This four-party case in admiralty presents a host of issues, some old, some new. We discuss the warranty owed by a ship undergoing major repairs, the basis of admiralty jurisdiction over torts, principles of negligence and legal cause in admiralty (more patrieularly, where shipowners and seamen are not involved), laches, indemnity, and contribution.
I.
Facts, Contentions, Conclusions.
Unlike the legal issues, the facts of the case are uncomplicated. The plaintiff, John Watz, was a pipefitter for Levingston Shipbuilding Company. He was injured on October 12, 1959, when a twenty-foot pipe fell on his leg and foot. At the time of the injury, Watz was helping to install the pipe in an exhaust system on Zapata Off-Shore Company’s vessel ÑOLA III. Zapata had purchased the vessel ÑOLA III as a drill tender barge and had hired Levingston to convert it to an over-the-side drilling barge.
The pipe fell on Watz because a one-ton hand hoist holding the pipe failed. Levingston owned this hoist; Eaton Yale & Towne, Inc., had assembled the device. The hoist failed because a link of the load chain on the hoist gave way. Eaton had purchased the chain from its manufacturer, Campbell Chain Company.
The link gave way for two reasons. First, the link suffered from an imperfect weld: only 42.6 percent of the welded area had been fused and even in that area foreign materials were present. Second, external damage from abuse of the chain had occurred after Eaton had sold the hoist in commerce. The district court found that neither of these two factors alone would have caused the failure, but that together they did so.
Watz recovered medical expenses and compensation benefits from Levingston on March 4, 1963. Not until May 15, 1967, did Watz sue Zapata. Travelers Insurance Company, compensation insurer for Levingston, is subrogated to $25,965.06 of the judgment. Zapata impleaded Eaton, the hoist-maker, and Watz thereupon also named Eaton as a defendant. Eaton impleaded Campbell, the chain-maker, but Watz did not name Campbell as a defendant.
The district court awarded Watz a joint and several judgment of $136,185.06 against Zapata and Eaton; awarded Zapata fifty percent indemnity against Eaton; and awarded Eaton fifty percent indemnity against Campbell. Thus, if Watz should execute his judgment against Zapata, the liabilities would be Zapata, fifty percent; Eaton, twenty-five percent; and Campbell, twenty-five percent.
These damages derive from the district court’s findings that Zapata owed Watz a warranty of seaworthiness with regard to the ÑOLA III and that the defective hoist breached this warranty; that Eaton was negligent in failing to discover the defective weld before it placed the hoist in commerce and that it breached its implied warranty of reasonable fitness for intended use; that Campbell negligently manufactured the *104chain and breached its implied warranty of reasonable fitness for intended use.
Zapata, Eaton, and Campbell appeal. Zapata contends that Watz’s claim is barred by laches, that it owed no warranty of seaworthiness to Watz, and that it should recover from Eaton one hundred percent of any judgment it must pay Watz. Eaton contends that it was not negligent, that its implied warranty was inoperative because of the external damage occurring to the chain, that Watz is barred by the Texas statute of limitations or laches in any event, that it should recover from Campbell one hundred percent of any judgment it must pay, and that it should pay no indemnity to Zapata. Campbell contends that the evidence does not prove its negligence, that the external damage to the chain was the real cause of the accident, that having met Eaton’s specifications on the chain it had no other duty, that Eaton’s claim was barred by the Pennsylvania statute of limitations or laches, that the district court had no jurisdiction of the fourth-party complaint against Campbell, and that the court erred in refusing Campbell’s jury demand.
We conclude that Zapata owed Watz no warranty of seaworthiness. Therefore, we do not reach the other issues raised by Zapata’s or Eaton’s argument against indemnity. We conclude that the district court had admiralty jurisdiction over Watz’s claim against Eaton and Eaton’s impleader of Campbell. It therefore properly denied Campbell’s jury demand. Eaton has not sustained its defense of laches, and the evidence supports the district court’s finding that Eaton was negligent. We do not reach the question of its breach of warranty. The evidence also supports the district court’s finding that Campbell was negligent, and we do not reach the question of its breach of warranty. We conclude that the district court correctly denied indemnity from Campbell to Eaton but that contribution was proper. Therefore, the court properly divided the damages between Campbell and Eaton. We remand, however, for further findings and conclusions with regard to Campbell’s defense of laches.
II.
The Warranty of Seaworthiness.
The problem of what warranty a ship offers while she undergoes repairs continues to trouble the courts notwithstanding such definitive utterances as that of Judge Learned Hand’s ten years ago that
It is now authoritatively settled, if indeed it was ever in doubt, that, when a ship has been withdrawn from navigation and while she is being reconditioned, she does not warrant her seaworthiness to those who work aboard her until she returns to active service.
Latus v. United States, 2 Cir. 1960, 277 F.2d 264, 266, cert. denied, 364 U.S. 827, 81 S.Ct. 65, 5 L.Ed.2d 55. The district court here, for example, found that at the time Watz was injured, he “was performing the duties of a seaman aboard the vessel, ÑOLA III, which was in navigable waters”. At one time, this determination would have settled the issue whether the vessel owed a warranty of seaworthiness. Courts “had held it sufficient merely to require proof that the vessel was in navigable waters and that the shore-based worker was engaged in seamen’s work”. Lawlor v. Socony-Vacuum Oil Company, 2 Cir. 1960, 275 F.2d 599, 603-604, cert. denied, 1960, 363 U.S. 844, 80 S.Ct. 1614, 4 L.Ed.2d 1728. Indeed, the district court here reached the “conclusion of law” “ [t]hat the ÑOLA III was not withdrawn from navigation and that the shipowner, Zapata Off-Shore Company, owed a warranty of seaworthiness to the plaintiff”.
Roper v. United States, 1961, 368 U.S. 20, 22-23, 82 S.Ct. 5, 7 L.Ed.2d 1, 3, however, has declared that it is a question of fact whether a vessel is in or out of navigation.1 As Judge Medina had *105observed a year earlier, the problem in these repair cases is that “resort to a mere phrase such as ‘out of navigation’ [does not get] us very far.” Lawlor, 275 F.2d at 602. “[Everything depends upon what we mean by ‘out of navigation’ in the context of the doctrine of unseaworthiness.” Id. at 603. That observation holds true for both the factfinder and the reviewing court.
Modern cases dealing with the issue all go back to West v. United States, 1959, 361 U.S. 118, 80 S.Ct. 189, 4 L.Ed. 2d 161, the opinion that prompted Judge Hand’s utterance. In West, the Government had relegated a/“Liberty” ship of World War II vintage to the “mothball fleet” at Norfolk,/Virginia. There it rested “in total deactivation for several years, with its pipes, boilers, and tanks completely drained, and an oil preservative injected through them to prevent rusting”. 361 U.S. at 119, 80 S.Ct. at 191, 4 L.Ed.2d at 163. In 1951, however, with the Korean conflict at hand, the Government decided it needed the ship. Accordingly, the Government hired an independent contractor to overhaul and reactivate her, but in order to inspect the work in progress placed six of its own men on board. The Government towed the ship to the contractor’s repair docks at Philadelphia and turned it over to the contractor. There, one of the contractor’s shore-based employees was injured by a condition that the contractor had been hired to correct. The Supreme Court, concluding that the warranty of seaworthiness did not apply, advanced a loosely defined standard: “It would appear that the focus should be upon the status of the ship, the pattern of the repairs, and the extensive nature of the work contracted to be done, rather than the specific type of work that each of the numerous shore-based workmen is doing on shipboard at the moment of injury.” 361 U.S. at 122, 80 S.Ct. at 192, 4 L.Ed.2d at 165. That much is clear. The import of the decision, however, has produced confusion.
