This motion arises out of the failure of the jury to agree upon a verdict after four days of trial and a substantial period of deliberation.
After oral argument, an order was entered dismissing the motion for judgment upon the whole record. Subsequently, respective counsel, with the approval of the court, filed additional briefs treating the legal questions involved more fully and with greater particularity. After careful study of the case, we have reached the conclusion that the motion for judgment should have been granted. We have given consideration to the advisability of vacating the original order and of the entry of one conforming with the conclusions reached. Upon reflection, we have deemed it advisable not to disturb the original order lest one or the other of the parties might be prejudiced by the technical application of the term “rule”. We have elected this *688procedure for the reasons given, especially since the questions raised on appeal will be presented without the technical complications involved in a modification of the original order and without the confusion that a shift in the identity of the appellant would create.
The case is one which may justly be denominated “hard” not only by reason of the gravity of the misfortune which was the subject of litigation but also because of the nature of the proofs adduced to support and contradict the asserted liability and the result evolved out of our consideration.
Four cases were consolidated for purposes of the trial. All arise out of the deaths of Robert Taylor, a minor, and his grandfather, Jacob Gluck, after they had eaten lamb purchased at a store owned and operated by defendant. In each instance the cause of death was “infective enteritis”.
The household struck by this tragedy consisted of Helen Taylor; her son, Robert; her mother and father, the latter being Jacob Gluck, and Mrs. Taylor’s brother, William Gluck. Helen Taylor regularly marketed and prepared the food for the household. On Friday, August 21, 1941, she visited the store of defendant, situate at 4445 Frankford Avenue, Philadelphia, at about 1:00 p.m. She purchased a stewing chicken and a large leg of lamb with the rack attached to it. The rack, otherwise known as the chops, was separated from the leg by the butcher at Mrs. Taylor’s request, and the two portions of lamb were wrapped separately. The meat then looked “all right” to her. She continued her shopping at defendant’s store as well as at a fruit market across the street and, her shopping concluded, she returned home about 3:45 p.m. or 4:00 p.m.
Mrs. Taylor washed her purchases, wrapped the meats in separate pieces of wax paper and put them into her electrical refrigerator, which was then and continuously thereafter in good working order. Fri*689day, as well as the succeeding days which intervened to the time of the tragedy, was a very hot day. However, the meats were not handled by anyone from the time they were put away until the Sunday following the purchase, at which time Mrs. Taylor removed the chops, prepared them for dinner and they were eaten by all the members of the family except Robert Taylor, who was away. No one suffered any visible discomfort as a result of that meal.
The leg of lamb was removed from the refrigerator by Mrs. Taylor for the first time on Monday afternoon, August 25,1941, at which time it was prepared by her for roasting. At the evening meal, at which the lamb was served, all the members of the household were present together with one William Livezey, an employe of William Gluck, who alone did not eat the lamb. The meal included mashed potatoes, succotash, lettuce and tomato salad, coffee, two types of pie and the lamb. With the exception of William Livezey, all became ill thereafter.
Robert Taylor, who first manifested untoward effects from the meal, became violently ill about midnight of the same night, and before the end of the morning following every member of the family was suffering. A physician, Dr. Long, was summoned, and he diagnosed the condition as food poisoning. On Wednesday morning, Jacob Gluck died. The other members of the family were removed to the Nazareth Hospital where Robert Taylor died Thursday morning, August 28th. Dr. William Wadsworth, coroner’s physician, performed a post-mortem on the body of Robert the same day and testified that the cause of death was “infective enteritis”, which he explained as follows: “Enteritis means an inflammation of the bowel, the lower, small bowel, and infective in this case I would signify as a septic condition or infection of the bowel itself, including the adjacent gland.”
*690There is no doubt that the cause of death was food poisoning of some kind as a result of eating the lamb. The doubt which arises, however, is founded on the uncertainty of the source and nature of the contamination from which the deaths resulted. A very grave doubt has been interjected as to whether the meat was contaminated when purchased or whether Mrs.. Taylor was infected with the bacilli salmonella schottmuelleri and she in turn contaminated the meat as a result of which the victims died. In our opinion the evidence on behalf of plaintiffs does not resolve that doubt with sufficient certainty to warrant the resubmission of the issue to a jury.
