The appellant, Pryor, was injured in the course of his employment on an oil well drilling rig, when the mast, sometimes called a derrick, collapsed under the tensile strain of a pipe-loosening and pulling operation. Having received workmen's compensation under the laws of Oklahoma, Pryor and his employer's compensation insurance carrier sued the appellee-manufacturer of the derrick as in MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050, L.R.A.1916F 696,1 claiming that a defective weld at the foot of one leg of the derrick caused it to collapse under the tensile strain. There was expert evidence to the effect that properly fused, the weld would be expected to have the same useful life as the parent metal, and that its failure indicated a defective weld at the point of failure.

The trial court, however, directed a verdict and entered judgment for the defendant-manufacturer, on the grounds that fifteen years of safe use for the purposes for which the equipment was intended foreclosed any probability that it was defectively or negligently made. The court followed Lynch v. International Harvester Co. of America, 10 Cir., 60 F.2d 223, 224, in which this court, while recognizing the MacPherson rule, held that five years of constant use of a threshing machine for its intended purposes was a "conclusive denial and contradiction of the allegation that the machine was imminently dangerous to life and limb when the defendant sold it."

The appellant would distinguish that case from ours for latent defects here and readily discoverable ones in the Lynch case. But the question whether mere lapse of time forecloses probability of negligent manufacture was squarely presented and decided in the Lynch case. The same question is presented here, and we must therefore now decide whether we will follow or recede from the Lynch doctrine. The Lynch case also arose in Oklahoma, but no Oklahoma case was cited in support of the rule. And, while Oklahoma has embraced the full vigor of MacPherson (i. e. see footnote 1), it has neither approved Lynch nor had occasion to consider the precise question. It has said only that if a device is suitable and safe for the purpose for which it is to be used when sold, the manufacturer has then discharged his duty to those who thereafter use it. Jamison v. Reda Pump Co., 190 Okl. 593, 126 P.2d 71. But this statement clearly presupposes nonnegligent manufacture, and is thus not helpful to a consideration whether any probability of negligent manufacture is legally foreclosed by prolonged safe use.

Only recently, the Fifth Circuit, following its conception of Texas law, deliberately refused to adopt a policy of foreclosing liability after prolonged use without defective failure. It did not think the lapse of seven years per se relieved a derrick manufacturer from liability for injuries caused by a defective weld. International Derrick & Equipment Co. v. Croix, 241 F.2d 216. Nor does the Second Circuit believe that mere passage of time confers immunity upon a negligent wrongdoer under the MacPherson doctrine. Fredericks v. American Export Lines, 2 Cir., 227 F.2d 450. The District of Columbia Court of Appeals has reached the conclusion that "the law itself does not say that the passage of nearly seven years, or the action of the elements * * * constituted an efficient intervening cause * * so that any causal connection between Gichner's negligence and the giving away was necessarily broken." Hanna v. Fletcher, 97 U.S.App.D.C. 310, 231 F.2d 469, 475, 58 A.L.R. 2d 847. All of the cases agree, however, that proximity of time and events is cogently relevant in the determination of the ultimate factual issue whether the negligent manufacture caused the harm. Hewitt v. General Tire & Rubber Co., 3 Utah 2d 354, 284 P.2d 471; Reed & Barton Corp. v. Mass, 1 Cir., 73 F.2d 359; Lill v. Murphy Door Bed Co. of Chicago, 290 Ill.App. 328, 8 N.E.2d 714. Our problem is not one of duty owing to the injured party under the MacPherson concept, rather it is one of causation which traditionally lies in the realm of fact, not law. Cf. Sitta v. American Steel & Wire Div. of U. S. Steel Corp., 6 Cir., 254 F.2d 12. The policy decision is therefore whether the defense of use without injury shall be a matter of law to be declared as such, or whether it shall be left to the trier of the facts, with the province to consider all of the relevant factors relating to legal cause. In a jury trial, the issue is for the court only when minds of reasonable men would not differ.

The proposition that prolonged safe use bars any inference of negligent manufacture has not gained wide acceptance in the application of the MacPherson doctrine. As far as we have been able to ascertain, the Lynch case has been followed and applied in only a few isolated cases involving this proposition. Solomon v. White Motor Co., D. C., 153 F. Supp. 917; Schindley v. Allen-Sherman-Hoff Co., 6 Cir., 157 F.2d 102; Sterchi Bros. Stores v. Castelberry, 28 Ala.App. 281, 182 So. 471; Willey v. Fyrogas Co., 363 Mo. 406, 251 S.W.2d 635. Certainly no firmly established body of tort law has grown up around it. A reappraisal of the problem in the light of subsequent decisions persuades us to recede from the rule in Lynch, and to hold that prolonged use of a manufactured article is but one factor, albeit an important one, in the determination of the factual issue whether the negligent manufacture proximately caused the harm.

Apparently quite apart from the time element as an intervening cause, the trial court did not think the plaintiff proved the issuable fact that the defective weld caused the derrick to collapse as it did. There was no direct evidence of proximate cause, and indeed there need not be, for proximate cause is often left to permissible inferences from established facts. Mealy-Wolfe Drilling Co. v. Lambert, 208 Okl. 624, 256 P.2d 818; Mid-Continent Coaches, Inc., v. Guthrie, 208 Okl. 533, 257 P.2d 829; City of Altus v. Martin, Okl., 268 P.2d 228; Fleming v. Hodgson, 199 Okl. 261, 185 P.2d 181. There was direct evidence to the effect that the southeast leg of the derrick was broken off where it was welded to its foot. The proof shows that in an ordinary drilling operation, such as ours, it is not unusual for the drill pipe to stick in the hole while being routinely pulled to the surface. Here the pipe had become temporarily fastened, but not immovably so, and the strain in this particular operation was not unusual or extraordinary. It was an incident which usually occurs during the drilling of any oil well. And, it was in these circumstances that the southeast leg of the derrick broke at the point of the weld. The derrick collapsed toward the northeast, causing the northeast leg of the derrick to buckle and bend outward. From these physical facts we think it was permissibly inferable that the failure of the weld caused the derrick to collapse under the ordinary pressure being applied in the routine operation. The question of proximate cause was therefore one of fact under all the related circumstances.

The case is accordingly reversed and remanded for a new trial.