The Second Circuit in Lawlor concluded from West that “at least in some cases, control of the vessel is a decisive factor”. 275 F.2d at 6042 This Court purportedly adopted the Lawlor reasoning, Moye v. Sioux City & New Orleans Barge Lines, 5 Cir. 1968, 402 F.2d 238, 240, cert. denied, 1969, 395 U.S. 913, 89 S.Ct. 1759, 23 L.Ed.2d 226, but with definite reservations:
It is a mistake * * * to read West as a holding that once all of the owner-charterers’ employees have left the scene the warrant [y] of seaworthiness * * * will evaporate. Rather —apart from the unique factor of a ship long withdrawn from service and the injury occurring during the process of breaking her out of lay-up status— West takes a sensible . but very restrictive stand as to the question of control. It is not approached in a doctrinaire or exclusively physical sense.
*106Id. at 241 (Brown, C. J., concurring).3 In Parker v. Cargill, 5 Cir. 1969, 417 F.2d 772, 774, we observed that “there may be instances in which a shipowner will be liable for injuries to a repairman caused by an unseaworthy condition even though he does not have physical possession of the vessel”.
With deference, we consider that Lawlor overemphasizes control as a “decisive” factor in these unseaworthiness cases and that this overemphasis is premised on a misreading of West. West used control as dispositive in dealing with the petitioner’s alternative argument to seaworthiness. The petitioner in West has contended that the shipowner, apart from the warranty of seaworthiness, had the duty to furnish him a safe place to work. He phrased this duty both in terms of strict liability and in terms of negligence. The Supreme Court concluded that shipowner’s lack of control, in conjunction with the fact that the subject of repair had caused the injury, disposed of these latter two contentions — separately from its discussion of seaworthiness. See Baum v. United States, 5 Cir. 1970, 427 F.2d 215. Cf. Parker, 417 F.2d at 774 n. 5.
That is not to say, however, that the degree of control plays no role in West’s analysis of the seaworthiness doctrine. Id. at 121. West analyzed the applicability of the warranty by reference to the paradigm of Seas Shipping Company v. Sieracki, 1946, 328 U.S. 85, 66 S.Ct. 872, 90 L.Ed. 1099, and its progeny. In concluding that repairman West could not receive the protection of those cases, the Court relied primarily on the historical scope of the doctrine. The vessels in the Sieraeki line of cases were “in the hands and under the control” of the owners and “instead of undergoing general repairs, were in active maritime service”. West, 361 U.S. at 121, 80 S.Ct. at 192, 4 L.Ed.2d at 164. Conversely, in West the vessel was not in maritime service but was undergoing major repairs and complete renovation. “[T]he work to be done was equivalent to ‘home port structural repairs.’ ” Id. West’s conclusion that “the focus should be upon the status of the ship, the pattern of the repairs, and the extensive nature of the work contracted to be done” is tied to its view that Sieraeki protection is appropriate only for “a seaman * * * doing a seaman’s work and incurring a seaman’s hazards”. 361 U.S. at 120, 80 S.Ct. at 191, 4 L.Ed.2d at 164. The inquiry cannot be limited to the “specific type of work that each of the numerous shore-based workmen is doing on shipboard at the moment of injury”, for that would produce “fortuitous results”. The consequence of focusing on the project as a whole is that in a repair project such as that in West “there could be no express or implied warranty of seaworthiness to any person”. 361 U.S. at 122, 80 S.Ct. at 192, 4 L.Ed.2d at 165 (emphasis supplied).4
Thus, West relies on the historical limits of the seaworthiness warranty.5 *107It appears to be a retreat from the wide ranging “humanitarian policy” approach that imposed loss on the shipowner since he could “distribute the loss in the shipping community which receives the service and should bear the cost”. Sieracki, 328 U.S. at 94, 66 S.Ct. at 877, 90 L.Ed. at 1106.6 The inquiry is not who best can bear the risk, but whether the repair project — not the specific task —is one that “seamen” historically have performed. The Supreme Court confirmed this position in Roper:
The view that a vessel not in navigation extends no warranty has often been expressed in the more familiar context of to whom does the warranty extend. Implicit within such cases is the reasoning that those working on vessels not in navigation are not seamen (or doing seamen’s work) and consequently not among those employees protected b'y the warranty of seaworthiness.
368 U.S. at 24 n. 3, 82 S.Ct. at 7, 7 L.Ed.2d at 4 (citation omitted; emphasis supplied).7
In applying West’s standards relating to the status of the ship, the pattern of the repairs, and the extent of the work contracted, to the situation here, we are largely unassisted by the district court’s findings. We are told only that the ÑOLA III was in navigable waters and that an official of Zapata, Mr. Taylor, was in almost constant attendance “overseeing the work that was being done”. Nor can we rely on the court’s “conclusion of law” that the ÑOLA III was in navigation, for it was premised on an incorrect standard.
Undisputed evidence, however, shows additionally that Zapata purchased a drill tender barge and delivered it to Levingston for conversion into an over-the-side drilling barge. This conversion involved drydocking and invading the watertight integrity of the vessel; widening the hull ten feet on each side; fabrication and installation of steel foundations for derrick and engine substructure, shale shaker, other drilling machinery, and material-handling cranes; installation of pipes and wires; and related work. The project originally was invoiced at $159,060. Extra work and the cost of installed machinery raised the cost to about $500,000 — to convert a barge Zapata had purchased for $202,500. The project lasted from two to three months. Work was still in progress at the time of the accident. The vessel had been set afloat again, but a considerable amount of work remained to be performed, and the vessel was not yet able to operate as an over-the-side drilling barge. Watz and a co-worker were still in the process of setting up exhaust pipes for engines on the ma*108chinery deck. Levingston men were working on most of the ship. The vessel had not yet been certificated in its new form by the Coast Guard. The vessel had crew’s quarters, but Zapata had no crew on board. It had a galley, but the galley was not in operation. In fact, the only Zapata representative present was Taylor. We have attempted to glean from the record what his “overseeing” amounted to. Zapata responded to a pretrial interrogatory (read into the record at trial) that “the NOLA 3 was not in the control of Zapata on October 12, 1959. The NOLA 3 was in the sole control, care, and custody of Levingston Shipbuilding Company on that date. Levingston operated a shipyard and was converting the NOLA 3 from its previous status as a tender to an over-the-side drilling barge”. Taylor testified, “I spent a large portion of my time in the vicinity of the shipyard”. When asked on cross-examination, “Was it your job to see they [Levingston Shipyard] did the job the way Zapata wanted it done?”, he replied, “That is correct”. Watz replied affirmatively to the question, “At the time of the accident you were working for Levingston Shipbuilding Company?”
You were not working for Zapata OffShore Company?
No, sir.
And so was your boss and co-worker, they were employees of Levingston Shipbuilding Company?
Yes, sir.
-x- * -x- * * *
There was no one from this Zapata Off-Shore Company directing you in your work that day [the day of the accident], was there?
No, sir.
There is no other pertinent evidence.
We now reexamine the facts as West frames them. (1) As to the status of the ship: the NOLA III, although once drydocked, had been set afloat; considerable work remained to be done, however, and she was neither ready to serve her functions nor certificated by the Coast Guard to do so; she was in the sole control and custody of Levingston Shipyard except for the scrutiny of one Zapata official. (2) As to the pattern of repairs: Substantial repairs were taking place throughout the ship; the very watertight integrity of the vessel was invaded and its hull widened to convert it to another type of vessel; repairs were not limited to a small area. (3) As to the extent of the work contracted for: the NOLA III was to be transformed into a new vessel; extensive changes were necessary, amounting in cost to two and one-half times the price of the vessel.
