This diversity action is based on the Pennsylvania wrongful death and survival statutes and seeks to recover for the death of plaintiff's husband, which was caused by a premature explosion of dynamite manufactured by defendant. At the close of plaintiff's case, defendant's motion to dismiss was granted, and plaintiff now asks us to decide that her evidence was sufficient to get her to the jury. We think it was.

Taking the evidence in the light most favorable to plaintiff and resolving all reasonable inferences in her favor, as we must, in view of the present status of the case, the following is a reconstruction of the event.

The deceased, Howard T. Hopkins, was an employee of C. J. Langenfelder and Son, Inc., a contracting firm then engaged in building part of the Pennsylvania Turnpike. On the day of the accident, the work involved blasting through a hill to bring the terrain down to the proposed grade of the highway. The dynamite and blasting caps used by Langenfelder were supplied by defendant. The blasting was to be done by the simultaneous detonation of a series of about three hundred boreholes five feet apart. The first step involved drilling the boreholes, which varied in depth, the fatal one being six feet deep. The next operation called for a crew of "loaders" who placed dynamite and blasting caps in the holes, tamped them down, filled the rest of the hole with dirt, tamped it down, and connected the wires from each hole to a trunk line so that the entire pattern could be detonated electrically.

On the morning of the accident, drilling began at about seven o'clock. There was no loading done, however, until about ten o'clock when the drillers had acquired a sufficient start. As the day progressed the loaders caught up to the drillers and throughout the afternoon worked right behind them. That is, there was an interval of approximately three or four minutes between the completion of the drilling of a hole and the beginning of the loading of it. Deceased was a loader. At about five or five-thirty in the afternoon, just before the explosion (one witness testified that it was a matter of seconds), he was seen working over a loaded and filled hole, coiling the wire around his hand. There was a drill operating within five feet of the hole. The dynamite in the hole exploded, and Hopkins was killed.

There was ample testimony to the effect that the general blasting method used was that customarily followed in construction work. The vein of rock in the area was very hard.1 Drilling in hard rock generates more heat than drilling in soft rock. Two of deceased's co-workers testified that in normal drilling operations on this job they had seen drill steels turn blue, one of them stating that this meant that the steel was very hot. None of the workers had any technical schooling on explosives. Their knowledge of the use and specific characteristics of dynamite was learned by experience in the field.

Plaintiff's expert, a consulting materials engineer, testified that when steel becomes blue it means that it has a temperature of from 500° to 900°, depending upon the precise shade of blue. Also, from drilling experiments he had made about two years after the accident, as near to the scene as possible (about 25 or 30 feet away), with practically identical drilling equipment and in the same vein of hard rock in which the fatal hole was drilled, and in weather conditions which, if anything, would lower the results of the readings, the expert found that drilling in the hard rock produced a steel temperature of 340° Fahrenheit and a borehole temperature of 200°, and in soft rock a steel temperature of 260° and a borehole temperature of 170°. He also found that the borehole would cool off at the rate of approximately 20° in 15 minutes. He testified that blasting caps are much more susceptible to impact detonation than is dynamite, but that when heated, the instability of both is greatly increased. That is, when heated, both the caps and dynamite will detonate because of vibration or impact much more readily than when at normal temperatures. His tests showed also that the vibration caused by the drilling was very substantial near the hole and could be felt 15 feet away. In answer to a hypothetical question, he testified that in his opinion the explosion was caused by heat in the borehole, or vibration from the drill being operated 5 feet away, or a combination of both heat and vibration.

Since diversity of citizenship is the basis of federal jurisdiction here, we apply Pennsylvania substantive law in accordance with Erie R. R. v. Tompkins, 1938, 304 U.S. 64, 58 S. Ct. 817, 82 L. Ed. 1188. Plaintiff makes no claim that defendant was negligent in the manufacture, inspection, packing, or distribution of its dynamite. It is plaintiff's contention that defendant knew or should have known that drilling through hard rock produces sufficient heat in the borehole to explode dynamite placed therein before the hole has cooled, and that the vibration caused by drills operating within a few feet of a loaded hole is sufficient, alone or in combination with that heat, to explode the dynamite; that defendant knew or should have known that in blasting operations the boreholes might be loaded soon after drilling them and that drilling might be done close to a loaded hole; that, because defendant knew that the men working with its dynamite are given no technical schooling as to the specific danger involved, it, in turn, knew that they would not realize the risk of loading the holes too soon or of drilling too close to a loaded hole; and, in the face of that knowledge, that defendant was negligent in failing to warn of the specific danger involved.

