineer reached the appellee, but that the car on which the appellee was sitting was still at the place where the injury was received. It also appears that there was an interval of time, although a very brief one, in which the engineer walked from his engine to the appellee, after the engine had been stopped. The appellee, as we have seen, estimates the time at not more than one or two minutes; so that, while there was an interval of time betweu the actual injury and the declarations, it was a brief one; but brief as it was, it was suificient to allow the engineer to traverse a short distance — a car’s length — from the place where the engine was stopped to the place where the car on which the appellee was sitting was standing. The statement of the witness,-that the engine had “gone down the track two or three hundred feet from where it was when the collision took place,” does not mean that the engine was that far from the appellee’s car when it was stopped; but what it means is, as the record shows, that the engine and the car on which the appellee was. stationed, although detached, continued in motion for that distance after the collision took place, so that the engineer did not walk that distance. On the contrary, he walked only the distance between the car and the engine, and that, as the witness says, was a car’s length. "We have, therefore, a case where there was a very brief interval between the collision and the declarations, and one in which the engineer walked only a few feet, after he had checked his engine, back to the car on which the injured person was sitting, moaning in pain.
We have no disposition to extend the rule respecting the competency of declarations of an agent, upon the ground that they constitute a part of the res gestea, for we are satisfied that an enlargement of the rule would very likely *250make it an instrument of evil; but, on the other hand, an undue limitation of. the rule might often prevent a party from availing himself of evidence to which he was entitled, and which would aid in establishing the truth. If we can ascertain the rule, as our decisions declare it, we shall deem it our duty to apply it without extending or narrowing it.
It is necessary to examine the decided cases, in order to obtain the means of solving the vexed question which faces us, and we begin this work by a reference to the cases which declare the general doctrine. One of the earliest of our cases is that of Bland v. State, 2 Ind. 608, wherein it was said:
“It has been decided that it is not competent for a prisoner indicted for murder to give in evidence his own account of the transaction related immediately after it occurred, though no third person was present when the homicide was committed.” The case from which we have quoted cites, as authority for the conclusion which it declares, the case of State v. Tilly, 3 Iredell’s L. 424, wherein it was said:
“ Unless the declarations form a part of the transaction they are not receivable in evidence.”
The case of Bland v. State, supra, has often been cited and approved; so that the doctrine of that case, in so far as it asserts that the declarations must be a part of the occurrence, act or transaction, is to be accepted as the law of this jurisdiction. The doctrine of that case, as we have stated it, is the prevailing one, for it is sanctioned by many able courts, State and Federal. Vicksburg, etc., R. R. Co. v. O’Brien, 119 U. S. 99; Durkee v. Central Pacific R. R. Co., 69 Cal. 533 (535); State v. Pomeroy, 25 Kans. 349; Michigan, etc., R. R. Co. v. Coleman, 28 Mich. 440 (446); Mayes v. State, 64 Miss. 329; Southerland v. Wilmington, etc., R. R. Co., 106 N. C. 100; Martin v. New York, etc., R. R. Co., 103 N. Y. 626; Waldele v. New York, etc., R. R. Co., 95 N. Y. 274; Lane v. Bryant, 9 Gray, 245; Luby v. Hudson River, etc., R. *251R. Co., 17 N. Y. 131; Williamson v. Cambridge, etc., R. R. Co., 144 Mass. 148; Chicago, etc., R. W. Co. v. Becker, 32 Fed. Rep. 849; Cincinnati, etc., R. R. Co. v. O’Mara, 26 Ohio St. 185; Adams v. Hannibal, etc., R. R., 74 Mo. 533; Memphis, etc., R. R. Co. v. Womack, 84 Ala. 149.
