CARTER, J.

Plaintiff, as administratrix of the estate of Joseph W. Showalter, deceased, brought this action to recover damages for the death of said deceased. Upon trial by jury, plaintiff had judgment in the sum of $18,000. Defendant appeals from said judgment and from the order denying its motion for judgment notwithstanding the verdict.

Plaintiff's complaint contained two counts. The first was predicated on negligence under the Federal Employers' Liability Act, 45 U.S.C.A. 51, et seq., and the second charged violation of the Safety Appliance Act, 45 U.S.C.A. 1, et seq. It was alleged in the complaint and admitted by the answer that deceased was a brakeman in the employ of the defendant; that both deceased and the defendant were engaged in interstate commerce at the time of the accident; that it was the duty of deceased at about 9 o'clock p. m. on the night in question to ride upon two freight cars in a switching operation at defendant's yard at Keddie, California, in which switching operation said cars were being dropped by the engine onto a track known as lower four in said yard; and that in said switching operation, deceased received the injuries which resulted in his death. It was alleged in the first count and denied by the answer that defendant "carelessly and negligently operated and controlled said cars" thereby causing said injuries and death. During the course of the trial plaintiff was permitted to amend by adding the words "and engine". It was alleged in the second count and denied by the answer that the handbrake on one of said cars was defective and that as a result of its defective condition said handbrake "suddenly and unexpectedly released and let go" thereby throwing deceased upon the tracks and causing said injuries and death. Defendant's answer affirmatively alleged that negligence of the deceased was the sole proximate cause of his death; that deceased was guilty of contributory negligence; that deceased assumed the risk.

At the close of plaintiff's case, defendant made a motion for nonsuit as to the second count. Said motion was granted, as plaintiff presented no evidence of any violation of the Safety Appliance Act. Defendant then rested, without offering any evidence, and moved for a directed verdict on the first cause of action. The motion was denied and the cause was submitted to the jury upon the evidence introduced by the plaintiff. The jury returned a verdict in favor of the plaintiff and judgment was entered thereon. Thereafter, defendant made a motion for judgment notwithstanding the verdict. Said motion was denied and this appeal followed.

The questions presented on this appeal are whether the evidence introduced by the plaintiff was sufficient to support a verdict in her favor; whether certain portions of said evidence were admissible; and whether the court erred in its instructions to the jury.

Before reviewing all the evidence with a view to determining its sufficiency, the question of whether certain portions thereof were properly received as evidence must be considered.

Over the objection of the defendant, there was admitted into evidence a statement made by the deceased to Parrott, a fellow brakeman. Parrott did not see the accident, but ran to the side of the deceased when he heard him "holler". He found the deceased lying across the tracks under the car which had run over him, cutting off his legs. Upon being asked if he then had any conversation with the deceased, Parrott answered, "I did--as I remember, I asked him, 'How in the world did it happen, Joe?' "--"He said, 'I got knocked off.' " Being asked if that were all the conversation, Parrott replied, "Well, he said, 'I am all done, both legs cut off.' "

It is plaintiff's position that these statements of the deceased were properly admissible as part of the res gestae.

Defendant contends that in actions under the Federal Employers' Liability Act substantive rights of the parties are governed by the act as interpreted by the federal courts; and that where an action is brought in a state court the law of such state may be applied to matters of practice and procedure. Asserting that rules as to the admissibility of evidence are matters of procedure, defendant concludes that under the law of the state of California the statements here in question are not part of the res gestae and that such has been the rule ever since the decision in the case of People v. Ah Lee, 60 Cal. 85 .

Plaintiff concedes that an action brought in a state court under the Federal Employers' Liability Act is governed as to substantive rights by the law as interpreted by the federal courts, and as to matters of practice and procedure by the state courts; but urges that the admissibility of the res gestae declaration was not purely a procedural matter to be governed by state law.

Plaintiff points to the statement in the case of Chicago, M. & St. P. R. Co. v. Coogan, 271 U.S. 472 [ 46 S.Ct. 564 , 70 L.Ed. 1041 ], that "proof of such negligence is necessary to recovery. The kind or amount of evidence required to establish it is not subject to the control of the several states". (Italics ours.)

Citing numerous authorities, plaintiff reasons that the "kind of evidence" goes to substance and not to procedure; that where a matter of substance is involved, a substantial right cannot be taken away under the guise of a rule of procedure; that the res gestae declaration is clearly admissible under the federal rules; that the res gestae statement here was the keystone of plaintiff's proof necessary to sustain her burden of proof; that any decision permitting or excluding it had to do "with the kind of evidence", hence no state court could take away plaintiff's right to produce it; that to permit its rejection would deprive plaintiff of a substantial right under the flimsy guise of procedure.

