England and Wales High Court (Commercial Court) Decisions

IN THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
LONDON CIRCUIT COMMERCIAL COURT (QBD)

		Rolls Building, 7 Rolls Buildings, Fetter Lane, London, EC4A 1NL
		23 October 2018

H.H. JUDGE KEYSER Q.C. :

Introduction

1. This is a claim for compensation arising from a contract for the carriage of cheese from France to the United Kingdom. The claimants say that they have suffered loss because the defendant, which had contracted with them to carry the consignment of cheese, allowed the temperature at which the cheese was being carried to rise too high, with the result that the cheese was damaged and was rejected by the customer that had consigned it. The defendant says that any damage to the cheese (none being admitted) was due to inherent vice or to the manner in which the load was stowed, neither of which was its responsibility.
2. The claim is a modest one; the compensation sought is only €69,287.31. However, it presents both factual and legal problems and was argued with vigour and thoroughness in a trial that lasted three days. I am grateful to Mr Bird and Mr Davidson, counsel for the claimants and the defendant respectively, for their helpful written and oral submissions and for the manner in which they conducted the trial.

The facts

3. The consignment of cheese was to be carried from the premises of la société LAÏTA ("LAÏTA") in Ancenis, France, to the premises of a related UK company, Eurilait Limited ("Eurilait"), in Evercreech, Somerset. It comprised approximately 11,140 kg, most of which was mould-ripened soft cheese (Camembert and Brie). Some of the cheese had been manufactured at LAÏTA's premises at Ancenis, but most of it had been delivered to the premises either from LAÏTA's other production sites or from other suppliers. The cheese manufactured by LAÏTA at Ancenis was stored in its cheese store, identified as Depot 442. The cheese from other sites was stored in a warehouse, identified as Depot 441. For convenience, I shall refer to the cheese that had been stored in these two places and subsequently formed part of the consignment to Eurilait as "the Depot 442 Cheese" and "the Depot 441 Cheese" respectively.
4. The Depot 442 Cheese was Brie that was manufactured on dates ranging from 22 April to 30 April 2015 and had Best Before dates ranging from 16 June to 24 June 2015. LAÏTA's records show that the temperature inside Depot 442 was measured at maxima of 3°C on 1 May, 4°C on 2 May, 5°C on 3 May and 5.4°C on 4 May 2015. The lowest recorded temperature during that period was 0.9°C. Records have not been produced for the period 22 April to 30 April 2015. LAÏTA's operating procedures required that each morning the temperature of six pallets of cheese in Depot 442, selected at random, be checked. It does not appear that any of the Depot 442 Cheese was subject to checking.
5. The Depot 441 Cheese arrived at LAÏTA's premises by means of 23 separate deliveries from eight sites of production. LAÏTA's operating procedures required that from each delivery two pallets would be selected at random and the temperature of their contents would be checked with a temperature probe. The temperature was deemed acceptable if it were within the temperature range mentioned on the product packaging. All deliveries were also subject to visual inspection by the warehouse staff. If the goods passed both the random checking and the visual inspection, they were taken into the warehouse and held there until they were loaded directly onto the trailer by forklift truck.
6. Production details are available for about two-thirds of the Depot 441 Cheese, showing that it was manufactured on dates ranging from 16 April to 30 April 2015 and had Best Before dates ranging from 5 June to 27 July 2015. For the remaining one-third of the Depot 441 Cheese, which came not from LAÏTA's other manufacturing sites but from different suppliers, there is no record of the date of production. The Best Before dates range from 19 June to 21 August 2015; in the absence of any further information, this tends to suggest that the cheese was not produced at any earlier date than 16 April, though that same absence of further information precludes any confident inference. However, I accept evidence given at trial to the effect that LAÏTA only ever agreed to buy young cheese from its suppliers. The deliveries containing the Depot 441 Cheese would have been subject to the reception process described above. LAÏTA's records show that only one pack in the Depot 441 Cheese—a pack of Brie Meaux from another supplier—had been selected for random checking. The sample that was checked had a temperature of 4.2°C, which fell within the acceptable range. Temperature records for Depot 441 show that the temperature was maintained between 3°C and 6.4°C during the period 27 April to 4 May 2015; the general high point did not exceed 6°C, though there was a single spike at 6.4°C. No records were produced for the period before 27 April 2015. The evidence does not establish whether any of the Depot 441 Cheese was received into Depot 441 before 27 April 2015.
7. LAÏTA engaged the claimants, French companies within the STEF Group, to carry the cheese to Eurilait's premises in Evercreech. (The distinction between the first claimant and the second claimant and the precise contractual arrangements between them are not important for the purposes of this case and I shall ignore them.) The terms of the contract between LAÏTA and the claimants are not in evidence, though on the basis of oral evidence given by Elisa Guillerm, LAÏTA's Quality Export Manager, I accept that they included a requirement that the cheese be transported at a temperature within a specified range, probably 0°C to 4°C. The claimants in turn engaged the defendant to carry the cheese. The defendant is a firm based in Scotland and, like the claimants, is engaged in the business of international carriage of goods by road. The evidence shows that the defendant had been carrying cheese for the claimants from LAÏTA's premises for a number of years on a weekly basis.
8. The claimants' instruction to the defendant, which the defendant accepted, was in the form of a transport confirmation (the "Transport Confirmation") that the defendant received on 1 May 2015. The Transport Confirmation included the following provisions:

"front compartment Temperature: +2/+4°C …"

"Temperature: The carrier take in charge controlled temperatures goods and commit to maintain them in the temperatures required and indicated during all the duration of the services. The vehicle must be pre-cooled and be at the appropriate temperature when loading commences. The temperature should be checked at loading and maintained throughout the transport. Periodic recording of the temperature during the loading and the transport is essential regardless of the distance (fresh, quick-frozen and frozen processed products, pharmaceutical products). The continuous settings of refrigeration unit is mandatory for all fresh product transports."

"General conditions: allowing the vehicle to be loaded implies full and formal acceptance of the above conditions."

