The policy sued on insured against “bodily injury sustained through accidental means and resulting directly, independently and exclusively of all other causes in * * * (d) death.” The insured died instantly on November 29, 1933, from a gunshot wound evidently from his own gun while he was hunting doves. The answer denied that the death was caused by accidental means as alleged. It then went further and set up a provision of the policy which excluded death by suicide, sane or insane, and averred that this death was suicide. At the conclusion of all the evidence, the defendant moved for a directed verdict, which was refused, and took exceptions to the charges and refusals of the court to charge touching the burden of proof and the presumption against suicide.
It is not necessary to state the evidence at length. There was no eyewitness of the infliction of the wound. The insured was hunting with his grown son, who at the moment was out of sight circling a hill to flush the doves, which were expected to fly past the stand taken by the insured. The son heard only one shot from the father’s vicinity, hut was himself shooting, as were others at a greater distance. Having made his circuit, he found his father lying dead near a fence at a point where the wires of the “hog wire” netting were fastened to two posts near together, thus making a sort of ladder over which persons customarily climbed from one field to the other; there being no gate. Above the netting was one strand of barbed wire about 42 inches from the ground. The insured was lying on his back, with feet towards the fence, shot through the left breast, the lower part of his heart being penetrated and the shot stopping in the spine. The gun was on the other side of the fence with butt about 12 inches from the fence and muzzle pointing away from the, fence and from the insured. There were bushes about the fence. One empty shell lay near the body, and another was found in the gun, which was a single-barrel “pump gun.” The remains of a dove were found a few days later 45 yards beyond the fence under a tuft of bear grass; no one having hunted there in the meantime. The son says that he found his father’s body lying perpendic*714ular to the fence and the feet were three feet from it. Another witness says the pool of .blood was two feet from the fence, and still another puts it three feet from the fence. The facts touching the life insurance, the health, the disposition, and the domestic and business circumstances of the insured in part tend to suggest suicide and in part to point strongly the other way. The request for a directed verdict rests principally on the contention that the physical facts are explainable only on the theory that insured, standing the length of the barrel from the fence, laid his gun on the barbed wire with trigger against the barb and muzzle against his 'breast and pulled the gun until it fired and he fell backward in his -tracks; the gun recoiling over the fence and falling beyond it. While this theory would explain the physical circumstances, we think it reasonably possible to believe a theory of accident, to wit, that the insured desired to cross the fence to get a dove he had shot, or, better, to use the fence row as a screen, and put his gun over it with butt down and held it by the muzzle as he climbed over the wires, that' he slipped and staggering back from the fence, dragged the gun by the muzzle so that the trigger hung on a bush or on the barbed wire and was discharged into his body; gun and body separating as the consequence of the shot as in the opposing theory. It is said that the feet at three feet from the fence are too far away for such a fall, but there are two replies to that. Two witnesses placed the pool of blood from the wound at two or three feet from the fence, so that the feet would have to be nearer according to them, making a conflict on the point of their exact location. Again, we find no evidence that tracks on the ground or other signs proved that the insured stood three feet from the fence and did not stagger back as the theory of accident supposes. Especially in view of the other circumstances, some of them strong arguments against the likelihood of suicide, we are of opinion that a question for the jury was made, and the judge did right in denying the motions for a directed verdict which were made by each side.
But we think the issue was not accurately and clearly submitted to the jury. In the portions of the original charge excepted to, the judge repeatedly told the jury that there was no doubt that the insured died,.by .external violent means, and that the law presumed an accident, and the burden of proof was on the defendant to show a suicide. At the end he said: “Now, as I stated to you when I opened this charge, the case of the plaintiff has been made out and she is entitled to recover unless suicide is established by the defendant company. When the proof was made of death by violent and external means the presumption of accidental death arose, and the defendant company must introduce evidence to you to prove to you that the insured, Mr. Driver, did commit suicide. Now, if 'the defendant has not convinced you by a preponderance of the evidence in the case that the insured committed suicide, then you ought to find for the plaintiff.” This charge is unfortunately worded in seeming to require the defendant company to introduce the evidence of suicide. Many of the circumstances which tend to show suicide were testified to only by the plaintiff’s witnesses. The jury, of course, could consider evidence on the point without regard to who introduced it. But the serious error lies in putting the burden on the defendant to show that the death was not accidental, but suicidal, in a suit on an accident policy. The plaintiff under such a policy can recover only on proof of a death by accident as set forth in the policy. A contention by the insurer that there was suicide is not the assertion of an exception which he has the burden of pleading and proving, but is only another way of saying that the death was not accidental. In such an issue the so-calle.d presumption against suicide is not a rigid presumption of law, but at most one of fact, the expression of a common observation and experience, so common that judges and jurors may utilize it in a trial without proof of it, that sane persons do not ordinarily take their own lives, but strive to the utmost to preserve them. This is but a part of the general background of common knowledge, against which’ every question of fact is tried. The likelihood of every occurrence and the credibility of all testimony are tested by it. Common knowledge very often supplies the gaps in the formal proofs. It is doubtless true that, if a death by a self-inflicted wound appears, and absolutely nothing more, judge and juror should adopt a reasonable theory of accident rather than one-of suicide in the light of this common experience. But such a case rarely happens. Generally there are circumstances which point one way or the other, and special motives to be considered. The formal proofs and the common knowledge must be taken *715together and honestly weighed by the appointed triors in an earnest effort to reach the actual truth. If in a suit upon an accident policy this is done and the triors are not convinced that the death was accident-til, the suit must fail, for the policy covers only a death by accident. The plaintiff has the burden of proving such a death. The presumption against suicide, which might better be termed the common knowledge on the subject, may in some cases of circumstantial evidence incline the scale of probability. In other cases the circumstances may outweigh it and indicate suicide as the reasonable truth. It is likely to confuse, the jury to refer to the matter at all as a presumption raised by law.
It is urged that the judge corrected the error by instructions later given on refusing defendant’s requested charges. He then said in part: “I do charge you that the burden of proving that Louis Frazier Driver came to his death directly, independently and exclusively of all other causes by bodily injury sustained by accidental means is upon the plaintiff, Elizabeth S. Driver; in other words, she must prove by a fair preponderance of the evidence that her husband’s death was accidental. She assumed the burden of that fact by bringing this suit and the burden of proving that fact remains with her throughout the case. The defendant has denied that his death was accidental, and therefore before Mrs. Driver can recover a verdict at your hands it is necessary that you first find from the evidence and by a preponderance of the evidence that the direct cause of the death of plaintiff’s husband resulted through accidental means.” This correct charge was nullified by adding: “However, while that is true as a general principle of law, I charge you that in this case the elements that are required to make out the case for the plaintiff and carry the burden so as to make a prima facie case have been established as a matter of law, so that it leaves open only the question of suicide. * * * I think all that, gentlemen of the jury, was covered in the charge which I gave you originally, and the one issue which I have stated to you is the controlling and the only issue for this jury to pass on.” We understand this to reaffirm the original charge touching suicide. Nothing of the original charge was expressly withdrawn. The error in it was not cured. The pleadings and the charges in this case are much like those in Travelers’ Insurance Co. v. Wilkes (C. C. A.) 76 F.(2d) 701, and what was said of the charge there applies here.
For error in the charge, the judgment is reversed, and the cause remanded for further proceedings not inconsistent with this opinion.