Fyffes Group Ltd v Reefer Express Lines Pty Ltd, [1996] 2 Lloyd's Rep. 171

171

Fyffes Group Ltd v Reefer Express Lines Pty Ltd

Queen's Bench Division (Commercial Court)

Mr. Justice Moore-Bick

Nov. 28, Dec. 4, 5, 6, 7, 11, 12, 13, 14, 15, 18,19, 20, 1995; Jan. 11, 12, 16, 17, 18; May 8,1996

[...]

172-174

[...]

Wednesday May 8, 1996 JUDGMENT Mr. Justice MOORE-BICK:

1. Introduction

The first plaintiffs in this action, Fyffes Group Ltd. ("Fyffes"), are well known as importers of bananas into the United Kingdom and the Continent from Central America and the Caribbean. In order to provide bananas for the U.K. and European markets they enter into long term contracts with state organizations or private co-operatives to purchase bananas which are grown by local farmers. At the time of the events with which this action is concerned Fyffes bought bananas grown in Belize under contracts with the Banana Control Board of Belize. They also purchase bananas from the Dominican Republic through the second plaintiffs, one of their affiliated companies, under contracts with El Proyecto La Cruz-Manzanillo and Asociación Dominica de Productores de Banano. Apart from these long term contracts Fyffes also entered into contracts with other banana exporters from time to time for the purchase of fruit in Central America. At the time in question they had one such contract with the American company Del Monte for the supply of bananas from Guatemala.

As part of the arrangements for shipping bananas to Europe Fyffes entered into a contract of affreightment ("the COA") with the first defendants, Reefer Express Lines Pty. Ltd. ("REL") on Nov. 26, 1990. Under that contract REL were to provide refrigerated vessels to load cargoes of fresh green bananas and other fruit at Big Creek in Belize, Puerto Cortés in Honduras and Manzanillo in the Dominican Republic for carriage to Portsmouth and Zeebrugge during the period from about Oct. 25, 1990 to Dec. 31, 1991. The COA incorporated the Carriage of Goods by Sea Act, 1971.

Kriti Rex is a refrigerated vessel built in 1978 and owned by the second defendants, Reefkrit Shipping Inc. ("the owners"). On Aug. 31, 1988 she was time chartered to REL for a period of three years on the Baltime form with various additions and amendments which included, in particular, a clause incorporating the United States clause paramount. In March, 1991 Kriti Rex was nominated by REL to load a cargo of bananas under the COA at Manzanillo, Santo Tom s, Puerto Cortés and Big Creek. She arrived at Manzanillo on Mar. 14 having experienced heavy weather during the latter part of the voyage from Zeebrugge which caused her to reduce speed. Loading at Manzanillo began on Mar. 14, and was completed on Mar. 15, and two bills of lading were issued on behalf of the master evidencing the shipment of a total of 16,530 boxes of green bananas. The vessel left Manzanillo at 07 30 hours on Mar. 15. On sailing Fyffes informed the master that her next port of loading would be Santo Tom s, Guatemala. He gave an expected time of arrival there of about 10 30 hours on Mar. 17. However, at about 19 00 hours on Mar. 16 while the vessel was at sea problems began to appear with the operation of the main engine to which I shall refer in more detail later. For the moment it is sufficient to say that the engine was stopped and on investigation serious damage was found to have occurred to No. 7 piston and cylinder liner and to Nos. 7 and 15 crankpin bearings. Repairs were undertaken at sea and the next morning, Mar. 17, the engine was restarted. Before long, however, the bearings were found to be running hot and the engine had to be stopped to enable a further attempt to be made to solve the problem. The engine was re-started at 11 00 hours on Mar. 18, but when the speed was increased Nos. 7 and 15 crankpin bearings again 175began to heat up rapidly. The engine was stopped for the third time at about 14 10 hours that day and on inspection those bearings were again found to be damaged. No. 8 main bearing was also found damaged and at that point the master decided to declare general average and proceed to Puerto Cortés as a port of refuge. Because of the damage to the main engine the vessel had to proceed at reduced speed; she eventually anchored at Puerto Cortés on the morning of Mar. 21. 

At the time the vessel left Manzanillo on Mar. 15 Fyffes were intending to load additional cargo at Santo Tom s, Puerto Cortés and Big Creek. Mr. Stickney, Fyffes' representative in the area who was responsible for organizing the cutting of bananas, had originally expected the vessel to arrive at Santo Tom s at about 06 00 hours on Mar. 17. That was probably somewhat optimistic, but in any event he arranged for cutting of the cargo due to be loaded at Santo Tom s to begin on the 16th in anticipation of the vessel's arrival. By the end of the day 14,530 boxes had been cut and brought down to the port in readiness for loading. Early in the morning of Mar. 17 the master informed Fyffes, REL and the agents at Santo Tom s that the vessel had stopped the previous evening to repair a seized piston in No. 7 cylinder but that he expected to sail within a few hours and would reach Santo Tom s by 10 00 hours on the 18th. A further 15,506 boxes of bananas were cut and brought down to the port that day. Arrangements had previously been made to start cutting bananas in Belize on Mar. 18 in anticipation of the vessel's arrival there on 19th. However, on Mar. 18 following the further problems with the engine the master sent another message to Fyffes warning them that his e.t.a Santo Tom s had been put back again to daybreak on the 19th. That led Mr. Templeman, Fyffes' shipping co-ordinator, to send instructions to Mr. Stickney to stop the cut in Belize. In the event 4009 boxes were cut and packed on Mar. 18 before the cut was stopped.

During Mar. 19 Mr. Templeman made various enquiries in an attempt to find alternative shipping space for the bananas which were by then waiting at Santo Tom s. He was unable to do so, and eventually he reached agreement with Del Monte for them to be carried in one of their ships, Harvester, to Galveston and sold there on Fyffes' behalf. Over the following days there were various communications between Fyffes, REL and the owners concerning the progress of repairs and the condition of the bananas which were on board. On Mar. 19 REL told Fyffes that it had another vessel, Iceland Rex, about to leave Philadelphia free of charge. She could be in the area by Mar. 24, but the indications REL had received from the owners were that Kriti Rex herself could be expected to have completed her repairs by that date so Mr. Templeman did not pursue the matter any further. On Mar. 28, Mr. Stickney and a representative of the cargo underwriters, Mr. Guzman, went to Puerto Cortés to inspect the bananas on board Kriti Rex. At that stage Mr. Stickney thought they would remain in marketable condition for only about another week. They were no longer in a condition to be carried to Europe; nor was it practicable or economic in his view to ship them to the United States. There was no market for the bananas locally and all those concerned agreed that they would have to be discharged and dumped. In the event, however, Fyffes were able to arrange for them to be donated to the Honduran Red Cross. Thus, although from a commercial point of view they were a total loss, they did not go to waste. 

The claims in this action fall under three heads. Fyffes claim damages from REL for breach of the COA in failing to tender the vessel for loading at Santo Tom s and Big Creek. They seek to recover damages based on the value at destination of the bananas that would have been loaded at those two ports, less the saving in expenses which would otherwise have been incurred in relation to those cargoes and the net proceeds of sale of the bananas which were carried on Harvester to Galveston. Caribbean Gold as holders of the bills of lading covering the cargo loaded at Manzanillo claim damages against the owners for the loss caused by their failure to carry those goods to their destinations. They likewise seek to recover damages based on the value of the bananas at Portsmouth and Zeebrugge. There is a counterclaim by the owners for a declaration that they are entitled to recover a contribution to general average expenses, but it is agreed that that depends simply on the outcome of the claim against them. Finally, REL has issued contribution proceedings against the owners seeking an indemnity against any liability which it may be held to have incurred towards Fyffes. Both plaintiffs originally made their claims in United States dollars, but their right to do so having been challenged by the defendants, they now accept that any damages which they are entitled to recover are to be measured and awarded in sterling. 2. The cause of the casualty

I turn first to the main factual dispute in this case which concerns the reason for the main engine failure which occurred on Mar. 16. This gave rise to an interesting conflict of opinion between the expert witnesses called on behalf of the plaintiffs and the owners which involved a close consideration of the nature and operation of main crankshaft bearings and the crankpin bearings and also of the lubrication system which serves them. In order to understand the nature of the issues which arise in relation to this part of the case it is necessary to give a brief 176 description of the relevant parts of the main engine. (a) A brief description of the main engine

The vessel is powered by a 16 cylinder PC2-5V BTC Pielstick four stroke trunk piston diesel engine of "V" configuration. The crankshaft has nine main bearings numbered 1-9 from aft to forward. The cylinder units are in pairs, each crankpin serving two units; the pairs of units are numbered from 1-9 aft to 8-16 forward, Nos. 1-8 being on the port side and Nos. 9-16 on the starboard side. No. 8 main bearing lies towards the forward end of the engine. The adjacent pairs of crankpin bearings are Nos. 7-15 (aft) and Nos. 8-16 (forward). Of these, it is Nos. 7 and 16 crankpin bearings which lie closest to No. 8 main bearing through which they receive their lubrication.

The main crankshaft bearings are of standard construction. Each contains two semi-circular shells which fit precisely within a housing. The bearing is supported beneath by a massive saddle and is closed on top by the cap or keep. The bearing shells themselves are steel lined with a running surface made up of the following layers: a layer of lead-tin alloy of a thickness of 30 to 50 μm often referred to as "white metal", "overlay" or "flash" coating; a nickel diffusion barrier typically of about 3 to 5 μm in thickness; and a layer of copper-lead alloy, sometimes referred to as "bronze" or simply "copper", of about 0.5 to 1mm in thickness which directly overlies the steel backing. The white metal is very soft and forms the main running surface of the bearing. It is used in bearings of this kind for a number of reasons: in particular, if for any reason there is a temporary breakdown in the oil film, it will provide a degree of dry lubrication to prevent damage to the shaft; also, its very softness enables it to entrap hard particles circulating in the lubricating oil which would otherwise cause abrasive wear to the bearing or shaft. This is known as "embedding". The function of the nickel diffusion barrier is to provide a form of chemical insulation between the [1996] 2 Lloyd's Rep. 171 Page 7 [1996] 2 Lloyd's Rep. 171 [1996] 2 Lloyd's Rep. 171 (Cite as: [1996] 2 Lloyd's Rep. 171) Copr. © West 2009 No Claim to Orig. Govt. Works metals on either side. I have mentioned it for completeness, but it does not play any significant part in the events with which I am concerned and I shall not refer to it again. The copper layer is much harder and inherently stronger than the white metal and it is also much thicker. It provides a satisfactory running surfacefor the journal after any initial want of conformity has been taken by the white metal. The soft lead component permits a small degree of "dry lubrication" but because this layer is much harder it has little ability to embed hard particles circulating in the oil. The crankpin bearings which link the connecting rods to the crankshafts are of similar construction.

Lubricating oil is supplied to the engine under pressure by the lubricating oil pump. Oil is drawn from the sump tank beneath the engine and was first passed through filters. These provide the primary protection to the engine against the damaging effect of hard combustion products in the lubricating oil and were designed to remove particles of 30μm or more in diameter. From time to time all filters need to be cleaned of accumulating debris. The filters on this vessel were designed to backflush automatically when the difference between the pressure on the inlet side and that on the outlet side reached a certain level. Backflushing causes debris to be washed from the surfaces of the filters and carried away to a separate drain tank. The vessel is also equipped with a centrifugal purifier which will remove water and particles down to 2μm in diameter, but its capacity is limited. Oil leaving the filters is supplied to the engine through the main supply line (also referred to as the "rail") which is about 5 in. in diameter. (This and all subsequent dimensions of the main lubricating oil supply system are given for illustrative purposes only; the precise dimensions do not matter.) At the time of the casualty an audible alarm was set to sound if the pressure at the inlet rail fell to 4.2 bar.

From the main supply rail individual feeder pipes of about 25mm in diameter run into the top of each of the main bearings where oil is fed into the upper clearance of the bearings through ports in the upper shell. In normal operation the rotation of the shaft causes the oil to be drawn beneath the journal to maintain a continuous film of lubrication. The lubricating oil also performs the very important function of cooling the bearing. The crankshaft itself contains a continuous central bore of about 25mm in diameter. At each of the main bearings, and at each of the crankpin bearings also, there is a channel of about 25mm diameter, leading from the surface of the journal to the central bore. Some of the oil which enters the main bearing passes into the crankshaft bore and via this bore to the crankpin bearing immediately adjacent on either side where it provides lubrication and cooling before leaving the bearing by way of a channel in the connecting rod. From the crankpin bearing the oil passes through the connecting rod up to the piston itself where it provides cooling and lubrication to the top bearing or gudgeon pin. Some oil is deposited on the walls of the liner to provide lubrication for the cylinder; the remainder returns by way of another channel in the connecting rod from which it is discharged into the crankcase and eventually returns to the sump tank. The fact that the lubricating oil provides the means of cooling both the pistols and the bearing is of some significance in this case as will appear in due course.

177(b) The circumstances of the casualty

The essential facts which emerge from the evidence of the chief engineer and the reports of the surveyors´who attended the vessel at Puerto Cortés are not in dispute. Together with a number of main bearing shells removed from the vessel they provide the primary basis for the various conclusions drawn by the expert witnesses and can be stated quite shortly.

At about 19 00 hours on Mar. 16 the chief engineer who was then on the bridge noticed a change in the note of the low pressure turbocharger. He went straight down to the engineroom and very shortly after he reached the control room the lubricating oil pressure alarm sounded. He immediately shut down the engine manually. It was agreed that the change in the sound of the turbocharger was the result of the seizure of No. 7 piston slowing down the engine, and since it was generally accepted that it would have taken the chief engineer about a minute to get from the bridge to the control room, it is reasonably certain that only about a minute to a minute and a half elapsed between the development of the seizure to a serious extent and the shutting down of the engine. Subsequent investigation showed that No. 7 unit had suffered most damage. The crankpin bearing had been completely wiped; it had lost all the white metal and bronze from the bearing shells and the shells themselves had rotated in the housing. No. 7 piston had seized to the extent that piston and liner had to be removed as a single unit at Puerto Cortés. When it was removed there were signs of scoring and heat discolouration of both the liner and piston, and the piston rings, although unbroken, were stuck in their grooves. The connection rod, however, was not seriously damaged, although it did show signs of discolouration by heating and there was some slight distortion of the bottom end. No. 15 unit also suffered quite severe damage. The crankpin bearings had failed completely, having also lost all the white metal and bronze, although unlike No. 7 the shells had not rotated. No. 15 piston was not damaged, but there was cracking at the top of No. 15 cylinder liner. Although No. 8 main bearing was found to be running hot after the casualty, it was not opened at sea and operated adequately, albeit at low speed, until the vessel reached Puerto Cortés. There it was found to be heavily damaged, having lost nearl y all of the bronze as well as the white metal from the lower shell and most of the white metal and some of the bronze from the upper. Both the bearing saddle and cap showed signs of discolouration by heat and the horns of the cap were distorted. The cap and saddle were sent to Houston for reconditioning. The crankshaft journal in way of No. 8 main bearing was scored and showed signs of heating, including a crack of about 0.3mm in depth. The No. 7-15 crankpin was more severely damaged in that it was heavily scored and showed signs of severe overheating with three cracks of up to 1mm in depth. Apart from No. 4 main bearing which also showed signs of overheating, other bearings were undamaged. In particular, no damage was found to No. 7 main bearing or to No. 16 crankpin bearing, each of which was adjacent to one of the damaged bearings. Throughout the period of the casualty the lubricating oil pump had continued to operate normally until the engine was shut down and there is no suggestion that the alarms and other protecting devices were not in good working order. 

The shells of Nos. 1, 3, 4, and 7 main bearings were removed from the vessel within a short time of the casualty. Fortunately they were preserved for later inspection and were therefore available for inspection by the various expert witnesses. They were produced for examination at the trial. It will be necessary at a later stage to deal with their condition in more detail, but it is sufficient for the moment to say that with the exception of No. 4 which showed some blackening of the exposed copper in the lower half of the shell, none of them were worn to a greater degree than might have been expected having regard to their periods in service, and No. 1 in particular showed very little sign of wear in that the white metal was intact over the whole area of the bottom half. However, all of them to different degrees showed signs of a grey "mottling" which on microscopic examination was shown to be caused by a large number of embedded particles.(c) The nature of the dispute

Against that background the primary dispute between the parties was whether the initiating cause of the casualty had been the failure of No. 8 main bearing or the seizure of No. 7 piston. If it was the failure of No. 8 main bearing, there was a secondary question whether that was caused by an excessive quantity of particles in the lubricating oil.