[1] We are not convinced by this argument. As a practical matter every ruling as to evidence affects to a greater or lesser degree the ultimate result of an action. Therefore, the fact that a particular piece of evidence is vital to a party's burden of proof will not alone transform it from a matter of procedure into a matter of substance. In cases brought under the Federal Employers' Liability Act it has been held that rules of evidence are governed by the law of the state in which the suit is brought. (Central Vermont R. R. Co. v. White, 238 U.S. 507, 511 [ 35 S.Ct. 865 , 59 L.Ed. 1433 , Ann. Cas. 1916B, 252]; New Orleans etc. Co. v. Jackson, 145 Miss. 702, 710 [ 110 So. 586 ]; Fleming v. Norfolk S. R. Co., 160 N. C. 196 [ 76 S.E. 212, 215 ]; Kansas City So. Ry. Co. v. Leslie, 112 Ark. 305 [ 167 S.W. 83, 91 , Ann. Cas. 1915B, 834].)

We are, however, of the opinion that the statement in question should be admissible under the rule of res gestae as applied in this state.

[2] Courts in general have been in considerable confusion as to the rule of res gestae. In this respect the courts of this state are not different. The rule which we deem to be correct was enunciated in the early case of People v. Vernon, 35 Cal. 49 [ 95 Am. Dec. 49 ]--namely, that declarations which are voluntary and spontaneous and made so near the time of the principle act as to preclude the idea of deliberate design, though not precisely concurrent in point of time therewith, are regarded as contemporaneous and admissible.

Numerous cases decided subsequently, however, failed to follow this decision and have erroneously held that the test of whether a statement is res gestae is not its proximity to principal act in point of time, but its utterance before the principal act itself has been completed. Of the cases adhering to this latter rule, the following are representative: People v. Ah Lee, 60 Cal. 85 ; Durkee v. Cent. Pac. R. R. Co., 69 Cal. 533 [ 11 P. 130 , 58 Am. Rep. 562 ]; Murphy v. Board of Police etc. Commissioners, 2 Cal.App. 468 [ 83 P. 577 ]; People v. Murphy, 28 Cal.App. 708 [ 153 P. 732 ]; Salvo v. Market St. Ry. Co., 116 Cal.App. 339 [2 PaCal.2d 585]; Zohner v. Sierra Nevada Life & Casualty Co., 114 Cal.App. 85 [ 299 P. 749 ]; Willard v. Valley Gas & Fuel Co., 171 Cal. 9 [ 151 P. 286 ]; People v. Wong Ark, 96 Cal. 125 [ 30 P. 1115 ].

The rule which they have applied is perhaps expressed most clearly in Williams v. S. P. Co., 133 Cal. 550 , at page 554 [ 65 P. 1100 ], where the court said, "Expressions of persons who are actors, made during the occurrence, may generally, but not always, be proved. If spontaneous, and caused by the event, they may nearly always be shown. But if, afterwards, no matter how shortly afterwards, there is an attempt to explain what has happened, or to account for it, or to defend oneself, or the like, it is incompetent, and inadmissible as res gestae." The court in said case stated also that "A narrative, even if given during the occurrence is inadmissible" for the reason that it is not a spontaneous outgrowth of the occurrence. As its authority for this statement the court cited Heckle v. S. P. Co., 123 Cal. 441 [ 56 P. 56 ].

A few cases in this jurisdiction have properly refused to apply the rule as thus stated and have negatived both time and completion of the principal occurrence as tests for determining admissibility. The test so laid down is whether or not the statement is made under such circumstances of physical shock or nervous excitement as preclude the likelihood of reflection and fabrication. The better view as set forth in the case of Lloyd v. Boulevard Express, 79 Cal.App. 406 , at page 411 [ 249 P. 837 ], and quoted with approval in the case of Coryell v. Clifford F. Reid, Inc., 117 Cal.App. 534 , at page 536 [4 PaCal.2d 295], is as follows:

" 'His statements, if true, indicate that they were made before the excitement attending the accident had subsided. Questions as to the admissibility of such questions as part of the res gestae depend upon whether such statements are the emanations of the acts done, before the excitement usually incident to an accident causing great injury has subsided, or are the statements made after the excitement has subsided, and the witness has time to calculate a policy to be pursued. Here the statements were made at the scene of the accident, and in the presence of those suffering from its results, and in the presence of the wreckage the accident created. Again we hold that the trial court before whom the witness appeared is better prepared to pass upon the evidence upon which the admissibility of such statements depends than is an appellate court, and his ruling should not be disturbed unless he has abused his discretion. We cannot say from the evidence appearing in the record that the declarations of Mitchell were inadmissible. If the statement of the employee is spontaneous, that is, made at the time of the accident and during the excitement attending it, before the employee has time to reflect upon the consequences of his statement, or to fabricate evidence against his employer, it is admissible against the employer.' (See, also, 10 R. C. L. 978, and 4 R. C. L., Perm. Supp., p. 2801.)"

Plaintiff has argued in the instant case that the statement in question was made before the completion of the occurrence and therefore admissible. As authority she points to the case of Heckle v. S. P. Co., supra, wherein this court held that it was error for the trial court to have refused to admit a statement of the decedent made while he was still under the wheel of the car which had run over him. It did not appear with any certainty exactly how long the statement was made after the deceased had been first caught by the wheel, but it did appear that the time was short. The court pointed out, however, that the time may have been so long that the statements of the deceased took on the character of narrative rather than spontaneity, and said that if that appeared to be the case it would be the duty of the trial court to strike them out.

