This appeal concerns the proper measure of damages to compensate a shipper for injury to a cargo of soybeans aboard a carrier’s vessel en route from Oklahoma to Louisiana. The shipper accepted the cargo of deteriorating soybeans, blended the beans with a higher grade of soybean, and sold the blend at a price equal to the original market value of the cargo. The shipper then sued the carrier for damages. The district court found the carrier liable for the cargo damage but held that the shipper could recover only the cost of blending the deteriorated soybeans with the higher grade beans. Because the shipper failed to present any evidence of its blending costs, all recovery was denied.
We hold that the district court applied the incorrect rule of damages. The usual measure of damages for injury to goods in the possession of a carrier is the difference between the fair market value of the cargo at the place of destination, in like condition as when it was shipped, and its actual value at the place of destination. This market value test should determine the shipper’s damages in this case. We therefore reverse the district court judgment denying recovery to the shipper, and order that damages be awarded in accordance with the traditional rule.
I. The Facts
The facts of the case are clearly set out in the district court’s Statement of the Case and Findings of Fact. Record at 176-81. In February, 1973, defendant carrier Upper Mississippi Towing Corporation (UMTC) contracted to carry from Wagoner, Oklahoma to Reserve, Louisiana a consignment of soybeans belonging to plaintiff shipper Cook Industries, Inc. (Cook). On February 27,1973, UMTC loaded the Barge UM-3081 with approximately 43,000 bushels of soybeans2 and issued its bill of lading to the order of Cook in Reserve, Louisiana. The bill of lading noted that the cargo was “in apparent good order” at the time of loading. Record at 5. The cargo was subsequently sampled and an Official Grain Inspection Certificate was issued and transmitted to Cook. The certificate classified the cargo as Grade No. 3 yellow soybeans *853with a moisture content of 14.5%.3 Plaintiff’s Exhibit 2.
Subsequent to loading, the Barge UM-308 was shifted from the loading facility to a fleeting area on the Verdigris River. Because UMTC did not operate its own towboats on the Verdigris River, it customarily arranged for other companies to tow its barges through the McClellan-Kerr Arkansas River Navigation System, of which the Verdigris River is a part. Generally, it took one to three days to obtain towing service for a barge awaiting transit to the Mississippi River. UMTC made daily efforts to secure a tow for the UM — 308, but before towage could be arranged the waters in the McClellan-Kerr Navigation System reached flood stage due to heavy rains, and the system was closed to traffic.
Because of the long delay, UMTC hired Muskogee Marine Service to periodically check the UM-308 to insure that her moorings were secure and her hatches tight and seaworthy. No provision, however, was made to inspect the cargo itself or to insure that it was properly ventilated. Although the shipper, Cook, neither requested UMTC to inspect the cargo nor notified the carrier of the soybeans’ moisture content, UMTC was aware of the nature of the cargo and knew that soybeans would spoil if left in an unventilated barge for an extended period of time.
In March, Cook notified UMTC by telex to reconsign the UM~308’s cargo to Cargill, Inc. at Baton Rouge. The barge remained in the Verdigris River fleeting area until May 12, 1973, seventy-four days after loading, when it was finally towed to the Mississippi River, and then to Baton Rouge, where it arrived on May 25 with hatch covers secure and undamaged.
Cargill, Inc., Cook’s consignee, inspected the cargo upon arrival and rejected it as damaged. Cook notified UMTC of the alleged damage and abandoned the entire cargo to the carrier for salvage. A cargo surveyor inspected the cargo for UMTC and found excessive heating4 but no damage from water leakage. His report indicated that the cargo damage resulted from the moisture content of the soybeans and the extended period of stowage in an unventilated hold. Although the surveyor made no chemical or weight analysis of the cargo, his report suggested that the sour odor of the soybeans automatically downgraded them from No. 3 to sample grade.
After initially rejecting the soybeans, Cook agreed to accept the cargo and arrange for its delivery to Reserve, Louisiana. There, a cargo sample was taken and an Official Certificate of Inspection was issued classifying the cargo as sample grade yellow soybeans. See Plaintiff’s Exhibit 3. Cook subsequently blended5 the entire cargo of damaged soybeans with other soybean shipments and sold the resulting mixture at a price at least equal to the original market value of the cargo in question.
