(dissenting).
I believe the majority’s decision in this case results in an erroneous disposition of this appeal on two grounds. First, it considers the breach of duty by defendant, if any, to have occurred in January, 1977, rather than in May, 1977, when plaintiff demanded return of the bail money deposited with the clerk. Because the breach occurred in May and suit was commenced in September, the action was brought within six months of the breach and is therefore timely regardless of whether or not we apply the “discovery” rule. Second, I believe the discovery rule should be applied, making the action timely regardless of the date of defendant’s breach of duty, so long as the section 613A.5 conditions are fulfilled within the required times after discovery of the breach, or after it should have been discovered in the exercise of reasonable diligence.
I. Nature of defendant’s duty.
A clerk of district court acts as a trustee of funds deposited in his office.1 He does not merely owe a duty to retain the funds intact, preventing negligent release of them as here claimed, but a duty to pay them out to the rightful claimant upon demand. When the demand is made, and the clerk fails to comply, the breach of duty occurs, and it has uniformly been held that this is when time-barring statutes commence to run, without any consideration of the “discovery” rule or expanded definition of “accrued.” In Washburn Land Co. v. Sanborn, 150 Wis. 562, 137 N.W. 782 (1912), former owners of land deposited security with an attorney, in lieu of a deposit with the clerk as required by law, in order to attack the validity of a tax sale of the property. The court said that:
The check was left with Sanborn as a deposit for a specific purpose, and, until the case was finally determined, it could not be withdrawn without loss of the *919right to defend the action. Sanborn really was in the position of trustee whose duty it was to pay the money to his former clients if they were beaten in the suit, or to McLeod if he was unsuccessful. So long as the litigation was in progress, neither of the parties could withdraw the money without jeopardizing their rights, and it was Sanborn’s duty to pay it to the party finally entitled thereto, and the statute of limitations did not begin to run until payment was demanded. It would hardly be contended that if this deposit had been made with the clerk of the court, as the statute requires, that officer could at the close of the litigation, regardless of any statute requiring him to pay the money, refuse to pay it over because it had been on deposit for more than six years, and there is little difference between the two situations. (Emphasis added.)
Sanborn, 150 Wis. at 568, 137 N.W. at 785.
In Farmers & Merchants Bank v. Duke, 112 Ind.App. 589, 44 N.E.2d 172 (1942), an owner of bonds deposited them in a bank for safekeeping. The deposit was made in 1920. In 1926 they were stolen. Twelve years later demand was made for their return. The court held that the six-year statute of limitations would not prevent the action brought for their recovery, saying that:
There was no breach until there was a demand by appellees or repudiation by appellants. The mere fact that the bonds were stolen did not constitute such a breach. They might have been recovered or substitution made before demand. The contract was a continuing one and the statute of limitations did not begin to run until it was breached.
Duke, 112 Ind.App. at 594, 44 N.E.2d at 174. The breach was held to have occurred in 1938, upon demand for the bonds’ return, despite the fact they were lost in 1926 and the owner knew of it immediately.
The court’s conclusion in Duke that a breach did not occur until demand and refusal, because the custodian might have recovered them after their loss, is in accordance with the general rule. At 26A C.J.S. Deposits in Court § 9d(l) the rule is stated that “[wjhere a fund in court is paid out to a person not entitled thereto, the court will order its restoration, as where the order was procured by fraud.” Section 9(d)(2) states that “[i]n case money is improperly paid out of, or withdrawn from, court, the person rightfully entitled thereto may proceed by motion to compel its restoration. The statutes of limitation are not a bar to such a proceeding . . . The last statement was quoted with approval by our court in State v. Rudolph, 240 Iowa 726, 732, 37 N.W.2d 483, 487 (1949). In that case, property was attached, then sold. The proceeds of sale were delivered to the clerk to hold in lieu of the attached property. The clerk paid it to a party who was not entitled to it. We said that:
The original property, after its attachment, was in the custody of the court. The money represented the attached property in a different form and was turned over by the sheriff to the clerk. That official, without authority, paid it to Richard D. Rudolph [the wrong party]. It was then in custodia legis.