Under the authority of this case it might be possible to regard the deceased's statement herein as having been made prior to the completion of the occurrence, nevertheless the character of the utterance would still remain in question.

We do not propose to place our decision on this basis, for in our opinion the rule to be followed is that where a declaration is made under the immediate influence of the occurrence to which it relates and so near the time of that occurrence as to negative any probability of fabrication, said declaration is admissible. Measured by this standard we conclude that the ruling of the trial court admitting the statement of the deceased herein was correct.

The cases cited hereinabove to the contrary, holding the view that the admissibility of such statements is dependent on proximity of time to the principal event or on the continuation of the occurrence to which they relate, are as a consequence hereby overruled.

The position of the courts as a whole with regard to declarations allegedly a part of a transaction is set forth briefly in 35 Harv. L. Rev., at p. 447, in the following manner:

"There are two theories on the admissibility of various declarations which are alleged to be part of the res gestae, or as James Bradley Thayer preferred to say, res gestae (transaction). Thayer's view is that the declarations must be substantially contemporaneous with the action of a material transaction, and must explain the acts. He bases their admission on the need of telling a complete story of the event. Wigmore's view goes farther and does not require contemporaneousness with an event. He favors the admission of statements or exclamations by an injured person, or by any other person present at any exciting occasion, immediately after the event, describing the circumstances as observed by him. In the cases the word 'immediately' undergoes a considerable stretching. The confusion in the decisions is illustrated by the position of the United States Supreme Court, which supported Wigmore's view in Insurance Company v. Mosley, and Thayer's view in Vicksburg etc. R. R. v. O'Brien. Recent discussion in the cases and law reviews inclines toward Wigmore's view. ... On either view, the declarations are excluded if the event has terminated and the nervous excitement has been succeeded by reflective powers."

[3] Professor Wigmore points out that to be admissible it is not necessary that such declarations be made precisely at the same time as the occurrence to which they relate, as is the case under the "Verbal Act" doctrine. These declarations are admissible not because they fall without the hearsay rule, as is the case with "verbal acts", but because they fall within an exception to that rule. The foundation for this exception is that if the declarations are made under the immediate influence of the occurrence to which they relate, they are deemed sufficiently trustworthy to be presented to the jury. (Wigmore on Evidence, [2d ed.], sec. 1747 et seq., and cases cited.)

The basis for this circumstantial probability of trustworthiness is "that in the stress of nervous excitement the reflective faculties may be stilled and the utterance may become the unreflecting and sincere expression of one's actual impressions and belief". [4] To render them admissible it is required that (1) there must be some occurrence startling enough to produce this nervous excitement and render the utterance spontaneous and unreflecting; (2) the utterance must have been before there has been time to contrive and misrepresent, i. e., while the nervous excitement may be supposed still to dominate and the reflective powers to be yet in abeyance; and (3) the utterance must relate to the circumstance of the occurrence preceding it. (Wigmore on Evidence, [2d ed.], sec. 1750.)

[5] The practice in this state heretofore has been to allow the trial court no discretion in determining the admissibility of alleged res gestae statements; however, under the view taken herein there is necessarily some element of discretion in the trial court.

As stated by the court in Roach v. Great Northern Ry. Co., 133 Minn. 257, 260 [ 158 N.W. 232 ]: "In passing upon the admissibility of testimony claimed to constitute a part of the res gestae, the trial court determines whether unsworn statements are so accredited that they may go to the jury and be weighed and valued by it, and in determining this it considers whether the statements are spontaneous, whether there was an opportunity of fabrication or a likelihood of it; the lapse of time between the act and the declaration relating to it; the attendant excitement; the mental and physical condition of the declarant, and other circumstances important in determining whether the trustworthiness of the statements is such that they may safely go to the jury."

[6] The circumstances surrounding the declarations of the decedent herein are such that we cannot say that the court erred in admitting them. The deceased had suffered a severe physical and mental shock; although the exact time between his receiving the injury and the utterance does not appear, it is apparent that the time was short; the statement was made while he was still lying midway between the trucks of the car which had run over him; the words were spoken almost in the same breath in which he said he "was done for"; the statement related to the circumstances causing the shock; it is therefore most reasonable to suppose that the utterance was spontaneous, unreflected, and dominated by the occurrence which had just preceded it.

A case so closely paralleling the case at bar that it requires comment is that of Chesapeake and Ohio R. Co. v. Mears, 64 Fed.2d 291 (certiorari denied, 293 U.S. 557 [ 55 S.Ct. 69 , 79 L.Ed. 659 ]). In said case, it was alleged that Mears was knocked from the top of a car, on which he was walking by another car placed too close on an adjoining track. No one saw the accident. He was last seen by witness Mills, and when the latter returned about ten minutes later, he found the deceased lying near track three with both legs cut off above the knees and rapidly bleeding to death. Mills asked what had happened, and deceased, pointed to the car on track three, replied. "That car knocked me off."