9. To perform the delivery, the defendant used a lorry with refrigerated trailer, which it had bought second-hand some time previously and which it had used many times to carry dairy goods from LAÏTA's premises to the UK. For that purpose the trailer was required to have a valid ATP Certificate, showing compliance with the Agreement on the International Carriage of Perishable Foodstuffs. However, the ATP Certificate for the defendant's trailer, which had been issued in February 2006, had expired in February 2012, more than three years before the events giving rise to this claim. It is not suggested that the absence of a valid ATP Certificate is either of causal relevance or itself indicative of malfunction of the refrigerated trailer, but it does constitute a failure to comply with legal requirements and it also means that there had been no recent inspection for the purposes of ATP compliance.
10. The trailer was 13.6m long and insulated with polyurethane foam. The refrigeration unit in the trailer was a Carrier Vector 1800 unit, of a kind conventional for road trailers. Cooling air was supplied to the top of the cargo space and returned to the refrigeration unit via ducts in the forward wall of the trailer. The unit had two relevant operational modes, namely the "Continuous" mode and the "Start/Stop" mode, either of which could be selected by the driver. In the Continuous mode, the refrigeration unit and air-circulation fans run continuously. In the Start/Stop mode, by contrast, they run for a specified Minimum Run Time—in this case, probably the manufacturer's default period of four minutes—and thereafter until the supply air temperature cools to the set point temperature; they then remain off until the temperature rises to a particular level, in the case of perishable goods usually 2°C above the set point temperature. The purpose of the Minimum Run Time, as stated by the manufacturer, is to ensure that the air inside the trailer has fully circulated and that the product temperature has reached the set point. The purpose of using the Start/Stop mode is to reduce fuel consumption. The integral data logger in the refrigeration unit shows that successive starts of the unit during the journey from Ancenis to Evercreech took place approximately every 17 minutes.
11. On 4 May the defendant's vehicle arrived at LAÏTA's premises during the late morning. The defendant's driver probably did not know the detailed terms of the instructions received from the claimants, but he did know the required temperature range for the trailer, and at about 1 p.m. local time (GMT + 1) he set the refrigeration unit to the Continuous mode and a set point temperature of 2°C. (The control parameter for the set point temperature was the supply air temperature rather than, as was the default setting, the return air temperature. It has not been suggested that there was anything inappropriate about this.) Loading of the trailer took place shortly after 3 p.m. local time and was carried out by LAÏTA's personnel. The contemporaneous documentation indicates that, immediately before loading the cheese, LAÏTA found the temperature of the trailer to be 5.2°C. Although that temperature exceeded the temperature range stipulated by LAÏTA for carriage, it loaded the cheese onto and stowed it in the trailer. (The recorded temperature of 5.2°C was within the permitted ATP maximum temperature of 6°C.) The Depot 442 Cheese was loaded first, and then the Depot 441 Cheese; therefore the cheese manufactured on site was probably towards the front of the trailer, near the refrigeration unit, with the larger amount of cheese from other sites extending towards the doors at the rear of the trailer. The total load comprised 38 pallets, of which 33 pallets were on the floor of the trailer. The pallets were closely stowed along the width and length of the trailer, save that at the rear, near the doors, there was a single pallet with nothing beside it.
12. The defendant's driver signed a CMR International Consignment Note (the "CMR Consignment Note"), which acknowledged receipt of 38 pallets of various cheeses with a gross weight of 11,139 kg. The CMR Consignment Note recorded, "Fridge Temperature set at +2 Degrees centigrade in accordance with Shipper's Instructions". Under the heading "Special agreements" was the following text, which had been pre-typed and was standard on the defendant's CMR Consignment Notes:

"Loading & Stowage – Full responsibility of Shipper.

The Driver not having opportunity to check and control the actual loading of the trailer."

13. The driver also signed two other consignment notes, one in respect of 10 pallets and the other in respect of 28 pallets. Both of these consignment notes were typewritten in French. Each of them recorded the vehicle temperature as being 5.2°C and contained text that means, in English, "The goods are shipped in good apparent condition and at a temperature which accords with regulations."
14. After loading had been completed, the driver changed the setting of the refrigeration unit from Continuous mode to Start/Stop mode; he left the temperature set point unchanged at 2°C. Then he left LAÏTA's premises and proceeded to Caen, where he boarded a ferry for Portsmouth. The ferry arrived in Portsmouth at about 6 or 6.30 a.m. (the records are not easy to reconcile) on 5 May and arrived at Eurilait's premises at about 11 a.m.
15. With one brief interruption, the refrigeration unit remained on Start/Stop mode, with a set point of 2°C, at all stages of the journey from LAÏTA's premises to Eurilait's premises. That interruption was for a period of about forty minutes immediately before the vehicle came off the ferry at Portsmouth: while on the ferry the refrigeration unit, which normally was powered by the lorry's diesel engine, ran by way of a connection to the ferry's power supply, but that supply was disconnected by the ferry crew in readiness for disembarkation. There is a complete set of the data recorded by the refrigeration unit of the so-called Supply Air Temperature, which is the measurement taken by a sensor near to the inlet duct that supplied the cool air. These show that from the time the vehicle left Ancenis until the unit was shut down as the ferry approached Portsmouth, the recorded temperatures ranged from 0.7°C to 5.4°C and were usually in the range 2°C to 5°C. During the period of the shutdown the temperature rose as high as 10.6°C, but within a few minutes after re-start it had reduced to 2.9°C. It then remained almost entirely within the range 2°C to 5°C until arrival at Eurilait's premises.
16. When the defendant's vehicle arrived at Eurilait's premises, Eurilait rejected the cheese. A manuscript endorsement placed on the CMR Consignment Note by Eurilait's Technical Manager at 11.10 a.m. reads: "Rejected for over temperature 10-11 deg. C". (I refer below to some additional information contained in an email from Eurilait on the following day.) The cheese was not unloaded, so it is a reasonable inference that the temperature, whether it be of goods or air, was taken near the doors of the trailer.
17. As soon as Eurilait rejected the cheese, the defendant's driver made a telephone call to Ross Fraser, the defendant's transport planner. Mr Fraser was alarmed at the rejection of the cheese and, in the hope that Eurilait could be persuaded to reverse its decision if the temperature of the cheese were reduced, instructed the driver to set the refrigeration unit to the Continuous mode at a set point of 0°C. The driver did so at about 11.37 a.m. At 4.25 p.m. that day the driver changed the set point temperature to 2°C; however, the unit remained in the Continuous mode.
18. Arrangements were made for an inspection of the cheese on the following day, 6 May, at the premises of Fowler Welch in Portsmouth. The cheese was taken there in the defendant's vehicle. CL Surveys attended the inspection on the instructions of the claimants' insurer. The defendant was invited to instruct someone to attend on its behalf but did not do so. CL Surveys inspected the cheese at 12.30 p.m. on 6 May. The following temperatures were recorded for the heart of the cheeses stowed at various positions in the trailer:

Trailer doors: +8.2°C to +8.9°C

¼ trailer: +5.2°C / +5.5°C / +8.7°C / +7.1°C

Middle of trailer: +5.2°C to +5.5°C

¾ trailer: +4.9°C / +4°C / +4.6°C / +6°C.