Rudolph, 240 Iowa at 731, 37 N.W.2d at 486.
In quoting from a North Dakota case, this court said:
“The power of the court over moneys in its custody continues until they are distributed pursuant to final decrees in the cases in which the moneys are paid, and if from any cause such moneys are previously withdrawn without authority of law, the court can, by summary proceedings, compel their restitution. Until a decree of distribution is made and enforced, the summary power of the court to compel restitution remains intact.”
Rudolph, 240 Iowa at 731-32, 37 N.W.2d at 486, quoting Agricultural Bonds & Credit Corp. v. Courtenay Farmers’ Cooperative Association, 66 N.D. 122, 131, 262 N.W. 453, 457-58 (1935).
In National Automobile & Casualty Insurance Co. v. Pitchess, 35 Cal.App.3d 62, 110 Cal.Rptr. 649 (1973), a sheriff held funds under garnishment and by error paid *920them out to a party not entitled to them in March, 1967. Plaintiff, who was entitled to the funds, made demand upon the sheriff for their return on October 9, 1969. The court held that the sheriff breached his fiduciary duty in 1967 when he paid the funds to the wrong party, but that the suit commenced on September 10, 1970, was timely under a one-year statute of limitations. It said:
When the sheriff received the money from Nelson & Belding on March 28, 1967, he became a trustee of that money and plaintiff was the beneficiary of that trust. [Case cited.] Although that money may have been commingled with other monies as to which the sheriff was also a trustee, it remained identifiable as a specific fund held for a specific purpose. [Cases cited.] When the sheriff paid the money involved to [the wrong party], he breached his fiduciary duty, in that he failed to safeguard the trust fund. But that breach, assuming that it may have given plaintiff a right to sue on the theory of an anticipatory breach [cases cited] did not require plaintiff to sue at that time and on that theory. A total repudiation by a trustee of his trust starts the running of the statute of limitations ., but a mere anticipatory breach does not. ... In the case at bench, the sheriff’s duty to turn over the trust fund received by him from Nelson & Belding did not accrue until the 9th of October, 1969, when plaintiff, having by then recovered its judgment against Johnson, made demand for the garnisheed funds. Any time limits involved, thus, ran from that date and not from April of 1967. (Emphasis added.)
Pitchess, 35 Cal.App.3d at 64-65, 110 Cal.Rptr. at 651.
In In re Estate of Nixon, 2 N.C.App. 422, 163 S.E.2d 274 (1968), money was ordered to be held by the clerk pending distribution to named individuals. The clerk, on his own initiative, paid the funds to the state university under their escheat statute. The defendant contended the claim was barred by a statute of limitations. The court said:
The clerk remained liable to account for these funds to the persons entitled thereto so long as the funds remained in his possession, and no statute of limitations would apply to bar an action by the beneficiaries for whose account he held the funds until they had made demand upon him and he had refused to honor the same.
Nixon, 2 N.C.App. at 429, 163 S.E.2d at 279.
Based upon these authorities, I believe the payment of the bail money to the wrong party here did not fix and limit the plaintiff’s right as of that time. The funds remained in the custody of the law, Rudolph, 240 Iowa at 731, 37 N.W.2d at 486, and theoretically could have been recovered by the county at any time. The breach of its trust duties occurred on May 1, 1977 when demand was made for the money by its owner. Suit, having been commenced on September 29, 1977, was therefore within the six-month limit of section 613A.5, The Code.