The court therein held the statement admissible and said:

"While a logical basis is not always given for the decisions, the cases almost uniformly hold that a statement made by an injured person as to the cause of his injury is admissible if the time which has elapsed since the injury is so short that he is still under the influence of the happening and his statement presumptively a spontaneous expression growing out of it and not the result of reason and reflection."

Said court also cited with approval the early case of Travelers' Ins. Co. v. Mosley, 8 Wall. 397 [ 19 L.Ed. 437 ].

We think the following statement made by Mr. Justice Swayne, at page 408 in the latter case worthy of repetition here. Said he:

"Here the principal fact is the bodily injury. The res gestae are the statements of the cause made by the assured almost contemporaneously with its occurrence, and those relating to the consequences made while the latter subsisted and were in progress. Where sickness or affection is the subject of inquiry, the sickness or affection is the principal fact. The res gestae are the declarations tending to show the reality of its existence, and its extent and character. The tendency of recent adjudications is to extend rather than to narrow, the scope of the doctrine. Rightly guarded in its practical application, there is no principle in the law of evidence more safe in its results. There is none which rests on a more solid basis of reason and authority. We think it was properly applied in the court below. In the ordinary concerns of life, no one would doubt the truth of these declarations, or hesitate to regard them, uncontradicted, as conclusive. Their probative force would not be questioned. Unlike much other evidence, equally cogent for all the purposes of moral conviction, they have the sanction of law as well as of reason. The want of this concurrence in the law is often deeply to be regretted. The weight of this reflection, in reference to the case under consideration, is increased by the fact, that what was said could not be received as 'dying declarations', although the person who made them was dead, and hence, could not be called as a witness."

Defendant complains of certain rulings made by the trial court with regard to the admission of certain expert testimony. We have examined these rulings and find them to be without prejudicial error.

We come now to the question of sufficiency of the evidence.

[7] It is well settled and the parties do not dispute the facts that the liability of a carrier under the Federal Employers' Liability Act (U.S.C.A., tit. 45, sec. 51) is predicated on negligence; that the burden of proving such negligence rests on the plaintiff in any action brought under said act; and that the sufficiency of the evidence to support a verdict is a question of substantive law as interpreted by the federal courts. (Chicago, M. & St. P. R. Co. v. Coogan, supra.)

[8] Likewise, it is clear under the federal decisions as under the decisions of this state that no mere scintilla of evidence but substantial evidence must be produced by the plaintiff in order to sustain the burden of proof. (Pennsylvania R. Co. v. Chamberlain, 288 U.S. 333 [ 53 S.Ct. 391 , 77 L.Ed. 819 ]; Small Co. v. Lamborn & Co., 267 U.S. 248 [ 45 S.Ct. 300 , 69 L.Ed. 597 ]; Gunning v. Cooley, 281 U.S. 90 [ 50 S.Ct. 231 , 74 L.Ed. 720 ]; Southern Ry. Co. v. Walters, 284 U.S. 190 [ 52 S.Ct. 58 , 76 L.Ed. 239 ]; Chicago G. W. R. Co. v. Rambo, 298 U.S. 99 [ 56 S.Ct. 693 , 80 L.Ed. 1066 ].) The evidence produced to show that negligence on the part of the carrier was the proximate cause of the accident may be either direct or circumstantial, but it must be substantial. Evidence which leaves the determination of these essential facts in the realm of mere speculation and conjecture is insufficient. (Pennsylvania R. Co. v. Chamberlain, supra; Chicago, M. & St. P. R. Co. v. Coogan, supra; Atchison, T. & S. F. R. Co. v. Toops, 281 U.S. 351 [ 50 S.Ct. 281 , 74 L.Ed. 896 ]; Northern R. Co. v. Page, 274 U.S. 65 [ 47 S.Ct. 491 , 71 L.Ed. 929 ]; St. Louis S. F. R. Co. v. Mills, 271 U.S. 344 [ 46 S.Ct. 520 , 70 L.Ed. 979 ]; Kansas City So. Ry. Co. v. Jones, 276 U.S. 303 [ 48 S.Ct. 308 , 72 L.Ed. 583 ]; Patton v. Texas & P. R. Co., 179 U.S. 658 [ 21 S.Ct. 275 , 45 L.Ed. 361 ].)

So far as the record discloses there were no eye-witnesses to the accident in question. Necessarily, therefore, proof of the facts must be based on circumstantial evidence alone. Three other employees of defendant were engaged in the switching operation in question. The defendant called no witnesses, and the plaintiff called only the brakeman, Parrott. There is no conflict in the evidence as the cause was submitted at the close of plaintiff's case. The following facts are either admitted facts or facts to which the witness Parrott testified.

The accident in question took place in the Keddie yard of defendant. The yard consists of four tracks, numbered one to four, from south to north, which run in an easterly and westerly direction parallel to the main track. The yard is on a one per cent grade extending from east to west. A series of tracks and switches known as the "snake-lead" connect the tracks. This "snake-lead" perhaps derives its name from the continuation of figure "S" curves it makes as it curves away from each track and straightens out again to cross the next one. By means of these "cross-overs" it is possible to run trains from the main line onto any one of the four tracks, or from any one of said tracks to another. The portion of each track to the east of the snake- lead is designated "upper", while the portion to the west is designated "lower".

