The plaintiff’s intestate met his death while in the defendant's employ, and the question involved in this case is whether his death was caused by actionable negligence of the defendant. The defendant is engaged in the manufacture of nitroglycerine in the county of Houghton, and, prior to the injury complained of in this suit, had been engaged in that business for 10 or 12 years. The method of manufacturing nitroglycerine at the defendant’s works is as follows: 1,200-pounds of sulphuric acid is first mixed with 1,650 pounds of nitric acid, after which 1,700 pounds more of sulphuric acid is added, and all is mixed up in a tank called the “agitator.” The agitator is an iron tank having about it a water jacket, through which cool water is caused to flow continually while the mixing process is going on, in order to keep down the temperature. After mixing the acids, the oil, or sweet glycerine, is run in slowly, and cold water kept running through the water jacket, so as to keep the temperature down; otherwise, it is liable to explode during the process of manufacture. It is considered necessary to keep the temperature down to about 80 deg. Fahrenheit. In this manner about 515 to 515 pounds of glycerine is introduced into, and chemically combined with, the quantity of nitric acid above mentioned. After mixing, the compound is allowed to stand and settle from one to two hours. The nitroglycerine, being lighter than the acids, rises to the top as it forms. Only.about 70 per cent., or less, of the nitric acid, has combined with the glycerine; about 15 per cent, remains mixed with the glycerine, but not chemically combined; and the remainder passes off in fumes. The sulphuric acid and whatever of the nitric acid is not mixed with the glycerine settles to *584the bottom of the agitator, being heavier than the nitroglycerine. The sulphuric acid, being heavier than either, is at the bottom. None of the sulphuric acid combines with the glycerine, and does not enter into the composition of nitroglycerine. It is used because of its strong affinity for water, and it absorbs or takes up the water that is formed by chemical action in the combination of the nitric acid with the glycerine. After the mixture has settled, the sulphuric acid is drawn off through a pipe to tanks some distance away, called “settlers.” In these settlers the nitroglycerine which becomes mixed with the sulphuric acid rises to the ■ top, and is skimmed , off, and washed like the other nitroglycerine, and the sulphuric acid is saved for another process. After the sulphuric acid, or most of it, is drawn out from under the nitroglycerine, the glycerine, together with the nitric acid mingled with it, and whatever of sulphuric acid remains, is drawn through a 2-inch pipe into a tank of water about 5£ feet deep, called the “drowning tank.” The specific gravhy of nitroglycerine is about 1 0-10, and, with whatever of nitric or sulphuric acid is mixed with it, it sinks directly to the bottom of the drowning tank. The drowning tank is not stirred or agitated except by a stream of water running into it from the cold-water jacket, which runs in part of the' time while the nitroglycerine is falling into and lies in the drowning tank. The substance that goes into the drowning tank is mainly nitroglycerine, but has mixed with it a quantity of free nitric acid and some sulphuric acid. Immediately after the mixture is dumped into the drowning tank, they commence to draw it off into pails, and carry it to the wash-house, about 200 feet distant, where it is washed in two or three waters, and finally in soda water, until the acid is entirely eliminated or neutralized. The bottom of the drowning tank has a slope of 10 inches towards the faucet where the nitroglycerine is drawn out into pails, and also towards the faucet where an attempt was made to draw it out through a hose.
*585On the 16th of May, 1895, the defendant endeavored to change the method of conveying the nitroglycerine from the drowning tank to the wash-house, or vault, as it is called. A rubber hose three-quarters of an' inch in diameter was procured, and laid so as to run on a level from a faucet in the drowning tank to the vault, a distance of about 200 feet. The hose had a drop of about 5 or 6 feet in the vault, so as to convey the glycerine into the bucket or other receptacle. Between 9 and 10 o’clock (the superintendent says half past 10) the faucet was opened, and the glycerine started to run. At first, for a few minutes, it ran somewhat freely, but very soon it began to run very slowly, and only drizzled. After trying about 15 minutes to run the glycerine through the hose, it was given up; and, by orders of the superintendent, the faucet at the drowning tank was closed, and the nitroglycerine immediately stopped running out of the hose. During this time only one, or, at most, two, pails of the compound had run through the hose. The employés were at once set to work carrying the nitroglycerine in pails, as they had done before; but the hose remained attached to the faucet at the drowning tank, and very little, if any, of the nitroglycerine that was in it, ran out. The hose lay in that condition, containing so much of the nitroglycerine as had not run out, until about 20 minutes after 1 o’clock in the afternoon, at which time the nitroglycerine remaining in the hose, and about 1,200 pounds more, stored in the vault or wash-house, exploded, and Fred Schoepper was thereby killed. The negligence imputed to defendant was the use of this hose in the manner in which it was used, and permitting the compound to remain in the hose until, by chemical action of the ingredients, an explosion occurred, which resulted in the death of the plaintiff’s intestate.
At the conclusion of the evidence, the learned circuit judge directed a verdict for defendant, on the ground that the plaintiff had not shown that the explosion was caused in the manner indicated in the declaration. This is the *586only question that need be considered, for, while the defendant makes the point that the plaintiff failed to show that the deceased was not guilty of contributory negligence, it is enough to say that, in addition to the presumption of due care which obtains in cases of death resulting from an injury, it is plainly apparent that deceased had nothing whatever to do with the work of running.this compound through the hose in question, and was in no> way responsible for its being left there ± and it follows that, if the explosion was caused by these acts of defendant, the deceased in no way contributed to the result.