The plaintiffs' case was that No. 8 main bearing failed as a result of an interruption of the oil film caused by an excessive amount of particulate matter in the lubricating oil. As the bearing failed small shards of white metal and copper were carried by the oil to No. 7 crankpin bearing where they either entered the bearing, interrupting the oil film there in turn, or gathered at the narrow point where the crankshaft bore opens into the bearing to form a dam which reduced the flow of oil into the bearing. For one or other of these reasons the lubrication of No. 7 crankpin bearing was interrupted, the bearing failed and in due course the shells rotated in their housing thereby blocking the flow of oil through the channel in the connecting rod to the piston. The 178 resulting loss of cooling and lubrication caused the piston to overheat leading to seizure. The failure of No. 15 crankpin bearing occurred while this was going on. Eventually the loss of pressure caused by the combined effects of the failure of the three bearings caused the pressure at the main inlet rail to fall and the alarm to sound.

The defendants argued that it happened the other way round. Their case was that the damage started with the seizure of No. 7 piston which they attributed to poor combustion in No. 7 cylinder during periods when the vessel encountered heavy weather on the voyage to Manzanillo. Poor combustion gave rise to excessive carbon deposits within the cylinder which caused the piston rings to become gummed up, thereby losing some of their sealing properties. That in turn allowed blow-past of hot gases which broke down the lubricating oil film on the walls of the liner and caused overheating and seizure of the piston. The increasing friction between the piston and the liner placed stress on the crankpin bearing causing it to fail. No. 15 crankpin bearing was affected by a sudden local loss of oil pressure as No. 7 failed, and perhaps by heating and vibration as well, and also by debris extruded from No. 7. It failed shortly after. The loss of oil pressure at Nos. 7 and 15 crankpin bearings caused a fall in the pressure within the central bore of the crankshaft and thereby within No. 8 main bearing, in particular at the point where the channel to the central bore opens into the bearing. That was sufficient to disrupt the oil film in the loaded part of No. 8 main bearing which consequently failed. The combined effect of the failure of all three bearings caused the pressure at the main inlet rail to fall and the alarm to sound.(d) Circumstances surrounding the casualty

Before considering these competing arguments it is necessary to say something more about certain matters which occurred before and after the casualty which in some respects form an important part of the basis for the experts' conclusions. 

It is apparent from the vessel's rough engineroom logs that for some months prior to the casualty those on board had been keeping a detailed record of main engine filter flushings. The standard form of engineroom log book in use on this vessel does not provide for keeping any record of this kind, but on this ship the engineers made a daily note of the times of each flushing. The original reason for their doing so may ultimately not be of great significance, but the practice had been going on for some months before the chief engineer who was on board at the time of the casualty joined the vessel at the end of October, 1990,and I am satisfied that the owners' superintendent, Mr. Kallergis, must have been aware of it and may well have initiated it. The logs show that during the period from January to March, 1991 the filters were rarely flushed fewer than five times a day under normal conditions and that on occasions flushing occurred much more often, sometimes rising to 10 or even more times a day. On the face of it that is surprising, but it was common ground that the frequency of flushing is affected by various circumstances which are inevitably encountered from time to time. In particular, carrying out work in the crankcase inevitably results in the generation of dirty particles which cannot be removed and which therefore increase the amount of debris brought up to the filters during succeeding days; and the movement of the ship in heavy weather can stir up sludge in the sump tank and so increase theamount of material carried to the filters. In the present case it was agreed that there is a correlation between the frequency of flushing recorded in the logs and the incidence of heavy weather. The fact that flushing occurs frequently indicates that the filters are working effectively, but it also provides an indication of the quantity of particulate matter in the oil being drawn from the sump tank.

On her approach voyage to Manzanillo in March, 1991 the vessel encountered very heavy weather on various occasions between Mar. 3 and 15 with sea states as high as 10. At various times she reduced engine revolutions and when doing so the engineers changed to using diesel oil in place of the normal heavy fuel oil. During that period filter flushing occurred on average over 11 times a day. During the day after the vessel left Manzanillo the number of flushings was unusually high (23 are recorded in the 24 hour period to noon on Mar. 16), but that was due in part to the fact that work had been carried out in the engine while the vessel was in port. Following the casualty a considerable amount of work was carried out in the engine both at sea and at Puerto Cortés which would inevitably have generated a lot of small debris. The Classification Society surveyor who attended the vessel recommended, inter alia, the cleaning of the lube oil sump tank as well as the crankcase, probably as a precautionary measure. The chief engineer confirmed that the system was flushed after completion of repairs by disconnecting the feeder pipes to the main bearings and circulating oil through the crankcase from the sump via the filters. The filters could be expected to clean up the oil over a period of time, but the log indicates that in this case the whole operation, including the disconnection and reconnection of the feeder pipes, occupied rather less than three hours in all. It is unlikely, therefore, that it was sufficient to remove all of the debris created by the work which had gone on. As far as the sump tank itself is concerned, the evidence of both Mr. Alexopoulos and Mr. Kallergis was that it cannot be entered for cleaning 

179 by hand. Although structural drawings were not available, it was agreed that this tank, which lies underneath the main engine, is part of the ship's structure. I have no doubt that it is a very confined space and one which could only be entered with difficulty, but the ship's witnesses agreed that it was provided with an access manhole and I am unable to accept that it would not be entered for cleaning and inspection if the occasion required.

On completion of repairs at Puerto Cortés the vessel sailed to Costa Rica and thence to Zeebrugge. Flushing of the filters continued to occur at a constantly high rate, averaging over 11 times a day. Further work was carried out in the engine at Zeebrugge and on that occasion the vessel also took on a substantial quantity of fresh lubricating oil. The vessel sailed from Zeebrugge for South Africa late in the evening of Apr. 25. On Apr. 27 and 28 the filters were flushed 23 and 27 times respectively in 24 hours, which is a considerable number, even allowing for the fact that a good deal of work had been carried out in the engine. More surprising, however, is the fact that from Apr. 30 the frequency of flushing plummeted for no immediately obvious reason to an average of only three a day over a period of 10 days. No very satisfactory explanation for that was provided. On June 13 the vessel arrived at the anchorage at Pedro Bank off Jamaica where she drifted waiting for orders until early in the morning on June 22. After she left the anchorage the number of flushings dropped to an average of about four a day for the remainder of the period for which records were available.(e) Did No. 8 main bearing fail first?

The experts agreed that the failure of a main bearing leading to failure of a crankpin bearing and consequent seizure of its piston is a recognized sequence of events. It is perhaps for that reason that the initial instinct of all the marine engineers who considered the case, including Mr. Bevis, was that the casualty originated with the failure of No. 8 main bearing. However, Mr. Glennie, Q.C. for the owners (whose submissions on this issue were adopted by Mr. Young, Q.C. on behalf of REL) argued that there are a number of insuperable objections to this conclusion: first, the absence of any plausible mechanism to explain the failure of No. 8 main bearing; second, the lack of any signs of unusual wear in any of the other main bearings; third, the absence of any damage to No. 16 crankpin bearing; and fourth, the timing of the incident as described by the chief engineer. 

It is convenient to take the last of these first. I am cautious of drawing too strong an inference simply from the timing of the whole incident, partly because there is an element of uncertainty about it - the chief engineer's account of events and the experts' estimate of the time it would have taken him to reach the control room are inevitably imprecise - and partly because neither of the experts was able to place any precise estimate on the times that would be required for the various processes, particularly that of bearing failures, to occur. Nonetheless, the overall time scale of about a minute and a half between the chief engineer's hearing the slowing of the turbocharger and his shutting down the engine after hearing the alarm is not in my view inconsistent with the plaintiffs' case. It is true that it requires No. 8 main bearing to have been running in a progressively failing condition for something in the order of two minutes, but I do not think that is inconsistent with the damage which was subsequently found. The lower bearing shell had been completely wiped of both white metal and copper and the copper layer in the upper shell had been extruded so as partially to cover the oil ports. 

Dr. Baker, a metallurgist who gave evidence for the plaintiffs, was of the opinion that this extrusion of the copper had been caused by the expansion of the crankshaft journal due to friction heating. In his view the "blueing" of the bearing cap and saddle and the distortion of the horns of the bearing cap described by the Salvage Association surveyor was indicative of considerable heating which could not have occurred in a matter of seconds. His opinion was not shared, however, by Mr. Deegan, a metallurgist called by the defendants. It was his view that the degree of cracking and hardening exhibited by the journal in the way of No. 8 main bearing was not consistent with the degree of heating necessary to close the clearance within the bearing. Basing himself on the amount by which the Classification Society had required the journal to be machined in order to recover its original hardness he produced an estimate of the temperature gradient within the shaft itself and by reference to it calculated the amount of heat which had been produced during the destruction of the bearing. It was not in his view sufficient to close the clearance in the bearing. The difficulty with this exercise, however, is that there was really insufficient firm data from which to draw reliable conclusions. Moreover, Mr. Deegan's conclusions have to be tested by reference to the only other mechanism by which it was suggested the extrusion of the copper could have been caused, namely the "hammering" effect of upward forces on the crankshaft generated by the seized No. 7 piston. I think that is unlikely to have been the cause, however. The crankshaft is a massive piece of metal which, although capable of deflection, is essentially very stiff. Any "hammering" force generated by No. 7 piston would have had to be sufficient to deflect the crankshaft and would have been borne to an almost equal degree by No. 7 main bearing. More significantly, I am satisfied by the 180 evidence of both Mr. Martin and Mr. Bevis that such forces would almost certainly have been sufficient to cause considerable damage to the bottom end of No. 7 connecting rod. The damage to the connecting rod was, however, slight. In my judgment it is unlikely that the extrusion of the copper in the upper shell of No. 8 bearing was caused in that way and taking all these factors into account I think the more likely explanation is that it was caused by heat expansion of the journal. This argues in favour of the longer time scale required by the plaintiffs' case.

The defendants understandably laid a good deal of emphasis on the absence of excessive wear to any of the main bearings examined at Puerto Cortés and Zeebrugge following the casualty. Insofar as that suggests that there was no excessive or significant contamination of the lubricating oil by fine particles it does, of course, assist their case. However, although none of the shells, with the possible exception of the lower shell of No. 4 bearing, showed signs of exceptional wear for the period in service, there was no evidence to indicate whether the degree of wear was greater or less than might ordinarily have been the case at these periods for this vessel. Insofar as the condition of these other bearings is relevant to the question whether there was an unusually large quantity of contaminating particles in the lubricating oil, in my judgment it favours the plaintiffs' rather than the defendants' case. Each of the bearings showed signs of the "mottling" to which I have already referred. Both Mr. Martin and Dr. Baker described it as unusual and I did not gain the impression from either Mr. Bevis or Mr. Deegan that in their experience it was usual to find embedded particles in these quantities. The absence of apparent scoring of the kind that is associated with damage caused to bearings by particles in the lubricating oil is probably to be explained by the fact that the particles in question were small, under 30 μm in diameter and thus able to pass through the filters. Although the effect would be similar to that of larger particles, because of their size it would appear as polishing rather than scoring. The tentative conclusion that I would draw from the evidence provided by these bearings is that there had been an unusually high level of particles in the lubricating oil, and that view is reinforced by the evidence in the vessel's logs of the frequency of filter flushing during the months preceding the casualty.

As I have already mentioned, the frequency of filter flushing is affected by a number of factors so that high levels of flushing, even for fairly prolonged periods, do not necessarily indicate unusually heavily contaminated oil. Nonetheless, the broad sweep of the evidence in relation to this vessel over the period from January to July, 1991 points clearly to the conclusion that during the months leading up to the casualty there was an unusually high level of particulate matter in the oil which was being drawn up from the sump tank to the filters. The manufacturers of this engine in the U.K. suggest that on average the filters can be expected to flush about three times a day under normal conditions, but I would not place too much weight on this evidence since the frequency of flushing may depend to some extent on which of the many different models of filter which are available is fitted. The chief engineer said that on this vessel the filters would normally flush on average five or six times a day, and although in the period following the vessel's time at anchorage off Jamaica the average fell to three a day and remained at that level for a long time thereafter, I am prepared to assume for present purposes that five or six times a day was more normal for this vessel. However, the high frequency with which flushing occurred both before and after the casualty cannot be explained simply by the effects of bad weather and work being carried out in the engine. It is in my view indicative of there being an unusually high level of particulate material or "sludge" in the sump tank. With varying degrees of enthusiasm both Mr. Martin and Mr. Bevis accepted that there was probably some correlation between the amount of "large" (that is, over 30 μm) particles and "small" (under 30 μm) particles being drawn up to the filters so that if there were an unusually high number of large particles being captured in the filters there were probably also an unusually large number of small particles passing through the filters and circulating in the oil reaching the engine.

No. 16 crankpin bearing is the next bearing forward of No. 8 main bearing from which it receives its lubrication. Some of the shards of metal created by a failure of No. 8 main bearing would be carried away from the bearing by the lubricating oil into the central bore of the crankshaft. There they would follow the flow of the oil, some towards No. 7 crankpin bearing and some towards No. 16. The absence of any damage to No. 16 crankpin bearing is difficult to explain if, as the plaintiffs contend, the casualty began with the failure of No. 8 main bearing. Mr. Martin accepted that on his view of events as much debris should have been carried to No. 16 crankpin bearing as to No. 7. The only explanation he could offer for the absence of damage was that debris may have started to build up at the entrance to No. 16 but had not reached the point where it reduced the flow of oil sufficiently to cause a failure before No. 7 crankpin bearing failed. The defendants said that if the casualty began with the seizure of No. 7 piston, on the other hand, the absence of damage to No. 16 crankpin bearing is easier to explain. The destruction of No. 7 crankpin 

181 bearing and the consequent increase in the clearances would have allowed oil to flow without interruption through that bearing, and the failure of No. 15 crankpin bearing would have contributed to that effect. That would have caused a bias in the flow of oil from No. 8 bearing in the direction of No. 7 crankpin bearing, and as No. 8 main bearing failed little of the debris would have been carried towards No. 16 crankpin bearing.

There is no doubt, as Mr. Martin agreed, that it is difficult to reconcile the absence of damage to No. 16 crankpin bearing with the plaintiffs' case and as such it is a telling piece of evidence against it. However, the precise manner in which the destruction of a bearing occurs and the behaviour of surrounding elements, including the flow of oil within the crankshaft and through adjacent bearings, is not susceptible of precise analysis. Apparently random effects occur and no doubt the precise course of events varies from case to case. Although this is undoubtedly an obstacle to the plaintiffs' case, it does not tell conclusively against it but must be put into the scales with all the other evidence. 

Finally I come to the question of the mechanism by which particles in the lubricating oil could have caused the failure of No. 8 main bearing. I think it is fair to say that Mr. Martin's opinion that the failure of the bearing was caused by an excessive concentration of hard particles in the lubricating oil was primarily based on two things: the evidence that there were unusually large quantities of small particles in the lubricating oil, to which I have already referred; and the practical experience of marine engineers over the years that nearly all main bearing failures are caused by contamination of the lubricating oil, for example, by water or fuel oil or foreign bodies. He regarded the link between contamination (including contamination by excessive quantities of small particles) and bearing failure as sufficiently well established not to require any further explanation, given that in his view there was no plausible initiating cause of the casualty other than the failure of No. 8 main bearing. However, although by his own admission he was reaching the limits of his expertise, he did suggest a mechanism by which that could have occurred in this case, namely, that the number of small particles circulating in the oil exceeded the ability of the dispersant chemical in the oil to keep them apart. Some particles, he suggested, agglomerated to form bodies of particulate matter held loosely together in matrices perhaps as large as a fingernail and up to 50 μm thick. If drawn into the load bearing part of the bearing (where the oil film is only some 2 - 5 μm thick) such a loose body of particles would be sufficient to disrupt the film and cause the bearing to fail.

This suggestion was heavily attacked by the defendants. Mr. Bevis had never heard of small particles agglomerating into bodies of the kind suggested by Mr. Martin and said that as far as he was aware there was no reference to any phenomenon of this kind in any of the technical literature. (Mr. Martin agreed that that was so.) He suggested that bodies of particles of the kind postulated by Mr. Martin would deform and break up under pressure within the bearings, or, if they did not, that their presence would be reflected in scoring damage on the running surface of the shells. Mr. Martin accepted that if he was right agglomerated bodies of the kind he was suggesting must have been present in the oil for some time prior to the casualty. The absence of such damage to any of the other main bearings therefore assists the defendants. These are powerful arguments against the plaintiffs' case. However, as a matter of ordinary experience it is sometimes possible to be satisfied, even, perhaps, to a high degree of probability, that a certain effect proceeds from a given cause even though one cannot identify the mechanism whereby it comes about. Again, therefore, I do not regard them as conclusive; they too have to be considered in the light of all the evidence.