We conclude from these considerations that the NOLA III repairs, like those in West, were the equivalent of “home port structural repairs”, not the sort of endeavor that seamen traditionally engage in. We reach this conclusion by carrying out West’s instruction to view the repair project as a whole rather than Watz’s specific assignment to fit pipes. With regard to Taylor’s “overseeing”, we note that in West, the Government placed six men on board the ship “to inspect the work and materials to insure compliance with the contract”. 361 U.S. at 119, 80 S.Ct. at 191, 4 L.Ed.2d at 163. This sort of “control” did not alter West’s conclusion that the vessel owed no warranty of seaworthiness. See Guenard v. United States, E.D.La.1968, 278 F.Supp. 310.
Finally, it is true that Roper construes the question of navigation as one of fact not to be reversed unless clearly erroneous. But in this case, the district court made the finding starting under a misapprehension of the factors to be considered in reaching its determination. We do not remand the case because the evidence under a West analysis permits no finding that the vessel was in navigation.8
*109III.
Watz v. Eaton:
Jurisdiction and Laches
Watz sustained his injuries October 12, 1959. The compensation award issued March 4, 1963, Watz filed suit against Zapata May 15, 1967. Zapata impleaded Eaton March 8, 1968. Watz named Eaton as a defendant July 3, 1968.
Eaton contends that Watz’s complaint against it is barred on two grounds: (1) Watz’s cause of action is not in admiralty but in a civil action based on diversity of citizenship, and therefore the Texas two-year statute of limitations9 for personal injuries bars Watz’s recovery; (2) even if the cause of action is in admiralty, it is barred by laches.
Watz originally sued Zapata in admiralty. When Watz amended his complaint to include Eaton as a defendant, he identified the claim as one in admiralty under the provisions of Rule 9(h), Fed.R.Civ.P. We therefore treat the claim in admiralty if we find admiralty jurisdiction.
Watz sustained his injuries on board a vessel afloat in navigable waters. The conduct of which he complains, however, occurred on land with no demonstrated relationship to maritime affairs other than its impact here on Watz. The Plymouth, 1866, 3 Wall. (70 U.S.) 20, 18 L.Ed. 125, affirmed for American courts the view that “the jurisdiction of the admiralty over maritime torts depends upon locality”. 18 L.Ed. at 127. There the Supreme Court dealt with the reverse of our case: the conduct occurred within admiralty jurisdiction, but the damages did not. A fire caused by negligence on board a ship in navigable waters had spread to a wharf and buildings. Although it might appear that admiralty should be concerned with such seaside conduct, the Court declared, “The negligence, of itself, furnishes no cause of action; it is damnum absque injuria”. Since the damages had occurred on shore, the fact that the cause of the injury originated within admiralty jurisdiction, according to the Court, “affords no ground for the exercise of the admiralty jurisdiction”, “the cause of action not being complete on navigable waters”. 18 L.Ed. at 128. Where the “substance and consummation of the wrong were on board the vessel”, on the other hand, admiralty jurisdiction would attach.
In Grant Smith-Porter Ship Company v. Rohde, 1922, 257 U.S. 469, 42 S.Ct. 157, 66 L.Ed. 321, 324, the Court announced, “The general doctrine that, in contract matters, admiralty jurisdiction depends upon the nature of the transaction, and in tort matters upon the locality, has been so frequently asserted by this court that it must now be treated as settled”. Several courts have found the conclusion inescapable that locality is the determining factor and that
it is not the place where the negligent acts occur that is all important, but rather the place where the tort occurs. If A tort is deemed to occur not where the wrongful act or omission has its inception, but where the impact of the act or omission produces such injury as to give rise to a cause of action.
McCall v. Susquehanna Electric Company, D.Md. 1968, 278 F.Supp. 209, 211 (negligent operation of a dam, an extension of land, which resulted in the capsizing of a boat, held within jurisdic*110tion).10 Perhaps most notable are the air disaster cases where courts have found admiralty jurisdiction because the airplane fortuitously crashed into the sea although the alleged negligence was shore-based. See, e. g., Weinstein v. Eastern Airlines, Inc., 3 Cir. 1963, 316 F.2d 758, cert. denied, 375 U.S. 940, 84 S.Ct. 343, 11 L.Ed.2d 271.
Certainly this case meets such a test. Although the negligence and breach of warranty charged against Eaton occurred on land and, at least so far as the proof shows, were not maritime-oriented activity, Watz sustained his injuries on board the NOLA III while it was afloat in navigable waters, the Sabine River. Our earlier conclusion that Zapata owed Watz no warranty of seaworthiness does not defeat admiralty jurisdiction. That related primarily to the responsibility of the shipowner. For example, it is generally accepted that an uncompleted vessel does not offer the warranty of seaworthiness, see, e. g., Frankel v. Bethlehem-Fairfield Shipyard, Inc., 4 Cir. 1942, 132 F.2d 634, cert. denied, 319 U.S. 746, 63 S.Ct. 1030, 87 L.Ed. 1702, yet the Supreme Court has held that admiralty jurisdiction attaches to a tort occurring on board such a vessel when in navigable waters. Grant Smith-Porter Ship Company v. Rohde, supra.
Commentators11 and some courts,12 however, have suggested that more than a mere locality test should be necessary to occasion admiralty jurisdiction. Insofar as the justification for admiralty jurisdiction is “the national interest in a uniform judicial supervision of the maritime industry”,13 it does seem anomalous to premise exercise of that jurisdiction solely on the locale of the injury. The American Law Institute has recommended that the statutory grant of admiralty jurisdiction, 28 U.S.C. § 1333, be amended to include the sentence, “Unless otherwise provided by Act of Congress, the admiralty and maritime jurisdiction does not include a claim merely because it arose on navigable waters”. ALI, Study of the Division of Jurisdiction Between State & Federal Courts, Admiralty & Maritime Jurisdiction § 1316(a), p. 34 (1969). In addition to their recommendation of a statutory amendment, however, the Reporters contend that judicially the question is still open, that the Supreme Court is not committed to the “locality alone” test. Id. at 232. They base this conclusion on two factors: (1) the Court’s locality language has been used only in cases excluding admiralty jurisdiction; (2) in Atlantic Transport. Company of West Virginia v. Imbrovek, 1914, 234 U.S. 52, 62, 34 S.Ct. 733, 735, 58 L.Ed. 1208, 1213, the Court raised doubt as to the requirement by stating:
If more is required than the locality of the wrong in order to give the court jurisdiction, the relation of the wrong to maritime service,' to navigation and to commerce on navigable waters, was quite sufficient.
*111The tort in this ease seems sufficiently related to maritime affairs and commerce to satisfy most demands.14 The mishap occurred to one repairing a ship in preparation for its return to active maritime service. In dealing with tortious activity it is common to focus on the injury in determining the relationship of a claim to a forum. See Restatement (Second) of Conflicts § 146 comment e (proposed official draft 1968). Of course a thorough reweighing of the appropriate limits of admiralty jurisdiction would require analysis of data regarding the principal orientation of the ship repair and conversion industry to determine admiralty’s interest in its regulation, see Black, supra, at 276, and might well raise questions concerning the appropriateness of admiralty jurisdiction over personal injuries. Id. at 278. It might indicate that state courts could adequately manage cases such as the case before us where the “culpable” conduct occurred away from the sea and navigable waters and where the conduct has no demonstrated orientation towards maritime affairs. On the other hand, it might show that uniform regulation by admiralty is necessary to protect workers in maritime-related activities.