II. Application of the “discovery rule.”
The time-limit provision of the Tort Claims Act, in section 613A.5, The Code, provides in part that:
Every person who claims damages from any municipality or any officer, employee or agent of a municipality for or on account of any wrongful death, loss or injury within the scope of section 613A.2 or section 613A.8 or under common law shall commence an action therefor within six months, unless said person shall cause to be presented to the governing body of the municipality within sixty days after the alleged wrongful death, loss or injury a written notice . . . . No action therefor shall be maintained unless such notice has been given and unless the action is commenced within two years after such notice. (Emphasis added.)
The “discovery rule” simply provides that a claim does not “accrue” under a statute of limitations until the claimant knew of it, or should have known of it, in the exercise of reasonable care. It was adopted by us to ameliorate the harsh results where a person *921was unaware of his claim until it was too late to pursue it. Chrischilles v. Griswold, 260 Iowa 453, 150 N.W.2d 94 (1967) (suit against architect). We said there:
If an injured party is wholly unaware of the nature of his injury and the cause of it, it is difficult to see how he may be charged with a lack of diligence or sleeping on his rights.
Chrischilles, 260 Iowa at 461, 150 N.W.2d at 100. The rule has since been applied by us in medical and legal malpractice cases. See Baines v. Blenderman, 223 N.W.2d 199 (Iowa 1974) and Cameron v. Montgomery, 225 N.W.2d 154 (Iowa 1975). It appears the rule is firmly established in our law.
Two bases are relied upon by the majority in concluding the rule is not applicable under section 613A.5, The Code: (1) The Tort Claims Act is a “statute of creation” with a built-in time limitation affecting the right to sue, not just the remedy; and (2) the Tort Claims Act does not refer to “accrual” of the cause of action but rather to the claimant’s “wrongful death, loss or injury,” and the act therefore lacks the “flexibility” of traditional statutes of limitations which have been tempered by application of the discovery rule.
We have said that our tort claims act in section 613A.5 limits the right, as well as the remedy, in a claim against the municipality. Sprung v. Rasmussen, 180 N.W.2d 430, 433 (Iowa 1970); Flynn v. Lucas County Memorial Hospital, 203 N.W.2d 613, 616 (Iowa 1973). We have never held, however, that a time bar of a statutorily created cause of action was so rigid as to permit no exceptions under any circumstances. We have said that statutes of limitations are not favored by the law in cases involving the “substantive” time bar of section 613A.5. See, e. g., Vermeer v. Sneller, 190 N.W.2d 389, 394 (Iowa 1971); Sprung, 180 N.W.2d at 433.
Other courts deciding the issue have almost uniformly held that statutorily created causes of action with built-in time limitations do have exceptions, mainly based upon the policy consideration that it is unfair to do otherwise. In the early case of Osbourne v. United States, 164 F.2d 767 (2d Cir. 1947), for example, Judge Hand writing for the second circuit held that a claimant could bring a “late” action under a federal statute containing time limitations as conditions precedent, because he had been a prisoner of war and it was physically impossible to commence his action. Other examples of “exceptions” to condition precedent time bars are myriad. In cases under the Federal Employers’ Liability Act,2 in statutory actions to set aside a patent,3 in statutory actions for wrongful death,4 and civil antitrust actions under the Clayton Act,5 various exceptions have been applied extending the time to commence actions despite the built-in time limits.
The Federal Tort Claims Act time limitation under 28 U.S.C. 2401(b) contains condition precedent language, as follows:
(b) A tort claim against the United States shall be forever barred unless it is presented in writing to the appropriate Federal agency within two years after such claim accrues or unless action is begun within six months after the date of mailing, by certified or registered mail, of notice of final denial of the claim by the agency to which it was presented.
The discovery rule has been applied under the federal act despite the fact it is regarded as a condition precedent, or time-bar provision. See, e. g., Jordan v. United States, 503 F.2d 620 (6th Cir. 1974); Quinton v. United States, 304 F.2d 234 (5th Cir. 1962); Casias v. United States, 532 F.2d 1339 (10th Cir. 1976); Reilly v. United States, 513 F.2d 147 (8th Cir. 1975); Ciccar*922one v. United States, 486 F.2d 253 (3d Cir. 1973).
