RUGG, C.J.

The deceased employee received an injury in the course of and arising out of his employment through a splash of molten lead into his eye on September 17, 1913. He was treated at a hospital until October 13, 1913, when, as was found by the Industrial Accident Board, "while insane, as a result of his injury, he threw himself from a window and was fatally injured." The board found further that "this insanity was brought about and resulted from the injury," and that, while the evidence was very close upon that point, the death "did result from `an uncontrollable impulse and without conscious volition to produce death,'" under Daniels v. New York, New Haven, Hartford Railroad, 183 Mass. 393, 400. The arbitration committee, whose findings were affirmed and adopted by the Industrial Accident Board, put it this way: "We find and decide as a fact that the accident injured the eyesight of the deceased, caused the loss of his eye, caused a nervous and mental derangement, caused insane hallucinations and caused him, while mentally deranged, in a state of insanity and under the influence of hallucination, by an irresistible impulse, to commit suicide, and that the accident was the sole, direct and proximate cause of the suicide."

The insurer contends that these findings are not warranted by the evidence. That question is open to it, for the substance of the evidence is reported. Pigeon's Case, 216 Mass. 51.

The burden of proving the essential facts necessary to establish a case warranting the payment of compensation rests upon the dependent in a case arising under the workmen's compensation act as much as it does upon a plaintiff in any proceeding at law. The dependent must go further than simply to show a state of facts which is as equally consistent with no right to compensation as it is with such right. They can no more prevail if factors necessary to support the claim are left to surmise, conjecture, guess or speculation, than can a plaintiff in the ordinary action in tort or contract. A sure foundation must be laid by a preponderance of evidence in support of the claim, before the dependents can succeed. The elements that need to be proved are quite different from those in the ordinary action at law or suit in equity, but, so far as these elements are essential, they must be proved by the same degree of probative evidence. Of course this does not mean, as was said by Lord Loreburn in Marshall v. Owners of Steamship Wild Rose, [1910] A.C. 486, "that he must demonstrate his case. It only means that if there is no evidence in his favor upon which a reasonable man can act he will fail." If the evidence, though slight, is yet sufficient to make a reasonable man conclude in his favor on the vital points, then his case is proved. But the rational mind must not be left in such uncertainty that these essential elements are not removed from the realm of fancy. Plumb v. Cobden Flour Mills Co. Ltd. [1914] A.C. 62. Barnabas v. Bersham Colliery Co. 4 B.W.C.C. 119. Fletcher v. Owners of Ship Duchess, [1911] A.C. 671. See also Childs v. American Express Co. 197 Mass. 337; Bigwood v. Boston Northern Street Railway, 209 Mass. 345. The board adopted rulings, and thereby instructed itself as matter of law, in accordance with the substance of these propositions as requested by the insurer, and no error is shown in this regard.

There was evidence tending to show that, although for a time after the injury the deceased was in his normal temperament which was hopeful and joyous, he then became silent and moody, and was depressed, and suffered from certain marked hallucinations. He did not appear affectionate as he always theretofore had been toward his wife and young children. There were two witnesses of the event which directly produced his death. One gave the following description: "That morning I was making my first visit to the ward. . . . Mr. Sponatski was sitting on the window sill, leaning against the frame and his feet were up against the other side. The window was open and he was looking out and I spoke to him and asked him to come down. He turned around and gave me a kind of wild look. I thought he was getting off the window sill. He let one foot down and raised up on the other knee and at that he got up on the window sill and leaped right out. . . . It happened very quick. . . . He had a wild look. He looked as if he was frightened. . . . He appeared as if he had just woke up out of a deep thought. Kind of wild." The other said that after he was spoken to "he hesitated a few minutes; he looked as blank; he was undecided what to do; [he had] a very wild glassy look. He didn't seem to act as though he heard at all just looked blank. . . . He had a vacant stare as though he didn't see you as though he was picturing things he didn't see; things in his imagination. He didn't pay a bit of attention to us at all just as if we were not there." The medical examiner who made a post mortem examination of the brain testified that the deceased "did not have any form of insanity, except possibly general paresis, but for any other form, I could not express an opinion." An alienist of experience testified that probably there was developed from the accident a "mental disturbance" accompanied by "delusions and hallucinations" and as a result committed suicide. After his death a letter, which the board decided was written by him, was found under his pillow, as follows: "my wife folks are not to blame for anything my wife was a pur woman when I married her she be is pure to this day it is all my own fault (signed) Martin Sponatski." Aside from this there was no evidence tending to show that he had contemplated suicide or that the jumping from the window was the exercise of even a "moderately intelligent power of choice." Daniels v. New York, New Haven, Hartford Railroad, 183 Mass. 393, 400.