In this case we are not furnished with either the data or the arguments so to rework the traditional bounds of admiralty jurisdiction. And although the ALI Reporters may be accurate in their measure of the Supreme Court’s commitment to “locality alone”,15 the language of its decisions leaves us less room to manoeuvre. Grant Smith-Porter Ship Company v. Rohde (post-Im-brovek) stated:
Construing the first question as meaning to inquire whether the general admiralty jurisdiction extends to a proceeding to recover damages resulting from a tort committed on a vessel in process of construction when lying on navigable waters within a state, we answer yes.
257 U.S. at 477-478, 42 S.Ct. at 159, 66 L.Ed. at 325 — this despite the fact that “neither Rohde’s general employment, nor his activities at the time, had any direct relation to navigation or commerce”. 257 U.S. at 476, 42 S.Ct. at 158, 66 L.Ed. at 324.
We conclude that the district court had admiralty jurisdiction of Watz’s complaint against Eaton. The Texas statute of limitations therefore does not apply. Instead, we look to the analogous limitations period of the Jones Act, which is three years. 46 U.S.C. § 688; Flowers v. Savannah Machine & Foundry Company, 5 Cir. 1962, 310 F.2d 135. Since the action was brought after this statute had run, Watz bears the burden of proof with regard to the two elements counteracting a laches defense: excuse for the delay and lack of prejudice to the defendant. McMahon v. Pan American World Airways, 5 Cir. 1962, 297 F.2d 268, 270.
The district court concluded:
That no harm having been shown by reason of the delay for filing the plaintiff’s cause of action and by reason that he did not prevent one witness or any evidence from being heard and by reason of the fact that all the defendants, Zapata Off-Shore *112Company and Eaton Yale & Towne, Ine., had notice within a matter of months after the occurrence that the doctrine of laches or limitations does not apply.
Travelers Insurance Company, carrier for both Zapata and Levingston, notified Eaton and its insurance company of the chain’s failure on April 25, 1960. Eaton had notice by August 1961 that a claim might be asserted against it. Travelers furnished Eaton the hoist and chain without the defective link in July 1961. Although Eaton did not obtain the defective link until one week before trial, the testimony indicates that apparently it made no attempt to do so. We cannot term the district court’s findings of no prejudice clearly erroneous. Rule 52(a), Fed.R.Civ.P.
The district court, however, made no finding regarding excuse for the delay in bringing suit. In Watz’s brief on appeal, he points to no evidence of such excuse except to say that the payment of compensation benefits encourages delay.
Three cases more recent than McMahon v. Pan American World Airways, however, have modified the importance of proving of an excuse for delay. In Molnar v. Gulfcoast Transit, 5 Cir. 1967, 371 F.2d 639, 642, this Court declared that “[ajlthough frequently classified as a separate element, the inexcusability of the delay is closely intertwined with the predominant factor of detriment or lack of detriment, and what is a thin excuse may turn out to be bearable because no harm is suffered” (citations omitted). Akers v. State Marine Lines, Inc., 5 Cir. 1965, 344 F.2d 217, 220, held that unexcused delay without prejudice would not bar a suit. Crews v. Arundel Corporation, 5 Cir. 1967, 386 F.2d 528, 530, concluded that it was settled that “claims based on unseaworthiness are barred only if the delay in filing the libel beyond the three-year Jones Act limitation period is inexcusable and if the delay has prejudiced the defense of the suit” (emphasis original). Since proof of either absence of prejudice or excuse for delay will repel a claim of laches, we conclude that the absence of prejudice in this case justifies the district court’s decision not to dismiss the suit for laches.
IV.
Watz v. Eaton:
Substantive Claims.
We proceed now to the substantive claims. Watz and Eaton couch their respective contentions primarily in terms of Texas law. Campbell advances Pennsylvania law in its brief. Traditionally, courts apply general principles of maritime law in eases subject to admiralty jurisdiction. Insofar as admiralty jurisdiction depends on locality, however, conflicts principles indicate that that conclusion need not inevitably follow. Although principles of negligence and legal cause are often approached in a general common law fashion in many American jurisdictions, we think it appropriate to indicate more specifically the factors that guide us here — in terms of what body of law applies and what that law is.
Conflicts principles suggest that “[i]n an action for a personal injury, the local law of the state where the injury occurred determines the rights and liabilities of the parties, unless, with respect to the particular issue, some other state has a more significant relationship to the occurrence and the parties * * ” Restatement (Second) of Conflicts § 146 (proposed off. draft 1968) (emphasis supplied). Where the tortious conduct and the injury occur in different jurisdictions, “the local law of the state of injury will usually be applied to determine most issues involving the tort”. Id., Comment e, at p. 146. By analogy, admiralty principles apply to this action seeking recovery for personal injuries sustained on board a ship in navigable waters. On this reasoning then, we find Kermarec v. Compagnie Generale Transatlantique, 1959, 358 U.S. 625, 79 S.Ct. 406, 3 L.Ed.2d 550, applicable. *113There, where both the injury and the tortious conduct occurred on board ship, the Supreme Court ruled that the legal rights and liabilities were “measurable by the standards of maritime law”. By the same token, that conclusion might be different should we find a significant policy of Texas — within whose boundaries the injury occurred — or of Pennsylvania — within whose boundaries the chain and hoist were respectively made and assembled — that would be frustrated by such an application.
Admiralty precedent lends partial support to these concerns of conflicts law. In Just v. Chambers, 1941, 312 U.S. 383, 390, 61 S.Ct. 687, 692, 85 L.Ed. 903, 908, the Supreme Court observed that “the maritime law was not a complete and perfect system and that in all maritime countries there is a considerable body of municipal law that underlies the maritime law as the basis of its administration”. Enlarging on this point, the court in The S. S. Samovar, N.D. Calif.1947, 72 F.Supp. 574, 584, said:
As stated by the Supreme Court in the latest of these cases: “With respect to maritime torts * * * the State may modify or supplement the maritime law by creating liability which a court of admiralty will recognize and enforce when * * * not hostile to the characteristic features' of the maritime law or inconsistent with federal legislation.” [citing Just v. Chambers]
By the same token, where the common law of a state in whose territorial waters a maritime tort occurs has been modified or supplemented so as to impose on manufacturers or builders a duty of care to persons not “in privity”, a court of admiralty should recognize that duty and enforce the liability proximately resulting from a breach thereof, if adoption of the state rule “works no material prejudice to * * * maritime law.”
We proceed first to discover what the maritime principles are. The source of substantive American admiralty law originally was the general maritime corpus that had grown up among the commercial nations. As time progressed, however, litigated problems narrowed to more peculiarly national issues. G. Gilmore & Black, The Law of Admiralty, § 1-16, pp. 41-42 (1957). Admiralty courts have felt free to cull what they considered the best principles from the decisions of various courts and from treatise and textwriters. See, e. g., Petition of Kinsman Transit Company, 2 Cir. 1964, 338 F.2d 708, cert. denied, Continental Grain Co. v. City of Buffalo, 380 U.S. 944, 85 S.Ct. 1026, 13 L.Ed.2d 963 (Friendly, J., taking from “all manner of ‘respectable’ authority, Bohlen, Prosser, Seavey, and cases from California, New York, Tennessee, Texas, Wisconsin, not to mention English cases. Some are maritime and some are not. In the end, the rule applied is substantially the common law rule”. J. P. Lucas, Admiralty: Cases & Materials, 328 (1969)). Thus, in Kermarec, the Supreme Court declared that the standard of care for a shipowner “must be decided in the performance of the Court’s function in declaring the general maritime law, free from inappropriate common-law concepts”. In fulfilling that function, the Supreme Court rejected the “semantic morass” of the common law and its “conceptual distinctions” “foreign to [admiralty’s] traditions of simplicity and practicality”. 358 U.S. at 630, 631, 79 S.Ct. at 410, 3 L.Ed.2d at 554, 555.