Specifically, as relevant here, a cause of action for conversion, brought under the Federal Tort Claims Act,' has been held not to accrue until discovery of the conversion. Missouri Bank South v. United States, 423 F.Supp. 571 (D.C.Mo.1976).
The early cases establishing exceptions to the previously rigid view of statutorily created rights have been referred to as “chinks in the armor” of the inflexible rule that “substantive” time limitations bar the rights created.6
Most important, the United States Supreme Court has rejected the view that time limits in statutes of creation are per se less flexible than “traditional” limitation statutes merely because of their origin. In Burnett v. New York Central Railroad, 380 U.S. 424, 85 S.Ct. 1050, 13 L.Ed.2d 941 (1965), the Supreme Court considered a claim under the Federal Employers’ Liability Act, which contained a “substantive” time limitation providing that “no action shall be maintained . . . unless commenced within three years from the day the cause of action accrued.” The claimant had filed an F.E.L.A. action in state court which was dismissed for lack of proper venue. Eight days after the dismissal, he filed his action in federal court. The district court dismissed the federal action because it was not commenced within the three-year limitation, and the court of appeals affirmed, rejecting the plaintiff’s claim that the F.E. L.A. limitation was tolled during the pend-ency of the state court action. The court of appeals said:
[t]he limitation in the Act is more than merely procedural. It is contained in an Act which created a new right and prescribed the remedy. The remedy is a part of the right and is a matter of substance. Failure to bring the action within the time prescribed extinguished the cause of action. (Emphasis added.)
Burnett v. New York Central Railroad, 332 F.2d 529, 530 (6th Cir. 1964). The court of appeals cited several authorities in support of this view and distinguished cases which had extended the time limits of several other “substantive” federal statutes. The court concluded by saying that “[w]e find nothing in [a previous supreme court case] indicating that the Supreme Court has overruled previous cases holding that the limitation in the Act was substantive and not procedural.” Id. at 531.
The Supreme Court on appeal reversed the court of appeals. It did not change the previous characterization of the time limit as “substantive;” it simply said the label was no longer determinative because the main issue is whether Congress intended a particular result. It said that:
The basic question to be answered in determining whether, under a given set of facts, a statute of limitations is to be tolled, is one “of legislative intent whether the right shall be enforceable . after the prescribed time.” [Cases cited.] Classification of such a provision as “substantive” rather than “procedural” does not determine whether or under what circumstances the limitation period may be extended. As this Court has expressly held, the FELA limitation period is not totally inflexible, but, under appropriate circumstances, it may be extended beyond three years. (Emphasis added.)
Burnett, 380 U.S. at 426-27, 85 S.Ct. at 1053-54, 13 L.Ed.2d at 944-45.
A footnote to the above statement said:
The distinction between substantive and procedural statutes of limitations appears to have arisen in cases involving conflicts of laws. [Authority cited.] While the embodiment of a limitation provision in the statute creating the right which it modifies might conceivably indicate a legislative intent that the right and limitation be applied together when the right is sued upon in a foreign forum, the fact that the right and limitation are written into the same statute does not indicate a legislative intent as to whether or when the statute of limitations should be tolled. Thus the “substantive” — “pro*923cedural” distinction would seem to be of little help in deciding questions of extending the limitation period. (Emphasis added.)
Id., 380 U.S. at 427, 85 S.Ct. at 1054, 13 L.E.d.2d at 944. See also Glus v. Brooklin Eastern Terminal, 359 U.S. 231, 234-35, 79 S.Ct. 760, 762-63, 3 L.Ed.2d 770, 773 (1959) (F.E.L.A, time bar extended).
The Supreme Court in Burnett held that extending the time under the Act, based upon the pendency of the state action, “effectuates the basic congressional purposes in enacting this humane and remedial [Federal Employers’ Liability] Act as well as those policies embodied in the Act’s limitation provision . . . .” Burnett, 380 U.S. at 427-28, 85 S.Ct. at 1054, 13 L.Ed.2d at 945.