Defendant’s counsel contend that the cause of this explosion is a matter of mere conjecture, and it is said by counsel that it is not enough for plaintiff to prove circumstances consistent with her theory, but that these circumstances, and each of them, must preclude any other rational conclusion. This we take to be but another way of stating the proposition that the proof must exclude all reasonable doubt. It is hardly necessary to say that no> such rule obtains in civil cases. It is true 'that where an injury occurs that cannot be accounted for, and where the occasion of it rests wholly in conjecture, the case may fail for want of proof. Robinson v. Charles Wright & Co., 94 Mich. 283; Redmond v. Lumber Co., 96 Mich. 545. But such cases are rare, and that rule should never be so extended as to result in a failure of justice, or in denying an injured person a right of action where there is room for balancing the probabilities, and for drawing-reasonable inferences better supported upon one side than the other. In this case there was no direct proof of any other probable producing cause of the explosion than such as was offered by the plaintiff.
It is true that defendant offered testimony to the effect that washing the nitroglycerine with very hot soda would be dangerous. But there was no evidence that deceased was engaged in washing this nitroglycerine in hot soda. It is true, one witness swears that on one occasion he was in the vault when the soda water that Schoepper was *587going to use was too hot. He testifies that he told Schoepper that it was too hot, and that Schoepper replied, “Let it stand awhile, and it will cool off.” Another witness gave similar testimony. But this does not show (certainly not conclusively) that on the day in question deceased was using soda water which was too hot. It was further in evidence that deceased had been known to drive a wooden faucet into a nitroglycerine barrel, and there is testimony that under certain conditions, if foreign substances were present, this might cause an explosion. But it further appeared that deceased had been told not to do this, and, since being so instructed, had not been known to do it. On the other hand, the testimony on the part of the plaintiff (and, in determining this case, the question must be whether the testimony of plaintiff was sufficient to make a prima facie case for the jury) was, in addition to the facts above stated, furnished by an expert witness, one Dr. Wheeler, and by a Mr. Johnson, who had had practical experience in the manufacture of nitroglycerine. Dr. Wheeler testified that, from his experience, he would not consider it safe to permit the compound to remain in the hose in the manner in which it was for the length of time that it was permitted to remain there; that the result would be that gases would form, and that it would be dangerous, — would result in an explosion, — and that he would not consider it safe to leave it there in that condition at all, unless it was running. He also testified that, when this compound leaves the drowning tank, the percentage of free acid is about 5 or 6 per cent.; that, the greater the per cent, of acid, the more dangerous it would be, but that any amount of free nitric acid in a compound is dangerous if the compound is stationary, and not running a steady stream, even 1 per cent.; that he would consider it dangerous to leave this compound, as it leaves the drowning tank, in a rubber hose 150 feet long, any length of time without running. “I wouldn’t want to experiment by leaving it there three hours, and I would consider it dangerous if it was unwashed nitroglycerine, *588containing 1 per cent, or more nitrous acid, because the gases would form there, and their pressure would cause an explosion.”
It is true that defendant’s testimony would tend to show that the pressure upon the hose, before the heat was generated sufficient to cause an explosion, would be greater than the hose would bear, and that the degree -of heat necessary to cause an explosion is supposed to be 380 deg. Fahrenheit, while the hose would stand no such pressure; and it is argued from this, as well as from the fact that the hose was open at one end, that, before the degree of heat that was necessary to cause an explosion could be generated, the compound would escape. But Dr. Wheeler testified that it might not escape at the open end of the hose, and that, with one end of the hose open, he would still consider it dangerous; and, further, that, if the hose was closed at both ends, a pressure sufficient to explode the pipe would probably explode the nitroglycerine. And he further testified that if the hose described was open, and a pressure of something like 500 gallons of water on one end of it, and the pressure was not sufficient to force the mixture through the hose, this would indicate that there must have been some obstruction; and, further, that he didn’t think that it was necessary, in order that nitroglycerine should explode, that the vessel should stand 375, pounds pressure to the square inch; that he didn’t think that it would be necessary to have anything stand such pressure to explode it. One William A. Dunn, who had had practical experience with nitroglycerine containing free acids, testified to actual explosions which had taken place of the material in squai’e tin cans, of ordinary sheet tin, and testified that such explosions occurred even though the cans would stand very little pressure, as” there was simply a cork put into them, and any small pressure would blow out the cork.
It is true, defendant’s theory and expert testimony is opposed to that offered by the plaintiff. It is the peculiar province of the jury to weigh this testimony. If the *589testimony offered in behalf of the plaintiff is credited, we have a case in which, on the first occasion on which an attempt was made to use this hose in the manner in which it was used, an explosion followed upon such a use of the hose as experience teaches is careless and unsafe. Under such circumstances, we think the question of what was the cause of this injury was one for the jury. Negligence, like any other fact, may be inferred from circumstances. Alpern v. Churchill, 53 Mich. 613; Barnowsky v. Helson, 89 Mich. 523. And, though the proof of plaintiff depended upon inference to establish the main fact, the question of whether the inference suggested by the plaintiff’s theory is the correct one, or whether it was sufficiently rebutted, was for the jury. Crosby v. Railway Co., 58 Mich. 458; Hagan v. Railroad Co., 86 Mich. 615; Woods v. Railway Co., 108 Mich. 396.
The judgment will be reversed, and a new trial ordered.
The other Justices concurred.