From the very inception of the proofs concerning the cause of death, plaintiffs’ case is marked with uncertainty. Miss Edna Void, employed by the Department of Health, City of Philadelphia, at the municipal laboratories as a bacteriologist, testified that she examined various foods taken from the refrigerator of the stricken household and made tests for anything that might cause food poisoning and found salmonella schottmuelleri, otherwise known as paratyphoid B, as well as enterotoxin, in the roast lamb. The tests on the other foods were negative. Over plaintiffs’ objection the same witness testified that she examined cultures of the throat and stools of Mrs. Taylor and Mrs. Gluck. The tests of Mrs. Gluck were negative and those of Mrs. Taylor showed positive for paratyphoid B for both the throat and the stool on one culture. Other tests made from the same subject were negative. We do not refer to the expression of opinion, elicited on the cross-examination of this witness, as to the cause of death, because we do not believe she was properly qualified. However, Dr. John A. Kolmar, a bacteriologist of unchallengeable standing, appearing on behalf of plaintiffs, first questioned that the organism was in fact salmonella schottmuelleri which caused the deaths, it being his opinion that the contamination resulted *691from a different organism, and he further was mystified at the results obtained from the cultures taken' from Mrs. Taylor for, as he put it, “I have been a bacteriologist for thirty years, and this is the first time I can remember finding paratyphoid B in a throat culture.”
From the testimony of Dr. Kolmar we derive the following explanation of the nature of the bacillus which allegedly contaminated the meat. The organism is known as salmonella schottmuelleri, being one of the paratyphoid group, which produces a toxin or enterotoxin. The living bacillus, he stated, is destroyed by the cooking of the meat, but not so the enterotoxin. In his opinion, assuming the meat to have been free from contamination on the day purchased and contaminated thereafter by the person who handled it, the bacillus would not produce sufficient enterotoxin to cause the deaths in the interval which elapsed from the date of purchase to the time of consumption. He did not deny the possibility but offered it as his opinion that such was improbable. As to the presence of salmonella schottmuelleri itself in the meat after the cooking, this expert could offer no explanation save the possibility that the bacillus escaped destruction by heat because of the nature of the medium, or the suggested possibility that it was re-contaminated by Mrs. Taylor after the cooking. Such an opinion, however, suggests in the alternative that the death was caused by either the bacillus or the enterotoxin. However, the definitive testimony of Dr. Wadsworth, who performed the autopsy, fixed the cause of death as “infective enteritis” and, from the explanation offered by Dr. Kolmar, we understand that “infective enteritis” is caused by the swallowing of the living organism itself and not from the consumption of the toxin produced by the bacillus. Nonetheless, and to further emphasize the uncertainty of plaintiff’s proofs, Dr. Kolmar offered as his opinion that the victims died as a result of *692swallowing the' enterotoxin, rather than the living organism, thus contradicting the testimony of Dr. Wadsworth as to the cause of death.
As to the presence of the bacillus in the throat specimens of Mrs. Taylor, Dr. Kolmar suggested several alternative possibilities; that somehow the bacillus escaped destruction by heat so that she became infected as a result of eating the lamb; that she became infected from handling the meat after she purchased it; or that she was a temporary or evanescent carrier and thus contaminated the meat. As to the latter, Dr. Kolmar’s opinion was most reserved — he conceded the possibility that Mrs. Taylor had contaminated the meat but regarded this as very unlikely.
Of course, Dr. Kolmar’s opinions must at all times be considered in the light of the other proofs adduced on behalf of plaintiff and the contradiction inherent therein. His testimony throws doubt on Dr. Wads-worth’s findings of the cause of death, for Dr. Kolmar believes it was a food poisoning of the toxin class rather than the infective type. Secondly, he -doubts that the bacillus in question was schottmuelleri at all but rather suggests it was of another type. Thirdly, he cannot understand the presence of the bacillus in the throat of Mrs. Taylor in the first instance and why, .if present once, it did not reappear on subsequent cultures.
On direct examination, Dr. Kolmar, answering a hypothetical question which assumed that Mrs. Taylor had contaminated the meat, stated that “it would take far more time than you specify for salmonella schottmuelleri — if we can grant that that was really the organism — to produce endroxins or enterotoxins in that period of time. . . .” This answer, however, assumes that the enteritis was of the toxin rather than the infective type, and is thus contradictory to Dr. Wadsworth’s findings. On a restatement of the same question, Dr. Kolmar answered, “I would say that that *693is possible, but highly improbable. \ . . I do not deny the possibility, but I do state it is highly improbable.” Expressing his opinion as to the cause of death he declared, “. . . I still maintain that that enterotoxin was, in my opinion, probably present in that mutton on the day that it was purchased. . . .”
Plaintiffs’ actions must here fail because of the failure to meet the burden of proof imposed upon them. Though the actions sound in assumpsit, being founded in tlie warranties raised by the law of sales so as to obviate the need for proof of negligence, the burden still is theirs to prove that the deaths resulted from the eating of contaminated meat and, more important, that the meat was contaminated at the time of purchase. As to the first element of proof, there is no dispute. As to the second element, however, the requirement of direct positive proof must be fully sustained by plaintiffs or their causes fall. A jury may not be permitted to guess or speculate on such an issue or the entire safeguard of the law, which demands an exact certainty in such matters, must fall. This rule, stringent and harsh though the results of its application may be, is of particular import in cases where the subject matter, as here, is one in which the jurors may not draw inferences or conclusions from the reservoir of their own experience to aid them because it is one which “requires special scientific knowledge for its solution, and, for that reason, the jurors are dependent on expert evidence to enable them intelligently to reach a decision . . . ”: Mudano v. Philadelphia Rapid Transit Co., 289 Pa. 51, 60 (1927).