The court in Burnett said that to determine legislative intent, the purposes and policies of time-barring statutes must be examined. It said that:
Statutes of limitations are primarily designed to assure fairness to defendants. Such statutes “promote justice by preventing surprises through the revival of claims that have been allowed to slumber until evidence has been lost, memories have faded, and witnesses have disappeared. . [Case cited.] Moreover, the courts ought to be relieved of the burden of trying stale claims when a plaintiff has slept on his rights.
This policy of repose, designed to protect defendants, is frequently outweighed, however, where the interests of justice require vindication of the plaintiff’s rights.
Id., 380 U.S. 428, 85 S.Ct. at 1054, 13 L.Ed.2d at 945.
With Burnett, the “chink in the armor” of the substantive-procedural dichotomy became a gaping hole. Legislative intent and policy, rather than origin, are now the touchstones for determining the adaptability of the time bar to exceptions previously limited to “traditional” statutes of limitations.
Applying these principles to the case at hand, certainly no one could contend a claimant has slept on his rights if he was neither aware of them, nor capable of being aware of them in the exercise of reasonable diligence. And what would be “fair” or “just” in denying a plaintiff’s claim under these circumstances merely because the defendant is a municipality?
I contend that the same policies underlying application of the discovery rule in Chrischilles apply equally here. The rule enunciated there should be extended to claims under the Tort Claims Act unless clear intent of the legislature appears to prevent it. I do not believe that such intent is established either by the wording of the statute or the history of our rule of sovereign immunity preceding it.
In Olsen v. Jones, 209 N.W.2d 64 (Iowa 1973), we faced the issue of whether timely notice under section 613A.5 was required in order to claim indemnity or contribution from a municipality. The act did not make it clear on its face. We concluded, however, that the notice requirement was intended to be operative only in an action by the injured party, not in claims for indemnity or contribution, saying that to hold otherwise “would not only thwart the intent of the statute but would also summarily abrogate the important equitable principles of contribution which were finally established — after long struggle — in Best v. Yerkes, 247 Iowa 800, 810, 77 N.W.2d 23, 29, 60 A.L.R.2d 1354 (1956). . . .” Olsen, 209 N.W.2d at 66.
The “long struggle” referred to was almost insignificant compared to the one involving sovereign immunity, as attested to by our Boyer case.7 Did our legislature intend to provide a time limit under the Tort Claims Act without sufficient flexibility to accommodate a reasonable discovery rule? If so we have only modified the old concept that “the king can do no wrong” to now provide he can still do no wrong if the victim f'oes not discover it right away.
*924The short claims period provided makes application of the discovery rule even more imperative in tort claims act cases than in others with longer periods provided. In Gates Rubber Co. v. USM Corp., (7th Cir.) 508 F.2d 603, 612-13, the court said:
[The] short statute provides a stronger case for a discovery rule than a long one. For the probability of injustice resulting from the failure to discover meritorious claims unquestionably diminishes as the statutory period increases. Moreover, the risk that a defendant might be unable to refute an unfounded claim is greater when a prospective plaintiff is allowed a period of five years after discovery of his claim than when he must act within two years of discovery. (Emphasis in original.)
It is true that the federal statutes discussed use the word “accrue,” and the Iowa Tort Claims Act does not. I disagree with the majority’s conclusion, however, that this denies the “flexibility” required to apply the discovery rule to this case. The federal cases extending the time period under the Tort Claims Act, and similar acts, have balanced the policy considerations of both parties, and concluded Congress did not intend to enact a law causing unwarranted hardships on plaintiffs. In none of them has the court said the relaxation of the time bar was a result of some special meaning to be given the use of the word “accrue.” This word simply means the claimant “has a right to institute and maintain a suit.” Chrischilles, 260 Iowa at 461, 150 N.W.2d at 99. Moreover, as discussed later, the Iowa Act refers to the claimant’s “injury,” which has been considered to permit a broad application in determining when a “substantive” time bar may be extended.