(It is likely that the measurement for "¾ trailer" either includes or would be similar to the measurement at the front of the trailer.) It was recorded that traces of condensation were noted on the inside of a bag of brie located near the doors of the trailer. There is also some evidence that the cheese gave off a strong smell; see below.

19. On the strength of the findings on the inspection and the reduction in the recorded temperatures, the claimants decided to seek to persuade LAÏTA and Eurilait that the cheese should be accepted. They therefore instructed the defendant to transport the goods to the premises of Peter Green Chilled, which are adjacent to the premises of Eurilait in Evercreech, for storage overnight. The claimants also sent an email to LAÏTA. It explained the printout of data from the refrigeration unit, showing the supply air temperature (SAT) and refrigeration unit temperature (DTT1), and concluded (in English translation): "We can see in this instance that the SAT and DTT1 are fine. So everything is normal. I hope that this information satisfies your requirements." Ms Guillerm replied as follows (again, in English translation; I have also corrected the dates and times on the refrigeration unit data, which are agreed to be wrongly set):

"We thank you for this information and inform you that we have on our side found two anomalies:

from 6:10 to 6:50 on 5 May 2015: temperatures above +4°C, and

from 11:40 on 5 May 2015: negative temperatures."

In addition, the note upon receipt by Eurilait of the non-conformity of the temperatures within the goods delivered confirms the problems encountered.

As the goods have experienced a rise in temperature, then negative temperatures, we cannot take the risk of accepting them: the risk of a quality issue is too high.

As such, the rise in humidity for soft cheese within its packaging caused by such variations in temperature will cause major presentation defects.

The maturing of our products will also be impacted and our product will not satisfy the expectations and demands of our customers.

The negative temperatures will also have given rise to a syneresis especially in our 'Fromages à Tartiner', its texture is affected when temperatures fall below 0°C.

We will under no circumstances agree to the commercialisation of the products in question."

20. The first anomaly mentioned by Ms Guillerm concerns the period when the refrigeration unit was turned off shortly before disembarkation from the ferry. Ms Guillerm said in cross-examination that she thought the extent and duration of the increase in temperature was sufficient to affect the cheese adversely and justify rejection of the cheese, though her opinion in that regard does not seem to me to be supported by the weight of the evidence, and I accept Mr Wilkie's opinion that the period in question would not have been causative of damage. As for the second anomaly, it became apparent in cross-examination that Ms Guillerm had been considering the readings from the probe in the refrigeration unit itself (DTT1); she said that she considered these to be relevant, but as they do not relate to the temperatures in the trailer I am satisfied that she is mistaken in that regard.
21. That same afternoon, Eurilait's Supply Chain Manager sent an email to LAÏTA:

"To reiterate[:] the state of the trailer was on arrival 11C. When opening the trailer the trailer was very smelly suggesting no temperature control for some time. The core temperature of some cheeses sampled was 17C. the driver was unable to provide any temperature control documentation or where the temperature is recorded.

We will not accept this vehicle load again under any circumstances. We have received a phone call requesting to book this in again now the temperature is ok[;] however due to the previously mentioned issue we will not accept the stock."

If the information in this email is correct, the reference to 10-11°C in the manuscript endorsement on the CMR Consignment Note is not to the temperature of the cheese (which had been generally assumed; cf. para 8.7 of the first report of Richard Lawton, below) but to the temperature of the trailer, and that endorsement makes no reference to the fact that a pulp temperature of up to 17°C had been measured. The state of the evidence regarding these matters is unsatisfactory.

22. On the morning of 8 or, possibly, 7 May, acting (as I think probable) on the claimants' instructions, Peter Green Chilled re-delivered the cheese to Eurilait for acceptance. A report from CL Surveys records what happened:

"At time of arrival of the means of transport on the Company EURILAIT premises, at the opening of the doors, the EURILAIT Company noted that a strong smell was released from the goods and that the temperature taken at the heart of some products was +17°C."

Accordingly, Eurilait maintained its rejection of the cheese on the ground that it had been presented above the stipulated temperature.

23. LAÏTA refused to take the cheese back or to allow it to be sold commercially. Its position was set out in an email sent on 11 May 2015:

"As noted above, these products cannot be sold as is by our subsidiary. We consider that this claim is at your charge. We suggest you head these products to a company which can undertake melting."

24. A further inspection of the cheese took place on 13 May at the premises of Peter Green Chilled, where it had been held in a cold store at 2°C. This time representatives of all interested parties were present; these included a surveyor instructed on behalf of the defendant. The report produced later by CL Surveys contained the following relevant text concerning this inspection:

"During our intervention on the 13/05/2015, the goods were a similar state in the previous survey dated on the 06/05/2015. The goods gave off a smell of the same intensity that smell on the first survey, in relation to the nature of the goods.

We noted the following temperatures on the surface of the products: +5.0°C / +6.8°C / +8.4°C / +7.0°C / +5.9°C / +6.6°C / +8.3°C / +7.9°C. Synthetically, the temperatures ranged from 5.0°C to +8.4°C.

Temperature remained high despite storage conditions. It reflected an evolution of cheese, ripening is an exothermic process …

We noted the presence of condensation on external packaging of some cheeses, showing the influence of thermal variations. It was more developed than in the first survey inspection."

25. On 20 May, the defendant's surveyors, CWH & Co, sent an email to Eurilait:

"Whilst there did not appear to be a vast amount of damage and deterioration suffered by the cheese at the time of our inspection, I do accept your reservation particularly regarding those cheeses made with raw milk and the fear of e coli, listeria and other nasty problems.

It is understood that you are making further enquiries with your Parent company Laita as to the best course of action to adopt to mitigate the loss and in this respect I should be grateful to receive your confirmation in writing as to whether this would simply be restricted to the cheeses made with raw milk, Brie de Meaux etc, or whether it would be involving the consignment as a whole."

26. CL Surveys asked CWH & Co for direct confirmation that they had no objection to the proposal that, in line with LAÏTA's suggestion, the cheese be reprocessed and sold for salvage. CWH & Co replied:

"As was agreed with your surveyor at time of survey we had no objection for the product to be reprocessed albeit that agreement at that time was on an entirely without prejudice basis and without any admission of liability.

It is for the goods owners to mitigate their loss and to take all actions they feel necessary to achieve that goal.

I trust therefore you will be guided accordingly and I will continue to complete my investigations into the cause of the loss/damage."

