*780Opinion by

I. The first complaint made by counsel for plaintiff error in their brief is, that the verdict of the jury is an illogical one. The line of reasoning by which they arrive at such conclusion is about this: The plaintiff below sought to hold the railroad company liable for the act of its agent, William Higgins, under the doctrine of respondeat superior. Higgins was joined with the railroad company as defendant in the action, both for this purpose and the additional reason that he was personally liable for the wrong committed. The liability of the railroad company is predicated upon that of its agent, for whose act it was responsible. Now as the jury found a verdict in favor of Higgins, thus saying that he committed no wrong, there is no liability of the agent to predicate that of the superior upon. If the major premise of this proposition is true, there seems to be no escape from the conclusion. The first inquiry then involves the construction of the pleadings in the case, to determine whether or not it is sought to hold the company liable solely for the act of its agent under the doctrine of respondeat superior. Before judgment the most unfavorable construction is to be given the pleadings, but after judgment that construction must be given them which will best harmonize with the whole record.

The allegations of the petition in this respect are as follows:

“That the defendants procured the arrest of the plaintiff by said William Higgins as the agent, and at the instance and request of the defendant company, making and filing with L. J. Webb, a justice of the peace, an affidavit,” etc.

In the answer of the railroad company it is alleged:

“Second, that on September 17, 1883, it did, through and by its proper officer, notify the board of railroad commissioners of the facts with reference to the plaintiff throwing a certain switch on the railroad of the defendant, and for which said plaintiff was subsequently arrested, at the time mentioned in the plaintiff’s amended petition, and that said defendant requested instructions from said board of commissioners; that on the 21st day of September, 1883, said board did, in writing, request and direct this defendant to proceed to capture the plaintiff' herein, and turn him over to the proper authori*781ties to be held accountable for his conduct, and in pursuance of such authority and direction of the board of railroad commissioners aforesaid, this defendant caused the complaint mentioned in plaintiff’s amended petition to be made and filed; and the defendant says that said arrest was made so as aforesaid solely at the suggestion and request of the said board of railroad commissioners.”

Higgins’s answer is in substance the same as the railroad company’s, with the additional averment “'that he filed the complaint and caused the warrant to issue at the instance and request of the company.”

' In this state of the pleadings, there can be no successful contention against the primary liability of the railroad company. The petition charges that the criminal prosecution was instituted at the instance and request of the defendant company. The defendant company states in its answer, “this defendant caused the complaint mentioned in plaintiff’s amended petition to be made and filed.” Higgins the agent says in his answer that he filed the complaint and caused the warrant to issue at the instance and request of the company. The fact that the prosecution was begun at the instance and request of the railroad company, and that the company caused it to be instituted, is admitted by the pleadings. A tort which one directs or advises another to commit he is always responsible •for. (Cooley on Torts, p. 534.)

The liability of the railroad company for the wrong in this case, if any wrong there was, is based upon its direct connection with the prosecution, not only as adviser, but because it directed the institution of it, caused it to be begun, and set it in motion. There is no question here of dependency of liability upon the subordination of the agent, because the agent acted under the express direction and in strict obedience to the orders of the company. The agent may be responsible for his participation in the wrong committed by the orders of the principal, but this is an independent question for the jury, and does not necessarily involve the liability of the principal. If in this case the railroad company had been the only defendant, and the petition and answer had contained the same allegations, *782could there be any doubt about the liability of the company for the commencement of the prosecution? Then the case does not fall within the ordinary doctrine of respondeat superior in the sense in which it is assumed to fall by counsel for the plaintiff in error. The verdict is not illogical or absurd, for a jury could consistently say on the facts admitted by the pleadings that the railroad company having caused the institution of these proceedings, and its agent having done only as he was expressly directed to do by his superior, the consequences shall rest on the company alone. Probably the strict view of this question is, that as a railroad corporation can only act by its agents, when the head of a department directs one of its subordinates to do an act from the performance of which injury is done a third party, abstract justice requires that the corporation shall suffer the consequences that follow the obedience of the subordinate to his superior.

1. Probable cause; for jury, when.

3 Criminal prosecution; conduct of complaint, how weighed

*783 2 Evidence-grouped in’ instructions.