Our cases interpreting the time limits of section 613A.5 have exhibited a desire to avoid unreasonable results and hardships and have not shown an inflexible application of its provisions. For example, Olsen, in holding that a party seeking contribution or indemnity against a municipality is not bound to the notice and time provisions of the Tort Claims Act, despite the lack of a statutory exception for it, said “[a]ny other result would lead to impractical and illogical consequences, which we should avoid in searching out legislative meaning.” Olsen, 209 N.W.2d at 67.
In Flynn a claimant sought to apply the discovery rule to his claim under chapter 613A. That case, in concluding the plaintiff had immediate knowledge of his cause of action and was thereby precluded from relying upon the discovery rule, clearly left the door open for future consideration of whether the rule should be adopted, stating that “[w]e do not in this case find it necessary to reach the question of the applicability of the discovery rule to the notice of claim requirements of section 613A.5.” Flynn, 203 N.W.2d at 616.
In Baines we discussed the applicability of the discovery rule in medical negligence cases and quoted from a Rhode Island case which said “[t]o require a man to seek a remedy before he knows of his rights, is palpably unjust.” Baines, 223 N.W.2d at 203. Baines involved a common law action for negligence, rather than a claim under the Tort Claims Act, but the palpable unfairness to a claimant is no less because the alleged wrongdoer is a municipality. This is the very sort of inconsistency characterized as “unjust” and “unsupported by any valid reason” by the four-member dissent in Boyer v. Iowa High School Athletic Association, 256 Iowa 337, 349, 127 N.W.2d 606, 613 (1964).
Several considerations demonstrate to me that precluding application of the discovery rule here would not be compatible with the policy considerations discussed above, and therefore could not be presumed to be intended by the legislature in the absence of express language to that effect. The words used in the Act are not only compatible with application of the discovery rule, I believe they affirmatively require it. Under Iowa’s Act, the time limitation commences to run upon the “wrongful death, loss or injury” of the claimant. Section 613A.5, The Code. It is significant, I believe, that words such as “act,” “negligence,” “occurrence” or “event” are not *925used. It is also significant that the legislature used “injury” as one of the trigger words, because this word has been broadly applied, at least as flexibly as “accrue” in other “substantive” limitation statutes. It is usually held, for example, that “injury” for purposes of commencing a period for presenting a worker’s compensation claim, does not occur until “discovered,” i. e., actually discovered or should have been discovered in the exercise of reasonable diligence. This rule is stated at 3 Larson, Workmen’s Compensation § 78.41, at 15-83 (rev.perm. ed.1976):
The usual statute merely dates the period from the time of injury, disability, or accident, saying nothing about time of discovery of the nature of the condition. Yet the great majority of the courts have been sufficiently impressed with the acute unfairness of a literal application of this language to read in an implied condition suspending the running of the statute until by reasonable care and diligence it is discoverable and apparent that a compensable injury has been sustained. (Emphasis added.)
Our worker’s compensation law is a legislatively created right, and its time bar could also be subjected to a narrow condition-precedent analysis. It has not been, however. We have interpreted the word “injury” in the workers’ compensation statute requiring notice to an employer of a potential claim to mean a discovered injury. Jacques v. Farmers’ Lumber & Supply Co., 242 Iowa 548, 47 N.W.2d 236 (1951). We said in that case that:
Since the legislature made disease com-pensable under its term “injury” then clearly it must have meant the “occurrence” of this type of “injury” was when the employee found out about the disease. To hold otherwise would defeat the obvious legislative purpose. The employee could hardly be held under a duty to notify his employer of a disease of which he had no knowledge. It would be unreasonable to conclude that the legislature intended a construction of “occurrence of the injury” — and we substitute “occurrence of the disease” — to mean a point of origin before the employee found out about his disease. (Emphasis added.)