27. In due course, accordingly, the cheese was sold for €5,570. The claimants paid €73,618.53 to LAÏTA in respect of compensation for the damage to the cheese. It incurred additional transportation and storage costs of €1,239.60. Against this total outlay of €74,858.13 it gives credit for the proceeds of the sale of the cheese. It claims the balance of €69,288.13 from the defendant.

The Legal Framework

28. It is common ground that there was a contract between the claimants and the defendant and that it was governed by and incorporated the terms of the Convention on the Contract for the International Carriage of Goods by Road (the "CMR Convention"), which is set out in the Schedule to the Carriage of Goods by Road Act 1965.
29. Section 1 of the Carriage of Goods by Road Act 1965 provides:

"Subject to the following provisions of this Act, the provisions of the Convention on the Contract for the International Carriage of Goods by Road (in this Act referred to as 'the Convention'), as set out in the Schedule to this Act, shall have the force of law in the United Kingdom so far as they relate to the rights and liabilities of persons concerned in the carriage of goods by road under a contract to which the Convention applies."

Section 14(2) provides, so far as relevant for present purposes:

"The persons who, for the purposes of this Act, are persons concerned in the carriage of goods by road under a contract to which the Convention applies are—

(a) the sender,

(b) the consignee,

(c) any carrier who, in accordance with article 34 in the Schedule to this Act or otherwise, is a party to the contract of carriage,"

It is common ground that, in the circumstances of this case, LAÏTA was the "sender", Eurilait was the "consignee" and the claimants and the defendant were "carriers" for the purpose of section 14(2) and "successive road carriers" for the purpose of article 34 (below).

30. The following provisions of the CMR Convention are relevant to this case.

"Article 8

1. On taking over the goods, the carrier shall check:

…

(b) the apparent condition of the goods and their packaging."

"Article 9

1. The consignment note shall be prima facie evidence of the making of the contract of carriage, the conditions of the contract and the receipt of the goods by the carrier.

2. If the consignment note contains no specific reservations by the carrier, it shall be presumed, unless the contrary is proved, that the goods and their packaging appeared to be in good condition when the carrier took them over and that the number of packages, their marks and numbers corresponded with the statements in the consignment note."

"Article 17

1. The carrier shall be liable for the total or partial loss of the goods and for damage thereto occurring between the time when he takes over the goods and the time of delivery, as well as for any delay in delivery.

2. The carrier shall however be relieved of liability if the loss, damage or delay was caused by the wrongful act or neglect of the claimant, by the instructions of the claimant given otherwise than as the result of a wrongful act or neglect on the part of the carrier, by inherent vice of the goods or through circumstances which the carrier could not avoid and the consequences of which he was unable to prevent.

…

4. Subject to article 18, paragraphs 2 to 5, the carrier shall be relieved of liability when the loss or damage arises from the special risks inherent in one or more of the following circumstances:

…

(c) handling, loading, stowage or unloading of the goods by the sender, the consignee or persons acting on behalf of the sender or the consignee;

(d) the nature of certain kinds of goods which particularly exposes them to total or partial loss or to damage, especially through breakage, rust, decay, desiccation, leakage, normal wastage, or the action of moth or vermin; …

5. Where under this article the carrier is not under any liability in respect of some of the factors causing the loss, damage or delay, he shall only be liable to the extent that those factors for which he is liable under this article have contributed to the loss, damage or delay."

"Article 18

1. The burden of proving that loss, damage or delay was due to one of the causes specified in article 17, paragraph 2, shall rest upon the carrier.

2. When the carrier establishes that in the circumstances of the case, the loss or damage could be attributed to one or more of the special risks referred to in article 17, paragraph 4, it shall be presumed that it was so caused. The claimant shall however be entitled to prove that the loss or damage was not, in fact, attributable either wholly or partly to one of these risks.

…

4. If the carriage is performed in vehicles specially equipped to protect the goods from the effects of heat, cold, variations in temperature or the humidity of the air, the carrier shall not be entitled to claim the benefit of article 17, paragraph 4(d), unless he proves that all steps incumbent upon him in the circumstances with respect to the choice, maintenance and use of such equipment were taken and that he complied with any special instructions issued to him."

"Article 25

1. In case of damage, the carrier shall be liable for the amount by which the goods have diminished in value, calculated by reference to the value of the goods fixed in accordance with article 23, paragraphs 1, 2 and 4.

2. The compensation may not, however, exceed:

(a) if the whole consignment has been damaged the amount payable in the case of total loss;

(b) if part only of the consignment has been damaged, the amount payable in the case of loss of the part affected."

"Article 34

If carriage governed by a single contract is performed by successive road carriers, each of them shall be responsible for the performance of the whole operation, the second carrier and each succeeding carrier becoming a party to the contract of carriage, under the terms of the consignment note, by reason of his acceptance of the goods and the consignment note."

"Article 35

…

2. The provisions of article 9 shall apply to the relations between successive carriers."

"Article 37

A carrier who has paid compensation in compliance with the provisions of this Convention, shall be entitled to recover such compensation, together with interest thereon and all costs and expenses incurred by reason of the claim, from the other carriers who have taken part in the carriage, subject to the following provisions:

(a) the carrier responsible for the loss or damage shall be solely liable for the compensation whether paid by himself or by another carrier;

(b) when the loss or damage has been caused by the action of two or more carriers, each of them shall pay an amount proportionate to his share of liability; should it be impossible to apportion the liability, each carrier shall be liable in proportion to the share of the payment for the carriage which is due to him;

(c) if it cannot be ascertained to which carriers liability is attributable for the loss or damage, the amount of the compensation shall be apportioned between all the carriers as laid down in (b) above."

"Article 40

Carriers shall be free to agree among themselves on provisions other than those laid down in articles 37 and 38."

31. For the purposes of Articles 8 and 9, the apparent condition of goods "refers to what is discernible externally on a reasonable examination, 'so far as meets the eye'": Clarke, International Carriage of Goods by Road: CMR (6^th edition, 2014), at section 25a. The text in that section continues:

"The carrier is not obliged to test the strength of ropes or packaging used by the sender, still less to assess the quality of the goods as such. In particular, the carrier is not required to check that goods have been consigned at an appropriate (core) temperature. However, apparent condition extends not only to goods but also to their packaging and external temperature."

There was some debate at trial as to the correctness of the final reference to external temperature and as to the degree of support provided by the reference given in the footnote to the text. However, if the basic meaning of apparent condition, as mentioned above, is borne in mind, I do not think that there is any great difficulty in the present case. The defendant's driver was not required to test the pulp temperature of the cheese. It is implausible to suppose that reasonable examination required him to touch the different packs of cheese. If he saw anything untoward on the pallets—it is not suggested that there was anything untoward to see—, he should record it on the consignment note. If, which is unlikely and is not a realistic possibility on the evidence, he sensed unusual heat being given off by the load, he should note that. Of greater relevance to the present case, apparent condition would concern what met the nose just as much as the eye; if the driver was aware of an abnormal smell from the load, he should record that.