The particular switching operation in which the crew--consisting of the deceased, a brakeman; Parrott, a brakeman; the engineer and fireman--was engaged involved three cars and the engine. At the beginning of the operation the engine and three cars were on lower track one. The order of the cars and engine from east to west was as follows: a water car, a gondola car, a refrigerator car and the engine. The pilot of the engine was headed east. In railroad parlance a gondola car is referred to as a "gon" and a refrigerator car is referred to as a "reefer". The object of the switching operation was to place the gon and the reefer on lower four coupled to a cut of twelve cars which were then located on lower four; and to place the water car on upper four. The method to be pursued in accomplishing this object was as follows:

The engine was first to push the three cars to upper one so that the entire unit would be clear of the switch located on track one; the water car was then to be uncoupled, left standing on upper one; the engine was then to give the reefer and gon a slight start and then uncouple and run back onto lower one; the switch to the snake-lead was to be thrown as soon as the engine crossed said switch and the reefer and gon were to be permitted to roll down across the snake-lead to a coupling with the twelve cars located on lower four; the switch on track one was then to be lined so that the engine could proceed from lower one to upper one and couple onto the water car; the switch on track one was to be then thrown again so that the engine and water car could proceed down the snake-lead to lower four from which point the engine was to push the water car into upper four. The cars were all approximately forty feet in length. It was apparently known to all members of the crew that the easterly end of the cut of twelve cars, standing on lower four was approximately one hundred feet west of the switch on track four at the end of the snake-lead. With the reefer and the gon added thereto, there would remain only approximately twenty feet between the easterly end of the gon and said switch, which would obviously leave insufficient space for the engine and water car to enter lower four before entering upper four without first coupling onto the fourteen cars, including the original twelve and the reefer and gon, and pushing them a short distance down lower four.

The deceased was the senior man of the crew and was in charge of this switching operation. He assigned Parrott to handle the switch on track one and to remain with the engine and water car during the balance of their movement. Deceased was to uncouple the engine from the reefer and gon after the engine had given them a rolling start on upper one and was then to ride said cars across the snake-lead and to a coupling with the twelve cars on lower four. The night was dark and it appears that the movement was intended to be and was carried out at a slow rate of speed. The deceased when last seen was rolling down with the two cars at a speed of perhaps two miles per hour. Parrott followed with the engine and water car, the pilot of the engine and its headlight being pointed away from the direction of the movement down the snake-lead to track four.

The first part of the movement was carried out without any unexpected happening. The deceased uncoupled the reefer and the gon from the engine on track one and climbed to the top of the reefer, which was the first of the two cars, as said two cars rolled slowly by the switch on track one and onto the snake-lead for the purpose of rolling over the snake-lead and coupling with the cars on lower four. The duty of deceased was to control said cars by means of the handbrake and to bring them to a normal coupling. He was last seen before the accident on top of the reefer at the west end thereof as the two cars passed Parrott who was standing at the switch on track one. Said cars rolled by and the balance of the crew proceeded with their duties in completing the switching operation.

Parrott then lined the switch on track one so that the engine could again proceed from lower one to upper one. The engine did so proceed and he rode the engine pilot to the place the water car was standing on upper one and coupled the water car to the pilot of the engine. He then went back to the switch on track one and threw said switch so as to permit the engine to proceed from upper one across the snake-lead to lower four. After throwing the switch, Parrott got on the footboard at the end of the tender, which was in front of the engine, at the left side or engineer's side, which was his proper position. The engine then proceeded to back over the snake-lead. Parrott did not see the lighted lantern of the deceased and there was nothing to prevent him from seeing it if it had been on the top of the cars.

While the engine was still backing and when the front of the tender upon which Parrott was riding had arrived at or just west of the switch on lower four track, Parrott first heard the deceased "holler". It was a holler as if "someone were in pain or distress in some way". If deceased had been "hollering" before that time, Parrott did not believe he could have heard him "because the engine was rolling, and because of the distance I was from him". When he heard the holler Parrott testified they were near the gondola, but whether they had made a coupling or not he couldn't remember. Parrott ran across the track to the north side and found deceased lying across the north rail of track four underneath the gon, about halfway between the trucks of the gon, possibly a little closer to the east end. He was about forty or fifty feet from the cross-over switch on track four.

The two cars had rolled down to a coupling with the cars on lower four. Deceased's head was toward the north and his feet were toward the inside of the track. He had been run over and while he was still conscious spoke to Parrott when Parrott arrived. He died shortly thereafter as the result of his injuries.

The remaining testimony throws but little light on the manner of the happening of the accident. Parrott was qualified as an expert and gave certain testimony regarding the proper and customary manner of riding cars under the circumstances which deceased rode the two cars. He testified that the proper place for a man to ride under the circumstances was on top of the reefer, the leading car, near the brake which was at the west end of said reefer. It is assumed by the parties, however, from other evidence that deceased fell to the ground between the east end of the reefer and the west end of the gon. This assumption seems to be made because apparently only the westerly trucks of the gon ran over the deceased; his lantern was found between the rails near his feet; there were certain marks found along the east end of the reefer; and on the following morning deceased's brake club was found by Parrott in the grab-iron at the west end of the gon.

