delivered the opinion of the court.
This was an action brought by appellee, Weiss, for damages against appellant on account of an injury *44received by him while acting as a switchman for appellant. The verdict and judgment rest on a count based on the Federal Safety Appliance Act. The negligence charged is in hauling and using a car with a coupler in a defective condition that necessitated appellee’s going between cars to uncouple them.
Appellee’s testimony was substantially as follows: Fie was one of a switching crew that was breaking up a train which just before the accident passed a switch where the rear or south car was to be “cut off” and “kicked” down a certain track. He dropped off at the switch, and as soon as the train passed it, gave both a stop and back-up signal to the engineer, threw the switch, and started forward on the east side of the train to cut off the rear car which stopped about half a car length north of the switch. The couplers were of the automatic type. The lever for uncoupling on the south car was operated from the east side and that on the north car from the west side. It is operated by raising a crank on the outside of the car. This causes the lever arm to lift and pull up a pin from the couplers high enough to permit the knuckles to pass under it and open. The pin hangs on a chain, one end of which is attached to the lifting arm and the other to the drawbar. Unless the pin is held up, the cars will lock automatically on impact when the knuckles are open. He took hold of the crank in the usual way to raise the pin and gave it a jerk. It came part way easily and then appeared to strike something solid. As he lowered the lever he heard the pin drop back. He repeated the effort about three times, with the same result. During that time the train was not moving. Failing to lift the pin by operating the lever, he stepped and reached in between the cars, took hold of the chain attached to the pin on the adjacent north car and lifted it. Thereupon the knuckle on the north car opened, and the south car moved away about eighteen inches. After the uncoupling was thus accomp*45listed, he reached over to close the knuckle so as to prevent the cars from recoupling on coming together in the next movement of the train, and while in the act of reaching in and pulling it towards him, the detached train, pursuant to the signal he had previously given, moved back against the uncoupled car, causing the injury to his hand and wrist complained of.
The testimony in behalf of appellant was substantially as follows: Appellee was immediately taken aboard the engine by the crew and started for a hospital. The engine returned in about an hour and a half. The conductor testified that he then found the cars where they had left them and that he examined the safety appliances and couplers on the cars in question, operated the levers on the ears both when coupled together and apart, that they worked perfectly, that he found no defect, and that they had no difficulty in uncoupling and switching the car. He testified that before leaving for the hospital he took the numbers and initials of the two cars and reported them and the accident to the yardmaster, who testified that he in turn reported said facts to the only inspector in the yard; that the latter, within half an hour from the time of the accident, went to examine the cars, and that no cars were moved until the engine came back. The inspector testified that there was nobody in the yard at the time during that night to repair the appliances but himself; that when he made the inspection he tried all the safety appliances, including the pin lifters, chain, lock and knuckles, and found all in good, working order; that he took out the knuckles and the lock, found nothing bent or out of order, put them back, tried the lever four or five times and it was in perfect working order. Another inspector testified that he examined the cars when delivered to appellant the night before and found nothing defective.
*46The two main questions of fact thus presented were, (1) whether the appliances were defective, and (2) whether appellee was injured in the act of uncoupling the cars.
As to the former, there was no proof except what might be inferred from appellee’s testimony of his inability to operate the lever. Against such inference is direct and positive testimony, corroborated in part by several witnesses, that, without repair, while essential conditions remained unchanged, the appliances were inspected, tested and found to operate perfectly and to be free from defects, and appellee’s own testimony is not inconsistent with that state of facts.
Appellee’s counsel argues, however, that there was no evidence that a repair may not have been made between the time the accident occurred and appliances were inspected. If the evidence, that in the interval said inspector was the only one in the yard to repair the appliances and that shortly after the accident he found them in perfect working order, is not sufficient answer to this contention, the fact remains that there was no proof of repair. This same argument was made in Union Pac. R. Co. v. Brady, 88 C. C. A. 579, 161 Fed. Rep. 719, upon a very similar state of facts and proof. The Court said: “It was suggested that a repair might have been made between the times of injury and inspection, but there was not the slightest proof of one.” In that case, as this, there was no attempt to prove a specific defect. The Court said: “We are of opinion that there was no substantial evidence of a defective condition of this appliance,” and significantly added, “But, if this were not so, it was plaintiff’s duty to use the lever on the other car.”