32. "In Article 17.1 of the CMR, damage means any change in the physical state of the goods which reduces their value": Clarke, op. cit., at section 57. Professor Clarke cites OLG Celle 13.1.75 as authority for the proposition that, for the purposes of the CMR, "damage is characterised by external or internal physical deterioration, which results in diminution of value". He continues: "Diminution of the value of goods without physical deterioration, however, is not damage; were it otherwise it would be too difficult to distinguish damage from loss of market, for which the carrier is not liable under the CMR."
33. The distinction between loss caused by "inherent vice of the goods" (Article 17, paragraph 2) and loss arising from the special risks inherent in "the nature of certain kinds of goods etc" (Article 17, paragraph 4(d)) is not entirely clear. Both can encompass damage from overheating of the goods: for inherent vice, see Ulster-Swift Ltd v Taunton Meat Haulage Ltd [1975] 2 Lloyd's Rep. 502, affirmed at [1977] 1 Lloyd's Rep. 346; for risks inherent in the nature of the goods, see Article 18, paragraph 4. That there is a distinction between the two causes of loss is evident from the different incidences of the burden of proof. Although the cases do not speak with one voice on the point, in my judgment the distinction is that risk inherent in the nature of goods concerns a quality existing in all goods of the same nature, whereas inherent vice is some inherent defect in the goods, but not inherent in their nature and not normally found in goods of the same kind, which renders them unfit to withstand the ordinary incidents of carriage: see Messent and Glass, "CMR: Contracts for the International Carriage of Goods by Road" (4^th edition, 2017), at paragraphs 6.25 to 6.27. Thus, there is a risk inherent in the natures of cheese and meat that they will deteriorate when warm, but this is not an inherent vice; but if particular cheese or meat is peculiarly susceptible to damage on account of overheating, because it is inherently rotten or because it was inappropriately warm when loaded into the refrigerated unit, it has an inherent vice.
34. Finally, in Article 17, paragraph 2, the expression "could be attributed to" probably "connotes something less than a balance of probabilities [but] something more than a remote possibility": Ulster-Swift Ltd and another v Taunton Meat Haulage Ltd [1977] 1 Lloyd's Rep. 346, at 354.

Summary of the case

35. The claimants' case in summary is as follows. They have paid compensation and incurred costs by reason of damage to the condition of the cheese. The defendant is the carrier responsible for that damage under article 17, paragraph 1, because it occurred between the time when the defendant took over the goods and the time of delivery to Eurilait. Therefore, the defendant is liable to pay to the claimants the amount of compensation and other expenses they have paid: article 37(a). (The claim is also put in contract, independently of the CMR Convention, though I do not think it necessary to consider that distinct way of putting the claim or even the extent to which a claim could lie outside the terms of the CMR Convention.) Factually, the claimants' case is that the damage to the cheese occurred because the defendant failed to maintain its temperature within acceptable levels, and that that failure resulted from two causes, namely the running of the refrigeration unit in the Start/Stop rather than the Continuous mode and an incorrect configuration of the trailer.
36. In response to this claim, the defendant contends, first, that there is insufficient evidence to justify a finding that the cheese was damaged during carriage. If, however, such damage is found to be proved, the defendant contends, second, that the claim fails because the claimants are unable to prove—the burden, it is said, being on them—that the damage was not due to either incorrect stowage or the perishable nature of the cargo. The availability of contractual or common-law rights of action outside the Convention is disputed. If any liability is found to exist, issue is taken with the quantum of the compensation payable by the defendant.
37. Accordingly, although the necessary analysis has a number of refinements, the central questions that fall to be answered are the following: (1) Did the goods suffer damage while in carriage from Ancenis to Evercreech? (2) If so, was the cause of the damage one for which the defendant is responsible? (3) If so, what is the extent of the damage for which the claimants can recover compensation from the defendant?
38. The factual evidence available to assist in answering these questions was mainly in the form of documentation, including a report written by CL Surveys in September 2015. I have summarised above what seem to me to be the most important matters established by the documents. No evidence was given by any witness who saw the cheese. Evidence of fact was given at trial for the claimants by Ms Guillerm, who is not based at Ancenis and whose evidence was for the most part based on the documents she was able to produce. The defendant's driver, Douglas Kirk, did not give evidence; it appears that he is too unwell to do so, and it was agreed that I should disregard a statement made in his name. The defendant's witness of fact was Ross Fraser, whose direct involvement in the matters giving rise to this claim was limited.
39. I received evidence from two expert witnesses: for the claimants, Richard Lawton, Technical Director of Cambridge Refrigeration Technology; for the defendant, Kevin Wilkie, a director of Endeavour Marine & Technical Ltd t/a Reefer Consult. Their evidence was principally directed to what I have identified as the second question (causation of damage). I have to say that I did not find the evidence of either expert to be especially compelling. Mr Lawton has considerable experience in carriage of goods by road and, in that context, of refrigeration, but he did not always demonstrate persuasive grounds for his opinions; to a degree, he was also hampered by the limited factual information available. Mr Wilkie, by contrast, has little experience in carriage of goods by road; his expertise principally concerns carriage at sea. This did not inhibit him from a large measure of dogmatism, coupled with a robustly dismissive attitude towards opinions differing from his own and an unwarranted tendency to attribute, sometimes by implication and sometimes expressly, bad motives to the claimants and those assisting them (see, for example, the unjustified statement in paragraph 11.7 of his first report that certain references made by the claimants to temperature readings were "deliberately misleading"). Nevertheless, Mr Wilkie's opinions have to be assessed on their merits, and certain parts of his reasoning seemed to me to have considerable force. It is convenient to summarise the main points in the expert evidence before turning to the specific issues in the case.