Although the process has afforded reasonably definable standards for shipowners, seamen, and others related to maritime contracts, the standards for a case such as ours are difficult “because cases brought in admiralty in which neither party is a shipowner or in some way related to a maritime contract are very rare”.16 Lucas at 328. Perhaps *114one of the reasons is that an action for unseaworthiness against a shipowner appeals to plaintiffs since negligence need not be proved. 54 Geo.L.J. 1439 n. 2 (1966). On the other hand, some generally applicable principles have emerged. Sinram v. Pennsylvania R. R., 2 Cir. 1932, 61 F.2d 767, 770, for example, incorporated into admiralty the doctrine of Palsgraf v. Long Island R. R., 1928, 248 N.Y. 339, 162 N.E. 99, that a tortfeasor was not chargeable for consequences to those to whom he owed no duty. Sieracki v. Seas Shipping Company, 3 Cir. 1945, 149 F.2d 98, aff’d, 1946, 328 U.S. 85, 66 S.Ct. 872, 90 L.Ed. 1099, adopted the principle of MacPherson v. Buick Motor Company, 1916, 217 N.Y. 382, 111 N.E. 1050, and the Restatement of Torts § 395 (1934) disposing of privity.
The first question we deal with is the standard of care. With regard to Watz’s claim against Eaton, the district court found that
[t]he chain hoist, after assembly, was tested by Yale & Towne Manufacturing Company and was found to be capable of lifting weights of one and one-half times its rated capacity, expressed in long tons, or 3360 pounds. Other tests, such as radiographic testing, ultra-sonic testing, dye-check testing, and magnafluxing testing could have revealed the imperfect weld. That the only test of the load chain and hoist sold by Yale & Towne Manufacturing Company was a straight lifting load test of one and one-half tons dead weight, and such test was not sufficient to reveal the defective weld in the link that failed.
•X -X- X -X- X -X-
That the link of the chain which was broken, broke above the hook of the chain hoist. That a one ton Yale & Towne Load King hand hoist is a dangerous object when a defect exists such as the defect in the link of chain, and resulted in an unreasonable risk by producing substantial bodily harm when used in the manner intended.
This case is in many respects similar to Sieracki as it was treated in the lower courts. In Sieracki a longshoreman sued the shipowner for unseaworthiness of the vessel. The shipowner impleaded two third-party defendants and the longshoreman thereupon amended his complaint to include a claim against them as well. These defendants were respectively the shipbuilder and the installer of a ten-ton boom and tackle. The boom and tackle had injured the longshoreman when the shackle supporting it broke. The third-party defendants contended that the evidence was insufficient to establish their negligence. They had purchased the shackle from another company (which in turn had purchased it from still another company). The third-party defendants also contended *115that they had taken off customary safeguards. But the Third Circuit concluded that “even if they had, that would not necessarily excuse them for while usage is relevant in determining whether reasonable care is exercised it is not conclusive in (with one or two exceptions) establishing compliance with the legal standard”. 149 F.2d at 100. As the district court did here, the Third Circuit noted that the shackle had been subjected only to a lifting test, and that an x-ray or a tapping test would have disclosed the defect. “This conclusion, as would be expected, was not received with enthusiasm by the [third-party defendants]. Nevertheless, we think it is sound.” Id.
Like the third-party defendants in Sieracki, Eaton does not receive the district court’s conclusions here with enthusiasm. It argues that it did sufficiently test the chain: “(a) random samples of the chain were tested and pulled to destruction; (b) each chain was cut and visually inspected before it was installed on the chain hoist; and finally (c) each chain and chain hoist were then tested to see if the chain and chain hoist could hold 3360 pounds. * * * ” Furthermore, Eaton contends that the evidence shows that other tests were too expensive or not reliable. But in all this Eaton faces the significant obstacle of showing the district court’s findings to have been clearly erroneous. Rule 52(a), Fed.R.Civ.P. There was specific expert testimony that other tests would have revealed a defective weld and that such tests are customarily used in the industry. We conclude that the district court’s finding that sufficient care had not been used is not clearly erroneous. The standard of care violated by Eaton would not in this case vary if we applied the law of one of the states. And of course MacPherson v. Buiek Motor Company extended the remedy for negligence beyond the immediate purchaser of a chattel. That case has become “so widely accepted as to be construed as a part of the general law of torts, maritime as well as common law”. Sieracki, 149 F.2d at 100. See also Restatement (Second) of Torts § 395 (1965). Watz therefore has a remedy against Eaton in admiralty. Furthermore, we agree with the Third Circuit in Sieracki that purchase of a component part from another source does not relieve a defendant from his liability as manufacturer.
Eaton contends, however, that the chain link was not defective. Partly this argument goes to the evidence and partly it goes to the question of legal cause. Eaton maintains that even if it be true that the link was only 42.6 percent fused, that factor did not produce the mishap and injury. It bases its argument on evidence that other links sustained their loads even when cut sixty percent through the weld, and on the district court’s finding that the “imperfection in the weld would not have caused the accident but for the accompanying damage and injury to the chain caused by the external forces”. There is ample evidence, however, to sustain the district court’s conclusion that the defective weld contributed to the failure of the chain in this instance.
With regard to legal cause, Eaton argues that the abuse to which the hoist was subjected constituted a “new and independent cause, breaking the link of causation” and was not foreseeable. The district court found
That the worn place caused by external damage in the link which failed contributed to cause the failure of the link at the time of plaintiff’s injury, but the failure at the time of plaintiff’s injury would not have occurred but for the defective weld in the link of the load chain.
Prosser writes of such circumstances:
On its face, the problem is one of whether the defendant is to be held liable for an injury to which he has in fact made a substantial contribution, when it is brought about by a later cause of independent origin, for which he is not responsible. In its essence, however, it becomes again a *116question of the extent of the defendant’s original obligation; and once more, the problem is not one of causation at all, since it does not arise until causation is established. It is rather one of the policy as to imposing legal responsibility. * * *
In general, this has been determined by asking whether the intervention or the later cause is a significant part of the risk involved in the defendant’s conduct, or is so reasonably connected with it that the responsibility should not be terminated. It is therefore said that the defendant is to be held liable if, but only if, the intervening cause is “foreseeable.”
Prosser on Torts, 3d ed. 1964, § 51, at 309-11. To deal with Eaton’s argument, then, we should have to decide whether the abuse to which the hoist was subjected was foreseeable. On occasion, we have required a defendant to contemplate the future negligence of others. See, e. g., Transcontinental Gas Pipe Line Corp. v. Mobile Drilling Barge, 5 Cir. 1970, 424 F.2d 684; Prosser at 313. But we cannot tell from this record whether hoistmakers can expect theii' hoists to be subjected to the abuse shown here.
It is not always necessary, however, to find intervening conduct specifically foreseeable. Courts have held tort-feasors responsible when their conduct threatens a particular sort of result and an unanticipated force intervenes to produce that result. Prosser at 326. Here, despite Eaton’s protestations as to the carrying capacity of a 42.6 percent fused link, there was testimony that such a link was dangerous, increasing the likelihood of failure of the link, and that a prudent manufacturer would discard the link as unsafe and unsuitable for lifting a weight of up to 2000 pounds (the capacity of the hoist). Thus, the danger created by Eaton’s conduct was the likelihood that this link would fail in lifting weights within the hoist’s capacity. It did fail in just such a manner and the result was thus within the scope of Eaton’s negligence.
This concept has already been incorporated into admiralty. In Kinsman Transit, Judge Friendly wrote:
We would find it difficult to understand why one who had failed to use the care required to protect others in the light of expectable forces should be exonerated when the very risks that rendered his conduct negligent produced other and more serious consequences to such persons than were fairly foreseeable when he fell short of what the law demanded.