The defendants' case that the casualty began with the seizure of No. 7 piston likewise faces some considerable evidential obstacles. The plaintiffs contended that the following are some of the most significant: the absence of any direct evidence of conditions which might account for a seizure of No. 7 piston; the absence of significant damage to the bottom end of No. 7 connecting rod; the evidence concerning the propagation of pressure changes throughout the lubrication system; the absence of damage to No. 7 main bearing; and the overall timing of the casualty.

The weather conditions which the vessel encountered on her approach voyage to Manzanillo were certainly severe and were of a kind in which poor combustion can occur if a vessel with this type of engine is not handled properly. Poor combustion and overloading is normally associated with higher than normal cylinder pressures, an increase in exhaust temperatures as well as the formation of carbon deposits. Mr. Bevis was of the opinion that the design of this engine with its two-stage turbocharger made it particularly susceptible to poor combustion, but he accepted that if that were so it was a characteristic which affected all units equally and one which would occur whenever the vessel encountered heavy weather. As such, he accepted that it would be one of which the owners and engineers would have been aware. In fact, however, there was no suggestion from any of the second defendants' witnesses that that was the case, and even though the chief engineer had drawn many of 182the pistons for routine maintenance during his time on board, no signs of such a problem appear to have been observed. 

The engineroom log covering the passage from Zeebrugge to Manzanillo gives no indication that the performance of the engine was in any respect abnormal during that period and nothing was said by the chief engineer to suggest that it was. I do not find that surprising. Although the weather was undoubtedly severe, no one suggested that it exceeded the limits within which the vessel and engine were designed to operate, and the change from fuel oil to diesel oil when the engine was running at reduced speed on various occasions should have avoided any problems that might otherwise have arisen from poor combustion. None of the units exhibited signs of overloading or poor combustion, either during the voyage or when subsequently examined. Likewise, there is little evidence to support the conclusion that blow-past occurred in No. 7 unit after the vessel left Manzanillo as a result of the rings of No. 7 piston having become fouled by carbon deposits and sticking in their grooves. If there had been an earlier build up of carbon of that kind it is difficult to see why significant blow-past should only have occurred at that stage. It is true that on arrival at Puerto Cortés the rings were found to be stuck in the grooves, but that was after the casualty and was probably due to the effects of friction and heating. More importantly, none of those who inspected the vessel at Puerto Cortés noted any signs of carbon deposits on the piston. Given the nature of the casualty both the surveyors and Mr. Kallergis can be expected to have taken a close interest in No. 7 unit. It is clear that Mr. Stewart who attended on behalf of Fyffes was looking for the presence of carbon deposits and no doubt the others were as well. It may be that by the time he took his photographs some cleaning of No. 7 piston had taken place, but the absence of any reference in any of the surveyors' reports to the presence of carbon deposits persuades me that there were none of any significance. 

The defendants' case proceeds on the footing that it was the forces generated by the seizure of No. 7 piston that led to the destruction of all three bearings. The experts agreed that failure of the crankpin bearing is a recognized consequence of a serious piston seizure and little imagination is required to see why that should be so. However, there was less agreement about whether failure of the crankpin bearing under those circumstances could be expected to lead to a main bearing failure. Mr. Martin considered it most unlikely. He said that the weakest point in the system was the bottom end of the connecting rod which took all the force of the piston dragging in the liner. It was not uncommon in his experience to find that the bolts holding the bottom end together had been stretched or even burst, but even under those conditions there had been no damage to the main bearings. Mr. Bevis said he had heard of cases in which main bearings had suffered damage as a result of piston seizures, but he

did not seriously dispute Mr. Martin's evidence concerning the likelihood of damage to the connecting rod and I was left with the clear impression that he did not regard damage to main bearings as by any means a common consequence of piston seizure. Again, that does not strike me as surprising in the light of the size and strength of the crankshaft in comparison with the connecting rods. In the present case, of course, No. 7 crankpin suffered very little damage. 

In this context the absence of any damage to No. 7 main bearing also presents difficulties for the defendants. No. 7 is the next main bearing aft of No. 15 crankpin bearing. As such it lies in the same position relative to No. 15 crankpin bearing as No. 8 main bearing does to No. 7 crankpin bearing. If the forces generated by the seizure of No. 7 piston were sufficient to cause damage to No. 8 main bearing, it is difficult to see why No. 7 should remain unscathed. Similar questions arise in connection with the effects of a drop in lubricating oil pressure following the failure of the crankpin bearings. It was common ground that the failure of Nos. 7 and 15 crankpin bearings would have had at least two effects: it would have caused the oil pressure at those bearings to fall to atmosphere and the pressure in the crankshaft bore to fall to the same or almost the same level; and it would have led to an increased flow of oil through Nos. 7 and 8 main bearings to the open bearings at No. 7-15 crankpin. There would be no significant difference between the pressure within the bore in way of No. 7 and No. 8 main bearings. Accordingly, if the casualty began with the seizure of No. 7 piston it is difficult to explain why No. 7 main bearing showed no signs of stress.

The means by which, in the opinion of Mr. Bevis, the failure of No. 7 crankpin bearing led to the failure of No. 8 main bearing involved severe localised falls in oil pressure, first at No. 15 crankpin bearing and then, after that failed, at No. 8 main bearing. Nos. 7 and 15 units are served by the same crankpin and the two bearings are only about 6mm apart. I have little difficulty, therefore, in accepting that a seizure of No. 7 piston could lead to the failure of No. 15 crankpin bearing as a result of a combination of loss of oil pressure, mechanical forces and heat transfer. To explain the damage to No. 8 main bearing is more difficult, however. Mr. Bevis postulates a drop in pressure within the central bore of the crankshaft from a working pressure of around 2 bar to something close to 

183 atmosphere as a result of the failures of Nos. 7 and 15 crankpin bearings. He suggested that a drop in pressure of that magnitude would be sufficient to disrupt the oil film within the loaded part of No. 8 main bearing causing it to fail.

Ultimately there was a very large measure of agreement between Mr. Bevis and Mr. Martin about the varying pressure within the lubricating oil system and the manner in which changes in pressure were propagated through the system. I am satisfied having heard their evidence that if the casualty began with the seizure of No. 7 piston, by the time No. 15 crankpin bearing had failed the oil pressure at that point would have fallen nearly to atmosphere. There would also have been a very rapid fall in pressure within the bore of the crankshaft which would for all practical purposes be equal throughout its length. Such a drop in pressure might be sufficient to disrupt the oil film in one or more of the main bearings, but would only be likely to do so if a bearing was operating close to the limit. I say that because neither of the experts suggested that main bearing failure is a normal consequence of the destruction of a crankpin bearing, or even one which is regularly encountered. The fact that none of the other main bearings (with the possible exception of No. 4) showed signs of stress tells against this conclusion.

Finally there is the matter of timing. The defendants' case necessarily requires that virtually all the damages occurred within the space of about one and a half minutes. Mr. Martin could not accept that. In his view that simply did not allow enough time for what he described as "the whole show" including the heating of No. 8 main bearing saddle and cap. Mr. Bevis said that it was the failure of No. 8 main bearing coming at the end of the sequence of events which caused the pressure back at the main inlet rail to drop and the alarm to sound, the engine then being shut down very soon after. However, in the view of Mr. Martin and Dr. Baker that would not have allowed No. 8 main bearing to run in a failed condition for long enough to generate the heat required to cause the damage noted to the bearing cap and saddle, particularly bearing in mind that it continued to receive a supply of coolant in the form of lubricating oil until the engine was shut down.(f) Conclusions

Having stated the competing considerations at some length I propose to state my conclusions on this part of the case fairly shortly. Only two events have been put forward as possible initiating causes of the casualty, but it is not simply a question of choosing between them; if the plaintiffs are to succeed it is for them to satisfy me on the balance of probabilities that the damage began with the failure of the No. 8 main bearing. If they succeed in that they must go on to show that the bearing probably failed because of contaminants in the lubricating oil.

I am at least able to conclude with some confidence that the casualty did not start with a seizure of No. 7 piston, principally because there is nothing in the evidence before me to support the suggestion that there was any overloading of No. 7 unit leading to blow-past. Moreover, the absence of carbon deposits, the absence of any serious damage to the connecting rod, and the absence of any damage to No. 7 main bearing all support the conclusion that the casualty did not begin in that way. Moreover, Mr. Bevis himself very fairly accepted that if one assumes that the surveyors had looked for, but failed to find, any carbon deposits on No. 7 piston (as I am satisfied they did), he could not provide any explanation for a piston seizure let alone suggest a mechanism by which it might have occurred. No-one suggested that pistons are apt to seize for no apparent reason. In those circumstances I am satisfied that this theory must be discarded. Apart from that, however, such conclusions as can be drawn from the timing of the incident as a whole and the absence of damage to other bearings, particularly No. 7 main bearing, also point away from a seizure as being the initiating cause. I am unable to accept that the extrusion of the copper on only one side of the upper half of No. 8 bearing shell was caused by forces generated by the seized piston. In my judgment it was caused by the heating and thermal expansion of the crankshaft journal and the time required for that to occur is consistent with the degree of heating required to cause the discolouration of No. 8 bearing cap and saddle. I do not think that a sufficient degree of heating would be likely to have occurred within the time span postulated by Mr. Bevis. This too supports my concl usion that the casualty did not occur in the manner he suggested.

Despite the difficulties to which I have already referred, I am persuaded that the casualty began with the failure of No. 8 main bearing and that that failure was caused by excessive amounts of small particles in the lubricating oil. I have reached that conclusion not simply because I have rejected the alternative theory (though the fact that I have done so inevitably points towards the failure of No. 8 main bearing as being the primary cause), but because it accords better with the evidence as a whole. The fact that Mr. Martin could not put forward a compelling explanation of the precise mechanism by which the failure came about is not in my judgment fatal to the plaintiffs' case. At least one publication in the field of marine engineering which was before me shows that it has been recognized for some years that even very small 

184 particles (5 μm and upwards in diameter) are capable of causing damage to highly loaded bearings of the kind in question. Although there is nothing in the published material which supports the specific mechanism proposed by Mr. Martin, this is not, therefore, a case of a wholly inexplicable failure; it is a case of a failure of a recognized kind which can be linked by other evidence to a cause of a recognized kind even though it may be difficult to describe the precise mechanism by which cause leads to effect. As I suggested earlier, practical experience may lead one to be satisfied that such a link exists even though it cannot be fully explained. I accept Mr. Martin's evidence, and the evidence of various publications put before me by the parties, that bearing failures are usually caused by some kind of contamination of the lubricating oil. (It is agreed that there is no question of any misalignment of the crankshaft in this case.) In this case there is evidence to which I have already referred that leads me to conclude that there was an unusually high level of particles in the lubricating oil circulating in the engine. Moreover, there is other evidence in this case which I shall mention in a moment that points clearly to the failure of No. 8 main bearing as the initiating cause, but no evidence of any likely cause of that failure other than contamination of the oil by excessive quantities of small particles.

The overall timing of the incident is in my judgment more consistent with the plaintiffs' case than with that of the defendants. I have already given some of the reasons for reaching that conclusion. The seizure of No. 7 piston was progressive and had clearly been going on for a while before it became serious enough to affect the speed of the turbocharger. Nonetheless, if the casualty began with the seizure it would have to have developed very rapidly in order to cause a failure of No. 8 main bearing within the short time available. Mr. Martin thought that that sequence of events could not occur in much less than five minutes. Having regard to the evidence of the degree of heating of No. 8 main bearing I think it unlikely that the damage could all have occurred within the time scale established by the chief engineer's evidence.

Finally there is the conundrum posed by the absence of damage to No. 16 crankpin bearing. None of the surveyors who attended the vessel at Puerto Cortes remarked on the presence of debris from the failed No. 8 main bearing within the bore of the crankshaft or in way of No. 16 crankpin bearing, but they did not remark on the presence of debris anywhere else either. That may just be because the existence of debris following multiple bearing failures was not remarkable. At all events, although this piece of evidence tells against the plaintiffs' case, it is not so strong in itself as to preclude a conclusion which in my judgment is otherwise supported by the rest of the evidence as a whole. I think it likely that some debris from No. 8 main bearing was carried towards No. 16 crankpin bearing but that the effect of the flow of oil from No. 9 main bearing (which, being at the extreme forward end of the crankshaft supplied oil only to Nos. 8 and 16 crankpin bearings) prevented such large quantities from reaching it and that although a blockage began to form it did not reach the point at which the flow of oil to the bearing was interrupted.

For these reasons I have reached the conclusion that the initiating cause of the casualty was the failure of No. 8 main bearing which was itself caused by an unusually high concentration of particulate matter in the lubricating oil.

Finally, before leaving this aspect of the matter I would like to pay tribute to the expert witnesses for the admirable way in which they gave their evidence which I found of the greatest assistance.3. Was the casualty caused by unseaworthiness of the vessel?

I was reminded by both Mr. Young and Mr. Glennie that the concept of seaworthiness is not absolute in the sense that a vessel will not be considered unseaworthy merely because with hindsight she can be seen to have been suffering from some deficiency which in due course led to a casualty. In F.C. Bradley & Sons Ltd. v. Federal Steam Navigation Co. Ltd., (1927) 27 Ll.L.Rep. 395 the point was put by Viscount Sumner in this way at p. 396: 

In the law of carriage by sea neither seaworthiness nor due diligence is absolute. Both are relative, among other things, to the state of knowledge and the standards prevailing at the material time . . . and in M.D.C. Ltd. v. N.V. Zeevart Maatschappij "Beursstraat", [1962] 1 Lloyd's Rep. 180 Mr. Justice McNair applied what he regarded as the most common test, and is still regarded as a good test, namely:

Would a prudent shipowner, if he had known of the defect, have sent the ship to sea in that condition? Another way of putting it is that the vessel must be fit in all respects to carry her cargo safely to its destination having regard to the ordinary perils to which such a cargo would be exposed on such a voyage (see Scrutton on Charterparties, 19th ed., art. 48). The defendants say that the vessel was fit to encounter the ordinary perils of the voyage and that a prudent shipowner would happily have sent her to sea even if he had known that there was an unusually large amount of sludge in the sump tank 

185 and in consequence that the lubricating oil circulating in the engine contained an unusually high level of particles small enough to pass through the filters. I am unable to accept that. The general nature of the effect which particles in the lube oil may have on bearings and of the ability of bearing surfaces to deal with them is well known and has been for some time, and one of the reasons for undertaking regular expert analysis of lubricating oil is to ascertain whether there is an excessive build up of particulate matter. Of course it can be said that provided the filters are in good working order, as they were in this case, there is some assurance that the particles which are small enough to pass through them can be safely accommodated by the bearings. In that sense they can be viewed as "design size" particles, to adopt Mr. Young's phrase, but I do not think the matter is as simple as that. The problem is one of degree. Mr. Martin did not consider that it was satisfactory simply to rely on the efficiency of the filters and Mr. Bevis did not dispute the suggestion that it was not good practice to have as much sludge in the sump tank as this vessel had, and one can see why: the greater the amount of particulate matter circulating with the oil, the greater the risk of damage to the bearings. I do not think that a prudent owner would simply rely on the filters and assume that the quantity of fine particles was a matter of indifference. Whether he would have foreseen the particular sequence of events which occurred in this case, or would have appreciated the mechanism by which they occurred, does not matter. For these reasons I am satisfied that when she set out from Zeebrugge on her approach voyage to Manzanillo the vessel was unseaworthy because of the amount of fine particulate matter in the lubricating oil. 4. The exercise of due diligence

The question whether the owners exercised due diligence to make the vessel seaworthy does not loom large in this case because they accepted that having failed to have regular analyses of the lubricating oil carried out it would be difficult for them to argue successfully that they had done all that they could reasonably have done to ensure that the oil was fit for service. In my judgment they were right to make that concession since regular independent analysis of the lubricating oil is a standard precaution against contamination by water and other foreign matter. Nonetheless, Mr. Glennie on behalf of the owners submitted that that was the only criticism that could be made of them and that their failure to exercise due diligence in that respect was not causative of the loss. In support of the proposition that they were therefore still entitled to rely on the protection given by art. IV of the Hague-Visby Rules he relied on the decisions in Union of India v. N.V. Reederij Amsterdam (the Amstelslot), [1963] 2 Lloyd's Rep. 223 and The Yamatogawa, [1990] 2 Lloyd's Rep. 39.

In The Amstelslot the issue was whether the surveyors who had inspected the vessel's reduction gearing ought to have carried out a more detailed kind of inspection which would have revealed the existence of the fatigue crack which led to the casualty. Mr. Justice McNair held that their failure to do so was not negligent and his finding on that point was upheld by the House of Lords. That was not a case, therefore, in which the Court found that there had been any failure on the part of the owners to exercise due diligence and it does not take the matter any further. The Yamatogawa, however, does support Mr. Glennie's proposition. That case also concerned a failure of the vessel's gearing. Mr. Justice Hobhouse found that there had been a failure to exercise due diligence in one respect when the gearing was inspected, but he also found that that failure had no causative effect on the casualty. The claim therefore failed.