In 1975 Cook instituted this admiralty and maritime action against UMTC for car*854go damage, invoking jurisdiction under 28 U.S.C. § 1338 (1976). The nonjury trial resulted in a pyrrhic victory for the plaintiff: the defendant carrier was held liable for the deterioration of the soybeans, but no damages were awarded. Both parties appeal the district court’s judgment: UMTC contests the finding of liability and the assessment of costs against it, while Cook takes issue with the court’s denial of damages. We affirm the finding of liability and the assessment of costs and reverse on the issue of damages, for reasons to be set forth infra.
II. The Liability Issue
The district court found that although UMTC did not know the precise grade of soybeans shipped aboard the UM-308, or the moisture content of the cargo, it knew or should have known that a delay in transit of seventy-four days would create a substantial likelihood of cargo deterioration. Trial testimony indicated that UMTC frequently handled shipments of soybeans and was familiar with the spoilage problems associated with this type of dry cargo. See Supplemental Record at 121-23. The court concluded that UMTC was negligent in failing to inspect the cargo aboard the Barge UM-308 during the unusually long delay on the Verdigris River, and that UMTC’s negligence was a proximate cause of the cargo damage. Had the cargo been inspected, said the court, remedial action, such as providing ventilation, could have been undertaken to protect the soybeans. Record at 180.
We have reviewed the record in the case, and we do not find the district court’s findings of fact as to UMTC’s negligence to be clearly erroneous. See Fed.R. Civ.P. 52(a); Chaney v. City of Galveston, 368 F.2d 774, 776 (5th Cir. 1966). Because deterioration is an inherent vice of a cargo of soybeans, and because even the better grades of soybeans are likely to begin to deteriorate after six to eight weeks when stored in the unventilated hold of a cargo barge, see Supplemental Record at 49-53, we agree that the carrier UMTC was .obligated to inspect the cargo during the seventy-four day delay in shipping. A common carrier is under a continuing duty to protect a shipper’s cargo against damage while the cargo remains in the carrier’s custody. This duty includes an obligation to inspect a cargo which may deteriorate because of unforeseen transit delays or because of inherent problems in the cargo. See United States v. Lykes Bros. Steamship Co., 511 F.2d 218 (5th Cir. 1975); United States v. Central Gulf Steamship Corp., 456 F.2d 1281 (5th Cir. 1972). Because UMTC breached its duty to Cook, UMTC is responsible for any damage resulting from its failure to inspect the soybeans aboard the UM-308. We hold that the district court’s finding of liability is not clearly erroneous, and we affirm on that issue.
III. The Damage Issue
The district court found that Cook had blended the entire cargo from the UM-308 with other shipments of soybeans, and that the damaged soybeans may have been blended with shipments of Grade No. 3 or higher. Since there was no evidence that any of the bushels of soybeans aboard the UM-308 had been discarded or were sold at less than their original market value, the court concluded that “the only actual loss that plaintiff may have sustained is the cost of the blending process itself.” Record at 181. The court denied any recovery because Cook presented no evidence of blending costs, thus failing to meet the burden of proving the existence and amount of damages sustained.
Generally, when a common carrier is held liable for loss or damage to goods transported, the measure of the shipper’s recovery is the difference between the market value of the cargo in the condition in which it would have arrived had the carrier performed properly, and the cargo’s market value in its damaged state on arrival at port of destination. The Ansaldo San Giorgio I v. Rheinstrom Bros. Co., 294 U.S. 494, 496, 55 S.Ct. 483, 484, 79 L.Ed. 1016 (1935); Holden v. S/S Kendall Fish, 262 F.Supp. 862 (E.D. La. 1966), aff’d, 395 F.2d 910 (5th Cir. 1968). *855The district court took note of this general rule; however, its conclusions of law were that:
7. The primary object in awarding damages is to indemnify the shipper for the loss sustained as a result of the carrier’s negligence. F. J. Walker Ltd. v. The M/V LEMONCORE, 561 F.2d 1138 (5th Cir. 1977).
8. Where the shipper is able to reconstitute the damaged cargo and sell it at its initial fair market value, the shipper is only entitled to such damages as he may prove resulted from the expense of reconditioning the damaged product. Weirton Steel Co. v. Isbrandtsen-Moller Co., 126 F.2d 593 (2d Cir. 1942), cited with approval in F. J. Walker, Ltd. v. The M/V LEMONCORE, supra.