*782II. The question of probable cause was left to the jury, and counsel for plaintiff in error claim that the evidence in this cáse was undisputed, and it was therefore for the court to determine whether probable cause existed or not. This involves an examination and determination as to what the facts are, and what are the reasonable deductions from them. If the facts are not in dispute, the question is for the court; if they are disputed, the jury must be left ^ pagg Upon the existence or want of probable cause. Now, in the determination of this question, two propositions must constantly be kept in view: The first is, that the burden of proving the want of probable cause in this action was upon the plaintiff who alleged it; the second is, that the conduct of Foulks, the claim agent of the railroad, and the officer who ordered a criminal prosecution against Watson, “must be weighed in view of what then appeared to him to be the acts and declarations of Watson, and not in the light of subsequently appearing facts.” (Stewart v. Sonneborn, 98 U. S. 194.) The very many things introduced for the purpose of establishing actual malice, and other issues, must not be taken into con*783sideration or allowed to have any bearing on the question of the existence or want of probable cause. The belief of Foulks as to the existence of probable cause, is to be determined by the state of facts existing before and up to the time of the arrest, and is not to be influenced by the other evidence in the case, or the state of facts developed subsequent to the arrest. Out of considerations of this character has grown an unbroken line of authorities establishing one of the most important and beneficial rules that govern in actions for malicious prosecutions. The rule is that in a case where there is a substantial dispute about facts, constituting the existence or want of probable cause, it is for the jury to determine what facts are proved, and for the court to say whether or not they amount Pr°bable cause. It is therefore generally the duty of the court in such a case, when evidence is given tending to prove or disprove the existence of probable cause, to submit to the jury its credibility, and what facts it proves, with instructions that the facts found amount to proof of probable cause, or that they do not. The court should group the facts in the instructions, which the evidence tends to prove, and then instruct the jury that if they find such facts have been established, they must find that there was or was not probable cause. (Johnson v. Miller, [Iowa,] 17 N. W. Rep. 34; Owen v. Owen, 22 Iowa, 271; Shaul v. Brown, 28 id. 37; Gee v. Culver, [Ore.] 6 Pac. Rep. 775; Hadrick v. Bishop, 12 O. B. 275; Castro v. De Uriarte, 16 Fed. Rep. 93; Stewart v. Sonneborn, 98 U. S. 187; Heyne v. Blair, 62 N. Y. 19; Sutton v. Johnson, 1 Term Repts. 493.) This rule must not be made a pretext by which a question primarily for the court, is transferred to the jury. There must be a substantial dispute about the existence of probable cause before it can properly go to the jury, and if about the facts that are claimed to prove or disprove probable cause, there can fairly be said to be a dispute, a conflict of testimony; irreconcilable statements of witnesses; a strong flavor of improbability, then the jury are the sole judges of these, as of every other material fact in the ease; but if the evidence on *784this question, fairly considered and impartially weighed, produces in the mind of the court a reasonable conviction of the existence or want of probable cause, then it is the clear duty of the court to instruct the jury accordingly. The dispute must be of such character as to compel the court to weigh evidence, and determine the credibility of witnesses, before it ceases to be a question of law for the court and becomes an issue of fact for the jury. Whenever the evidence of the existence or want of probable cause produces in the mind of the court a reasonable doubt as to its proper determination, then it should be submitted to the jury. It is said in the case of Stewart v. Sonneborn, 98 U. S. 187, that in all cases in which the question of the defendant’s belief of the facts relied on to prove want of probable cause is involved, what that belief was, is always for the jury to determine.