In my judgment the present case is rather different. Regular independent analysis of the lubricating oil is a standard precaution against contamination by water and other foreign matter and would probably have shown that there was excessive particulate matter in the oil. However, I do not think that criticism of the owners can be confined to their failure to have such analyses regularly carried out. The unusually high frequency of filter flushings which had been a continuous feature of this engine's operation prior to the voyage was sufficient to indicate that there was a large amount of sludge in the sump tank which ought to have been cleaned. I accept that the sump tank was not easy to enter because of its size and construction, but I am not satisfied, as I have said, that it was completely inaccessible, much less that it was impossible to remove sludge from it by one means or another. In these circumstances I am satisfied that the owners' did fail to exercise due diligence in the respect I have mentioned and that their failure to do so caused or contributed to the casualty.

In the light of these findings I turn to consider the various claims in the action.5. Fyffes' claim against REL

The COA between Fyffes and REL contained the following terms:

1.

It is this day . . . . . mutually agreed between [REL], Owners, Disponent Owners, or Operators of refrigerated vessels, hereinafter called Owners, and [Fyffes], hereinafter called Charterers, that the Owners will provide first class refrigerated vessels and the Charterers will ship cargoes 

186 of fresh green bananas in cartons . . . . . . and any other suitable products that they may from time to time transport . . . . .

2A. LOAD

At the inception of the service the said vessels shall proceed to a safe port, or ports, out of Big Creek, Belize, and/or Pto Cortés, Honduras, and/or Manzanillo, Dominican Republic, with the loading rotation and number of ports being at Charterers' option . . . . . and there load cargoes . . . . . and being so loaded shall proceed to the discharge port(s) 

2B. DISCHARGE

At the inception of the service the vessels will proceed to Portsmouth, England, as ordered on signing Bills of Lading . . . . . Charterers have the right to discharge Zeebrugge . . . . . with rotation of discharge ports at Charterers' option . . . . 

12. CARGO DAMAGE

. . . 

Loading, stowage and discharge to be always the responsibility and risk of the Charterers . . .

16. BILL OF LADING

Charterers' Bills of lading attached hereto are to be used. All other terms, clauses, exceptions, and conditions not expressed in this Agreement to be as per Charterers' Bills of Lading. Any conflicts existing between the language contained in this Agreement and that of the Bills of Lading shall be governed and controlled by this Agreement. This Agreement is a private contract of carriage and it is agreed that no negotiable Bill of Lading or similar document of title will be issued under this Agreement. . . .

29. NOTICES

. . . Owners shall provide the following reports while at sea:

. . .

b) daily Eta to next port

c) any change in Eta

. . .

31. GOODWILL CLAUSE

Both Charterers and Owners will use their best endeavours to ensure both parties' interests are looked after and agree to work together to overcome any operational and production problems that may arise during this contract period. . . .

32. UNITED KINGDOM CARRIAGE OF GOODS BY SEA ACT 1971

This Contract of Affreightment shall have effect subject to the provisions of the United Kingdom Carriage of Goods by Sea Act 1971 which shall be deemed to be incorporated herein . . .

Fyffes put their case against REL in two ways. Their primary submission was that the REL as carrier under the COA was in breach of its duty under art. III r. 1 of the Hague-Visby Rules to exercise due diligence to make the vessel seaworthy. Alternatively, they submitted that REL as carrier was in breach of the implied obligation to proceed on the voyage with all reasonable despatch.(a) Was REL obliged to load at Santo Tom s?

Before turning to the detail of the parties' arguments concerning the obligations of REL under the COA, however, it is convenient to deal with a preliminary argument raised by Mr. Young. The casualty occurred as the vessel was on her way from her first port of loading, Manzanillo, to Santo Tom s, Guatemala where she was due to load a further parcel of bananas before sailing to Big Creek, Belize to complete. Orders to load at these three ports had been accepted by REL and the owners apparently without question. Big Creek and Manzanillo were two of the ports named in cl. 2A of the COA. Santo Tom s, however, was not so named and Mr. Young submitted that in the absence of a variation of the COA (which he said had not been pleaded or proved) there was no obligation on REL to send the vessel to Santo Tom s at all. Accordingly, REL could be under no liability as a result of the vessel's failing to load there.

Although it was possible when looking back to see that this argument had been foreshadowed in an oblique way during Mr. Young's opening remarks, I think it is fair to say that it did not feature with any prominence at any stage in the proceedings until he came to make his final submissions. In the circumstances I therefore considered it right to allow Fyffes to amend their points of reply at that late stage in order to allege that the COA had been varied by agreement between the parties or that REL was estopped from contending that Santo Tom s was not a contractual loading port.

For reasons which I have indicated this question was not investigated with the witnesses as thoroughly as might have been appropriate. Nonetheless, I am satisfied on the evidence before me that there had been agreement between Fyffes and REL to extend the service under the COA to include ports in Guatemala, specifically Santo Tom s. The exchanges between the parties which were before the Court do not go back beyond Feb. 19, 1991, but it is apparent that even before that date there had been discussions about lifting cargoes from Santo Tom s. By that time REL had already become concerned about the amount of fruit Fyffes were making available for shipment. Claims for deadfreight had been made and adjustments to the COA were being considered. Both Fyffes and REL were 

187 keen to increase the volume of cargo available and by the middle of February the possibility of obtaining bananas from Del Monte at Santo Tom s had been explored. Santo Tom s lies between Puerto Cortes and Big Creek and it was not suggested that there were any disadvantages from REL's point of view in loading there. An internal memorandum written by Mr. Templeman of Fyffes on Apr. 19, 1991 comparing their original plan with the position as it had by then developed refers to Fyffes having obtained the flexibility to load also in Guatemala. The orders to load at Santo Tom s appear to have been conveyed by Fyffes direct to the vessel soon after she sailed from Manzanillo. On receiving them the masterinformed REL asking for its confirmation which he received shortly after. It appears that REL may have already heard from Fyffes that orders had been given to the vessel; at all events there was never any indication that that came as anything of a surprise or that REL did not regarded them as good orders under the contract. Mr. Young accepted that they were orders which REL was quite willing to comply with; the only question, as he accepted, was whether that part of the adventure was being performed under the COA or outside it under some voluntary ad hoc arrangement (although REL would naturally have expected to be paid freight, presumably on a quantum meruit basis).

The suggestion that REL was willing to load at Santo Tom s without comment of any kind on an ad hoc basis outside the terms of the COA strikes me in all the circumstances as fanciful. Mr. Ira Heisler, a senior vice-president of REL responsible for chartering of vessels, had already embarked on direct negotiations with Fyffes over the problems caused by the shortage of fruit which he described in a written statement served at a late stage during the trial. It is significant, in my view, that nowhere in that statement did he suggest that REL had no obligation to load at Santo Tom s or that it had been willing to allow the vessel to load there on some ad hoc basis outside the terms of the COA. I think the reason is clear enough: REL had already agreed that Fyffes should have the flexibility to load in Guatemala as reflected in Mr. Templeman's memorandum. I [1996] 2 Lloyd's Rep. 171 Page 20 [1996] 2 Lloyd's Rep. 171 [1996] 2 Lloyd's Rep. 171 (Cite as: [1996] 2 Lloyd's Rep. 171) reject Mr. Young's submission that there was no consideration to support a variation of that kind (never an attractive argument at the best of times). Both parties wanted to ship more fruit and each of them could expect to gain by mutually agreeing to extend the COA to cover loading in Guatemala. For these reasons, therefore, I am satisfied that well before the voyage with which I am concerned the COA had been varied to include Santo Tom s as a contractual port of loading and that accordingly REL was obliged to load cargo there. (b) Breach of duty under art. III, r. 1 of the Hague-Visby Rules 

I come, then, to the first of the way in which Fyffes put their case against REL. Basing himself on Adamastos Shipping Co. Ltd. v. Anglo-Saxon Petroleum Co. Ltd. (the Saxon Star), [1957] 1 Lloyd's Rep. 271; [1957] 2 Q.B. 233 ; [1958] 1 Lloyd's Rep. 73; [1959] A.C. 133 Mr. Flaux on behalf of Fyffes submitted that the effect of cl. 32 of the COA was to impose on REL as carrier an obligation to exercise due diligence before and at the beginning of every voyage performed under the COA to make the ship by which it was performed seaworthy in accordance with art. III, r. 1. That obligation, he submitted, extended not just to cargo- carrying voyages, but to ballast voyages as well. Accordingly, REL was in breach of the COA inasmuch as at all material times after she left Zeebrugge on her voyage to Manzanillo Kriti Rex was unseaworthy by reason of the owners' failure to exercise due diligence and it was that unseaworthiness which caused the voyage to be abandoned. (c) Is the COA a contract of carriage?

In response Mr. Young made several interrelated submissions, two of which are of particular importance because they provided the basis for a number of other ways in which he developed his arguments on this part of the case. The first of these was that the COA did not represent (in his own words) a contract of carriage between REL and Fyffes (except in cases where a vessel owned by REL itself was employed), but was in truth a contract for carriage, that is a contract under which REL merely undertook to procure contracts of carriage between shipowners and one or other of the plaintiffs on terms of Fyffes' standard form bill of lading. Thus, although he accepted that some clauses concerning the vessel's movements, in particular cl. 29, did impose an absolute personal obligation on REL, its obligations in relation to the seaworthiness of the vessel in the present case were limited to using its personal best endeavours in accordance with cl. 31 to procure a vessel which was seaworthy.

I am unable to accept this view of the COA. It is, of course, necessary to look at the COA as a whole against its commercial background in order to ascertain the essential nature of the contract which it embodies, but in this case both practical commercial considerations and the language which the parties used point unequivocally to the conclusion that the COA is a contract of carriage. This was a new venture between Fyffes and REL which had been the subject of extensive discussions and negotiations, and although the parties were unsure how some aspects of the business would work in practice, it was not in either party's interest for the 188 arrangement to be any less well defined than necessary. An agreement of the rather nebulous kind suggested by Mr. Young would in my view provide a most unlikely commercial vehicle from the point of view of either party. However, the language of the COA puts the matter beyond any doubt. Although not conclusive, the agreement is described as a "Contract of Affreightment" which in the context of carriage by sea is an agreement of a well-recognized kind, and consistently with that REL is described in cl. 1 as "Owners, Disponent Owners or Operators" and Fyffes as "Charterers". Clause 2A is couched in language which reflects standard charter-party obligations in relation to the loading, carriage and discharge of cargo, but most telling is cl. 16 in which the agreement is described in terms as "a private contract of carriage". Numerous other indications throughout the agreement point the same way and the very fact of the incorporation of the Carriage of Goods by Sea Act, 1971 supports the conclusion that the parties understood themselves to be entering into a contract of a kind to which the provisions of the Act would be relevant. In truth, there is nothing in the language of the contract or its background which supports Mr. Young's contention. (d) One voyage or three?

The second point was that in a case such as the present where the vessel was ordered to three loading ports there were three separate voyages for the purposes of the COA. The essential nature of the service to be provided under the COA can be found in cll. 1, 2A, 2B and 3. These clauses indicate that vessels owned or operated by REL under time charter will undertake voyages to Central America where they will load cargoes at one or more ports and carry them to Portsmouth or Zeebrugge. That itself suggests that the parties regarded the whole trip from commencement of the approach voyage to the first loading port to completion of discharge at the final discharging port as constituting a single voyage for the purposes of this contract, a view which is fortified by the indication in cl. 8 that REL's intention was to operate vessels on a round voyage trip basis. It would make little commercial sense to treat a voyage on which the vessel was ordered to load at three ports instead of one as three separate voyages for the purposes of this agreement, and I am confident that that is not what the parties had in mind. (e) Did the Hague-Visby Rules apply at the time of the casualty?

I turn, then, to Mr. Young's specific points in relation to the impact of the Hague-Visby Rules. His first point was that since the Hague-Visby Rules are incorporated into the COA only by contract, it is necessary to examine the contract as a whole to see which aspects of performance they were intended to regulate. Thus far I agree with him and I do not think it necessary to refer to the authorities which he cited in support of the proposition. His next point was that the COA in this case is essentially different from a consecutive voyage charter for a specific vessel of the kind under consideration in The Saxon Star. In particular, because the COA does not identify any specific vessel to perform any particular voyage and contains no machinery for nominating vessels, REL was entitled to fulfil its obligations by providing any vessel to load a particular cargo provided it presented at the loading port at the prescribed time. It was therefore impossible to identify any point at which the vessel performing the voyage had to be seaworthy other than the moment when loading began, and so no obligation to exercise due diligence could sensibly arise, he submitted, until the vessel had actually begun to load cargo. That, he said, was consistent with other terms of the contract. The COA incorporated the terms of the plaintiffs' form of bill of lading which provided, inter alia, as follows:

1. . . . . . The carrier shall not be liable in any capacity whatsoever for any delay, non-delivery, or misdelivery, or loss of or damage to the goods occurring while the goods are not in the actual custody of the carrier. The carrier does not undertake that the goods shall arrive at the Port of Discharge at any particular time or to meet any particular market or use and the carrier shall in no circumstances whatsoever and howsoever arising be liable for direct or consequential loss or damage caused by delay or, save as expressly provided herein, otherwise. Accordingly, there was a clear provision that REL accepted no responsibility until such time as goods were actually on board. An agreement to restrict the carrier's liability in that way, he submitted, does not conflict with the Hague-Visby Rules since by their own terms they relate only to cargo carrying voyages (see art. I(b), (c) and (e) and art. II), and in respect of such voyages the parties are free to determine what responsibilities the carrier undertakes to which the Hague-Visby Rules then apply: see G.H. Renton v. Palmyra Trading Corporation, [1956] 2 Lloyd's Rep. 379; [1957] A.C. 149 . In the present case the parties have agreed that loading and discharging are Fyffes' responsibility and by virtue of the incorporation of the bill of lading terms, in particular in this context cl. 1, they have made their intention that REL is to be under no responsibility before cargo is loaded, other than to use its best endeavours in accordance with cl. 31. Each loading port gave rise to a separate voyage for the purposes of the COA, and REL could have withdrawn the vessel after it had loaded at 

189Manzanillo, if it had wished to do so, and unilaterally substituted one or more different vessels to load at Santo Tom s and Big Creek. Accordingly, since the voyages to Santo Tom s and Big Creek were separate voyages under which no cargo was ever loaded, REL was not liable for breach of any duty under the Hague-Visby Rules, and having itself exercised its best endeavours to provide a vessel it was not in breach of its obligation under cl. 31.

This is an ingenious argument, but I cannot accept Mr. Young's construction of the COA or the conclusions which are said to follow from it. In my judgment it is clear, for the reasons I have already given, that even though the vessel may be ordered to load at more than one port, as she was in this case, there is only one voyage for the purposes of the COA. Since the COA gives REL considerable freedom to choose the vessels by which the service is to be performed and makes no provision for the nomination of vessels, there is clearly room for argument as to exactly when, if at all, in the course of her approach to the first port of loading the vessel enters upon the voyage. I am satisfied that during the negotiations Fyffes made REL aware of the general requirements of the banana trade and that it was primarily for that reason that cl. 29 was included in the COA. Clause 9 indicates that dates established by the giving of such notices were intended to have contractual significance and in my view the general scheme of the COA contemplates that a vessel will be committed to the voyage from the time that REL accepts orders for her to proceed to a port of loading. If necessary, therefore I would hold that the vessel begins the voyage at that time, although if I am right in holding that there was in this case only one voyage a decision on that point is unnecessary since on any view of the matter the vessel must have begun the voyage at the latest when she started to load cargo at Manzanillo. If I am wrong on that point I would hold that the voyages to Santo Tom s and Big Creek began at the latest when REL accepted orders on Mar. 15 to load at those ports. Accordingly, it follows that in the present case the vessel was already engaged in performing the voyage to each of those ports under the COA when she suffered the casualty.