338 F.2d at 723-724.”17 The Restatement provides that “[i]f the actor’s conduct is a substantial factor in bringing about harm to another, the fact that the actor neither foresaw nor should have foreseen the extent of that harm or the manner in which it occurred does not prevent him from being liable”. Restatement (Second) of Torts § 435(1) (1965). Perhaps this principle has not previously been stated so specifically in admiralty.18 But in our view, Eaton’s negligence is established, and we consider that the risk it created encompassed the sort of injury that occurred. We therefore do not impose on Eaton after-the-fact standards for its primary conduct. Finally, we see no significant policies of other states to be frustrated by this result.
*117We would reach the same result, moreover, by applying Texas law, the state in which this accident occurred. In Biggers v. Continental Bus System, Inc., 1957, 157 Tex. 351, 303 S.W.2d 359, 365, the Texas Supreme Court reaffirmed its rule that a tortfeasor “should have anticipated the danger to others created by his negligent act, and the rule does not require that he anticipate just how injuries will grow out of that dangerous situation”. (Quoting from Sullivan v. Flores, 134 Tex. 55, 132 S.W.2d 110, 111). In Biggers, the tortfeasor was driving a bus at an excessive rate of speed in its own lane of traffic. An automobile proceeding in the opposite direction negligently knocked another automobile into the path of the bus. In the absence of either of the two factors, the accident would not have occurred. But the Texas Supreme Court concluded that “[i]f negligent and excessive speed [of the bus] proximately contributed to cause the collision it is immaterial to [its] liability that the negligence of [the negligent automobile driver] also contributed thereto or that such concurring cause may not have been reasonably foreseeable. It is a sufficient predicate of liability that the excessive speed of the bus was a substantial factor in bringing about the collision”. 303 S.W.2d at 367 (citations omitted).19
Because we conclude that the district court properly held Eaton liable on the basis of its negligence we do not reach the issue of what warranty it owed Watz with respect to its hoist.20
V.
Eaton v. Campbell
On the fourth-party complaint of Eaton against Campbell we confront first a jurisdictional challenge. Campbell maintains that "the fourth party complaint, predicated on theories of negligence and breach of implied warranty arising out a sale taking place in the State of Pennsylvania, is not within the admiralty jurisdiction of the court”.
Rule 14(c), Fed.R.Civ.P., provides that “[w]hen a plaintiff asserts an admiralty or maritime claim within the meaning of Rule 9(h), the defendant or claimant, as a third-party plaintiff may bring in a third-party defendant who may be wholly or partly liable, either to the plaintiff or to the third-party plaintiff, by way of remedy over, contribution, or otherwise on account of the same transaction, occurrence, or series of transactions or occurrences”. In this case, Watz has asserted an admiralty claim under Rule 9(h) against Eaton and Eaton has brought in Campbell. Campbell’s liability, if proved, arises out of the same occurrence as Eaton's: the injury to Watz on board the ÑOLA III.
Campbell bases its contention that jurisdiction does not exist on cases holding that Rule 14(c) carries forward the judicially-imposed requirement of Admiralty Rule 56 that a third-party action be *118maritime in character and that ancillary jurisdiction alone will not suffice. See, e. g., McCann v. Falgout Boat Company, S.D.Tex.1968, 44 F.R.D. 34. Without expressing any view on the correctness of a decision such as McCann, we note that the third-party complaint there was distinctly not maritime. In McCann, a seaman sued the shipowner for an injury sustained on board a ship. The shipowner impleaded the doctor who had later treated the injury in San Antonio, Texas, for negligence and malpractice in treatment of the injury. Using a locality test, the district court found that no admiralty jurisdiction existed over the impleader because the medical malpractice tort arose solely in San Antonio. In our case, on the other hand, the consequence of Campbell’s conduct was injury on board the ÑOLA III. The same reasoning that led us to conclude that admiralty jurisdiction existed over Watz’s claim against Eaton sustains admiralty jurisdiction over Eaton’s claim against Campbell. And since the district court had admiralty jurisdiction, Campbell's claim to a jury trial was properly denied.
Campbell raises another threshold contention: that Eaton’s action is barred by laches.21 In this regard, the district court found:
Yale & Towne Manufacturing Company was notified of the failure of the chain hoist in question on or about April 25, 1960, and received notice by August of 1961 that claim might be asserted against them by reason of the failure of the chain hoist. Yale & Towne Manufacturing Company did not give notice of any kind [to] Campbell Chain Company until on or about the 22nd day of April 1968.
* * * * * -X-
At the time Eaton Yale & Towne, Inc. gave notice to Campbell Chain Company, the records of Yale & Towne Manufacturing Company with respect to the purchase of the chain had been destroyed in the ordinary course of business and no invoices or other records concerning such purchase were available. Likewise, at such time, the records of Campbell Chain Company with respect to the sale of the chain to Yale & Towne Manufacturing Company had been destroyed, as well as the specific inspections of chain sold in 1957, as well as detailed records with reference to the manufacture of the chain or any unusual incidents which might have occurred during the course of such manufacture.
These findings of fact suggest that the district court intended to hold that Campbell had sustained its defense of laches on Eaton’s part. Yet although the district court found specifically that Zapata and Eaton had failed in their lach-es defenses, it made no finding with regard to Campbell and instead awarded Eaton recovery against Campbell. The picture is further complicated by the fact that the district court had approved a pre-trial stipulation by Eaton and Campbell that all issues with respect to Campbell’s claim of “delayed notice, laches and limitations, and the effect thereof” would be postponed until after the trial of the other issues and that the district court would fix a date for taking relevant evidence when the “issues become material and decision required thereon”.
We conclude in this context that we can make no decision on the question of laches. We vacate the district court’s order of recovery against Campbell and remand for further proceedings. Unless the district court should find the parties waived the right to a hearing on these issues, they should be permitted to develop the evidence before the court.
Not knowing what will develop on the laches defense, however, we shall deal with the other issues in Eaton’s claim against Campbell while the case is here.
The district court made the conclusion of law that “[t]he load chain was negli*119gently manufactured. * * * Such negligence was a proximate cause of the plaintiff’s injuries and damages”. The parties have construed this finding as directed to negligence on the part of Campbell.
Thus the negligence found against Campbell differs to some extent from that found against Eaton. The district court found that Eaton “was negligent in failing to perform other tests which would have revealed the defective weld. * * * ” No such specific finding was made with regard to Campbell. The pertinent factual findings with regard to Eaton and Campbell are contained in the same paragraph:
The chain and link in question were tested by Campbell Chain Company and found to be capable of bearing weights in excess of 8500 pounds. The rated capacity of the chain hoist was 2000 pounds. The chain hoist after assembly, was tested by Yale & Towne Manufacturing Company and was found to be capable of lifting weights of one and one-half times its rated capacity, expressed in long tons, or 3360 pounds. Other tests, such as radio-graphic testing, ultra-sonic testing, dye-check testing, and magnafluxing testing could have revealed the imperfect weld. That the only test of the load chain and hoist sold by Yale & Towne Manufacturing Company was a straight lifting load test of one and one-half tons dead weight, and such test was not sufficient to reveal the defective weld in the link that failed.
The apparent explanation for the district court’s failure to make specific findings on the number of tests performed by Campbell and the adequacy of its tests is the court’s finding, already quoted, that Campbell had destroyed the relevant records in the ordinary course of business. In other words, specific evidence of the sort relating to Eaton was not available regarding Campbell’s procedures.22
Evidence did show and the district court did find, however, that the original weld on the link in question was defective. We have already approved the court’s finding of a substantial connection between that defect and Watz’s injuries. It is Campbell's argument, however, that no evidence shows in what manner Campbell was negligent either in its manufacturing or testing processes.