Proceeding to ascertain from the record whether or not there was such a substantial dispute about the facts tending to prove a want of probable cause, we shall first state what the evidence shows to be .the material facts in the whole case, and then examine such facts as are alleged to have produced in the mind of Foulks an honest belief of the guilt of Watson of willfully and maliciously throwing the switch. The main facts are, that on the first day of September, 1883, at Osage City, on the line of the Atchison, Topeka & Santa Fé Railroad, and about two o’clock in the morning of that day, a passenger train going west on the main track left that track and ran into and collided with an extra freight train that was standing on a side track and headed to the east, severely injuring several of the employés of the company, and destroying some of the property. While the general.course of the line of railroad is from the east to the west, at the depot and yards at Osage City, the course of the main track was from the northeast to the southwest, parallel with the main track for quite a distance, and on both sides of it were side tracks, one to the northwest called the north siding, and one to the southwest called the south siding. The main track and each of the side tracks were connected about midway of the side tracks by a *785cross or spur track running diagonally across the main track from the side tracks, and connected with the main track by switches. This spur track was located some distance east of the depot. The general course of the line being east and west, trains going over the road were called “east-bound” and “west-bound” trains.. The extra freight train to whose crew Watson belonged was running from Emporia to Topeka. It had left Emporia with orders to side track at Osage City to let three extra freight trains pass, and to get orders about the passenger train. Arriving, it had stopped on the north siding, close to the spur or cross track, awaiting signals from the conductor to go ahead. Watson was the head brakeman on this train, and one of his duties was to throw switches ahead of it. He stated that—

“As soon as the three extras had passed on the main line, he went up to the switch and unlocked it, and took the pin out of the lever and sat down on the switch-block, facing the depot.”

The depot was located some distance west of the switch-block upon which he was sitting, and hence when he sat down on the switch-block facing the depot he was facing west. This switch is located on the south side of the main track, and the freight train being on the north siding he could plainly see the depot, or at least that train would not obstruct his view. He sat down there on the switch-block facing the depot, waiting for his conductor to give the signal to the engineer of the freight train to move ahead, and to do- this the train would have to leave the north siding, go onto the spur track, and from that onto the main track. He states that—

“While sitting there I went to sleep and must have turned over, for I was leaning back this way; I must have turned clear over; the first thing I heard, I never heard no train whistle; I don’t know whether the train whistled or not; I heard a noise, and looked up and saw a headlight coming; Well, I jumped up — I thought it was our train pulling out— and throwed the switch. Well, after I did it, I could see the lights of the passenger train, and knew what I did, and tried my best to throw the switch under the baggage car, so that *786they would go on to the ties and stop. I was dazed and frightened.”

Throwing the switch let the passenger train from the main track onto the spur, and from there onto the north siding, when it collided with the freight train and did great damage and severely wounded the fireman on the passenger train, and hurt some other persons. The distance from the switch to the north siding where the freight train stood headed to the east, was about three car-lengths. Immediately after the collision occurred Watson ran toward the.wreck and told both the engineer of the passenger and the conductor of the freight that he had thrown the switch under the supposition that the approaching train was the freight. He assisted to carry the wounded man to the station house, hunted up a surgeon to attend him, and then disappeared. Martin Myers, the engineer on the passenger train, says:

“Approaching Osage City yards, I saw the switch lights were all right for the main track; they were all burning all right for the main track. As I neared the middle of the north siding there was a train standing on the siding, and about there is a spur or cross-track. The engine of the freight train was standing near the spur, and as I approached the engine I saw that gentleman [Watson] there start from the engine and run directly across the track in front of me, about one or two car-lengths in front of me, and he grabbed the switch and threw it around and ran me right directly into the freight train that was standing on the siding. I was unable to hold the train in such a short distance. I applied the air and reversed the engine, but I was too close to the freight on the siding and went into them. He was not at the switch as I approached. He ran across the track from the opposite side of the switch, from the same side his train was standing on; he ran across the track with a lamp in his hand. There was a rule in force then, and is now, and is on all roads that I am acquainted with in the United States, instructing brakemen that when the approaching train is expected, to keep entirely away from the switches and not unlock them until after the approaching train has passed.”

Perdue, the conductor of the freight train, says:

“ I had been to the office to get the orders for my train. I *787got the order, and started out to take it to my engineer, when this passenger train came in sight in the Osage City yards. I got opposite my way car, when I saw this brakeman give this passenger train signals to stop. When he did that, I stopped too; I knew there was something wrong; but I could not any more than say ‘ Jack Robinson ’ before they struck. It did not seem more than two seconds to me — just as quick as could be done. I did not see anybody throw a switch there that night, just before the passenger train came in. I saw the signal to stop. It looked to me like the light had left about the front of my engine or train. It looked like it had left my engine, but it was on the main track when signal was given. I was back of the engine about twenty cars.”