The general doctrine that the declarations must he a part of the act or occurrence is asserted without substantial diversity of opinion by the text-writers. Abbott’s Trial Ev. 51; Wood’s Pr. Ev. 469; 1 Wharton Ev. (3d ed.), section 259; Taylor on Ev. (8th Eng. ed.), section 602; 1 Rice on Ev. 375. "We assume, therefore, that the declarations of an agent or servant are not competent, unless they are part of the principal act, occurrence or transaction; but in ascertaining the general doctrine, we do not complete our work, for we have still to ascertain and decide whether the declarations of the engineer can he deemed part of the occurrence in which the appellee was injured; and, in order to reach a correct conclusion, it is necessary to examine the authorities with some care, not, however, for the purpose of ascertaining the general rule, hut for the purpose of ascertaining what the cases declare to be part of the res gestae. In Binns v. State, 57 Ind. 46, the doctrine of Bland v. State was held to govern a case where the witness reached the woman who had been shot after he had run a distance of two or three hundred yards, and arrived at the place where the shooting was done within a minute and a half after she had been wounded, and the judgment of the court was that her declarations were not part of the res gestee. A very similar application of the rule was made in Dukes v. State, 11 Ind. 557 (564). The question arose in the case of the Bellefontaine, etc., R. W. Co. v. Hunter, 33 Ind. 335, upon this state of facts: The body of the man who had been killed was on the train; it had been carried to the town of Lanesville, some miles distant from the place where the accident occurred, and the fireman of the engine Avhich ran over the deceased made statements while the *252body was being removed from the train; these statements were held to have been erroneously admitted, the court citing as authority for its conclusion Luby v. Hudson River, etc., Co., supra; Moore v. Meacham, 10 N. Y. 207; Lane v. Bryant, 9 Gray, 245. The case of Pittsburgh, etc., R. R. Co. v. Theobald, 51 Ind. 246, asserts the general doctrine that declarations of trainmen are incompetent unless made at the time of the occurrence, but it does not assert what shall be deemed a pai’t of the occurrence, nor does the opinion show how much time had elapsed between the performance of the agent’s act and the time of making the declarations. The statements of the agent, which were held rightly excluded in Pittsburgh, etc., R. W. Co. v. Wright, 80 Ind. 182, were made at a place different from where the injury was received, and thirty minutes or more after the occurrence which caused it. In Stephenson v. State, 110 Ind. 358 (372), the declarations excluded were made after the deceased had left the place where he was wounded, after the accused had left the spot, and after the deceased had gone into a saloon and remained for some time, so that a very considerable interval of time had elapsed. The court held, in Jones v. State, 71 Ind. 66 (81), that statements made by the deceased after he was shot, naming the person who shot him, and narrating a past occurrence, were not competent. So, in Doles v. State, 97 Ind. 555, the evidence was held to be properly excluded because, as the court said, “ It was merely a narrative of a past transaction, and not part of the res gestee.”
Rone of the cases we have cited precisely fits the one before us as to the point under immediate discussion. In the class represented by such cases as Burns v. State, one of the actors in the occurrence was absent, so that the declarations could not have been a part of the res gestee. We suppose it clear that where one of the principal actors in a transaction goes from the place where the transaction took place, what subsequently occurs can not ordinarily be re*253garded as part of the res gestee. In the class of cases of which the Bellefontaine, etc., R. W. Co. v. Hunter is a type, the declarations were made at a place and at a time different from that at which the transaction took place. The case at our bar differs from those cited in essential particulars, for here the declarations were made at the time and place where the collision they referred to occurred, and they illustrated the event, and wore made while all who participated in it were present. We may, therefore, well adjudge that there was no error in overruling the appellant’s objections without denying the doctrines asserted in our cases.
The latest decision of our court upon the question before us is that given in the case of the Louisville, etc., R. W. Co. v. Buck, 116 Ind. 566. In that ease the conductor of the train on which the intestate of the plaintiff' was employed as a brakeman was on the “ caboose ” when he received notice that the deceased had been injured while coupling cars, and he immediately ran forward and found the deceased under the rear end of the second car from the engine. The conductor, -when he took the deceased from under the car, asked, “ Ilow did this happen ? ” and the deceased fully described the cause of the accident. The court held that this testimony was competent, and cited many cases in support of its conclusion. We think the •doctrine declared in that case decides the point here under direct consideration against the appellant.