9. Plaintiff has failed to prove by a preponderance of evidence any actual damage or loss as a result of the cargo’s deterioration.
Record at 182. The district court thus rejected the shipper’s contention that the usual rule of damages should be applied, turning instead to the “reconditioning” rule of Weirton Steel Co. v. Isbrandtsen-Moller Co., 126 F.2d 593 (2d Cir. 1942). The court reasoned that since Cook had blended the deteriorated sample grade soybeans with higher grade beans and was able to sell the resulting blend as Grade No. 3, it had suffered no damages other than its “reconditioning” costs, the amount of which was not proven.
We find the district court’s reliance on Weirton to be misplaced. In that case, the carrier had allowed the shipper’s cargo of steel plate to be damaged by liquified resin. The consignee accepted the shipment and, after reconditioning, was able to use the damaged plate to make five-gallon steel cans as originally intended. Although the reconditioning process left the steel plates dull and imperfect in appearance, the proof showed that no reduction in market price was necessitated by the imperfection. Noting that recovery of damages is meant to indemnify the injured party, the court awarded the plaintiff only the cost of reconditioning the metal plates rather than the difference in value measured before and after the damage. Weirton Steel Co. v. Isbrandtsen-Moller Co., 126 F.2d 593 (2d Cir. 1942), cited with approval in F. J. Walker, Ltd. v. The Motor Vessel “Lemoncore ”, 561 F.2d 1138, 1147 (5th Cir. 1977).
We do not differ with Weirton ; its holding is simply inapposite to the facts proven here. The reconditioning of the cargo in Weirton returned the damaged steel plate to its original condition and character so that it could be used as initially intended. Unlike reconditioning, the blending process used by the shipper in the instant case could not restore the damaged soybeans to their former grade. Instead, Cook was merely able to mask the defect of the damaged cargo by adding to it overwhelming quantities of a superior grade blending stock. While the “reconditioning” in Weirton rejuvenated the damaged plate, the blending process used by Cook could not effect a similar result. The cargo remained sample grade; the sour, moldy, discolored beans were simply diluted with enough good soybeans to sell the resulting mixture as Grade No. 3.
Cook argues, and we agree, that blending did not restore the damaged cargo so as to give Cook all the possible uses and advantages of No. 3 soybeans. If the shipper had wanted to blend bad beans with good, he could have purchased sample grade beans directly instead of waiting for the carrier’s negligence to transmute its No. 3 cargo, valued at more than ten dollars per bushel on date of delivery at destination, to sample grade beans worth $7.56 per bushel. See Record at 178-79. By purchasing sample grade beans initially, Cook could have saved the difference between the sample grade market price and that of No. 3 beans while still obtaining the same blending results. The award of blending costs alone, if such were awarded, would not compensate the shipper for this potential savings, i. e., the loss realized by Cook in using its No. 3 beans as sample grade.
UMTC insists that application of the general “difference in market value before and *856after damage” rule in this case will enable the shipper to reap a windfall profit since the damaged cargo was in fact sold at the price of Grade No. 3 soybeans. This argument ignores the reality of the profit lost by Cook in having to use Grade No. 3 priced beans as sample grade in the blending process. In addition, a damage award of Cook’s blending costs alone, if such had been proven, could not fully compensate the shipper for items such as the loss of storage space in which to store other cargos,6 the opportunity to blend other beans into the blending stock utilized to absorb the sample grade cargo, and the expense of purchasing additional blending stocks to mix with the cargo.
The carrier reminds us that “[t]he test of market value is at best but a convenient means of getting at the loss suffered. It may be discarded and other more accurate means resorted to if, for special reasons, it is not exact or otherwise not applicable.” Illinois Central Railroad Co. v. Crail, 281 U.S. 57, 64-65, 50 S.Ct. 180, 181, 74 L.Ed. 699 (1930). In this instance, we find the market value test to be both a convenient and accurate means of measuring the shipper’s damages. We therefore reverse the judgment denying any recovery to the plaintiff-appellant Cook and remand the case to the district court with directions to award damages in accordance with the traditional market value rule. See F. J. Walker Ltd. v. The Motor Vessel “Lemoncore ”, 561 F.2d 1138, 1146-47 (5th Cir. 1977).