This was all of the material evidence of the circumstances connected with the immediate cause of the wreck.

C. M. Foulks was the claim agent of the defendant railway company. It was a part of his duty as such claim agent to investigate the cause of all wrecks, look after the wounded, take charge of the damaged property, and report to the company how they occurred, their cause and results. He made an investigation in the case, and found this state of facts:

First: That Leon Watson was the employé who misplaced the switch.

Second: That a day or two prior to the wreck Watson had called for his time.

Third: That immediately after the wreck Watson had disappeared, and could not be found, although he caused a search to be instituted and caused inquiries to be made of other roads “if such a man was in their employ?”

Fourth: Myers, the engineer of the passenger train, had given to Foulks his version of the circumstances attending the misplacement of the switch, substantially as he testified.

Fifth: Perdue, the conductor of the freight train, had made a report about the wreck, and the cause of it, substantially as he testified.

Sixth: The witness, Frank Brown, had made a statement in presence of Foulks of the declaration of Watson on the street that “he had done the company up at Osage City, and would do it again,” as Brown testified at the trial.

*788Seventh: Foulks had investigated the train sheets at division headquarters, and alleged to have found by these that the claim of Watson that the cause of falling asleep on the switch-block was by reason of continuous work which gave him no time to rest and sleep, was not supported by the reports.

Foulks, on the witness stand, swore that at the time he ordered Higgins to file a complaint and have Watson arrested, he had an honest belief of the guilt of Watson, produced by the facts above stated. On the other hand, it is in evidence that before the arrest, Watson’s deposition had been taken in the case of Amick against the railroad company, in which action Amick, who was a fireman on the passenger train, sued the company for damages for an injury he received at the time the collision occurred at Osage City, and in that deposition he had given his version of the cause of the wreck in substance the same as his testimony on this trial.

The theory of the counsel for Watson was, in the pleadings and at the trial, that the search and arrest of Watson was to prevent him from testifying in the case of Amick v. A. T. & S. F. Rld. Co., and not belief in his guilt. The correspondence between the railroad company and the board of railroad commissioners, shortly after the occurrence, is relied upon by both parties. The counsel for Watson claim that it presents a strong indication of the belief in the mind of Foulks at the time he addressed the letter to the board, as to the circumstances that caused Watson to misplace the switch, because in that written communication Foulks says, “An accident occurred caused by a brakeman by the name of Watson carelessly and negligently throwing a switch.” The counsel for the railroad company claim that, as the correspondence stated that “Watson had escaped,” and that “the railroad company would endeavor to effect a capture and turn him over to the board of railroad commissioners, to be dealt with according to law,” and as the railroad commissioners advised “that the absconding brakeman be captured and turned over to the proper authorities of Osage county, to be held accountable for his conduct,” the *789fair construction of the correspondence is, that while Foulks had not technically described a criminal offense as having been committed by Watson, that was his evident meaning. There are perhaps other facts and deductions, relied upon by both parties, to prove the want and the existence of probable cause; but these are sufficient to show that there was such a substantial dispute about it that, under the rule hereinbefore cited, the question as to what facts were established by the evidence was for the jury to determine. But the facts were nowhere in the instructions grouped, and the jury were not told what facts they should consider in the determination of the question of probable cause. They were not instructed that the conduct of Eoulks in this respect must be weighed by the facts that came to his knowledge from responsible and reliable sources before arrest, and not in the light of all the facts subsequently developed on the trial of the case; and hence we think that the question of the want or existence of probable cause was not fairly submitted to the jury, with the limitations and restrictions imposed by the rule, and that was material error to the prejudice of the plaintiff in error. The trial judge gives an admirable definition of probable cause, but fails to segregate the evidence on that question from the other evidence in the case, and leaves the jury to determine it in the light oí the facts and circumstances subsequent to the arrest. This is error; and because of it, we recommend that the case be reversed, and remanded to the district court for a new trial.

By the Court: It is so ordered.