Plaintiffs, tacitly conceding the requirement of the strict proofs necessary in such a matter, argue that the requirement is limited to a determination of causal relation and that here the rule has been satisfied in that there is no uncertainty concerning the fact that the deaths were caused by the contaminated meat. We cannot accept such a restricted understanding of *694“cause” and consequent limitation on the rule concerning expert opinion. While the law does not indulge in the philosophic pastime of teleological contemplations so as to search out the sequence of each succeeding event culminating in a liability-creating incident, it does go beyond an inquiry which would be limited to determining the last activating step in the chain of causation. To reject the inquiry into the cause of the contamination of this meat as part of the essential proofs of liability would be as absurd as to go to the other extreme and say that the cause of death was not the meat but the eating of it. We cannot avoid the strict rules of proof called for here by adroit rationalization. Those who philosophize' upon the abstract postulates of the law may dismiss the rejection of expert opinions of possibility and probability as a legal aberration (see 7 Wigmore on Evidence (3rd ed.), sec. 1796), but to a court of law in which the fact to be proved is heard the. requirement of certainty of proof offers the one safeguard against this exception to the hearsay rule running rampant on the field of liability. Here it is not sufficient that the jury be enabled to grope its way in the shrouded darkness of ignorance in a field unfamiliar to it; the testimony of the witness who is qualified and permitted to express an opinion must be clear and unequivocal: Vorbnoff v. Mesta Machine Co. et al., 286 Pa. 199 (1926). Where, however, the opinion is variable, indefinite, and uncertain, it neutralizes itself and loses all probative value.
In the case of Elonis v. Lytle Coal Co., 134 Pa. Superior Ct. 264 (1939), the following rule, applicable to the matter before us, was expressed at page 270:
“There was a period of time between May 9, 1921, and April 12, 1926, during which a statement by an expert that in his opinion a given result ‘most probably’ came from an assigned cause would, under the rulings of our Supreme Court, support a finding of causal connection. That ruling was made in Fink v. *695Sheldon Axle & Spring Co., supra, decided May 9,1921, but was modified and explained in McCrosson v. Phila. Rapid Transit Co., 283 Pa. 492, 129 A. 568, decided May 25, 1925, and definitely changed in Vorbnoff v. Mesta Machine Co. et al., 286 Pa. 199, 206,133 A. 256, decided April 12, 1926. Naturally, a number of cases will be found in both appellate courts in which the Fink case was followed, but it is no longer the law. The rule as laid down in the Vorbnoff case (page 206) is, ‘. . . the expert would have to state plainly the professional view that the accident had materially “contributed” (Farran v. Curtis, 276 Pa. 553, 556; see also Clark v. Lehigh V. C. Co., 264 Pa. 529, 533) to the ailments from which claimant suffers, in the sense of being, if not the sole cause, then, at least, the “superinducing cause” thereof: Jones v. Phila. & Reading C. & I. Co., 285 Pa. 317, 320. That is to say, the witness would have to testify, not that the condition of claimant might have, or even probably did, come from the accident, but that “in his professional opinion the result in question came from the cause alleged”; for, according to our latest pronouncement on this subject, a less direct expression of opinion would fall below the required standard of proof, and therefore would not constitute legally competent evidence.’ ”
That the rule last enunciated was developed out of the field of workmen’s compensation cases does not, in our opinion, limit it to such cases. The fact to be established in the case under consideration was that of “cause” and the applicable rule to guide us in evaluating the sufficiency of the proofs should not vary because of the category of the law wherein the fact issue arises. Until there be a contrary expression from our appellate courts, we regard the principle of the Elonis case as controlling.
The previous reference to the testimony of Dr. Kolmar, rejecting as we do for purposes of this disposition the opinion expressed by Miss Void, leads us to the con*696elusion that it is insufficient when measured by this test. To hold otherwise would be to encourage the jury to sift the opinions of Dr. Kolmar and Dr. Wadsworth, accept that which is favorable and create anew a hypothesis of the cause of death which has not been definitely expressed by either expert. This could not be permitted to be done. Liability cannot be predicated on “possible” conduct or action and, .under the present state of the law as we understand it, it cannot be allowed on an assertion of “probability” when that “probability” is attached to so fundamental an element as causation.
And now, to wit, September 15, 1943, defendant’s motion for judgment upon the whole record under the Act of April 20, 1911, P. L. 70, sec. 1, 12 PS §684, in the above cases is granted, and an exception is hereby allowed plaintiffs to this order.