My conclusion that there was in this case only one voyage which had begun before the vessel left Manzanillo makes it unnecessary to decide whether by virtue of cl. 32 of the COA the Hague-Visby Rules apply in relation to part or all of the vessel's approach voyage to the first or only loading port. However, if I am wrong on that point the question does arise in relation to Santo Tom s and Big Creek, and for that reason, and since it was in any event fully argued, I think it right to express my conclusions on it. Mr. Flaux, Q.C. submitted that this case was essentially very similar to The Saxon Star in that what the parties had in mind was a series of round voyages, each comprising a ballast and a loaded leg, to be performed by a small number of vessels continuously employed in the service which is described in cll. 1 to 4A of the COA. That is certainly consistent with the terms of the COA generally, and express support for it can be found in cl. 8. It remains the case, however, as Mr. Young pointed out, that there is nothing in the COA which compels REL to perform it in that way. Its only obligation is to provide suitable vessels at appropriate intervals ready to load at ports designated by Fyffes. He submitted that this was a significant ground of distinction between this case and The Saxon Star

In The Saxon Star the question arose whether the incorporation of the Hague Rules into a consecutive voyage charter was apt to render them applicable to ballast voyages. A majority of the House of Lords considered that it was natural in a case where there might be many ballast voyages that the parties should have intended the Hague Rules to apply over the whole of the vessel's employment (see, in particular, per Viscount Simonds at p. 82, col. 2; p. 156 and per Lord Somervell of Harrow at pp. 99-100; pp. 185-186). That conclusion is perhaps now seen as unremarkable inasmuch as the charter was for consecutive voyages by a specific vessel so that the ballast voyages would be performed under the charter just as much as the loaded voyages. Similar kinds of questions have since arisen under contracts of rather different kinds: see in particular The Satya Kailash, [1984] 1 Lloyd's Rep. 588 . In my opinion, having regard to the nature of the service which the parties had in mind and the manner in which they expected it to be performed, very similar considerations would exist in the present case to the extent that any part of the ballast passages, for example from the U.K./Continent to Central America, formed part of the contractual service.

Mr. Young submitted that the terms of the COA itself when read together with those incorporated by reference to Fyffes' bill of lading demonstrate the parties' intention to limit the scope of the application of the Hague-Visby Rules to cargo-carrying voyages only. I fully accept that when a contract incorporates by reference terms drawn from other sources it is necessary to construe all the terms of the contract, including the incorporated terms, together to ascertain the parties' intention and in doing so to give meaning and effect to them all, as far as that is possible. It is also necessary, however, not to lose sight of the overall thrust of the contract as a whole. In this case I do not think that the terms of Fyffes' bill of lading taken together with cl. 12 of the COA itself will bear the weight which Mr. 

190 Young seeks to put on them. In addition to the provisions cited above, cl. 1 of the bill of lading contains a paramount provision and also includes the following sentence:

The provisions stated in the Act and the rules shall govern before the goods are loaded on and after they are discharged from the ship and throughout the entire time the goods are in the custody of the carrier . . . and although the passage relied on by Mr. Young does relate to damage occurring while the goods are not in the custody of the carrier, it concerns loss, damage or delay to the goods. Its incorporation into the COA along with the other terms of the bill of lading does not in my view support the conclusion that REL was to have no responsibility at all for the seaworthiness of the vessel except while cargo was actually on board. 

Having regard to the nature of the COA, I cannot accept that by incorporating the Carriage of Goods by Sea Act the parties intended that the Hague-Visby Rules should only apply to cargo-carrying voyages but not to ballast voyages insofar as they formed part of the contractual service. Whether all or part of such a voyage does form part of the contractual service probably depends on when the vessel becomes committed to the particular cargo-carrying voyage. For reasons I have already given I would hold that the vessel became committed to the voyage to a particular port from the time that REL accepted orders to load at that port. In the present case the vessel had loaded at the first port and was proceeding under orders to the second and third loading ports. She was therefore performing the service contemplated by the contract, and that being so, I would hold for the reasons given above that the Hague-Visby Rules applied to regulate REL's obligations during that approach voyage. The casualty having been caused by the unseaworthiness of the vessel resulting from a failure on the part of those for whom REL is responsible, REL is liable to Fyffes for such loss as they have suffered by reason of the vessels not loading at Santo Tom s or Big Creek. (f) The obligation to proceed with reasonable despatch

In the alternative Mr. Flaux submitted that even if REL was under no obligation under the Hague-Visby Rules to exercise due diligence to make the vessel seaworthy at the material time, the failure of the vessel to go to Santo Tom s or Big Creek constituted a breach of an implied term of the COA that the vessel would proceed forthwith or with all due despatch to the nominated loading ports and having loaded a cargo would perform the voyage with all due despatch. In support of the existence of such an implied obligation he relied on the language of cl. 2A and 2B of the COA and the line of authority which runs from Monroe Brothers v. Ryan, (1935) 51 Ll.L.Rep. 179; [1935] 2 K.B. 28 by way of Evera S.A. Commercial v. North Shipping Co., [1956] 2 Lloyd's Rep. 357 to The Baleares, [1993] 1 Lloyd's Rep. 215.

Mr. Young put forward a number of arguments in support of his submission that no such obligation arose in the present case. These were mostly directed to showing that there were insufficient grounds for implying the terms on which Fyffes sought to rely. Basing himself on Chitty on Contracts, 27th ed., pars. 13-004 - 13-008 and the cases of Trollope & Colls Ltd. v. North West Metropolitan Hospital Board, [1973] 1 W.L.R. 601 , Liverpool City Council v. Irwin, [1977] A.C. 239 and Shirlaw v. Southern Foundries Ltd., [1939] 2 K.B. 206 he submitted that a term cannot be implied unless it is necessary, reasonable, not inconsistent with the express terms of the contract and capable of being formulated with reasonable precision, and that the terms on which Fyffes relied satisfied none of those tests.

Central to his argument in all its aspects was the existence of the "Goodwill" clause in the COA, cl. 31. In Mr. Young's submission that clause was adequate to cover any situation for which the parties had not made specific provision and thereby rendered any implied term wholly unnecessary. He pointed out, correctly, that this COA is unusual in failing to provide for the regular nomination of vessels (see cl. 5) or the shipment of minimum quantities of cargo on any vessel (see cll. 2 and 6). These were not subject to any implied terms but were, he said, examples of matters which the parties had agreed should be dealt with by co-operation under cl. 31. Similarly, there was no need to imply a term that the vessel would meet any particular schedule or proceed with despatch to or between loading ports because that too was all encompassed by the obligation to use best endeavours. Once cargo had been loaded bills of lading would be issued and obligations relating to the custody and carriage of the goods would arise between the shipowners and owners of cargo. Since the COA was not a contract of carriage between REL and Fyffes but only a contract for carriage, there was no need to imply any obligation on REL that the vessel would perform the loaded voyage with reasonable despatch, nor, for the same reason, would it be reasonable to do so. Indeed, Mr. Young submitted that an implied term requiring the vessel to proceed with reasonable despatch would be essentially inconsistent with the express terms of the COA which do not impose any precise obligations in relation to the scheduling or arrival of vessels. The only thing that makes the timing of arrivals critical, he submitted, is Fyffes' insistence for their own 191 purposes on cutting fruit in advance, but that was not something which the COA contemplated or provided for. 

As Mr. Young acknowledged, all these arguments ultimately depend on his establishing that the COA was not a contract of carriage and that REL was not a carrier. I say that because, as he himself recognized, if the COA is a contract of carriage, REL as carrier was directly responsible for the vessel's performance. In fact, for reasons I have already given I have no doubt that the COA in this case is a contract of carriage by sea, and that very fact of itself renders Mr. Young's whole approach to the question of an implied term wrong. Although the authorities relied upon by Mr. Flaux are primarily concerned with the time at which the vessel must begin her approach voyage, they depend upon the existence of an obligation to perform it with reasonable despatch. In many cases that is expressed in terms such as "shall proceed with all convenient speed", but if the charter contains no express term to that effect one will ordinarily be implied: see Louis Dreyfus & Co. v. Lauro, (1938) 60 Ll.L.Rep. 94 ; Scrutton at p. 121; and Voyage Charters (Cooke et al.) pp. 65-66. The terms of cl. 2A and 2B reflect language commonly to be found in voyage charters. The main difficulty in the way of Mr. Young's submissions is that the implied obligation to proceed with reasonable despatch arises from the nature of the contract and is necessary in order to give it commercial efficacy. Its existence is by now so well established that it can be regarded as an ordinary incident of any contract of carriage by sea which exists unless the parties have expressly or by implication provided otherwise. This means that rather than it being necessary for Fyffes to show that such a term is to be implied in accordance with the criteria established in the authorities to which Mr. Young refers, it is for REL to show that in this case the parties agreed that such an obligation should be excluded. However, in view of the nature of this contract if would be surprising if the parties had decided to dispense with that obligation and I can find nothing in the COA to suggest that they intended to do so. Clause 31 on which Mr. Young laid so much emphasis seems to me to have been intended to fill the gaps only in the sense that it expresses the parties intention to co-operate in overcoming difficulties which they could not specifically foresee at the time they entered into the COA. This was a new venture and neither party was sure how it would work in practice; in that context such an undertaking makes good sense. Its inclusion does not begin to suggest, however, that in relation to such well understood aspects of the adventure as sailing between ports of loading that the ordinary incidents of carriage by sea were not to apply. For these reasons I have no difficulty in holding that it was an implied term of the COA that the vessel would proceed between ports of loading with reasonable despatch which, of course, she failed to do. Accordingly, if I am wrong in holding that REL was in breach of obligations imposed on it the Hague-Visby Rules , Fyffes are in my judgment entitled to succeed on this alternative basis of claim. 6.Caribbean Gold's claim against the Owners

This part of the case is more straightforward. It is common ground that the bills of lading under which the bananas shipped at Manzanillo were being carried were governed by the provisions of the Carriage of Goods by Sea Act, 1971 . Accordingly, Mr. Glennie accepted that only two questions arose: was there a loss caused by unseaworthiness? If so, was that unseaworthiness caused by the owners' failure to exercise due diligence to make the vessel seaworthy prior to loading at Manzanillo? For reasons I have already given both these questions must be answered in the affirmative and accordingly the owners are liable to Caribbean Gold for such loss as was caused by their failure to carry the bananas to their destination. 7.Damages - issues arising between Fyffes and REL 

In this case a large number of issues were raised in relation to damages, most of them by REL in connection with the claim made against it by Fyffes' in which, of course the owners were indirectly interested by reason of the contribution proceedings and on which they adopted REL's submissions. These included issues concerning remoteness of damage, causation, mitigation and straightforward questions of quantum.(a) Remoteness

Mr. Young's primary submission was that the loss claimed by Fyffes, namely, a loss of the value of the goods at destination was too remote. He submitted that the correct measure of damage in respect of REL's failure to load at Santo Tom s and Big Creek was simply the additional cost of obtaining alternative tonnage, in support of which he relied on Scrutton at pp. 400-402 and Voyage Charters pp. 455-456. Many other vessels, he said, were available during the week or so following the casualty which could have carried bananas to the U.K.; the real problem was that the cargo due to be loaded at Santo Tom s could not wait that long for shipment and that difficulties transporting container barges from Big Creek to Puerto Cortés prevented shipment of bananas from Belize. None of those problems would ordinarily have been within REL's contemplation at the time the charter was made and therefore any loss based on the value in Europe of 

192 the bananas that would otherwise have been carried from those ports could only be recoverable under the second limb of the rule in Hadley v. Baxendale, (1854) 9 Ex. 341 . Since Fyffes could not establish the facts necessary to succeed under that limb their claim must fail because the loss was too remote.

In my judgment this argument fails both as a matter of fact and as a matter of law. Bananas have been carried from the Caribbean and Central America to Europe for many years and I accept the evidence of Mr. Templeman that it is the practice to cut fruit in advance of the vessel's arrival to ensure that loading can begin without delay and be completed as quickly as possible.

I should be very slow to accept that REL, who are experienced operators of reefer vessels which are used for the carriage of a variety of fresh fruit, including bananas, were not well aware of the general nature of the trade and how it is conducted. Not surprisingly, there was no evidence from REL that it was unaware of these matters, and such evidence as there is suggests that it was and that the importance which Fyffes attached to regular and accurate information about vessels' movements and the reasons for it were explained to REL during the negotiations which preceded the contract. In my view it must have been within REL's contemplation that if, having given the notices called for by cl. 29, a vessel broke down shortly before she was due to arrive at a loading port, there might well be a loss of cargo because of its perishable nature.

The position in relation to the cargo to be shipped at Big Creek is slightly different inasmuch as the vessel was due to load there after she had completed at Santo Tom s. In those circumstances a total loss of cargo may well have appeared less likely given that fruit might not have been cut and that more time was available for arranging alternative transport. Nonetheless, I do not think that it was beyond the contemplation of REL that Fyffes might be unable to obtain shipping space and so suffer a loss of the cargo. That being the case, I do not think that the precise manner in which that came about is of any significance as far as the question of remoteness is concerned (see Nissho Co. Ltd. v. N.G. Livanos, (1941) 69 Ll.L.Rep. 125 at p. 130 to which I shall refer in more detail shortly), although it may be relevant to questions of causation and mitigation. In fact, by 1991 it had become difficult if not impossible for practical purposes to ship bananas from Belize otherwise than directly on to ocean-going vessels through the port at Big Creek. In earlier times fruit had been loaded into refrigerated containers which were towed on barges to Puerto Cortés for transfer into liner vessels. That required a stock of containers and a crane to handle them as well as the necessary tug and barges, but after the port at Big Creek was constructed these facilities fell into disuse. By March, 1991 the crane was no longer in working order and the tug was apparently at risk of being arrested in Puerto Cortés. Shipment in that manner from Big Creek was no longer a practical proposition, at any rate without plenty of warning and the provision of new equipment.

In support of this part of his argument Mr. Young sought to draw support from cl. 1 of the bill of lading which in his submission specifically excluded loss of the market value of the cargo at destination. The relevant part of cl. 1 provides as follows:

The carrier does not undertake that the goods shall arrive at the Port of Discharge at any particular time or to meet any particular market or use and the carrier shall in no circumstances whatsoever and howsoever arising be liable for direct, indirect or consequential loss or damage caused by delay or, save as is expressly provided herein, otherwise.

The short answer to this point is that this part of cl. 1 is not directed to the present situation. It is concerned with the situation in which the goods are carried to their destination under the contract but, by reason of some event for which the carrier is responsible, arrive there later than would otherwise have been the case. I do not think that it can properly be read as applying to the case where the carrier fails to carry the goods at all. Fyffes' claim in the present case is not for damages caused by the late arrival of the goods, but for damages for failure to carry them at all and does not fall within this provision.

Mr. Flaux submitted that where a shipowner has failed to load goods, as REL did in this case, damages may be assessed on alternative bases: the additional cost of transport by other means, that is, the difference between the market rate and the charter rate of freight; or the cost of replacing the goods at the port of discharge at the time when they ought to have arrived (normally equal to the sound arrived value of the goods), less any value which the goods have at the port of loading, freight and other savings. In support of that proposition he referred me to Scrutton at p. 402, and McGregor on Damages, 15th ed., pars. 1142-1147, and also to the cases of Nissho Co. Ltd. v. N.G. Livanos, (1941) 69 Ll.L.Rep. 125 and A/s D/s Heimdal v. Questier & Co. Ltd., (1949) 82 Ll.L.Rep. 452 .

I start from the principle that Fyffes are entitled to recover the loss which flows naturally in the ordinary course of events from the breach of contract. In the case of a contract for the carriage of goods by sea the natural and obvious consequence of the shipowner's failure to load and carry the cargo is that the owner of the goods is deprived of 

193 the benefit of having them at the agreed destination when they ought to have arrived. Prima facie, therefore, the loss he suffers is represented by the market value of the goods at that time and place. However, he may be able to avoid or reduce that loss in one of two ways. If alternative shipping space can be obtained he may be able to mitigate his loss by having the goods carried by another vessel, in which case his loss will be confined to any additional cost of carriage and other expenses as well as any loss caused by the delay to the goods reaching their destination. Alternatively, he may be able to obtain substitute goods at the port of destination, in which case his loss will be measured by the cost of so doing, less the value of the goods left at the port of loading and expenses saved in connection with their transport. That does not mean, however, that in either case the sound arrived value of the goods does not remain the starting point for assessing damages, subject to the duty to mitigate.

The authorities to which I was referred bear out that approach. In McGregor on Damages at pars. 1142 and 1143 the learned editor puts the matter in this way:

Where the defendant has failed to carry the goods at all. . .two alternative measures appear to be open to the plaintiff to put him in the position he would have been in had the carriage contract been performed. First, he may engage substituted transport to get the goods to the contractual place for delivery and claim as damages the cost of doing so less the price he would have paid under the contract with the defendant, i.e. the market rate of freight less contract rate of freight, and in addition the amount by which the price at the place of delivery has fallen between the contractual time for delivery and the arrival of the goods by the substitute transport when this is necessarily later. Or, secondly, he may buy similar goods at the contractual place for delivery and claim the cost of so replacing less the sum of the value of the goods at the place of loading, the amount of freight and theamount of insurance.

If the circumstances allow of either course of action on the plaintiff's part. . .it would seem that on general principles of mitigation he will be bound to adopt that which is less costly, subject always to the consideration that all that is required of him is to act reasonably.

