Richards, a levee contractor, owning a number of mules and desiring to protect them from anthrax, purchased at a Memphis retail store ten bottles of anthrax vaccine, manufactured by the Mulford Company, which is engaged on a large scale in the manufacture of serums, vaccines, antitoxins, etc. Having had some experience in giving similar treatments, Richards proceeded himself to treat his mules by injecting this vaccine hypodermically. He poured two bottles into a cup, and, with the contents, filled his syringe and used it; he then emptied four bottles into the cup, and took therefrom two syringefuls, and then used the remaining four bottles in the same *678way. Within a short time all of the mules which had been treated with vaccine from the second cupful died from tetanus; none of the others were affected. Richards brought this action against the Mulford Company on the theory that, in the course of manufacture, it had permitted the vaccine in one of these bottles to become infected with the tetanus germ, whereby it became unfit for its intended use, and by negligence and by implied warranty the Mulford Company became liable for the value of the mules thus killed. The District Court directed a verdict for defendant, upon the ground that there was no evidence which supported plaintiff’s theory and which was of that character which justified submission to a jury. The plaintiff brings this writ of error.
The question is of very narrow compass. There is, for the purposes of this review, no dispute that the mules died because they were infected with tetanus at the time this vaccine was injected. There was no evidence .whatever directly tending to show negligence in the manufacture or unfitness when sold. The defendant’s evidence, undisputed and apparently not subject to doubt (unless inferentially), showed that the manufacture of this vaccine was carried on with the utmost skill and care, and with the most perfect precautions known to science. There was no reason to suspect the existence of tetanus germs anywhere around the Mulford factories, save that such germs were used in the manufacture of tetanus antitoxin; but this was at a place a mile distant from where the vaccine was made, and the employés were separate; and that the fatal germ may have come into the vaccine from this source is mere surmise; it is possible, but highly improbable. Counsel do not suggest any plausible explanation of how it could happen.
[ 1 ] On the other hand, the evidence that the trouble probably came from Richards’ own carelessness in the operation of administering the remedy is very strong. It appears, without dispute, that stable manure and the surface soil around a stable form the favorite habitat of this germ, and that it is commonly carried therefrom by flies or blown around by the wind; that many thousands might be carried on a fly’s foot or a speck of dust hardly visible to the naked eye; that Richards used this cup in a feed tent adjacent to his own corral, in which a hundred mules were kept on the levee bank, and where manure was scattered about and had accumulated; that the doors were open, it was a dry, hot day, the wind was blowing through, and the flies were thick; and the open cup from which he filled his syringe stood all the time exposed in this tent. It is manifest to us that the weight of the evidence is against the plaintiff’s theory; nevertheless, and even though the trial judge might rightly think it was his duty to set aside a verdict for the plaintiff, if rendered, and award a new trial — at least once— the case should have been submitted to the jury if there was any substantial evidence tending to support each of the steps essential to a recovery.
[2, 3] Plaintiff does not question the general rule that a case should not be submitted if a verdict for plaintiff would rest on conjecture or surmise, as distinguished from evidence, or that the plaintiff must fail where the case shows merely- a possibility that the foundation of the ac*679tion exists. Of course, the criterion, is not whether the evidence, hi the court’s judgment, will equally support either conclusion; many disputed issues of fact respond to that definition; the critical test of this class of verdict, directed for defendant, must be whether the court can say that all reasonable men must agree in finding the evidence insufficient to raise the inference upon which plaintiff relies. This court has pointed out that in awarding a new trial because the verdict was against the weight of evidence, the trial judge is applying his own judgment to the problem; in directing a verdict, he is finding that no reasonable man can deem the plaintiffs evidence fit to induce conviction. Mt. Adams Co. v. Lowery, 74 Fed. 463, 476, 20 C. C. A. 596; Big Brushy Co. v. Williams, 176 Fed. 529, 532, 99 C. C. A. 102; Jenkins v. Alpena Co., 147 Fed. 641, 77 C. C. A. 625; Nelson v. Ohio Co., 188 Red. 620, 628, 112 C. C. A. 394.
Upon either the theory of negligence or of breach of warranty, plaintiffs first step is to establish that the vaccine was infected when he bought it, and if he fails in that step, it becomes immaterial whether there is an implied warranty of fitness in the sale of such an article as this.
To put his case beyond the bounds of mere surmise, plaintiff relies upon several circumstances, of which we need mention only three: Rirst, that tetanus 'germs were later found in an empty bottle; second, that all the mules inoculated from the second cup, and none of those inoculated from the others, were affected; third, that this poison in the vaccine would have-been more virulent than germs from the stable, and that the extreme rapidity with which the disease developed indicated inoculation with poison of the highest potency.