Counsel argue, with plausibility, that the doctrine of the case cited does not apply to the case before us. One of the reasons assigned in support of their position is, that the declarations admitted in that case were those of the injured person, while the declarations admitted in this instance were those of the agent or servant. A complete and effective answer to this argument is that if the declarations were, as the case referred to adjudges, part of the res gestee, they were competent, no matter by. whom they were made. *254Baker v. Gausin, 76 Ind. 317; 1 Whart. Ev. (3d ed.), sections 259, 260, 261.
Our conclusion receives support from the familiar rule that where part of a conversation is competent the whole is admissible, unless some part of it is excluded by other rules of law.
In the case before us, the interval of time that elapsed between the actual injury and the time of making the declaration is not so great as it was in Louisville, etc., R. W. Co. v. Buck, supra. Here, as there, the declarations were made while the injured man was at the place where he was hurt, and the declarations were made to him by one who had taken part in the thing done. Here there is even a clearer and stronger line of causal connection between the direct-injury and the declarations of the agent, or servant, than there was in the case of Louisville, etc., R. W. Co. v. Buck.
In Stephenson v State, supra, it was said: “The time is not always so essential; ” and so we say here. The brief time that elapsed after the engineer stopped his engine and reached the ear where the appellee was sitting was not so essential as to break the line of connection that binds the acts together. We are strongly inclined to the opinion that where an employe is injured in a collision, all that is done toward stopping the train and relieving the injured employe from a dangerous position, forms part of one occurrence,but without authoritatively affirming this, we do affirm that,, where there is such a continuous chain of acts and events as there was in this case, all are part of the res gestee.. It is true, as Mr. Wharton says, that “Immediateness is tested by closeness, not of time, but of causal relation.” 1 Whart. Ev. (3d ed.), section 262.
-This conclusion we regard as involved in the principle thus stated in the case of the Louisville, etc., R. W. Co. v. Buck:
“It is not always easy to determine when declarations having reference to an act or transaction should be received *255as part of the res gestee, and much difficulty lias been experienced in the effort to formulate general rules applicable to the subject. This much may, however, be safely said, that declarations which are the natural emanations or outgrowths of the act or occurrence in litigation, although not precisely concurrent in point of time, if they were yet voluntarily and spontaneously made, so nearly contemporaneous as to be in the presence of the transaction which they illustrate and explain, and where made under such circumstances as necessarily to exclude the idea of design or deliberation, must, upon the clearest principles of justice, be admissible as part of the act or transaction itself.”
Our conclusion is that there was no error in admitting declarations of the engineer that did not refer to acts done, or matters which happened prior to the collision which caused injury to the appellee.
The other branch of the general question of the competency of the declarations of the engineer, that resting on the motion of the appellant to strike out, requires only very brief mention. The motion asked the court to strike out the statement of the appellee that the engineer said: “If that man, last night, would have fixed that cylinder-cock as I told him, you would never have been hurt.” This declaration related to the past, and was a narrative of what had been done at an entirely different time and place. It was, indeed, a combination of an opinion and a narrative of things that had passed, for it was a statement of the engineer’s opinion that if, on the night before, something had been done which he had then directed, the collision could not have taken place. It is too well settled to excuse the reference to authorities that neither narratives of past occurrences, nor matter of opinion, can be placed before a jury by proving the declarations of an agent or servant. IF or the error in overruling the motion to strike out the objectionable testimony, the judgment must be reversed; and, although many other questions are discussed by counsel, *256we do not deem it necessary to consider or decide them, as-they may not arise oh another trial.
Filed May 14, 1892.Judgment reversed.