In Nissho v. Livanos, (1941) 69 Ll.L.Rep. 125 the claimants chartered Eugenie Livanos or Evi Livanos for the carriage of a cargo of salt from Port Said (Port Sudan was later substituted) to Japan. The respondents nominated Eugenie Livanos to perform the charter and the claimants in their turn sold the cargo on a c. & f. basis to buyers in Japan and nominated Eugenie Livanos as the carrying vessel. Eugenie Livanos was unable to carry the cargo and in due course the respondents offered to perform the contract with Evi Livanos instead. However, a permit naming the vessel was required to obtain the credit necessary to enable the importation to be carried out. The claimants had obtained a permit naming Eugenie Livanos and the Japanese government refused to alter it to allow Evi Livanos to be substituted. The contract could therefore not be performed and the claimants claimed damages for its repudiation. A question arose whether the damages should be measured by reference to the difference between the charter and the market rates of freight, the claimants' loss of profits, or the cost of replacing the cargo in Japan less the value of the cargo at the port of loading, freight and insurance.

In relation to this question Lord Caldecote, C.J. began by pointing out that there was in fact no alternative vessel available to perform the charter and so no market rate was available which would enable a calculation of the difference between market rate and charter rate to be made. He continued as follows (p. 129): 

As I have said, [the difference between the market rate and the charter rate of freight] is not a proper measure in this case. In those circumstances, it appears to me that the ordinary rule applies, which takes the cost of the replacement of the cargo at the port of discharge after deducting the value, plus the freight and insurance, at Port Sudan. That is in accordance with what is laid down in Scrutton on Charterparties, 14th ed., at p. 453, where, having first dealt with the case where other ships could be obtained for the purpose of fulfilling the charter, the learned author deals with the position which arises when a substitute ship cannot be procured. The passage is as follows:

When, however, a substituted ship cannot be procured, the measure of damages payable to the charterer is the cost of replacing the goods at their port of destination at the time when they ought to have arrived, less the value of the goods at the port of shipment and the amount of freight and insurance upon them. This passage appears in the same terms at p. 402 of the current (19th) edition of Scrutton.

It is interesting to note that the respondents in that case raised an argument similar to that advanced by Mr. Young here, namely, that the loss measured by reference to the value of the goods at the port of discharge was too remote because (in that case) the respondents did not have in contemplation the refusal of the Japanese government to allow the alteration of the permit which made the 

194 difference in freight measure inapplicable. That argument was rejected on the grounds that the parties must have had it in contemplation that the measure of damages would be the increased cost of freight of another ship was available, or, if none were available, that the damages would be measured by reference to the value of the goods at the port of discharge. It did not matter how the circumstances came about which rendered one or other of those measures appropriate in the instant case. In my view that is fatal to Mr. Young's argument that a loss measured by reference to the value at the port of discharge is too remote, however, it may have come about.

In A/s D/s Heimdal v. Questier Co. Ltd., (1949) 82 Ll.L.Rep. 452 a vessel had been chartered to carry a cargo of logs from Port Gentil to Cape Town. The charterers made a claim against the owners alleging a failure to load a full cargo and succeeded in part. No alternative shipping space was available for the cargo that was shut out and in those circumstances Mr. Justice Morris held that the charterers were entitled to recover damages based on the replacement value of the cargo in Cape Town. It is worth noting that in assessing that value the Court did not regard itself as bound by the price at which the charterers had agreed to sell the logs to their buyers. In his judgment Mr. Justice Morris said at p. 472:

. . . . . if the parties at the time of the charter had given consideration to what would happen if the owners wrongly refused to carry 51 tons of okoume logs from Port Gentil to Cape Town, what the parties would mutually have had in mind would be that in the first place the charterers would seek to put those 51 tons on some other vessel, . . but . . . that, if a vessel could not be found, then loss would result at Cape Town. . .

If the 51 tons had been carried the defendants would have had 51 tons in Cape Town which would then have been of a certain value, namely, the cost, or replacement value or the value in Cape Town but disregarding the price paid under a particular contract of sale.

In my judgment these decisions support Mr. Flaux's submission that damages for failure to carry a cargo may be measured either by the difference between market and charter rates of freight or by reference to the value of the goods at the port of discharge (with appropriate deductions). Both measures of damage must be taken to have been within the contemplation of the parties because they are different ways of measuring the loss which the charterer suffers by not having his goods carried to the contractual destination. Which will be the appropriate measure in any given case will no doubt depend on the circumstances of the case, given that the plaintiff is under a duty to act reasonably to mitigate his loss. I do not consider that the passages in Scrutton and Voyage Charters to which I was referred are inconsistent with this view. They simply reflect the fact that in many cases alternative shipping space will be available, and, if it is, it will probably be incumbent on the charterer to mitigate his loss by making use of it. I do not accept, therefore, that the charterer must be able to show that alternative space was not available before he can recover damages measured by reference to the value of the goods at the port of discharge, as Mr. Young submitted. That does not, however, prevent the shipowner from seeking to establish that the charterer failed to act properly in mitigation, and I shall come to consider this aspect of the matter a little later.

In principle, therefore, I consider that Fyffes are entitled to recover damages assessed by reference to the replacement value of the goods at the time when they would otherwise have arrived at their destination, less any residual value of  the goods at the port of loading and savings on freight, insurance and other voyage expenses. (b) Causation

Mr. Young raised two issues under the general heading of causation: first, he submitted that there was no satisfactory evidence that Fyffes had actually suffered a loss at all; secondly, he submitted that there was no satisfactory evidence that any loss which Fyffes did suffer was caused by a breach of contract on the part of REL.(i) No evidence of loss 

As to the first of these, different arguments are advanced in relation to the cargo to be loaded at Santo Tom s and Big Creek.

As far as the Santo Tom s cargo is concerned, two rather different points were made. The first was that in the absence of a statement of some kind from Del Monte or original supporting documents there was no reliable evidence that any bananas had been cut and sold to Fyffes to provide a cargo for loading on Kriti Rex. I cannot accept that. There was evidence from Mr. Templeman that the bananas which Fyffes intended to ship from Santo Tom s were to be bought from Del Monte to meet the European specification (which is not the same as the United States specification) and packed by them in Fyffes' boxes for the European market. Mr. Young submitted that there was no evidence that any bananas had in fact been cut, packed or sold by Del Monte to Fyffes. I disagree. Between Mar. 17 and 19 Del Monte and their agents at Santo Tom s were in frequent communication with Mr. Templeman and Mr. Stickney to discuss what should be done with bananas which had been cut and packed 195 for loading by Fyffes on Kriti Rex. These exchanges culminated in an agreement made on the telephone between Mr. Templeman and Mr. Harrah of Del Monte (confirmed in a fax sent by Mr. Harrah) under which Del Monte, who were best placed to know the true position, agreed to sell 30,000 boxes of bananas for Fyffes in the United States. Later, on Mar. 25, Del Monte sent Fyffes a fax in which they summarized the events leading up to the shipment of 29,424 boxes of bananas on Harvester for Galveston for the account of Fyffes. It would be consistent with normal practice for the bananas, due to be shipped at Santo Tom s on a vessel arriving on Mar. 17, to be cut on Mar. 16 and 17. I see no reason to doubt that the position was as the parties to the agreement understood it to be, and as it was described in Del Monte's fax, namely, that 30,036 boxes of bananas had been cut and packed by Del Monte for Fyffes for loading on Kriti Rex for which Fyffes were liable to pay, and that nearly all of those bananas were shipped by Del Monte to Galveston for Fyffes' account in accordance with the agreement between them.

The second point was rather different. In their fax of Mar. 25, 1991 Del Monte set out the expenses which they said they had incurred in connection with the sale of the bananas and which they intended to deduct from the proceeds of sale. Fyffes accepted them without demur, not even asking to see any supporting documents. Mr. Young submitted that REL should not be required to pay damages on the basis of figures that were not properly evidenced and which Fyffes themselves had not even bothered to check.

If Fyffes were seeking to recover these expenses as part of their claim there might be some force in this point since, although I have no reason to think that the figures set out in Del Monte's fax of Mar. 25, cannot be sustained, they are not supported by any vouchers or underlying documents as one might expect. In fact, however, they do not form part of the claim in any sense since the net proceeds of sale for which Fyffes have given credit have been calculated without taking into account the expenses of which Mr. Young complains. No application has been made to amend the points of claim in that respect so those expenses will continue to be borne by Fyffes in any event.

The arguments in relation to the shipment at Big Creek are different again. Mr. Young submitted that the evidence as to the number of bananas which were cut was unsatisfactory, but his main argument was directed to the terms on which Fyffes bought bananas in Belize and the savings which they were in a position to make as a result of the cargo not being shipped. 

The first of these points can be dealt with quite shortly. Evidence about the arrangements for cutting bananas for loading on Kriti Rex was provided in the form of a statement from Mr. Francisco Cruz of the Banana Control Board in Belize and also by Mr. Stickney who at the time was Fyffes' representative in Central America. Mr. Cruz's evidence was that 4009 boxes had been packed by the time the cut was suspended on Mar. 18 and that the remainder of cargo which would have been cut (estimated at the equivalent of 24,652 large (40lb) boxes) was left to rot. That is consistent with other evidence about the way in which bananas are grown and harvested in Belize and I see no reason not to accept it. 

Fyffes bought bananas in Belize under an agreement with the Banana Control Board ("BCB") dated June 2, 1986. Under that agreement BCB agreed to sell Fyffes all the bananas of a certain type and quality which BCB could deliver. The agreement provided for BCB to give Fyffes weekly estimates of the amount of fruit that would become available over the following eight weeks, and by the Friday of the week prior to loading of each cargo Fyffes was required to confirm its acceptance to BCB thereby undertaking liability to accept delivery. These confirmed estimates formed the basis of the parties' obligations to deliver and accept fruit under the agreement. Clause 12.1 of the agreement provided as follows:

If except for reasons attributable to force majeure according to section 15 hereof [Fyffes] should ever fail to purchase the full amount of fruit according to the confirmed acceptance . . . . . [Fyffes] will pay the BCB compensation per box for such fruit which meets the specifications as per Annexe D hereof. . . . . . In the case of unharvested and non-packed bunches the average stem box ratio for the preceding two loadings will be the basis for a per box payment. and cl. 15.3 provided as follow:

The term "force majeure" as used herein shall include: Acts of God, strikes, lockouts, acts of public enemy, wars, mobilization, blockades, piracy, insurrections, riots, epidemics, landslides, earthquakes, fires, storms, floods, washouts, restraint of rulers, civil disturbances, explosions, shipwreck, unreachability on account of ship's safety and other events beyond the control of the parties which affect the production, sale, shipping, exportation. . .of bananas either to, at, or from the place of exportation . . . The sum payable for "left back fruit" under Annex D was BZ$6.26 per 40 lb box.

Fyffes' claim in respect of the failure to load bananas at Big Creek is calculated on the basis that they had become liable to pay for the cargo which 196 in the event had no residual value. Mr. Young submitted, however, that Fyffes were entitled to rely on the protection of cl. 12.1 because the breakdown of Kriti Rex fell within the expression "shipwreck", or at any rate within the expression "other events beyond the control of the parties", in cl. 15.3, and were not bound to pay anything for the cargo which was not loaded. Accordingly, he submitted, credit should be given against the market value on arrival in the U.K. for the full purchase price which would otherwise have been incurred. 

Mr. Flaux drew my attention to the commentary on force majeure clauses to be found in Chitty on Contracts, 27th ed., pars. 14-130 - 14-131 and in Benjamin's Sale of Goods, 4th ed., pars. 8-090 - 8-091 and to the well-known case of Lebeaupin v. Crispin & Co., [1920] 2 K.B. 714 in which Mr. Justice McCardie considered the principles applicable to clauses of this kind. Clause 12.1 in the present case is not a simple force majeure clause of the kind which was before the Court in that case because its scope is defined by cl. 15.3. That begins by identifying certain specific circumstances which constitute force majeure before coming to the general words "other events beyond the control of the parties". In general I think it fair to approach such clauses with the presumption that the expression force majeure is likely to be restricted to supervening events which arise without the fault of either party and for which neither of them has undertaken responsibility. However, as Mr. Justice McCardie pointed out, the question is one of construction and in each case close attention should be given to the language of the clause itself while paying due regard to the nature and terms of the contract as a whole.

Applying that principle I cannot accept that "shipwreck" in cl. 15.3 is apt to cover the breakdown suffered by Kriti Rex in this case, if only because there was in fact no shipwreck within the ordinary meaning of the term. Nor in my judgment can the facts be brought within the general words "other events beyond the control of the parties". The specific examples of force majeure events indicate that what the parties had in mind were supervening events which did not arise out of aspects of the performance of the contract for which they were directly or indirectly responsible. The bananas were sold on f.o.b. terms and as between themselves and BCB it was Fyffes' responsibility to obtain a ship and tender her for loading at the appropriate time and place. As in Lebeaupin v. Crispin , I do not think that the parties intended cl. 12.1 to apply to the case where the vessel's failure to arrive was due to some default on the part of those who were responsible for her maintenance or navigation any more than if it had been due to a failure on the part of Fyffes to make the arrange- ments necessary to ensure that a ship arrived at the right time and place. (ii) Loss not caused by REL's breach of contract

Next Mr. Young submitted that the loss of the cargoes due to be loaded at both Santo Tom s and Big Creek were caused by Fyffes' independent decision to cut bananas in accordance with their own assumptions about the vessel's movements rather than any breach of contract on the part of REL.

The vessel left Manzanillo at 07 30 on Mar. 15 and shortly afterwards the master gave her e.t.a. at Santo Tom s as 10 30 on Mar. 17. In accordance with what I find to be normal practice cutting of bananas began on Mar. 16 in anticipation of the vessel's arrival. About 14,530 boxes were cut and packed that day. On Mar. 16 the master gave a revised e.t.a. of 12 30 on Mar. 17, but the change was not significant enough to warrant making any change in the arrangements for cutting. Cutting continued at Santo Tom s on Mar. 17 which was a Sunday. Early in the morning of that day the master sent a message to Fyffes informing them that the vessel had stopped at 19 00 the previous day to replace No. 7 piston and giving a revised e.t.a. Santo Tom s of 10 00 on Mar. 18. A similar message was sent to the agents at Santo Tom s. The master's message did not come to the attention of Fyffes' personnel in the U.K. until the Monday morning, but the agents in Santo Tom s did relay the information they had received to Mr. Stickney in Belize by telex sent at 18 40 on Mar. 17. In the event he did not see that until the next day. A further 15,506 boxes were cut and packed at Santo Tom s on Mar. 17. On the morning of Mar. 18 the master sent a further message t o Fyffes putting his e.t.a. Santo Tom s back to daybreak on Mar. 19.

It is apparent from this brief summary of events that much of the fruit intended for the vessel at Santo Tom s had been cut before the master gave notice of the change in the e.t.a. from Mar. 17 to 18, and that all the cutting had been completed by the time the e.t.a. was put back from Mar. 18 to 19. Mr. Young accepted that nothing could have been done under these circumstances to stop the cut which was in progress on the 16th, but he submitted that cutting could have been stopped early on 17th if the information had been passed to Mr. Stickney quickly and acted on promptly. If that had been done, he submitted, the best part of the second day's cut could have been avoided and the fruit would not have been lost. The fruit cut on Mar. 17 was, he said, lost because Mr. Stickney could not be contacted early that day.

Between 1980 and 1987 Mr. Stickney had been Fyffes' representative in Surinam and between 1987 and 1991 he had been their representative in 197 Belize. At the time when these events occurred he was about to take up a new post as country manager for Honduras and was spending a certain amount of time travelling between Belize and Honduras. Fruit shipped by Fyffes from Santo Tom s was obtained from Del Monte; cutting was carried out by Del Monte in accordance with a schedule of ships' arrivals provided by Fyffes from London and was not under his control, though no doubt Del Monte would consult him and take account of any suggestions he made. Mr. Stickney himself had considerable experience of cutting and shipping bananas in Central America. In his view the change in the vessel's e.t.a. from Mar. 17 to 18 did not necessarily make it wise to stop cutting fruit on Mar. 17 because there were advantages in having all the cargo ready for loading as soon as the vessel arrived during the morning of Mar.18, as it was then expected she would. Even if the message had got through to Mr. Stickney on Mar. 17, therefore, I think it unlikely in view of his evidence that he would have asked Del Monte to stop cutting at that stage. The advantages an disadvantages of continuing to cut on Mar. 17 are matters that Del Monte themselves can also be taken to have had well in mind, and I do not consider that the evidence supports the suggestion that it was unreasonable to continue cutting fruit in anticipation of the vessel's arrival during the morning of 18th. In my view there is no substance in the argument that part of the loss sustained in relation to the failure to load at Santo Tom s was caused by Fyffes' own conduct.