It appears that the movement of the cars could have been controlled from the brake platform of the gon and, of the deceased, Parrott said, "If he found that the movement could be made properly, it was his discretion to take any position on the cars that he deemed consistent with the safe movement of the cars. That would be entirely up to him, as he was riding the cars. If he wanted to control the movement of the cars from the handbrake on the gondola car, that would be also up to him." This statement was corroborated by other expert testimony. Parrott further testified, however, that "a man could not see very easily where he was going on the brake platform of a gondola"; and "I do not know of anything that would make it advisable to do so. I do not know of any reason at all why he should have done so."

[9] It is the contention of the defendant that where the only evidence of negligence is circumstantial, an inference of liability can be drawn only where this is the only reasonable inference to be drawn. A reading of the cases cited by the defendant as its authorities, reveals that they support the correct rule that it is where the evidence tends equally to sustain either of two inconsistent propositions that neither of them can be said to have been established by legitimate proof. (Wheelock v. Freiwald, 66 Fed.2d 694-698; Pennsylvania R. Co. v. Chamberlain, supra; United States etc. Co. v. Des Moines Natl. Bank, 145 Fed. 273, 279 .) The rule by which a verdict is measured in order to determine its ability to stand is laid down in Wheelock v. Freiwald, supra:

"A verdict cannot be permitted to stand, which rests upon conjecture, surmise, or speculation, but plaintiff must produce substantial affirmative proof that the negligence of the carrier caused the injury, and 'where proven facts give equal support to each of two inconsistent inferences; in which event, neither of them being established, judgment, as a matter of law, must go against the party upon whom rests the necessity of sustaining one of these inferences as against the other, before he is entitled to recover'."

Although the defendant in this case offered no evidence it contends that that produced by the plaintiff showed no negligence on the part of the defendant, but, on the contrary indicated that plaintiff fell from the cars while engaged in a normal switching operation.

A contention much the same was made in the recent similar case of Weiand v. S. P. Co., 34 Cal.App.2d 500 [93 PaCal.2d 1023]. The court at page 558 therein stated that the rule in the Wheelock case quoted above is predicated upon the theory that neither of two inconsistent inferences has been established; that "If plaintiff has failed to make out a case for the jury on either of two theories, no recovery can be had. If, however, plaintiff has proven sufficient facts to justify a verdict upon one theory, the fact that there may be one or more other seemingly rational explanations of the episode in no manner precludes a recovery or invalidates the verdict. These are mere matters of argument to be presented to the jury."

The test, as laid down in another recent decision, Sherman v. Southern Pacific Co., 34 Cal.App.2d 490 [93 PaCal.2d 812], is "whether or not the inferences which may be drawn from the evidence to sustain the verdict are reasonable".

[10] It is admitted that the deceased was acting within the scope of his employment when the accident occurred. The evidence is undisputed that Parrott did not see deceased's lantern, and that to couple onto cars upon which a fellow employee is riding without knowing his whereabouts is a violation of railroad custom. Violation of a known custom admittedly constitutes negligence. (St. Louis & S. F. Ry. Co. v. Jeffries, 276 Fed. 73 , p. 75 and cases cited.)

The sole question then is whether or not plaintiff proved sufficient facts so that the jury might reasonably infer that the injury to the deceased was proximately caused by the negligence of the defendant and that the facts proved do not equally support another theory as to the cause of the death of the deceased.

It is plaintiff's theory of the case that the deceased was riding the reefer and gon slowly down to a coupling on track four; that since he was found approximately halfway between the trucks of the gon and only its westerly set had run over him, he was evidently operating the brake on the westerly end of the gon when he fell; the facts that his lantern was found at his feet and his brake club in the grab-iron of the gon at its west end support this premise. From the fact that he was found approximately forty feet from the switch on track four, and the car was forty feet long, it is evident that the end of said car was at or near the switch when deceased fell; the train crew intended to make a coupling with the cars on which deceased was riding; Parrott testified only that he did not remember whether the rear of the engine coupled (plaintiff asserts that in railroad parlance this term includes any contact and the briefs of both parties seem to use it in this manner) or collided with the westerly end of the gon although he was riding on the side of said engine; Parrott heard the deceased holler when the engine was still moving; he heard said holler when the engine was at or just west of the switch on track four. Plaintiff contends that these facts considered together with deceased's statement that "I got knocked off", and Parrott's testimony that he could not see deceased's lantern, clearly demonstrate that the cause of the death of the deceased was that the crew of the engine, contrary to admitted custom and practice, coupled with the gon at a time when the crew of the engine was not aware of the exact whereabouts of their fellow employee, the deceased.