Id., 242 Iowa at 552-53, 47 N.W.2d at 239. We also said in that case:
[I]t would defeat the very purpose of the workmen’s compensation law to first hold a disease an injury and therefore compen-sable and then hold the occurrence of the injury could be before the workman discovered the disease.
Id., 242 Iowa at 554, 47 N.W.2d at 240.
Jacques was distinguished in the later case of Mousel v. Bituminous Material & Supply Co., 169 N.W.2d 763, 766 (Iowa 1969) upon the basis that Jacques concerned use of the word “injury” in a section 85.23 (The Code 1966) notice of injury to an employer, while Mousel concerned use of the word in a time limitation for commencement of a proceeding for collection under section 85.26. However, the opinion did not say “injury” was erroneously defined in Jacques. In fact, it is interesting to note that it quoted one authority stating that “most courts with statutes dating the limitations period from the injury have found it possible to interpret the statute as running from the time a reasonable claimant should have known the compensable character of his claim.” Id., at 767, quoting from an earlier edition of 2 Larson, Workmen’s Compensation § 78.52, at 278.
The opinion then quotes, with apparent approval, the Chrischilles case language adopting the discovery rule, and says that “[w]e are not persuaded on considerations just suggested the commissioner and district court should be reversed under this record.” (Emphasis added.) Mousel, 169 N.W.2d at 767. The court then discusses the evidence showing plaintiff actually knew of his condition in time to make a claim within the prescribed period. Larson in a later edition, cites Mousel as a case in which “the failure [to apply the discovery rule] has . . . been on the facts rather than on the law; that is the court . has concluded that the claimant’s conduct in the light of his knowledge and all the other circumstances did not in fact meet the test *926of reasonableness.” 3 Larson, Workmen’s Compensation § 78.41, at 15-87 (rev.perm. ed.1976).
The flexible application of “injury” under Jacques still stands in Iowa. In addition the United States Supreme Court quoted with approval from a California case which said that: “[t]he afflicted employee can be held to be ‘injured’ only when the accumulated affects of the deleterious substance manifest themselves.” Urie v. Thompson, 337 U.S. 163, 170, 69 S.Ct. 1018, 1025, 93 L.Ed. 1282, 1293 (1949). The court said this statement of principle “seems to us applicable in every relevant particular” to the time-bar provisions of the F.E.L.A. Id.
Pennsylvania has taken a similar approach with a statute providing that a personal injury suit must be brought within two years “ ‘from the time when the injury was done and not afterwards. . . . ’ ” (Emphasis added.) In a medical malpractice context, the court held that the “injury” commencing the limitation period did not occur until discovered by the plaintiff. Ayers v. Morgan, 397 Pa. 282, 154 A.2d 788 (1959).
The flexibility allowed by the use of the word “injury” is recognized in 100 C.J.S. Workmen’s Compensation § 468(8), at 370-71, where the differences in words used to commence periods of limitations in compensation claims are discussed. It states there that:
[wjhere the claim is required to be made within a specified time after the “accident,” the time begins to run against such claim from the date of the occurrence of the accident, and not from the time of the development, discovery, or first manifestation of the injury.
It states, however, that “injury” does permit some latitude, saying that:
[S]ince an intent to bar compensation claims before they arise cannot fairly be imputed to the legislature, “injury” within such a provision has been construed to mean a compensable injury . . . ; and the time for the filing of the claim begins to run as of the time when the right to compensation accrues, which may be the time the disability or incapacity occurs, or which may be the time the disability or incapacity develops, or becomes apparent, and not necessarily the time of the occurrence of the accident. (Emphasis added.)