Summary of the expert evidence

40. Mr Lawton identified two causes for the high temperatures recorded on delivery to Eurilait. The primary cause was that the refrigeration unit had been turned to Start/Stop mode instead of Continuous mode. This meant that the cooling air was flowing for only about 55% of the duration of the carriage, with the result that the cheese not only would not be cooled for substantial amounts of time but would heat up by reason of its own metabolic processes. "Running in Continuous mode would have significantly slowed the natural increase of temperature of the cheese due to the exothermic ripening process, as there would always be airflow." The consequence was that the temperature of the cheese would have been on average higher throughout the period of carriage than if Continuous mode had been used. One objection to this theory is that the manufacturer's handbook for the refrigeration unit contains a section on recommended transport temperatures, which includes for Dairy Products a set-point range of 2°C to 6°C and, as regards the operating mode, states: "Auto-Start/Stop or continuous". Mr Lawton pointed to the accompanying text in the handbook, which said of the recommended transport temperatures and operating modes: "These are included for reference only and should not be considered pre-emptive of the setpoint required by the shipper or receiver." He also pointed to another section of the handbook, which deals with the Start/Stop mode and states: "Caution: During unit shut-downs, the evaporator fans also stop. Only use this operating mode for products which tolerate shut-downs of this kind." Mr Lawton expressed the opinion that the cheese was incapable of tolerating a shut-down of the refrigeration equipment and the air-circulation fans. However, the principal reason he gave for that opinion in cross-examination was simply that the temperature of the cheese on delivery showed that it had not tolerated storage in the Start/Stop mode; the risk of circularity in that reasoning is obvious. Mr Lawton did say, further, that the distinction between cheese and other dairy products such as milk was that cheese produces its own heat. That seems to me to be capable in principle of being a material distinction, though it does not follow that the Start/Stop mode is unsuitable for cheese. Mr Lawton confirmed that he had not been able to find any documentation or literature that supported his contention that the Start/Stop mode was unsuitable for cheese.
41. The second reason given by Mr Lawton for the high temperatures was that the internal configuration of the trailer was inappropriate, in that it was not such as to ensure that the cooling air introduced by the refrigeration unit circulated throughout the trailer compartment. His specific complaint concerned the false bulkhead at the front of the trailer. The experts agreed that the primary purpose of the false bulkhead was to prevent the cargo from pressing against the forward bulkhead and thereby blocking the air return to the evaporator. However, Mr Lawton was of the view that the interlocking plates or baffles of which this particular false bulkhead was comprised ought to have extended from a position just below the supply air duct near the top of the trailer down to a position just above the floor. Such a configuration would have meant that the return air vents, a little below the supply air duct, would be behind the baffles instead of (as was the case) above them. The effect of the return air vents being above instead of behind the baffles was, said Mr Lawton, that some of the cool air would have short-circuited between the supply and return air ducts at the front of the trailer, and the amount of cool air that reached the rear of the trailer would thereby be reduced. If the baffles were not to be placed in front of the return air vents, an alternative method of ensuring that the cool air reached the rear of the compartment was by the use of one or more "air socks", running along the ceiling of the trailer from the supply air duct to the rear of the trailer. No such air sock was fitted. As I have noted, Mr Lawton was clear that the absence of an air sock was not itself a point of criticism; rather, he criticised the missing baffles and said that, if the baffles were not to be replaced, an air sock would have compensated. His evidence on this point was, as it seems to me, weakened by the fact that he was unable to point to any support in the documentation or literature for the supposed importance of positioning baffles in front of the return air vents to prevent short-circuiting of the flow of cool air. Further, in cross-examination Mr Lawton said that, although in his opinion the absence of the baffles was a contributory factor to the high temperatures, he had not carried out any tests with a view to ascertaining the extent of the effect it would have had, and he did not know how much it would have affected air-flow.
42. By way of support for his conclusions, Mr Lawton pointed to two matters in connection with the temperatures recorded at the premises of Fowler Welch on 6 May. First, the temperature of the cheese at the door end was then no higher than 8.9°C, whereas it had been 11°C on arrival at Eurilait; this indicated that the Continuous mode was effective in reducing air temperature. Second, the temperatures recorded at different parts of the trailer showed that the cheese was progressively cooler towards the cooling machinery at the front of the trailer; this was consistent with ineffective maintenance of temperature caused by inadequate air-flow. Mr Lawton was unable to explain how the two things identified by him as relevant causes, namely incorrect operating mode and incorrect configuration of the trailer, could account for a recorded pulp temperature as high as 17°C, in the light of the readings from the data logger, though as he pointed out there is very little information as to how or where that reading was taken. More importantly, Mr Lawton's first point overlooks the fact that the refrigeration unit was programmed to a set point of 0°C for nearly five hours after its departure from Evercreech on 5 May.
43. Mr Lawton considered and dismissed two alternative explanations advanced by the defendant. First, he rejected the suggestion that the method of stowage of the cheese caused or contributed to damage: he considered the stowage, as it appeared from photographic evidence, to be normal and acceptable. Second, he rejected the suggestion that the damage to the goods was inevitable because of the temperature at which they had been loaded or because of some inherent vice: the loading temperature was within range and the temperature upon delivery was significantly outside the range, which it would not have been if the cheese had been stored at a proper temperature during carriage.
44. In Mr Wilkie's first report, paragraph 7.7 shows not only where his opinion differs from that of Mr Lawton but also where they agree on the potential significance of the temperatures recorded at different parts of the trailer by CL Surveys on 6 May 2015:

"This [i.e. the record of the temperatures] is quite strong evidence to show that the high pulp temperatures nearest the doors are more a product of poor air circulation due [to] short circuiting in this area, and/or cargo in this area was loaded in warm / incorrectly pre-cooled condition, and/or the cargo in this area had already started to decay and was self-warming, rather than as a result of any lack of cooling arising from the use of the 'Start/Stop' control mode for at least part of the journey."