The position in relation to Big Creek is rather different. On receipt of the master's message putting back his e.t.a. Santo Tom s to daybreak on Mar. 19 Mr. Templeman instructed Mr. Stickney to stop the cut in Belize. By the time that instruction had been relayed to the growers 4009 boxes had already been cut and packed. Mr. Young submitted that in Belize the bananas were being cut too far ahead of the vessel's arrival in any event, but that even so, if Mr. Stickney had acted more promptly the cut could have been stopped before any fruit at all had been harvested.

According to Mr. Stickney about 36 hours would normally be required after arrival at Santo Tom s to complete loading and sail to Big Creek. When she left Manzanillo the vessel had expected to arrive at Santo Tom s during the morning of Mar. 17. On that basis she could be expected to arrive in Big Creek on the morning of Mar. 19 (allowing for daylight navigation into port), so in accordance with normal practice cutting would begin on 18th. The change of e.t.a. notified to Fyffes and the agents at Santo Tom s on Mar. 17 meant that the e.t.a. Big Creek would go back to Mar. 20. Since, as Mr. Stickney confirmed, fruit should not be cut more than 36 hours before loading, cutting in Belize ought to have been deferred to 19th. The reason it was not is probably that it proved impossible to pass the information about the vessel's revised e.t.a. to Mr. Stickney during the weekend. In any event, at about 7.30 local time on Mar. 18 he spoke to Mr. Templeman by telephone and was told to stop the cut, but there then appears to have been a delay of some two and a half hours before the order to stop cutting was actually given. Mr. Stickney was unable to explain how that had come about, but it seems likely that that delay was enough to allow the cutting and packing of the 4009 boxes of fruit to which Mr. Cruz referred in his evidence.

In the light of Mr. Stickney's evidence I think Mr. Young is right in saying that the cut in Belize could have been stopped before the 4009 boxes of bananas had been cut, but I do not think that carries him very far. The fact is that the remainder of the cargo due to be shipped at Big Creek had to be cut and left to rot because it was too far advanced to enable it to be harvested and shipped on a later vessel. Bananas grown for export from Central America to Europe have to withstand a lengthy period of carriage and in order for them to reach their destination in sound condition it is essential for them to be harvested at the optimum stage of growth. In order to ensure that that is done stems are individually marked with coloured ribbons. The rate at which the fruit ripens after harvesting depends to a considerable degree on the age at which it is cut; fruit which has passed the appropriate age cannot safely withstand the voyage to Europe and cannot be shipped. There is no alternative market for it so it can only be discarded. As was the case with the fruit which had not been cut, the fruit which had been cut and boxed had reached the age at which it had to be harvested if it was to be shipped at all. It could not have been left on the plant in the expectation that it could be harvested for loading on a later vessel. In the absence of a vessel on which it could be loaded it therefore became a total loss, whether it had been cut or not.

That being the case, Mr. Young submitted that the loss of the cargo due to be shipped from Big Creek was caused not by the breakdown of Kriti Rex but by Fyffes' failure to obtain alternative shipping space. This he advanced as a matter of causation, but in my judgment the issue is properly to be viewed as one of mitigation for the reasons which I have already endeavoured to state and I therefore propose to consider it under that heading. (c) Mitigation

Mr. Young on behalf of REL submitted that Fyffes had failed in a number of respects to take 198reasonable steps to mitigate their loss. The first and perhaps most important matter he relied on is their failure to make any alternative arrangements to ship the cargo due to be loaded in Big Creek. Mr. Templeman who was responsible for organizing Fyffes' shipping arrangements learned on the morning of Mar. 18 that the vessel had broken down and that her e.t.a. Santo Tom s had been put back to daybreak on the 19th. He knew, of course, that she was carrying the cargo loaded at Manzanillo. At that stage he attempted to obtain more reliable information about the vessel's circumstances from REL and the brokers, but he took the precaution of telling Mr. Stickney to stop the cut in Belize which was due to start that morning. The master's telex message informing them that further problems had been encountered with the engines was received in Fyffes' offices at Eastleigh in the early hours of Mar. 19 and later they also received the master's message that he was proceeding to Puerto Cortés for repairs. These messages were passed to Mr. Templeman in London as soon as the office opened.

In the early afternoon of Mar. 19 REL sent Fyffes a telex in the following terms:

AS YOU KNOW VESSEL IS PROCEEDING TO PTO CORTES FOR REPAIRS. LATEST WORD FROM OWNERS IS THAT REPAIRS WILL TAKE UNTIL 24TH AND OWNERS FEEL THAT THEY KNOW THE EXTENT OF THE DAMAGE AND THUS BARRING ANY UNEXPECTED DISASTERS THIS IS A GOOD ESTIMATE.

EARLIEST WE COULD HAVE A SUBSTITUTE WOULD ALSO BE 24TH AS OUR ICELAND REX ONLY SAILING PHILADELPHIA LATE TONITE AND WILL NEED 4 DAYS TO GET DOWN . .

Mr. Templeman's primary concern at that time was the cargo at Santo Tom s which had been cut on Mar. 16 and 17 and had inevitably experienced high ambient temperatures while awaiting shipment. Under these circumstances fruit suffers stress whichaffects the speed of ripening and its ability to withstand a lengthy sea voyage and it was therefore essential to make immediate arrangements for it to be shipped if it were not to be lost. He made various attempts to find alternative shipping space for Europe with a number of other banana companies, but without success. However, Del Monte had a ship, Harvester, sailing for Galveston on which the fruit could be loaded at once and he therefore made an agreement with Del Monte for them to take the cargo there and sell it on behalf of Fyffes. Loading began the same afternoon and was completed next day. The position in Belize was less critical because the cut had been stopped and most of the fruit remained on the plants where it would not deteriorate over the course of a few days. Mr. Templeman therefore had to consider whether it would be better to take Iceland Rex, knowing that she would not be available to load in Belize until Mar. 24 at the earliest, or to continue with Kriti Rex in the hope that she would have completed her repairs by that date. In the event he decided not to take Iceland Rex.

Although Mr. Templeman said that he did not consider that REL was formally putting Iceland Rex forward as a substitute under the contract, he did accept that it was apparently willing to make the vessel available to Fyffes to take over from Kriti Rex, as it seems to me it plainly was. With hindsight it can be seen that Mr. Templeman would have done better to pursue the opportunity to take Iceland Rex because, as it turned out, Kriti Rex did not complete her repairs until Apr. 4, but at the time the position was far from clear. Mr. Young submitted that by turning down the offer of Iceland Rex Fyffes failed to act reasonably in mitigation of their loss, but I am not satisfied that that is so. Mr. Templeman was faced with a very tricky problem: the latest information he had about Kriti Rex was that the owners felt that they knew what was wrong and were confident that the date they had given for the completion of repairs was reliable. Moreover, the tenor of REL's message was that Iceland Rex did not offer any great advantage since she was unlikely to arrive any sooner. In the circumstances I do not think that Mr. Templeman acted unreasonably in letting Iceland Rex go, and by Mar. 21 it had been fixed under another charter.

The enquiries which Mr. Templeman made of other banana shippers on Mar. 19 were only directed to loading the cargo which was waiting at Santo Tom s. In the event none of the companies had ships bound for Europe until the end of the week at the earliest. That would have been too late, even if space had been available. No steps, however, appear to have been taken either then or later to find alternative shipping space for the cargo due to be shipped from Belize. Mr. Templeman was unable to explain why that was so, but I think some light was shed on that by Mr. Stickney. It appears from his evidence that the cut in Belize which had been cancelled on Mar. 18 had been rescheduled to begin on the 24th in anticipation of the vessel's arrival on the 25th. At that stage it was not thought that there would be any difficulty in shipping the cargo from Big Creek on Kriti Rex and I do not think that in all the circumstances Fyffes acted unreasonably in acting on the footing that she would be able to maintain that schedule. However, on Mar. 22 Fyffes were informed that the repairs would take a week longer than had previously been expected and would not be completed until Mar. 31. 199 That meant that the ship would not arrive before the beginning of April, at least 12 days later than originally expected and only a few days before harvesting for the next shipment was scheduled to begin. By then the next growth would be ready for harvesting and the previous growth would be well past its best. It is clear that following that change to the vessel's schedule Mr. Templeman did not try to find alternative tonnage to load the cargo in Belize, but it is for REL to show that Fyffes acted unreasonably in failing to take such steps as were open to them to reduce their loss following the vessel's breakdown. In fact there was no reliable´evidence that by that stage alternative shipping space could have been obtained before the fruit had become too old to travel safely to Europe. Accordingly, I reject this part of Mr. Young's submissions.

Mr. Young also criticised Fyffes for failing to take more active steps to scrutinize the account which Del Monte submitted in respect of the fruit which was carried on Harvester to Galveston, but again, in the absence of some evidence which tended to show the expenses claimed were excessive (and in fact there was none), such criticisms do not take the matter any further. However, since, as I have already pointed out, the expenses to which his criticisms were directed do not form part of Fyffes' claim, the issue is of no practical significance.

For these various reasons I reject REL's submission that Fyffes failed to mitigate their loss. (d) Quantum 

Mr. Stickney described in detail the procedure for harvesting and packing bananas in Belize and the indications are that the position was broadly similar in Guatemala. As one would expect, there is a good deal of wastage between the plantation and the ship's hold which he estimated as on average something in excess of 20 per cent. overall for one reason or another. Up to half of that (5-10 per cent.) represents wastage on the farm, and of the stems that reach the packing shed about 12 per cent. on average are not boxed. Then there is a further small amount of wastage between the packing shed and the ship due to losses transit (normally less than 1 per cent.), and some further losses occur during the course of loading. Mr. Stickney did not put any figure on these, but I infer from what he said that they would normally be small, probably less than 1 per cent.

Fyffes' claim against REL is based on the assumption that 30,036 40 lb boxes of bananas would have been loaded at Santo Tom s and 24,652 boxes at Big Creek for carriage to Portsmouth. As far as the Santo Tom s cargo is concerned there is evidence that 30,036 boxes were packed and transported to the port in preparation for loading. The documents emanating from Del Monte relating to the shipment on Harvester indicate that 29,424 boxes were loaded and charged to Fyffes, the difference being accounted for by damage and shortage of shipping space. No separate figure was given for the number of damaged boxes, but this evidence taken in conjunction with Mr. Stickney's account indicates a loss at loading of around 2 per cent. which I do not find surprising given that the fruit in containers had been transferred into the port warehouse for better ventilation and had therefore been subjected to additional handling. If it had been loaded on Kriti Rex as originally intended on Mar. 17, I think it likely that the wastage in the course of loading would have been kept down to about 1 per cent. or 300 boxes. I think it likely, therefore, that 29,700 40 lb boxes or the equivalent would have been shipped on Kriti Rex from Santo Tom s.

Because the cut in Belize was stopped early on Mar. 18 the number of boxes that would have been available for shipment can only be assessed on the basis of the estimates provided by the farms. These estimates are made by applying the box/stem ratio based on the last shipment to the number of stems ready for harvesting. This box/stem ratio reflects the number of boxes which emerged from the packing shed and so takes account of wastage in harvesting and packing. The estimates provided by the individual farms for the cut scheduled to take place on Mar. 18-20 amounted to 24,652 boxes. Although the final quantity could obviously have varied one way or the other from that figure I see no reason not to accept it as essentially reliable. There would, however, have been some losses between the packing shed and the ship's hold, and I accept Mr. Young's submission that about 2 per cent. ought to be allowed for that. On that basis I find that 24,150 40 lb boxes or the equivalent would have been shipped on Kriti Rex from Big Creek.

When the bananas reach Portsmouth they are inspected cursorily as they are discharged from the vessel and sample boxes are taken for thorough inspection by quality control inspectors. At this stage the bananas should still be completely green; boxes in which any fruit shows signs of turning yellow ("ripe and turning") are rejected. There is always liable to be some wastage at outturn, though it is usually below 2 per cent. and would not cause problems even if it were as high as 2½ per cent. In this case it was agreed between the parties' experts that 1 per cent. of bananas from Guatemala and 4 per cent. of bananas from Belize would have been rejected. 

After discharge the bananas are despatched in lorries to the ripening sheds. Fyffes' records of boxes manifested, boxes rejected on grounds of quality and boxes despatched from the port to ripeners during the weeks preceding the casualty 200 show that there was often a considerable discrepancy between the number of boxes manifested and the total number rejected and despatched to the ripeners. On average the number manifested was about 10 per cent. greater than could be accounted for in those ways. Mr. Young submitted that some allowance ought to be made for that, but I am not persuaded that it represents a real loss. Much of the cargo is shipped in break bulk form and boxes are tallied on and off the vessel. The scope for discrepancies in tallying is obvious, and in the absence of any evidence that substantial numbers of boxes were regularly lost in the course of carriage (which seems inherently unlikely) I am not persuaded that these discrepancies reflected actual losses rather than errors in tallying. I do not accept, therefore, that any further allowance should be made for that. 

On this basis I find that 29,400 boxes (544 tons) of Guatemalan bananas and 23,180 boxes (429 tons) of Belize bananas would have been landed in good condition in Portsmouth during the week beginning Apr. 1 (week 14) if Kriti Rex had loaded as intended.

The failure of Kriti Rex to carry the Santo Tom s and Big Creek cargoes deprived Fyffes of the benefit of having those bananas available for sale in the U.K. That loss is to be measured by the market value of the bananas, that is, the replacement cost, rather than the prices which Fyffes could actually command, insofar as that might be different (see Heimdal v. Questier, (1949) 82 Ll.L.Rep. 452) . Fyffes seek to recover damages by reference to the "Green Boat Price" of £700 per ton in the case of bananas from Guatemala and £610 per ton in the case of bananas from Belize.

Most of the bananas imported into the U.K. are sold to U.K. distributors for ripening. Some, but not all, of the ripeners are owned by companies in the Fyffes group. The Green Boat Price is the price at which Fyffes sell to ripeners in the U.K. and includes the cost of road carriage from the port to the ripening centre. The price is the same for all distributors regardless of location and is set by Fyffes taking into account market factors including, in the case of "dollar" bananas, (that is, bananas coming from outside the Commonwealth) prices quoted in Northern European ports. Fyffes were not themselves able to import all the bananas which they required to meet the demands of the U.K. market and they therefore regularly bought dollar bananas from Continental importers to make up the difference. Such purchases were normally made on a Thursday for delivery the following week. Records kept by Fyffes show that prices at which fruit was available for purchase remained fairly stable at this time. Guatemalan bananas are dollar bananas for which the average price quoted on the Continent in week 13 was DM30.44 per box. At then current rates of exchange that was equivalent to £10.26 per box or £554.04 per ton. Accepting, as I do, Mr. Shields' evidence that the average cost of transport from the Continent to the U.K. was about £55 per ton, this suggests that the cost of buying in dollar bananas to replace those that should have been carried from Santo Tom s was about £609 a ton. Evidence of purchases for week 14 itself, however, suggests that the cost would have been somewhat higher. The average price paid by Fyffes for bananas landed at various Irish and Continental ports between Mar. 28 and Apr. 2 together with the costs of carriage was £652.86 per ton. In the light of the evidence as a whole I consider that a figure of £640 a ton is a fair assessment of the replacement cost of the Santo Tom s cargo. Fruit from Belize was not available for purchase on the Continent and it was not disputed that £610 per ton was a fair value to take for the purposes of assessing damages in relation to the cargo that should have been loaded in Big Creek.

It was common ground that credit should be given for the net proceeds of sale of the bananas shipped to Galveston and any saving in expenses which Fyffes made. These, it was accepted, included freight, port charges, import duties and discharge and distribution expenses. It is necessary for me to consider the question of freight separately and I shall come to that in a moment. As far as the other items are concerned, although Mr. Young challenged Fyffes' evidence in relation to them, he failed to persuade me that insufficient credit had been given for the various items involved. The details of the savings are set out in the schedule to this judgment. I accept, however, that there was in addition a small net saving on insurance which at 0.3 per cent. of turnover amounted to £873.23 in relation to the Santo Tom s cargo and £785.07 in relation to the Big Creek cargo. I do not accept that there was any additional saving in respect of the costs of the normal quality control operations which on the evidence were part of the general overheads.