In our opinion the facts presented reasonably support the theory of the plaintiff. Correspondingly, we are unable to find that said facts equally support the theory of the defendant that the deceased fell from the train through unavoidable accident for which neither he nor defendant were at fault or through his own negligence. The facts presented show that deceased was "knocked off" at a time when the end of the gondola car was at the cross-over switch on track four. Defendant suggests that the jostling of the car as it rolled over the switch might have caused deceased to fall; that deceased may have tripped or lost his balance for any number of reasons. No evidence appears to substantiate any of these theories. Furthermore, deceased was an experienced brakeman and it is unlikely that the normal movement of the cars alone would cause him to fall.

[11] As was said by the court in Wabash Screen Door Co. v. Black, 126 Fed. 721 , 725: "Doubtless a jury ought not to be permitted to speculate, in the sense of guess between causes, when no reasonable explanation of the injury can be found in the testimony. But, in the absence of direct testimony, the simple suggestion of theories by the defense does not reduce the jury to mere speculation, and disqualify it from determining the cause of the injury complained of. The theories suggested may be forced and fanciful, finding no reasonable foundation in the facts proved. They may be explanations which do not explain; which the common sense of the jury, when applied to the testimony, would instantly reject." (Italics ours.)

The jury may not base a verdict on mere speculation, but neither should this court indulge in mere speculation as to a possible defense unsupported by any proven facts. (Sears, Roebuck & Co. v. Peterson, 76 Fed.2d 243, 247.) The trial court submitted to the jury the question as to whether defendant, or its employees, collided or coupled with the cars on which deceased was riding, and whether this constituted negligence proximately causing the death of the deceased. We are of the view that it cannot be said that there is no substantial evidence to sustain the verdict on these issues.

[12] Plaintiff contends that the jury could properly accept and properly reject portions of the testimony of the witness Parrott. Parrott, although not a party to this action, is an employee of the defendant. Furthermore, expert testimony was given at the trial to the effect that the brakeman who signals the engineer to move the train or a portion of it is responsible for the movements thereof. If, therefore, the deceased's death were caused by defendant's negligence, Parrott would be the employee at fault. Consequently, although not a party to the action he could be deemed interested in the result. Generally, the credibility of a witness who is interested in an action is for the jury. If the evidence is possible of contradiction in the circumstances, if its truthfulness or accuracy is open to a reasonable doubt upon the face of the case, and the interest of the witness furnishes a proper ground for hesitating to accept his statements, it is a necessary and just rule that the jury should pass upon it. (Chesapeake & Ohio Ry. v. Martin, 283 U.S. 209, 218 [ 51 S.Ct. 453 , 75 L.Ed. 983 ].) Certainly the same rule should apply here where the witness gave a clear, definite and exact account of the happenings immediately preceding the accident, but could not remember whether the rear of the engine on the side of which he was riding collided or coupled with the cars on which the deceased was riding, and said that nothing "unusual" happened. It has been held in this state that where the testimony of a witness is itself contradictory and inconsistent, the jury may believe certain portions thereof and reject others. (Firth v. Southern Pacific Co., 44 Cal.App. 511 [ 186 P. 815 ].) The verdict of the jury in this case indicates that the jury believed a portion of the witness' testimony to be contradicted by the facts and hence were justified in disregarding it.

[13] It has been held in other cases brought under the Federal Employers' Liability Act that the jury is authorized to accept part of the testimony of certain witnesses and reject other portions thereof. (Mobile & O. Ry. Co. v. Hedgecoth, 215 Ala. 291 [ 110 So. 44 ]; Southern Ry. v. Newman, 187 Ga. 132 [ 199 S.E. 753 ].) The jury is not permitted arbitrarily to disregard the uncontradicted testimony of an employee of a railroad company, but where such evidence is contradicted by proven facts as well as where it is contradicted by the testimony of other employees, as in the cases above cited, the same rule should apply.

The instructions given to the jury taken together as a whole properly advised the jury as to the law and we do not find that the defendant was in any way prejudiced thereby.

A careful review of the record convinces us that the evidence is sufficient to justify the verdict, and the judgment is therefore affirmed.

Shenk, J., Gibson, C.J., Marks, J., pro tem., Moore, J., pro tem., and Curtis, J., concurred.

EDMONDS, J.,

Dissenting.

As I read the record in this case, the evidence, even with the statement of Showalter, which I agree was part of the res gestae, does not support the verdict. My reasons are based upon these facts, which stand undisputed:

Showalter, the deceased, was the senior man of the train crew and in the absence of the conductor was directing their movements. As the two cars started down the snake-lead, Showalter boarded the refrigerator car, the leading one of the two, at its forward end and went to the top. He had his light and his brake club with him. His purpose was to control the movement of those cars by using the brake. When last seen, he was on top of the refrigerator car with his light. Neither he nor his light was seen again until he was found lying underneath the gondola car, half way between the two rails, and run over by the leading truck of that car.

As the two cars moved along the snake-lead Parrott, another member of the crew, got on the footboard of the tender and the engine backed slowly down the track following the cars. There is no claim that this was an improper movement. Parrott had his light, and was at the position where the tender would collide with anything which was on the track ahead of it. Concerning a coupling of the engine and the gondola, he said that he could not remember whether one was made.