45. As for the operating mode of the refrigeration unit, Mr Wilkie noted the terms of the guidance in the manufacturer's handbook (see paragraph 40 above), which he thought should be taken at face value. He found no support for the contention that cheese was especially vulnerable to fluctuations of temperature, provided it were kept within proper limits. The information from the data logger showed that the refrigeration unit was operating correctly; it would therefore have maintained the temperature within the trailer within the proper range. Mr Wilkie said that the data logger would have recorded not only the temperature of the supply air but also that of the return air; this potentially useful piece of information had not been downloaded and was not available in evidence. He considered that the temperature of the air around the pallets would not have risen significantly above 4C°. He pointed out that a carrier can control the temperature of the air around the load but not the pulp temperature of the goods, albeit that the former will have an effect on the latter.
46. As for the internal configuration of the trailer, Mr Wilkie robustly rejected any explanation in terms of missing baffle plates: "Such an unfounded and unsupported allegation is fundamentally flawed, is lacking in simple logic and defies basic common sense." He disagreed strongly with the suggestion that the false bulkhead ought to have been higher and that a main purpose of having it was to prevent short-circuiting of the air. In this respect, he pointed to the fact that the false bulkhead consisted of a complex of interlocking plates; if the purpose were as suggested by Mr Lawton, he said, it would make more sense to have either a single removable bulkhead or to fit a fixed baffle between the supply and return ducts. I do not find this objection very persuasive: the advantage of a single removal bulkhead could apply equally if the sole purpose of the bulkhead were to prevent the cargo from pressing against the forward bulkhead; further, the trailer and the refrigeration unit were manufactured by different manufacturers, and it is not clearly established on the evidence that different refrigeration units, perhaps of different configurations, could not be fitted in the trailer.
47. Mr Wilkie, accordingly, posited alternative causes of the high temperatures: either the cheese that was stowed at the back of the trailer was in a poor condition when loaded, whether because of poor storage before it was loaded or for some other reason; or the air circulation in the trailer was impeded by the manner in which the goods were stowed by LAÏTA's personnel. For the former explanation—stated by Mr Wilkie to be "the most probable cause of the alleged high pulp temperatures"—there was (as it seems to me) no evidence other than the fact of the deterioration of the cheese and the lack of a plausible alternative explanation. Mr Wilkie accepted in cross-examination that there was no positive evidence in the documentation or elsewhere that indicated that any of the cheese had been incorrectly stored or that it was in any way defective. However, he pointed to the gaps in the records as showing that reliance could not be placed on the available documentation as establishing that the cheese was in good condition when loaded. As for the suggestion that the high temperatures were due to the manner of stowage, Mr Wilkie's reasoning was set out in paragraphs 7.4 and 11.14 of his first report (and substantially repeated in his oral evidence):

"The CL Report at Page 15 includes photographs of when the trailer doors were opened. It is important to note that immediately adjacent to the doors there is one (low) pallet on the left hand side of the trailer. The floor space on the right hand side in this area is open with no pallet being stowed. This open space will lead to short circuiting of the cooling air which is of course supplied to the top of the container. Cooling air in that area near the door will tend to by-pass the pallet on the LHS and indeed the pallets immediately forward of that. This will result in a lack of cooling of cargo in this area and hence will result in relatively higher temperatures."

"Photographs taken during the joint survey on 6^th May show a large open floor space nearest the trailer doors. This would have resulted in short circuiting of the cooling air and poor cooling of the cargo in that area, which was apparent when comparing the pulp temperatures taken of the cargo in this area when compared to the cargo further inside the container."

The Issues

48. The central issues are identified in paragraph 37 above. It is helpful to distinguish the first issue clearly from the second issue, not only for the sake of clarity but because it is important to be clear at all stages of the analysis where the burden of proof lies. Similarly, the second issue ramifies in accordance with the provisions of Articles 17 and 18.

(1) Was there damage to the goods in transit?

49. The first question is whether damage occurred to the goods between the time when the defendant took them over (that is, when they were loaded into the trailer at Ancenis) and the time of delivery (that is, when the goods arrived at Evercreech). This involves a comparison of the condition of the goods at those two times, though in principle one must, as Mr Davidson observed, have regard to all the evidence before reaching a conclusion on either part of the comparison; this includes the evidence relevant primarily to the causation issue, though for convenience of exposition that evidence is for the most part discussed separately in this judgment. The claimants' case, in a nutshell, is that the cheese was clearly in a damaged condition when it reached Evercreech and that all of the evidence suggests that it was not in a damaged condition when it was loaded at Ancenis; therefore, the goods were damaged at the time mentioned in Article 17, paragraph 1, and, subject only to possible relief from liability by reason of other provisions of Article 17 and Article 18, the defendant is liable for the damage in accordance with Article 17, paragraph 1.
50. In my judgment, there was damage to the goods. (I consider later the question of the extent of the damage.)
51. It is convenient to begin with the condition of the goods on arrival at Evercreech. I find that the goods were in a substandard condition at that time; the matters indicating this, established on a balance of probabilities on the evidence, have been set out above and may be summarised shortly:

1) The cheese gave off a strong smell. This appears from the email from Eurilait's Supply Chain Manager on 6 May (paragraph 21 above), which said that the trailer was "very smelly" when it was opened on 5 May. That the cheese smelt abnormally upon delivery tends to be confirmed by the report from CL Surveys (paragraph 24 above), which implies that the goods gave off an intense smell both on 6 May and on 13 May, though they had been stored at very low temperatures since rejection on 5 May. It does not assist the defendant to observe that cheese tends to be smelly: neither Eurilait nor CL Surveys would be likely to comment on the smell unless it were significant and unusual.

2) The temperature of the cheese upon delivery, whether 10°C - 11°C or up to 17°C, was significantly higher than (a) was acceptable to Eurilait (see the endorsement on the CMR Consignment Note: paragraph 16 above), (b) the upper figure of 6°C provided on the packaging (as recorded by CL Surveys), (c) was provided for in the ATP Agreement, and (d) the 2°C - 4°C range stipulated in the instructions on the Transport Confirmation (paragraph 8 above). On 6 May, even after the refrigeration unit had been running in the Continuous Mode at 0°C - 2°C, the temperatures towards the rear of the trailer were significantly high, though the maximum recorded temperature was reduced to 8.9°C.

3) The cheese was rejected by Eurilait and sold for salvage. While not accepting the defendant's liability for the condition of the cheese, CWH & Co, acting on behalf of the defendant, acknowledged the merit of Eurilait's position and raised no objection to the sale of the cheese for melting (see paragraphs 25 and 26 above).

4) The fact that the cheese was sold for salvage and the price at which it was sold are some evidence of the fact of damage.

5) I mention also that on 6 May there were traces of condensation on a bag of brie near the trailer doors, which tends to indicate that the temperature of the cheese had varied significantly. This particular observation, given its date and the circumstances in which it was made, cannot be regarded as being of great independent significance; when taken with the evidence of abnormal temperatures and smell, however, it is relevant.