1. Upon the first point, it appeared that, after emptying each bottle into the cup, Richards put back the cork and threw the bottle on the ground outside. Rive or six days afterwards, all the bottles that could be found were” picked up and put, unwrapped, on a .shelf. A few days later they were taken away and examined. The evidence has some tendency to show that, when thus examined, the interior of one of them carried a large colony of these germs, and that the interiors of all the others were sterile. We think no substantial inference that the bottles were in the same relative condition before they were first opened can reasonably rest upon this testimony. The high probability that a bottle, so exposed and not tightly scaled, might have become infected in this way during the five days is conceded, and that one cork might have been put in loosely and the others more tightly, is one of several reasons, any one of which is enough to account for the difference in their later relative condition. Indeed, that only one bottle was infected, and others, filled at the factory at the same time from the same container, were not, goes far to disprove plaintiff’s theory.
2. The second point is of similar character. If there were no other probable explanation, it might indicate that the contents of one of the second batch of bottles were infected when that bottle was emptied into the cup, and that, for this reason, the contents of the first and third cupfuls were not; but it may as well indicate several other things. The cup, when used the second time, had been standing there about 20 minutes. It was then merely rinsed in a carbolic solution, too weak *680to be effective, and existing infection might have survived this rinsing, or there may have been a fresh infection in the cup during the few minutes before the next four bottles were all emptied, and the agitation by pouring the bottles into the cup or by filling the syringe would naturally have dissipated the germs throughout the cup, while the next rinsing may have left the cup clean. The danger from the exposed cup was constant and extreme. No inference of prior infection in the bottle can safely rest on this ..condition of the second cupful when finally used.
3. The third point has impressed us as more serious. The evidence shows that the disease and the symptoms which indicate the disease are not produced by the direct effect of the bacillus, or germ, but by the toxins which the germs throw off in the multiplying process which takes place when the germs have found a suitable location. If the vaccine had been infected at the time of manufacture, it might have been highly charged with these toxins at the time of use, and if these toxins were injected in the manner followed by plaintiff, s3unptoms of the poisoning would be expected to appear in three or four days, while, if the infection comes from the entry of the germs into an abrasion on the skin, symptoms are not expected within eight or ten days. The injection was Saturday morning; one of the.mules, had a stiff neck late Monday evening (during the third 24-hour period); two showed distinct symptoms on Tuesday; nearly all on Wednesday; and all 41 died within the. week. It is strongly urged that the appearance of symptoms as early as the third day distinctly tends to show that there was an injection of developed toxins, and so to show that the vaccine was infected when bought.
This contention overlooks the undisputed testimony from every witness,’ whose attention was called to it, that if the bacilli were injected beneath the skin, instead of lodging in a surface abrasion, and if this injection were accompanied by the conditions of an anthrax vaccination, the development of toxins would begin promptly and symptoms would appear or should be expected in three or four days. It is obvious that the matter of development will vary considerably according to the condition of the subject, and the lodgment the germ happens to take, and the extent of the' infection which happens to take place, and in view of such inevitable individual variations, there remains no substantial difference between the shortest possible time fixed by any witness with reference to the direct injection of toxins, viz., “two or three days,” and the time fixed by all the witnesses as a natural time for development, if the injection was of the bacilli under all the conditions here existing, viz., “three or four days.” In the balancing of these things, as applied to a case where there were some instances of development on the third day, many on the fourth, and others not until the fifth, we see nothing fairly and substantially tending to show that the vaccine contained developed toxins when it was used. As plaintiff’s expert witness frankly said, when asked what conditions led to quickest symptoms, “Well, it’s all speculation.”
We must think that plaintiff’s proof falls within the class pronounced insufficient by the Supreme Court in Patton v. Railway, 179 U. S. 658, 21 Sup. Ct. 275, 45 L. Ed. 361, and by this court in Virginia Ry. v. *681Hawk, 160 Fed. 348, 87 C. C. A. 300, for we have here not even a “quantitative probability.” See, also, our discussions in Toledo R. R. v. Howe, 191 Fed. 776, 112 C. C. A. 262, and Cincinnati Ry. v. Jones, 192 Fed. 769, 113 C. C. A. 55, 47 L. R. A. (N. S.) 483.
With the failure of this contention, plaintiff’s case must wholly,, fail. Nothing remains to support it, save conjecture — and conjecture which, upon the whole, is essentially improbable. In our judgment, all reasonable men must agree that an inference of defendant’s fault cannot safely rest on such premises. To permit a verdict for plaintiff would be to reward his negligence and penalize defendant’s care.
The judgment is affirmed.