Certainly Eaton bore the burden of proof to show Campbell’s negligence. But once it was proved that a defective weld had occurred during the manufacture of the chain by Campbell, we believe that the district court sitting as a finder of fact could reasonably infer negligence from that circumstantial evidence. Campbell objects that the pleadings and evidence did not specifically raise the doctrine of res ipsa loquitur and the district court did not refer to it. The evidence credibly established that responsible chain manufacturers attempt to avoid defective welds in the knowledge that they are dangerous. The finder of fact could reasonably infer that a defective weld would ordinarily not occur in the absence of negligence. We see no reason to invoke the Latin phrase here. We simply apply a rule of circumstantial evidence, not changing the burden of proof or casting presumptions against the defendant. See Restatement (Second) of Torts § 328D & Comment (1965). For that matter, in viewing the district court’s order as a whole, we think its finding that certain tests would have revealed a defective weld supports its conclusion of Campbell’s negligence.
Our conclusion that Eaton’s negligence was a legal cause of Watz’s injury was based on the defective link’s role in the mishap. Since we affirm the district court’s finding that Campbell was negligent, we may also conclude that Campbell’s negligence was a legal cause of Watz’s injuries unless Campbell owed no duty to Watz or Eaton’s negligence exonerates it. We believe that Watz was *120within the scope of the risk of Campbell’s conduct. In the sense of foreseeability, a chain manufacturer selling chain to a hoistmaker must be held to knowing that one danger resulting from a negligently made chain is that it will break on the hoist thus dropping its load to the damage of the load and to the injury of anyone beneath it. Thus, Pals-graf, adopted into admiralty by Sinram, does not foreclose Campbell’s liability. Our earlier discussion of legal cause disposes of any idea that Eaton’s negligence should cure Campbell’s.
Because we agree with the district court that on the evidence Campbell was negligent and that it thereby caused Watz’s injury, we do not reach the issues whether it breached a warranty of reasonable fitness, whether such a warranty exists in admiralty, and if so to whom it extends, or the effect of the specifications on any such warranty.
We turn finally to the contribution and indemnity claim of Eaton against Campbell. Since we have concluded that both Eaton and Campbell were negligent, we have a case of joint tortfeasors. This Court has already held that the apparent prohibition of Halcyon Lines v. Haenn Ship Ceiling & Refitting Corporation, 1952, 342 U.S. 282, 72 S.Ct. 277, 96 L.Ed. 318, against contribution between joint tortfeasors in admiralty does not apply to a situation where Watz as plaintiff could have proceeded against both Eaton and Campbell and could have recovered damages from both. Horton & Horton, Inc. v. T/S J. E. Dyer, 5 Cir. 1970, 428 F.2d 1131 [No. 28,706]. Halcyon is distinguished by the fact that there a statute precluded recovery by the plaintiff against the impleaded party.23 With Halcyon inapplicable, we agree with the district court that contribution was proper here. But Eaton argues for complete indemnity. It contends that its own negligence in failing to perform sufficient tests was passive whereas Campbell’s negligence in manufacturing a defective chain was active.
We are hard put to label Eaton’s negligence passive. The active-passive distinction, when made, usually contrasts the conduct of one who simply had custody of the cause of injury and failed to discover the danger created by another. But Eaton did not simply fail to discover a defect in Campbell’s product. Here, as the district court found, Eaton constructed a hand hoist of which the chain was a component part. Eaton had an affirmative duty safely to manufacture the hoist, and the hoist included the chain. As the district court found, “a one ton Yale & Towne Load King hand hoist is a dangerous object when a defect exists such as the defect in the link of chain”. Eaton’s negligence was not simply its failure to perform sufficient tests, but “in placing in trade and commerce the hoist with load chain attached with the defective weld” (emphasis supplied). Consequently, the district court properly denied the claim of complete indemnity. See Transcontinental Gas Pipe Line Corp. v. Mobile Drilling Barge, 5 Cir. 1970, 424 F.2d 684.
Affirmed in part; reversed in part; remanded in part.
. Roper involved a “mothballed” Liberty ship. “[H]er supplies, stores, nautical instruments, cargo gear and tackle were removed; her pipes and machinery were drained and prepared for storage; and her rudder, tail shaft and propeller were se*105cured.” Both her Coast Guard safety certificate and her license to operate had been withdrawn. In 1954, however, the Government decided to use some of the Liberty ship’s holds for grain storage. The vessels were towed to loading facilities, filled with grain, and returned to the deactivated fleet. Then in 1956, this particular vessel was towed back to the loading facility and unloaded. Roper, an employee of Continental Grain, sustained injury during the unloading operations from a defective part of Continental’s unloading equipment. The Supreme Court viewed the issue whether the ship was in navigation as a question of fact and refused to overrule the trial court’s finding that it was out of navigation.
. In Lawlor, the vessel was undergoing an annual overhaul. She had been in dry-dock but was moored in navigable waters at the time the accident occurred. Electric power and water were supplied from shore. A full crew of officers and men were aboard performing seamen’s duties. Xo major or substantial changes or repairs had been required, and the shipowner was in general control of the vessel although the shipyard had control of the particular repair area where the mishap occurred. The Second Circuit found that Lawlor was protected by the warranty of seaworthiness.
. In Moye, the shipyard had exclusive control of a barge. No representative of the owner nor crew was present. Although no major repairs or structural changes were involved, the barge required drydocking for repairs in its side and deckplating and hatch covers. We concluded that it owed no warranty of seaworthiness.
. The TVest Court buttressed its conclusion by emphasizing that there was no uncertainty regarding the status of the ship. The vessel,
as anyone could see, was not in maritime service. She was undergoing major repairs and complete renovation, as the petitioner knew. Furthermore, he took his orders from the contractor, not the shipowner. He knew who was in control. This undertaking was not “ship’s work” but a complete overhaul of such nature, magnitude, and importance as to require the vessel to be turned over to a ship repair contractor and docked at its pier for the sole purpose of making her seaworthy. It would be an unfair contradiction to say that the owner held the vessel out as seaworthy in such a case.
361 U.S. at 122, 80 S.Ct. at 192, 4 L.Ed. 2d at 165 (emphasis supplied).
. For cases considered outside the bounds of the historical limits see Comment, Seamen & the Warranty of Seaworthiness in Maritime Injuries — Sieracki Today, 34 Tulane L.Rev. 572, 578-85 (1960).
. Some commentators find Roper even more of a retreat. See, e. g., H. R. Baer, Admiralty Law of the Supreme Court § 6-11, at p. 174 (2d ed.1969); Comment, A New Look at the Unseaworthiness Doctrine : The Roper Case, 29 U.Chi.L.Rev. 519 (1962).
. The position does find support in Sierac-ki, however, for the Sieracki Court justified its extension of the warranty to longshoremen on the grounds that “[historically the work of loading and unloading is the work of the ship’s service, performed until recent times by members of the crew”. 328 U.S. at 96, 66 S.Ct. at 878, 90 L.Ed. at 1107. “[T]he owner should not be free to nullify [the liability] by parcelling out his operations to intermediary employers whose sole business is to take over portions of the ship’s work or by other devices which would strip the men performing its service of their historic protection.” 328 U.S. at 95, 66 S.Ct. at 877, 90 L.Ed. at 1106. We too have said that “the limiting factor of tradition * * * requires that the injured shore-based worker be engaged in work traditionally that of a seaman, excluding those persons performing such tasks as making major repairs requiring drydocking or special skills”. Atkins v. Greenville Shipbuilding Corp., 5 Cir., 1969, 411 F.2d 279, 282 Cert. denied, 396 U.S. 846, 90 S.Ct. 105, 24 L.Ed.2d 96. (Citations omitted).