The burden of proof was on plaintiff to show the appliance was defective; and it has been held that where one makes a mere prima facie case that is completely met by proof of a contradictory state of facts, sworn to by several unimpeached witnesses, “and there *47are no elements of probability to turn the scale,” the verdict will be set aside. Peaslee v. Glass, 61 Ill. 94. If anything, the elements of probability seem to be against appellee. It is conceded that the lever does not work when there is tension on the pin. Plaintiff testified, “when the slack is all out, we cannot raise the pin. ’ ’ While he testified on direct examination that the “slack was all in,” yet he also testified that the moment he lifted the pin, the released car moved eighteen inches away, and when asked what caused it, said, “that was slack run out.” As the train was stationary and remained so during his operations, such movement can apparently be accounted for only by the force of gravity, after release from tension on the pin occasioned by “slack run out.” This presents a reason for his inability to work the lever.
Appellee’s counsel suggests, however, that the movement may have resulted from a recoil of the springs back of the drawbars. If such theory did not lead to mere conjecture, yet against it is the fact that the action of springs would necessarily ensue at once upon the stopping of the train, and, presumably, have been spent in the interval taken by appellee to work the lever and uncouple the cars in the manner testified to by him. The immediate movement of only the rear ear when the pin was lifted renders the theory of gravity more plausible than that of recoil.
But appellee’s counsel further urges that if there were tension on one pin of the connecting couplers there would also be on the other, and if the one could not be moved by the lever, the other could not be by hand. Again we are led to the realm of conjecture. This depends upon details of construction, as to which the record is barren. Appellee, however, contends that as a model of some kind was used at the trial for illustration (not put in evidence, nor certified to us), a presumption in his favor will be supplied from the use "of such model before the jury. In the case cited *48to support such contention (North Chicago St. R. Co. v. Eldridge, 51 Ill. App. 430), the jury were permitted to inspect a car as to the existence of a certain alleged condition, a photograph of which was received in evidence and deemed sufficient to prove it. The Court added however, that “whatever, if anything, was lacking in the other evidence * * * we must presume was supplied by such inspection. ’ ’ In the case at bar there was no evidence whatever on the subject. Where the entire evidence is before us any essential matter of fact not inferable therefrom will not be supplied by presumption.
But, if there be doubt whether the record will support the finding of a defective appliance, there can be none, we think, that appellee’s injury did not result from the act of uncoupling. He was asked, “When you went to close that knuckle, you had completed uncoupling those cars?” He answered, “Yes, sir.” He was further asked, “When those cars separated eighteen inches, after they were uncoupled, you could then have stepped out on the outside, and not got caught between there, couldn’t you?” and answered, “Yes, sir.” It is apparent, therefore, that he had accomplished the uncoupling, and stepped forward and remained between the cars to adjust the knuckle and prevent their recoupling on the next movement of the train. It was while engaged in the latter act that he was injured. Such act was not necessitated by reason of a defective condition of the appliance, if any existed.
Contrary to appellee’s testimony, however, his counsel argues that he had not completed the uncoupling, claiming that the act to prevent recoupling was a part of the act of uncoupling the cars. To this we cannot assent. There can be no controversy over the ordinary meaning of the word “uncoupling,” or as used in the Federal Safety Appliance Act. Two things are required by section 2 of that act: (1) That the couplers shall couple automatically by impact; and (2) that *49they can be uncoupled without the necessity of men going between the ends of the cars. This action is predicated wholly upon an alleged violation of the latter requirement. It is significant that while each count of plaintiff’s declaration alleges that the cars could not be uncoupled without his going between them, and that he went between them for that purpose, it is also alleged that his injuries were received “while engaged in the work of uncoupling them and preventing them from reconciling themselves.” The declaration itself recognizes a distinction between uncoupling and preventing recoupling, and that the injury did not ensue alone, at least, from the act of uncoupling.