IV. Conclusions and Costs
The district court’s findings of fact re UMTC’s responsibility for the damage to Cook’s cargo of soybeans are not clearly erroneous. We therefore affirm the judgment holding UMTC liable for the cargo damage. We reverse the judgment insofar as it denies any recovery to Cook, because we find that the district court applied an incorrect rule of damages in this case. Holding that the market value test is the proper measure of plaintiff’s damages, we remand the case to the district court with directions to award Cook the difference between the market value of Grade No. 3 yellow soybeans on date of delivery at destination, and the market value of sample grade soybeans, determined as of same date and location.7
Finally, we note that defendant UMTC takes issue with the entry of judgment against it for costs. See Record at 183. The carrier asserts that it is the “prevailing party” in the action for purposes of Rule 54(d) of the Federal Rules of Civil Procedure, which states that “costs shall be allowed as of course to the prevailing party unless the court otherwise directs. . .” Because we are reversing the district court’s judgment denying recovery to plaintiff Cook, we need not decide the merits of the carrier’s claim. We affirm the district court’s assessment of costs.
AFFIRMED IN PART, REVERSED IN PART AND REMANDED.
. The Barge UM-308 is a dry cargo barge, 195 feet long and 35 feet wide. The barge was relatively new at the time of the voyage in this case. Findings of Fact, Record at 179.
. The carrier’s bill of lading, Plaintiffs Exhibit 1(a) indicates the cargo weight as 2,700,000 pounds, 1350 net tons, and “45,000 BW”. The complaint states that approximately 45,000 bushels of No. 3 yellow soybeans were delivered to the carrier for shipment. Record at 2. The official grain weight certificate issued by the Destrehan Board of Trade, Plaintiffs Exhibit 4, indicates a weight of 2,618,180 pounds, or 43,636.2 bushels. The district court found that 43,000 bushels of soybeans were loaded aboard the UM-308. Record at 179. The parties do not contest this finding.
. The United States Official Grain Standards classifies soybeans into five grades. Grade designation is determined by weight per bushel, moisture content, the percentage of foreign material, and the percentages of split, heat-damaged, and discolored soybeans. Soybean grades range from U.S. No. 1, the highest grade, to U.S. No. 4; U.S. sample grade soybeans are “soybeans which do not meet the requirements for any of the grades from U.S. No. 1 to U.S. No. 4, inclusive; or which are musty, sour, or heating; or which have only commercially objectionable foreign odor; or which contain stones; or which are otherwise of distinctly low quality.” 7 C.F.R. § 26.603; Finding of Fact No. 7, Record at 179-80.
. At trial, a cargo surveyor explained the heating process. “The moisture [in the soybeans] generates heat, and as the cargo commences to heat, . the heating will cause the soybeans to give off a musty odor, and that’s when the cargo commences to heat, early stages of deterioration.” Supplemental Record at 172. The surveyor further explained that the heating first produces a musty odor, which turns sour; further heating causes the beans to discolor. Id. at 173.
. Blending is a process in which an inferior grade of soybeans is mixed with a shipment of superior quality soybeans in such small proportions that the dilution does not affect the grade classification of the higher quality commodity. Record at 178.
. It was apparently necessary for the shipper to store the damaged beans for a substantial period of time until they could be mixed in small quantities with larger amounts of blending stock.
. The record contains uncontradicted testimony about the market value of Grade No. 3 soybeans as of May 25, 1973 (the date of initial rejection of the cargo as damaged at Baton Rouge, Louisiana); June 8, 1973 (the date Cook agreed to accept the cargo); June 11, 1973 (the next day of trading after delivery to Cook at Reserve, Louisiana); and June 22, 1973 (a date of uncertain relevance).
There is also evidence of the market value of sample grade soybeans as of June 26, 1973 (the date of a salvage sale of somewhat similar sample grade soybeans). Although the district court’s statement of the case includes the market value evidence presented, at trial the court made no findings of fact in this area. See Record at 178-81. We hesitate to award damages to Cook based on the uncontradicted testimony of one expert because we are unsure whether the issue of market value was thoroughly litigated or whether the parties agree on the proper date and location for determining market value. We thus remand the case to the district court to make the necessary findings of fact before awarding damages in accordance with our holding in this case.