The letter does not seem to us necessarily indicative of a suicidal purpose. It was not signed by the name of the deceased, which was Charles J. Sponatski. It apparently was wholly the product of a disordered intellect. It is as consistent with some other phantom of an unbalanced imagination as it is with a volition to end his life. The circumstances of the leap from the window as narrated by all the eyewitnesses, point rather to ungovernable lunacy than to the volition even of a diseased mind. The finding in this respect, although hanging on a rather slender thread of evidence, is not unsupported. Therefore it must stand.

This decision rests upon the rule established in Daniels v. New York, New Haven, Hartford Railroad, supra. That rule applies to cases arising under the workmen's compensation act. It is that where there follows as the direct result of a physical injury an insanity of such violence as to cause the victim to take his own life through an uncontrollable impulse or in a delirium of frenzy "without conscious volition to produce death, having knowledge of the physical nature and consequences of the act," then there is a direct and unbroken causal connection between the physical injury and the death. But where the resulting insanity is such as to cause suicide through a voluntary wilful choice determined by a moderately intelligent mental power which knows the purpose and the physical effect of the suicidal act even though choice is dominated and ruled by a disordered mind, then there is a new and independent agency which breaks the chain of causation arising from the injury. See McDonald v. Snelling, 14 Allen, 290.

The Industrial Accident Board was in error in adopting a ruling and thereby instructing itself "that the rule laid down in the Daniels case is not the rule to be followed under the workmen's compensation act. In other words, the question is not whether the consequence is a reasonable and probable one, but whether the consequence resulted from the injury." No question of negligence in its common law sense, or of reasonable and probable consequence was involved or discussed in the Daniels case. That was an action brought under Pub. Sts. c. 112, § 213, now St. 1906, c. 463, Part II, § 245, to recover damages for conscious suffering and death caused by failure on the part of the defendant railroad to give the statutory signals of warning where a railroad crossed a highway at grade. Under that statute the liability of the railroad is made out when the fact of failure to give the statutory signals is established (unless a special defence prevails).

The inquiry as to reasonable and probable consequences did not arise in the Daniels case; but it does arise in actions at common law and under some other statutes in order to decide whether there has been negligence. Even then the question is not whether "the consequence is a reasonable and probable one," but whether harm to some one of the same general kind as that sustained by the plaintiff was a reasonable and probable result of the act complained of, as bearing upon the ultimate question whether there was negligence on the part of the defendant. Negligence may be found even though the particular result of the act may not have been susceptible of being foreseen. See Ogden v. Aspinwall, ante, 100, and cases there collected; Larson v. Boston Elevated Railway, 212 Mass. 262; Wiemert v. Boston Elevated Railway, 216 Mass. 598; Brightman's Case, 220 Mass. 17.

Other instances where liability is not predicated upon negligence, and where therefore there is no occasion to consider in any aspect natural and probable consequences, are actions to recover damages arising from fires set by locomotive engines, Bowen v. Boston Albany Railroad, 179 Mass. 524; from a vicious animal knowingly kept, Marble v. Ross, 124 Mass. 44; from dogs, Pressey v. Wirth, 3 Allen, 191; or from the breaking away of impounded waters, Rylands v. Fletcher, L.R. 3 H.L. 330. So far as concerns conduct of defendants, liability follows absolutely in such cases when the particular decisive fact is shown to exist.

The obligation to pay compensation under the workmen's compensation act equally is absolute when the fact is established that the injury has arisen "out of and in the course of" the employment. Part II, § 1. It is of no significance whether the precise physical harm was the natural and probable or the abnormal and inconceivable consequence of the employment. The single inquiry is whether in truth it did arise out of and in the course of that employment. If death ensues, it is immaterial whether that was the reasonable and likely consequence or not; the only question is whether in fact death "results from the injury." Part II, § 6. When that is established as the cause, then the right to compensation is made out. If the connection between the injury as the cause and the death as the effect is proven, then the dependents are entitled to recover even though such a result before that time may never have been heard of and might have seemed impossible. The inquiry relates solely to the chain of causation between the injury and the death. Dunham v. Clare, [1902] 2 K.B. 292. Ystradowen Colliery Co. Ltd. v. Griffiths, [1909] 2 K.B. 533. See also Southall v. Cheshire County News Co. Ltd. 5 B.W.C.C. 251; Malone v. Cayzer, Irvine Co. 45 Sc. L.R. 351. In deciding whether the chain of causation between the injury and the death is broken by the intervention of some independent agency, the rule of Daniels v. New York, New Haven, Hartford Railroad, 183 Mass. 393, is to be followed under the workmen's compensation act as well as in other cases to which the rule is applicable. There is no difference between the rule laid down in the Daniels case and that in the English cases just cited.

But this error in law did not affect the result reached by the Industrial Accident Board. The decision of the board rests upon the rule of the Daniels case and hence need not be disturbed.

What has been said disposes of all the requests for rulings presented by the insurer. It does not appear that the board misdirected itself in any matter of law material to its decision on the facts found.

Decree affirmed.