We think that the case of Maize v. Atlantic Refining Co., 1945, 352 Pa. 51, 41 A.2d 850, 160 A.L.R. 449, is controlling here. There defendant put out a cleaning fluid, 45 per cent of which was made up of carbon tetrachloride, the fumes of which are deadly. Deceased was killed by inhaling the fumes while cleaning a rug. The name of the fluid, "Safety-Kleen," appeared on the four sides of the can, and, in smaller print on two sides, was placed the warning: "Caution. Do not Inhale Fumes. Use only in well-ventilated place." The issue left for the jury was whether defendant had given an adequate warning to probable users of the danger inherent in the fluid's normal use. The jury found for the plaintiff, and the Supreme Court of Pennsylvania affirmed, stating that it was clear that the trial court could not say as a matter of law that defendant had fulfilled its duty to warn. The Maize case stands for the proposition that one who supplies an instrumentality for use by others, and who knows or should know that the foreseeable use is dangerous to human life unless certain precautions are taken, and who realizes or should realize that the probable user will not recognize the danger, is under a duty to warn the user of such danger and to advise the proper precautions.2 The Maize case is even stronger than the present case because there the defendant actually did warn of the specific danger involved. The question was whether the warning was sufficiently prominent. Here there was no warning at all of the specific risk.

Defendant attempts to parry the thrust of the rule of the Maize case by saying that it is inapplicable here. Defendant tells us that everybody knows that dynamite is dangerous and that there is no need to warn against the obvious. But plaintiff's theory does not go to the generally dangerous character of dynamite. She contends that what is not known by blasting workers is that there is enough heat and vibration in a freshly drilled hole, when a drill is being operated a few feet away, to explode the dynamite placed in the hole. Everybody knows that dynamite should not be thrown in a fire, but apparently most construction workers do not know that it should not be placed in a hole under the conditions existent in this case.3 One Keller, a blasting foreman for Langenfelder, with twenty-five years' field experience in working with dynamite but with no technical schooling on explosives, testified that he had never seen an explosion such as occurred here. From this, defendant argues that the accident was not foreseeable and, therefore, there was no negligence. As the testimony of plaintiff's expert discloses, a simple test would have demonstrated the danger of an explosion resulting from a combination of blasting caps, dynamite, and a certain degree of heat and vibration. The jury could have found that defendant, because of its technical knowledge, should have foreseen the danger, while the workers, without such technical information, might not have foreseen it. Keller's testimony could be considered by the jury as weakening the opinions of plaintiff's expert, but it is not of such weight as to warrant taking the case from the jury.

It is true that defendant did publish a safety list which accompanied each case of dynamite. This is a list of sixty-three "Don'ts," advising what not to do in the many phases of blasting operations. The only one even remotely touching the specific danger involved here, however, is "Don't" No. 21. It says, "DON'T load a sprung bore hole with another charge of explosives until it has cooled sufficiently." A sprung hole is one usually employed in quarrying operations, in which the bottom of a plain borehole is enlarged by exploding a preliminary charge and then a larger, final charge is placed in the resulting cavity and is exploded. Naturally, the first explosion produces heat, and "Don't" No. 21 warns against loading the final charge in the sprung hole before that heat has dissipated. Thus, the warning against heat in a sprung hole could well mislead the user into concluding that there is no risk from heat in a freshly drilled plain hole. Consequently, we feel that the jury could have found that the warning was inadequate.

Next we are told that negligence may not be inferred from the mere happening of an accident; that there was no competent proof as to the cause of the accident; and that the jury may not be permitted to guess. The obvious answer is that if the jury were to believe plaintiff's expert, and we must assume that it would, the cause would be heat or vibration or a combination of the two. This would not be a guess or an inference from the mere happening of the accident. Proof to the degree of mathematical precision is not required. Before the explosion, deceased was seen near the loaded and tamped hole coiling the wire around his hand. He had completed, apparently carefully, the only actions in which his own carelessness could have caused the explosion. Furthermore, the Pennsylvania presumption that deceased used due care is in plaintiff's favor.4 The record is barren of evidence which would rebut that presumption.

Defendant asserts that there may have been other causes of the accident, for some of which it would not be liable. That there may have been other causes is true. But none of these fancied causes are fairly deducible from the evidence. This court, stating Pennsylvania law, has said that it is not necessary to eliminate every possible cause other than the one on which plaintiff relies, but only such other causes as fairly arise from the evidence. "It is not the rule that circumstantial evidence need exclude everything which the ingenuity of counsel may suggest as having possibly caused or contributed to the death." Williams v. Reading Co., 3 Cir., 1949, 175 F.2d 32, 34.

Plaintiff's expert testified largely on the basis of his post-accident tests. Defendant argues that conditions existing during the tests were so materially different from those existing at the time of the accident that the expert's testimony was worthless in determining the cause of the accident. Granted that the tests were made two years after the accident and twenty-five or thirty feet away from where it occurred, nevertheless, they were made with practically the same kind of drilling equipment and in the same vein of rock. True, the atmospheric temperature was five degrees higher than it was when the accident occurred, but there was a cold rain which could easily offset that factor. There were further differences, but these would, if anything, lower the test readings. Therefore, on the whole, we feel that the conditions were sufficiently similar to allow a logically relevant inference. The differences go only to the weight of the testimony. In any event, the admission of testimony based on experiments is a matter resting largely within the discretion of the trial court.5 That discretion has not been abused here.

We have considered carefully all the other contentions made by defendant and find them to be without merit.

For the reasons stated, the judgment will be reversed and the cause remanded for a new trial.