Fyffes have given credit in their claim for a saving of freight at U.S.$2.77 per box in accordance with the COA. During the months leading up to the voyage in question the quantity of bananas shipped under the agreement had fallen below the parties' original expectations and there had been negotiations which REL hoped would lead to a revised rate of freight which would go some way towards compensating it for past losses. The negotiations continued over much of the first six months of 1991 and culminated in agreement in July under which Fyffes agreed to pay a lump sum of U.S.$876,000 by way of "deadfreight" and a further sum based on an adjustment of the freight to 

201 U.S.$2.86 per box in respect of quantities actually loaded. In the case of Kriti Rex voyage 20/15 the quantity to which the figure was applied was just the 15,424 boxes loaded at Manzanillo before the breakdown occurred. The rate of U.S.$2.86 was agreed as part of a new freight structure and was calculate as an average weighted rate which took account of a variety of factors to reach a final figure which was satisfactory to both parties. As far as REL was concerned the agreement as a whole represented a satisfactory means of restoring the profitable position which it had expected it would enjoy under the COA.

Mr. Young submitted that Fyffes should give credit for freight saved in the sum of U.S.$3.04 per box instead of U.S.$2.77 on the grounds that if the voyage had been completed it would have served to increase REL's losses to the point where it would have been necessary to increase the freight from U.S.$2.86 to U.S.$3.04 to achieve the same result. I cannot accept that. All the indications are that the agreement made in July represented an acceptable compromise between differing positions. Although it appears that REL would have incurred a loss on the voyage at the original freight rate if it had been completed, it does not follow that there would have been a corresponding increase in either the deadfreight or the adjusted rate of freight agreed in July 1991. Of course, it is possible that in those circumstances the negotiations might have taken a different turn, but whether the outcome would have been more favourable to REL, and if so to what extent, is in my judgment simply a matter for speculation. However, I can see no reason to think that the agreement would have been any less favourable to REL and I am satisfied that Fyffes would have paid freight at the revised rate of U.S.$2.86. 

Fyffes' pleaded claim to recover damages in United States dollars was not pursued. Accordingly I hold that Fyffes' claim against REL succeeds in the sum of £321,931.84. The details of the calculation of that sum are set out in the schedule to this judgment. 8.Damages - issues arising between Caribbean Gold and the owners

If the voyage had been performed the cargo loaded in Manzanillo would have been discharge partly in Portsmouth and partly in Zeebrugge. In the ordinary way the vessel would have left Big Creek on Mar. 20 and could be expected to have reached Portsmouth about 12 days later, that is on Apr. 1. Allowing for the time required to discharge the majority of the cargo at Portsmouth before shifting to Zeebrugge, the cargo loaded at Manzanillo would probably not have completed discharging until some time on Apr. 5. This assumes that all would have gone smoothly, but it is possible that the vessel might have suffered some delay in the course of the voyage and that discharging in Zeebrugge would not have been completed until a little later. In the light of Mr. Shields' evidence I accept that the value of sound Dominican Republic fruit in Portsmouth in week 14 was £5.72 per 28 lb box; in Zeebrugge it was DM24.50 per 40 lb box.

The cargo on board Kriti Rex was first inspected by Mr. Stickney on Mar. 28 at Puerto Cortés. He found that although it was generally in a reasonable condition, it had been cut at maximum permissible maturity and was showing signs of ripening. He thought that it would remain in an acceptable condition for marketing for a further week, but thereafter it would begin to deteriorate quickly and there was a serious danger that within three or four days after that it would ripen to the point of becoming unmarketable. It could not be shipped to Europe or the United States and had already become a total loss in commercial terms so that the only thing to do was to discharge it and dispose of it locally. There is no dispute that that was the right course to take and in the event it was donated to the local Red Cross.

Mr. Stickney inspected the fruit again on Apr. 5 after it had been discharged. It was still mainly green, but had reached the point of turning and some ripe bananas were to be found in scattered boxes. Some fungal infection was setting in on finger stubs. The proportion of ripe and turning fruit at that stage was about 2 per cent. so if the fruit had been landed in Europe at that time the number of ripe and turning boxes would not have been sufficient to cause a real problem. However, the indications were that within a further two or three days the process would accelerate; the proportion of ripe and turning fruit would rise rapidly to the point where within a few days it would reach 20-30 per cent. and the cargo would become a total loss. 

In the light of Mr. Stickney's evidence I think it likely that that part of the Manzanillo cargo which was destined for Zeebrugge (12,363 40 lb boxes) would have reached Europe with about 2 per cent. of the boxes showing ripe and turning fruit even if there had been no loss of time on the voyage. However, after leaving Manzanillo the vessel had to load at both Santo Tom s and Big Creek (where there are certain restrictions on navigation) and would have discharged first at Portsmouth before completing at Zeebrugge. In my judgment there is a real possibility, therefore, that the vessel would not have completed at Zeebrugge by Apr. 5 and that the deterioration of that part of the cargo would have increased quite significantly. Mr. Glennie submitted that the cargo might have deteriorated to the point of its becoming worthless and that a further allowance of 3-8 per cent. ought to be made for that contingency. I think it unlikely that the cargo would 

202 have deteriorated to that extent, but I accept that there is a real possibility that it would have deteriorated to a considerable degree and that some allowance should be made for that contingency. I consider that 5 per cent. is appropriate.

Finally, Fyffes have made some saving in respect of insurance, port charges and the costs of discharging and distribution at Portsmouth and Zeebrugge for which credit has to be given. Taking all these into account as shown in the attached schedule I hold that Fyffes are entitled to recover damages from the owners in the sum of £104,194.11. 9.The owners' counterclaim against Caribbean Gold

It was not disputed that the counterclaim must fail if, as I have held, the owners failed in their defence to the claim of Caribbean Gold.10.REL's claim against the owners 

In contribution proceedings issued pursuant to R.S.C., O. 16, r. 8 REL seeks to recover damages from the owners amounting to a complete indemnity against its liability to Fyffes under the COA. Kriti Rex was chartered to REL under a three year time charter on the Baltime, 1939 form with a number of additions and amendments dated Aug. 31, 1988. It contained the following clauses as typed additions to the printed form:

39. The . . . . . U.S. Clause Paramount . . . . . [is] deemed to form part of this Charter Party . . . . . 

64. Owners will at all times make their best endeavours to maintain the vessel in good condition and her machinery and equipment in good operating order.

It was accepted, therefore, that for all practical purposes the owners' obligations towards REL in relation to the seaworthiness of the vessel were no less onerous or extensive than those which REL owed to Fyffes under the COA and that if, as I have held, REL is liable to Fyffes the owners are liable in damages to REL for breach of the time charter. However, the owners did raise a number of matters of argument both in relation to remoteness and the quantum of damages.(a) Remoteness 

REL's claim against the owners in the present case is to recover damages sufficient to indemnify it against the liability it has incurred to its subcharterers. Mr. Glennie submitted that that was not the normal measure of damages in such cases and that REL could only succeed if it could show, pursuant to the second limb of the rule in Hadley v. Baxendale, that this loss was of a type which the owners must have had in contemplation at the time they entered into the charter in 1988. In fact, he submitted, the owners had no knowledge at that time of the particular trades in which REL intended to employ the vessel and certainly had no reason to contemplate a loss of the kind which REL sought to recover. 

The formulation of a test for remoteness of damage in contract which provides just the right degree of foresight has not proved easy. Mr. Glennie submitted that there has of late been a movement towards a more restrictive test of "a very substantial degree of probability". In support of that he referred me to the recent case of Balfour Beatty Construction (Scotland) Ltd. v. Scottish Power Plc., 1994 S.L.T. 807 in which the House of Lords approved and applied a passage in the speech of Lord Reid Czarnikow Ltd. v. Koufos, [1967] 2 Lloyd's Rep. 457 at p. 465, col. 2; [1969] 1 A.C. 350 at p. 388E . There, having referred to the speeches of Viscount Dunedin and Lord Shaw of Dunfermline in R. & H. Hall Ltd. v. W.H. Pim (Junior) & Co. Ltd., (1928) 30 Ll.L.Rep. 159; (1928) Com. Cas. 324 , Lord Reid said: 

I would agree with Lord Shaw that it is generally sufficient that the event would have appeared to the defendant as not unlikely to occur. It is hardly ever possible in this matter to assess probabilities with any degree of mathematical accuracy. But I do not find in that case or in cases which preceded it any warrant for regarding as within the contemplation of the parties any event which would not have appeared to the defendant, had he thought about it, to have a very substantial degree of probability. Mr. Glennie understandably placed considerable reliance on the adoption of the expression "a very substantial degree of probability" and submitted that in this case at the time of entering into this charter in August, 1988 the owners had no reason to think that there was a very substantial degree of probability that a failure to exercise due diligence to maintain the vessel's seaworthiness would cause the charterer to incur liability to a subcharterer of the kind which it incurred in this case.

It is important with all statements of principle not to take the words used out of context. In the immediately preceding part of the same paragraph Lord Reid had said:

. . . . . but I think that Hall's case must be taken to have established that damages are not to be regarded as too remote merely because, on the knowledge available to the defendant when the contract was made, the chance of the occurrence of the event which caused the damage would have appeared to him to be rather less than an even chance. In referring to "a very substantial degree of probability", therefore, Lord Reid was clearly not intending to suggest that the defendant must have 

203contemplated that the event causing damage was more likely than not to occur, and I did not understand Mr. Glennie to be submitting otherwise. In a later passage in his speech Lord Reid considered and rejected a number of phrases including "liable to result", "a serious possibility" and "a real danger" as not adequately expressing the degree of likelihood which he considered necessary; ultimately he favoured the expression "not unlikely to occur" and other expressions found favour with others of their Lordships in that case.

I do not understand their Lordships in Balfour Beatty v. Scottish Power to be intending to modify the principle in Czarnikow Ltd. v. Koufos . As I read their speeches, both Lord Reid and Lord Jauncey of Tullichettle were using the expression "a very substantial degree of probability" simply to emphasize that the defendant must have realized that there was a real prospect as opposed to a mere chance that the circumstances giving rise to the loss would occur, which is what Lord Reid meant by saying that it is "not unlikely to occur". Bearing that in mind, I prefer, therefore, to frame the question in the following way:

At the time of entering into this charter in August 1988 did the Owners have reason to consider that a failure  to exercise due diligence to maintain the vessel's seaworthiness was not unlikely to cause the charterers to incur a liability to a sub-charterer of the kind which they incurred in this case?

When a vessel is time chartered for a period of three years for worldwide trading the owners will almost inevitably contemplate that it may be sub-chartered for some if not all of the period of the charter. Many charter forms make specific provision for sub-letting as the charter in this case does by cl. 22. The existence of sub-charter contracts is, therefore, clearly within the owners' contemplation, as is the likelihood that a breach of charter on their part which renders the vessel unseaworthy will cause the charterers in turn to incur liability to the owners of cargo. What is said to render the loss in the present case too remote is the peculiar nature of the banana trade in which fruit is cut ahead of the vessel's arrival and is unable to withstand prolonged delay waiting for alternative transport. 

In this context Mr. Glennie reminded me of another passage in Czarnikow Ltd. v. Koufos , this time from the speech of Lord Upjohn at p. 485, col. 2; p. 424C where he said:

Lord Wright pointed out in The Monarch that each must be taken to understand the ordinary practices and exigencies of the other's trade, but it must be remembered when dealing with the case of a carrier of goods by land, sea or air, he is not carrying on the same trade as the consignor of the goods and his knowledge of the practices and exigencies of the other's trade may be limited and less than between buyer and seller of goods who probably know far more about one another's business. This passage was also cited with approval in Balfour Beatty v. Scottish Power and the warning which it contains is no doubt an important matter to keep in mind. However, the carriage of green bananas from tropical ports to Europe and America has been going on for many years. They are one of the well-established cargoes for refrigerated vessels and a number of clauses in the charter make it clear that bananas were one of the cargoes the parties had primarily in mind. I should be surprised if owners of such vessels were not aware of the general nature and demands of the trade, and certainly there was no evidence from the owners in this case that they were not. I do not accept, therefore, simply as a matter of fact that the owners in this case were unaware of the particular nature of this trade or the consequences that might follow if the vessel were unable to load a cargo as a result of a breakdown of the kind which occurred. Moreover, if a loss measured by the replacement cost of the cargo at the port of destination is to be regarded as within the contemplation of a carrier by sea regardless of precisely how it comes about (as was held to be the case in Nissho v. Livanos ), it is difficult to see that a loss of that kind sustained by the charterer under his subcharter can be outside the owner's contemplation. As regards the position in Belize, the only oddity is that Fyffes have to pay for the fruit whether it is cut or not. That is only like buying it on the plant. It is not unusual to buy goods before the point of shipment and if the goods are then lost because they cannot be shipped the owner of the goods suffers the loss. In my judgment the loss in this case cannot be said to be too remote. (b) Mitigation

At a late stage in the trial I gave the owners leave to amend their defence to the contribution proceedings in order to raise the question whether REL was bound to give credit against its claim for certain savings which the owners alleged it had made as a direct result of the abandonment of the voyage at Puerto Cortés. It was not disputed that if the voyage had been completed REL would have incurred hire and fuel costs in relation to the remainder of the voyage amounting to U.S.$217,380. Moreover, the overwhelming likelihood is that following discharge at Zeebrugge the vessel would have been ballasted back to the Caribbean to continue the service under the COA. That would have given rise to further expenses for REL in respect of hire and fuel amounting in all to U.S.$141,564. Mr. Glennie submitted that REL would have recovered freight in respect of the cargoes loaded at Santo Tom s and 204 Big Creek at the rate of U.S.$2.77 per box provided for in the COA which, on a total of 53,850 boxes which I have held would have been loaded, would have amounted to U.S.$149,164.50. That would have given rise to a loss of U.S.$68,215.50 on the cargo-carrying leg alone and a loss of U.S.$207,541.34 overall. Since the voyage was abandoned as a result of the owners' breach of charter he submitted that REL must give credit against its claim for that saving. 

Mr. Young put forward two arguments in answer to that. First, he relied on the agreement made between REL and Fyffes in July, 1991, to which I have already referred, under which Fyffes paid a lump sum by way of deadfreight and the rate of freight was retrospectively increased to U.S.$2.86 per box on cargoes carried prior to that date. He submitted that if the voyage had been completed at a loss to REL, the agreement would have been more generous to REL so as to extinguish that loss. Accordingly, REL would not, in the event, have suffered any loss at all. Secondly, he submitted that the relationship between REL and the owners under the charter was in law a collateral matter or res inter alios acta as far as the loss represented by REL's liability to Fyffes under the COA was concerned and could not therefore be taken into account. 

As to the first of these points, it is possible that if Kriti Rex had completed the voyage and returned straightaway to the Caribbean the agreement might have been concluded in terms more favourable to REL, but the evidence falls far short of persuading me that it would, or what the terms would then have been. I have already said that I regard the agreement between Fyffes and REL as representing an acceptable compromise between differing positions and in my opinion, although it is unlikely that the agreement would have been any less favourable to REL, one can only speculate about whether the outcome would have been more favourable if the circumstances had been different in that respect.

As to the second point, it is well-established that the plaintiff must give credit for any benefits which he has obtained as a direct result of the breach of contract on which he sues. These usually take the form of savings in expense as a result of not having to perform the contract. The plaintiff can only recover the net loss caused by the defendant's breach: see generally Chitty on Contracts, 27th ed., par. 26-001. It is also well-established, however, that benefits which accrue to a plaintiff from independent transactions are not to be regarded as arising out of the defendant's breach of contract, even though it is his breach of contract which causes the plaintiff to obtain them. The benefit which REL

205

Big Creek Cargo obtained as a result of the owners' breach was not a saving of expense under the time charter itself but a saving which resulted from its not being able to perform that particular voyage under the COA in respect of which the claim against it is made. In other words, the owners' breach gave rise to REL's liability to Fyffes, but at the same time it prevented REL from incurring a loss in relation to the same voyage. That is just as much a direct consequence of the owners' breach of the charter as is the liability to Fyffes and in my judgment is a benefit which falls to be taken into account when assessing the net loss suffered by REL as a result of the breach. If it were otherwise REL would be better off as a result of the owners' breach than if the charter had been performed. Mr. Young submitted that even if that were true as regards the carrying voyage to Europe, it did not apply to any subsequent voyage since REL was not bound to keep the vessel in continuous service under the COA. I think there is force in that argument. Although credit must be given for losses avoided as a direct result of the owners' breach, the effects of the owners' breach are limited to the losses saved in not completing the voyage on which the vessel was then engaged. The employment of the vessel after she had discharged her cargo was entirely a matter for REL, and in my judgment such profits or losses as might have been made thereafter are to be viewed solely as a consequence of its decision in that respect. I therefore hold that REL is entitled to recover damages from the owners in the amount of its liability to Fyffes, less the sum of U.S.$65,977.34.

"KRITI REX"

SCHEDULE OF DAMAGES

First Plaintiffs' claim against the First Defendants

Santo Tom s cargo

Second Plaintiff' claimagainst the Second Defendants