At the close of the evidence, it was stipulated that standard automatic couplers are not constructed in such a manner that the knuckles thereof can be closed automatically after the cars have separated, or by the use of the lever or any other device in connection with such coupler, but that the same can only be closed by hand. At the close of the evidence, therefore, if not before, the fact was uncontroverted that the closing of a knuckle has no connection with the equipment for coupling or uncoupling. The statute in question was designed to prevent injury in the act of coupling or uncoupling, and requires equipment that will enable each to be accomplished without going between the ends of the cars. Had the cars come together without the pin being held up, they would have coupled automatically as required by the act. This would have happened even though the uncoupling appliances were free from defect. Had appellee not gone between the cars or remained there to engage in another distinct act, namely, to pull the knuckle so as to prevent recoupling, he would not have been injured, even if it had been, necessary to go between the cars to effect the uncoupling. The statute in question does not require automatic couplers to be so constructed that the knuckles thereof, when once opened, can be closed otherwise than by hand so as to prevent cars from recoupling. *50No one would question that if, when the rear car had thus separated from, the train, plaintiff had stepped out without adjusting the knuckle, the car would have been actually uncoupled, whether the next movement was forward or backward. His remaining between the cars to perform an act that would facilitate the next movement, that of “kicking” the car on another track, was the cause of his injury. He admitted that after accomplishing the uncoupling he could have stepped out and escaped injury. If, therefore, appellee’s injury was not the proximate result of uncoupling, then the failure to comply with the statute with respect to equipment therefor would not entitle appellee to recover.
At the close of plaintiff’s evidence and again at the close of the entire evidence in the case, appellant asked for an instructed verdict of not guilty. In our judgment, it was entitled to the instruction, as the uncontroverted facts showed that a failure to comply with the statute, if any, was not the proximate cause of the injury. The facts as to the manner of receiving the injury being uncontroverted, the question of proximate cause was a question of law. Devine v. Chicago & C. River R. Co., 259 Ill. 449.
In the case last cited, the cause of action was also predicated on the statute in question. It was claimed that the defective appliance necessitated the injured party’s remaining between cars to accomplish the very end sought in the case at bar,—to prevent a recoupling in a “kicking” movement. The injured party went between the locomotive and car to uncouple them and stood on the running board of the locomotive to hold up a coupling pin so as to prevent recoupling in the “kicking” operation. The Court said that when he was injured by being thrown therefrom “the purpose for which he had gotten upon it had been accomplished, —the cars had been uncoupled,” and that while the appliance failed to comply with the requirements of the statute, and such failure required the party to go *51between the ends of the cars to effect an uncoupling, yet his presence on the footboard when injured was not due to such failure, but because there was no other practicable way in which the operation of “kicking” could be performed, and that, therefore, such failure was not the proximate cause of the injury. It further said that, conceding that there was a device that would hold up the pin, yet “any standard coupler which couples automatically by impact and which can be uncoupled without requiring men to go between the ends of cars would satisfy the statute,” and that “if such a device is known, neither the Federal nor State statute requires its use.”
It appears from the evidence that if the pin is held up, either by hand or by holding the lever, as may be done, the- cars will not recouple on coming together. It was to obviate the necessity of holding up this pin by hand, or using the other lever, or stopping to give the engineer a counter signal, that led plaintiff to engage in the act in which he was injured, namely, closing the knuckle by hand, so the automatic device would not operate and recouple the cars. When, therefore, appellee was injured, he was not engaged in the act of uncoupling, against the dangers of which the statute is aimed, but in preventing recoupling, a matter that is not within the purview of the statute. We must, therefore, hold that on the uncontroverted facts, as a question of law, the proximate cause of appellee’s injury was not the failure, if any, to comply with the statute upon which his action is predicated, and must accordingly reverse the judgment.
Reversed.