Tetreault, Seamen, Seaworthiness, & the Rights of Harbor Workers, 39 Cornell L.Q. 381 (1954), seriously attacks the Sieracki Court’s reading of history. But West appears simply to take the Sieracki history as it stands.
. This Circuit appears to have developed another doctrine with regard to repair contracts that would also exempt Zapata. Patterson v. Humble Oil & Refining Co., 5 Cir. 1970, 423 F.2d 883 states that *109the owner’s warranty of seaworthiness does not extend to noncrew members regarding transitory conditions created by an outside repair crew during the course of substantial repairs to an existing unseaworthy condition when the transitory condition relates to the subject matter of the repair contract. See also Parker v. Cargill, 5 Cir. 1969, 417 F.2d 772. We think it preferable, however, to rely here on West’s broader principles directing attention to the status of the ship, the extent of the work, and the pattern of the repairs.
. Art. 5526, Vernon’s Ann.Tex.Civ.Stat.
. Weinstein v. Eastern Airlines, 3 Cir. 1963, 316 F.2d 758, 765 (airplane crash into sea because of negligence in inspection and maintenance on land is within admiralty jurisdiction) ; Smith v. Lampe, 6 Cir. 1933, 64 F.2d 201, cert. denied, 1933, 289 U.S. 751, 53 S.Ct. 695, 77 L.Ed. 1496 (blowing of automobile horn on land causing damage to barge when tug believed horn to be fog signals held, within admiralty jurisdiction) ; Southern Bell Telephone & Telegraph Co. v. Burke, 5 Cir. 1933, 62 F.2d 1015 (negligently maintained telegraph wires not in or covered by navigable waters damaged ship’s smokestack; held, within admiralty jurisdiction) .
. See, e. g., 1 Benedict, Admiralty 350-51 (6 Knauth ed. 1940)'. Other commentators are mentioned in ALI, Study- of the Division of Jurisdiction Between State & Federal Courts, p. 232 (1969).
. See, e. g., Smith v. Guerrant, S.D.Tex. 1968, 290 F.Supp. 111, noted in 44 Tulane L.Rev. 166 (1969); Chapman v. City of Grosse Pointe Farms, 6 Cir. 1967, 385 F.2d 962; Thomson v. Chesapeake Yacht Club, Inc., D.Md. 1965, 255 F.Supp. 555; McGuire v. City of New York, S.D.N.Y. 1961, 192 F.Supp. 866.
. Black, Admiralty Jurisdiction: Critique & Suggestions, 50 Colum.L.Rev. 259, 261 (1950).
. Those cases which have demanded some maritime connection see note 12 supra, have demanded only a minimum. They have refused to assume admiralty jurisdiction in cases involving swimmers, a person falling off a dock, a crane breaking and dropping a lift-truck into the water.
. The apparent fact that ‘‘locality alone” has been used only to exclude admiralty jurisdiction is not particularly persuasive, since exclusion of admiralty jurisdiction on that basis seems no more sensible than its extension. See, e. g., Swain, Yes, Virginia, There is an Admiralty: The Rodrique Case, 16 Loyola L.Rev. 43, 46 (1969); ALI, Study of the Division of Jurisdiction, supra, at p. 230. And Imbrovek’s invitation to doubt the test of locality alone seems qualified by Rohde.
. Of three analogous cases which we have discovered, only one refers to the problem. In Simpson Timber Co. v. Parks, a longshoreman was injured on board ship when he stepped through the packaging of a bundle of doors. He sued the shipowner *114and the manufacturer of the doors. In that case’s long journey through a Ninth Circuit three-judge panel (opinion and dissent), 9 Cir. 1966, A.M.C. 1081, en banc (opinion and dissent), 369 F.2d 324, the Supreme Court (per curiam vacating of the judgment), 1967, 388 U.S. 459, 87 S.Ct. 2115, 18 L.Ed.2d 1319, two law review comments, 66 Colum.L.Rev. 1190 (1961); 54 Geo.L.J. 1439 (1960) and remand 390 F.2d 353, cert. denied, 393 U.S. 858, 89 S.Ct. 126, 21 L.Ed.2d 127, the issue was not mentioned. Nor was the issue framed specifically in terms of what admiralty law said. The case appears to have proceeded on undifferentiated common law principles.
In McFall v. Compagnie Maritime Belge, 1952, 304 N.Y. 314, 107 N.E.2d 463, a longshoreman was injured by carbon tetrachloride fumes in a ship’s hold. He sued the vessel’s bareboat charterer and Dow Chemical, the manufacturer and shipper of the carbon tetrachloride. In discussing Dow’s negligence, the court made no reference specifically to maritime or any other law.
But in Todd Shipyards Corp. v. United States, D.Me.1947, 69 F.Supp. 609, a shipyard sued the manufacturer of a boom block for damage caused to its property when the block broke. Concluding that a maritime tort had occurred, the district court said “that the applicable principles of the law involved are a part of the general law of torts, maritime as well as common law”.
. Ours is not the tenuous relationship that Kinsman Transit admitted might be ex- . eluded. 338 F.2d at 725.
. But cf. Horton & Horton, Inc. v. T/S J. E. Dyer, 5 Cir. 1970, 428 F.2d 1131 [No. 28,706]:
Moreover, in determining causation in maritime matters, the applicability of such doctrines as last clear chance and the interruption of negligence by a subsequent intervening negligence is questionable. “[T]lie maritime court has been less ready than the shore courts to find that a subsequent wrongful act by one party breaks the chain of causation connecting the accident with the prior negligence of the other party.” Gilmore and Black, The Law of Admiralty, p. 404 (1957). See also Commercial Transport. Corp. v. Martin Oil Service, Inc., 7 Cir. 1967, 374 F.2d 813, 817.
. See Restatement (Second) of Torts § 442B (1905) :
Where the negligent conduct of the actor creates or increases the risk of a particular harm and is a substantial factor in causing that harm, the fact that the harm is brought about through the intervention of another force does not relieve the actor of liability, except where the harm is intentionally caused by a third person and is not within the scope of the risk created by the actor’s conduct.
Comment b, at p. 470:
This is to say that any harm which is in itself foreseeable, ns to which the actor has created or increased the recognizable risk, is always “proximate,” no matter how it is brought about, except where there is * * * intentionally tortious or criminal intervention, and it is not within the scope of the risk created by the original negligent conduct.
. Compare Schaeffer v. Michigan-Ohio Navigation Co., 6 Cir. 1969, 416 F.2d 217 with Noel v. United Aircraft Corp., D.Del. 1962, 204 F.Supp. 929; See Montgomery v. Goodyear Tire & Rubber Co., S.D.N.Y. 1964, 231 F.Supp. 447; McCune, Maritime Products Liability, 18 Hastings L.J. 831 (1967).
. Campbell also argues that the Pennsylvania Statute of Limitations bars Eaton's action. Since we have found admiralty jurisdietion, however, laches is the applicable standard. See text, part III, supra.
. Prejudice, if any, deriving from this factor should more properly be considered on Campbell’s laches defense.
. To the extent that Hartford Accident & Indemnity Co. v. Gulf Refining Co., 5 Cir. 1956, 230 F.2d 346, 354, cert. denied, Gulf Refining Co. v. Black Warrior Towing Co., 352 U.S. 832, 77 S.Ct. 49, 1 L.Ed.2d 52, is inconsistent with this result, we must treat it as controlled by Horton & Horton.