When Parrott found Showalter under the car he went up to him and said, "How in the world did it happen, Joe?" Showalter replied: "I got knocked off. I am all done, both legs cut off." There was nothing else said. It is significant that this question was asked by the man who was responsible for the movements of the engine and was controlling it by his signals. He was the first one to see Showalter. The spontaneous exclamation made by the man who would know most about the car movements indicates that Parrott did not know what had happened. The only reply by Showalter to his question was, "I got knocked off."

The respondent construes this reply to mean that the locomotive knocked him off. However, Showalter knew that Parrott was a member of the crew, but there was not one word of accusation. He did not say that the engine had bumped the cars and knocked him off or that the accident was the fault of any member of the crew.

Unquestionably, Showalter left his position on the moving refrigerator car, walked along the top of that car and climbed down or fell down between the refrigerator car and the gondola car. Did he move from the forward end of the refrigerator for the purpose of going to the brake stand on the gondola car, or did he walk across the car for some other reason? Whatever his purpose, he fell between the two cars being controlled by him. How he fell is entirely a matter of speculation.

Parrott's testimony concerning his first knowledge that an accident had occurred is most significant. He said that he heard Showalter shout; it was not a shout of surprise, but that of a man who had been already injured, and was in pain. Moreover, his testimony is that he heard a shout while the locomotive was backing. He also said that only one pair of trucks had passed over him.

Respondent contends that certain measurements prove that the engine following the two cars collided violently with them, knocking Showalter off. She argues that although Parrott was unwilling to testify that he violated his duty so as to permit the engine to contact the cars ahead, he was not unwilling to testify to facts which conclusively demonstrate that there must have been such a collision. But an examination of the record does not indicate that the only reasonable conclusion to be drawn from Parrott's testimony points to a contact. He testified: "We were near the gondola, but whether we had made a coupling into the gondola or not, I cannot remember at this time, the step I was standing on did not get much beyond the cross-over switch; assuming that the gondola and refrigerator cars had made a coupling with the cars, the east end of the gondola would not be as far as 40 to 50 feet from the switch. It would possibly be 8 or 10 feet from the switch; I said I found Showalter under the center of the car. ... With relation to the cross-over and lower 4 track at the time I heard that holler I was near the switch, or just west of the switch of lower 4 track." This is substantially all the evidence as to the position of the cars at the time Showalter was injured.

Respondent argues that because Showalter must have gotten under the car twenty feet before it came to rest, that the east end of the gondola car must necessarily have been in front of or a few feet to the east of the switch, and because Parrott testified that when he heard the cry of pain, the west end of the tender was near or just west of the switch, it follows that there must have been a collision. However the measurements were approximate and there is no assurance that Parrott heard the first cry of pain or that if he did, whether Showalter had called as he was falling or after he had been run over.

The burden of proof is on the plaintiff to show negligence, but there is no substantial evidence that the engine and tender were anywhere near the two cars. The words used by Showalter do not prove that he was knocked off by the engine. Indeed the evidence shows other facts which make his explanation entirely consistent with the happening of the accident in other ways. For example, it was the practice in the yard to use clubs in working the brakes. It is entirely possible that the spinning of a brake wheel or a lurching of the cars as they passed over the frogs and switches, may have knocked Showalter off the car. Certainly it is just as consistent with the evidence that Showalter was knocked off by the brake or a movement of the cars as it is that he was knocked off by the contact of the engine and under the Federal Employers' Liability Act, where the evidence is wholly circumstantial, to support a verdict it must exclude all other reasonable inferences. If there are other equally reasonable inferences, no one of them can be said to preponderate over the others.

This rule was succinctly stated by the Supreme Court of the United States in Patton v. Texas & Pacific Railway Co., 179 U.S. 658, 663 [ 21 S.Ct. 275 , 45 L.Ed. 361 ], where it was said: "... it is not sufficient for the employee to show that the employer may have been guilty of negligence--the evidence must point to the fact that he was. And where the testimony leaves the matter uncertain and shows that any one of half a dozen things may have brought about the injury, for some of which the employer is responsible and for some of which he is not, it is not for the jury to guess between these half a dozen causes and find that the negligence of the employer was the real cause, when there is no satisfactory foundation in the testimony for that conclusion. If the employee is unable to adduce sufficient evidence to show negligence on the part of the employer, it is only one of the many cases in which the plaintiff fails in his testimony, and no mere sympathy for the unfortunate victim of an accident justifies any departure from settled rules of proof resting upon all plaintiffs." The rule was also stated in Tucker Stevedoring Co. v. W. H. Gahagan, Inc., 6 Fed.2d 407, at 410, as follows: "Yet sound reason, as well as settled law, demands that a conclusion shall not be based upon circumstantial evidence alone, unless the facts relied upon are of such a nature and are so related to each other as to exclude every other fair and reasonable hypothesis. If the facts are consistent with either of two opposing theories, they prove neither." (See, also, Atchison, T. & S. F. R. Co. v. Saxon, 284 U.S. 458 [ 52 S.Ct. 229 , 76 L.Ed. 397 ]; United States F. & G. Co. v. Des Moines Nat. Bank, 145 Fed. 273 .)

In my opinion, the decision of the District Court of Appeal was correct and the judgment should be reversed.