The statute also refers to the claimant’s “loss” as commencing the running of the limitation period. This word, also, has been interpreted to mean a discovered loss. See, e. g., Federal Deposit Insurance Corp. v. Aetna Casualty & Surety Co., 426 F.2d 729, 739 (5th Cir. 1970) (word “loss” in notice provision of fidelity bond means date of discovery of fraud of bank director).
I do not believe we can attribute to the legislature an intent to bar a claimant before his injury is known to him by enactment of section 613A.5. In effect, a plaintiff would be out of court before he knows he is in it. This situation has been recognized by the federal courts to involve “blameless ignorance.” Urie, 337 U.S. at 170, 69 S.Ct. at 1025, 93 L.Ed. at 1292; Exnicious v. United States, 563 F.2d 418, 425 (10th Cir. 1977).
We must assume the legislature’s use of words such as “injury,” which has been given flexible interpretation in analogous contexts, was intentional. It could have used such words as “act,” “occurrence,” or other more restrictive terms but chose not to. We have said that, as to the proyisions of section 613A.5, “[a] rigid interpretation of this statute under circumstances presented here does nothing to meet the purposes for which it was enacted but further closes the door on unwary claimants.” Vermeer v. Sneller, 190 N.W.2d 389, 394-95 (Iowa 1971) (issue of whether section 613A.5 notice must be given in suit against employees in absence of specific statutory provision).
In United States v. Reid, 251 F.2d 691 (5th Cir. 1958), the court considered the timeliness of a claim under the Federal Tort Claims Act. It applied the discovery rule, despite the “substantive” nature of the time limitation in the Act. The court said:
At the outset, the matter should be determined as it would were we constru*927ing another similar Federal statute prescribing the time in which suit must be filed. We do not approach it with any begrudging niggardliness from a not too well concealed attachment for the vestigial strict construction of the sovereign’s waiver of immunity — a concept which has been expressly repudiated almost whenever and wherever this and similar legislation has been before the Supreme Court. [Cases cited.]
Id., at 693.
I believe we should take a similar view of our Tort Claims Act. A narrow approach in considering the circumstances under which the sovereign will “permit” suits against it is not required by the words or tenor of the act; it vitiates the policy considerations prompting its enactment and will, as here, result in harsh consequences which were not intended by the legislature. Other courts, in similar circumstances, have applied various concepts, including the discovery rule, to extend “substantive” time-bar limits. We need not, as we have on other occasions, await further legislation to deal with this matter. The wording of the statue permits application of the discovery rule to chapter 613A; policy considerations demand it.
I would reverse and remand the case for trial for the reasons stated in division I. In the alternative, under division II, I would reverse and remand for a factual determination on the issue of when plaintiff knew of her claim, or should have in the exercise of reasonable diligence. I would submit that issue to the fact finder with the other evidence in the case.
REYNOLDSON, C. J., and HARRIS, J., join this dissent.. 14 C.J.S. Clerks of Court § 40, at 1246. Cf. National Auto. & Cas. Ins. Co. v. Pitchess, 35 Cal.App.3d 62, 64-65, 110 Cal.Rptr. 649, 651 (1973) (funds held by sheriff paid to wrong party).
. See, e. g., Burnett v. New York Cent. R. R., 380 U.S. 424, 85 S.Ct. 1050, 13 L.Ed.2d 941 (1965).
. See, e. g., Exploration Co. v. United States, 247 U.S. 435, 38 S.Ct. 571, 62 L.Ed. 1200 (1918).
. See, e. g., Maryland v. United States, 165 F.2d 869, 873 (4th Cir. 1947) (substantive-remedial dichotomy is “to the say the best of it, technical and legalistic reasoning”).
. See, e. g., Kansas City, Mo. v. Federal Pacific Electric Co., 310 F.2d 271 (8th Cir. 1962).
. Scarborough v. Atlantic Coast Line R.R., 178 F.2d 253, 259 (4th Cir. 1949).
. Boyer v. Iowa High School Athletic Ass'n 256 Iowa 337, 349, 127 N.W.2d 606, 613 (1964).