52. For the defendant, Mr Davidson submitted that the evidence of the condition of the goods on delivery was "flimsy" and did not suffice to discharge the burden of proof on the claimants to show that the cheese was in a substandard condition. It is true that it is possible to imagine more rigorous evidence of condition upon delivery than was adduced at trial. Nevertheless, the evidence that was adduced satisfies me on the balance of probabilities that the cheese was substandard when it arrived at Evercreech. It may be noted, further, that, even though Mr Fraser and the driver were clearly concerned by Eurilait's rejection of the cheese—Mr Fraser seems to me to have been positively alarmed—, they did not seek to challenge the reason given for the rejection, namely the high temperature of the cheese, but rather took immediate steps to try to remedy it. They also appear to have decided not to instruct a surveyor to inspect the cheese on the following day, despite being given opportunity to do so. Mr Davidson pointed to the facts, first, that the claimants sought to persuade LAÏTA to accept the cheese, thereby (it is said) acknowledging that the cheese was in acceptable condition, and, second, that the reasons given by Ms Guillerm for refusing to do so were inadequate. The former point does have some force; however, it is perhaps not surprising that the claimants attempted to avoid a position in which the goods were rejected, and it is notable that the matters they relied on concerned not the condition of the cheese but the readings from the data logger on the defendant's trailer. The latter point is not, in my view, of any real significance: Ms Guillerm was, of course, representing the sender, LAÏTA, not the consignee; and, although the balance of the expert evidence indicates that the 40-minute period of high temperatures on the ferry would not in itself cause damage to the cheese and that there were no negative temperatures during transit, she was responding to the claimants' observations concerning the data logger readings rather than dealing with the condition of the cheese.
53. As for the condition of the cheese upon loading in Ancenis, Mr Davidson observed that the evidence on this matter was again limited: there are only partial records of the production dates; there are no records of delivery dates for cheese produced off site; there are temperature readings for the two storerooms for only part of the period during which cheese might have been stored there; there are no records of the precise positions within those storerooms of the cheese that comprised the consignment; only a single pack (not pallet) of the cheese that comprised the consignment had had its temperature taken on arrival at LAÏTA's premises; and, most importantly, there are no records of the pulp temperature of the cheese at the time of loading. Mr Davidson also relied on the positive evidence of the temperatures recording in transit via the data logger on the defendant's vehicle; this, he said, was relevant not only to the explanation of the condition of the cheese upon arrival at Evercreech but to the inferences to be drawn as to its condition in Ancenis, because evidence that the refrigeration unit had worked satisfactorily might itself be evidence that cheese in a bad condition when it reached Evercreech had been in a bad condition when it left Ancenis.
54. Despite these points, in my judgment the evidence establishes on the balance of probabilities that the condition of the cheese upon delivery had deteriorated from what it was upon loading and that it had thereby suffered damage. (The question whether it had some inherent vice is a different question, properly kept distinct and addressed in connection with the second issue.) I refer in particular to the following points that have been mentioned above.

1) The available evidence tends to indicate that the cheese was young cheese. In the case of the Depot 442 Cheese, this is directly supported by the records of the dates of manufacture; it has not been suggested that these show anything amiss. Although production dates are unavailable for about one-third of the cheese, Ms Guillerm's evidence, which I accept, was that LAÏTA purchased only young cheese from third-party producers. The long Best Before dates for the cheese at least lend no support to any possibility that this practice was departed from on this occasion. Mr Wilkie argued that it was possible that some of the cheese in Depot 441 was in an unusually advanced state of maturity. If that possibility exists, it is at least not supported by the documents or by any observation made at the point of loading.

2) Such evidence as there is tends to indicate that the cheese was stored properly by LAÏTA. It is correct to note that, as I have mentioned, the records have not been demonstrated to cover the entire period for which cheese might have been kept in the cheese store or the warehouse; it has not, however, been demonstrated that they do not cover that entire period, and I have no reason to suppose that full and proper disclosure has not been made of all documents within the scope of standard disclosure. Mr Wilkie accepted that there was nothing in the disclosed temperature records to indicate that the cheese was not stored at a correct temperature, though he did complain about the extent of the records.

3) The documentation is properly to be considered in conjunction with the evidence of LAÏTA's standard practices and procedures (see paragraphs 4 and 5 above). These do not prove "conclusively" (Mr Davidson's word: paragraph 11 of his written submissions) that cheese delivered from other sites was not received at an excessive temperature or that cheese stored in the cheese store or in the warehouse was never stored at excessive temperatures; however, they tend to indicate that it is unlikely that it was. (This point is primarily relevant to questions of causation, but it also has some relevance to the question whether the physical condition of the cheese changed for the worse between Ancenis and Evercreech.)

4) The inference that the cheese was in good condition when loaded gains some support from the lack of any recorded reservation in that regard, whether in the CMR Consignment Note or otherwise. The CMR Consignment Note made clear that loading and stowage were the responsibility of LAÏTA, not the defendant, and that the defendant's driver did not have opportunity "to check and control the actual loading of the trailer". In my view, that is not a "specific reservation" as to the apparent condition of the goods and their packaging for the purposes of article 9, paragraph 2, of the Convention; it says nothing about the condition of the goods or the packaging and does not state or imply that the driver had no ability to discharge the defendant's obligation under article 8 to check the apparent condition of the goods and their packaging or, indeed, that he did not do so. Therefore, by reason of article 9, paragraph 2, "it shall be presumed, unless the contrary is proved, that the goods and their packaging appeared to be in good condition when the carrier took them". This presumption is limited: it goes only to the apparent good condition of the goods, and it is rebuttable. But, in the absence of contrary evidence, it seems to me to justify the conclusion that the cheese gave off no unusual or unusually strong smell. (I can also infer that the cheese appeared normal to appearance and did not give off unusual heat sensible to the observer. This does not seem to me to be significant, however, as one would expect nothing else.) The two French consignment notes (paragraph 13 above) lend some further support to these inferences and to the inference that the cheese was at a proper temperature, though I bear in mind that it is at least possible that the driver did not understand the text and that anyway he would not have taken the temperature of the cheese.

55. Accordingly, it is probable and I find that during the period between the loading of the cheese onto the defendant's vehicle and its delivery at Evercreech the cheese underwent damage, in the sense of physical deterioration that resulted in diminution of value.
56. Up to this point, however, I have referred loosely to damage to "the cheese". It is necessary, however, to be more precise as to how much cheese was damaged, for two reasons: first, Article 25, paragraph 2(b), limits compensation where only part of the load has been damaged; second, the extent of damage is potentially relevant to identifying the cause of the damage. It is here that the limited nature of the evidence is especially problematic. The only temperatures taken in the UK were those at Eurilait's premises (which are recorded very briefly and without adequate explanations and are not supported by witness evidence at trial) and those taken on 6 May and 13 May and set out in the CL Survey report. The latter readings came after the storage conditions had altered and after lesser or greater lapses of time since delivery at Evercreech. There is little evidence to inform a conclusion as to the impact that either of these factors would have had on the temperature of the cheese. Further, there is an element of circularity in the problem, because depending on the cause of the high temperatures recorded at Evercreech the changed storage conditions may have had different effects on the temperatures of the cheeses at various points in the trailer.
57. That being the state of the evidence, I am satisfied on the balance of probabilities that the pallet nearest the trailer doors was damaged and that the pallets on the left side of the row immediately in front of that pallet were also damaged. Mr Davidson suggests that this amounts to five pallets in total; that seems to be right, though there is no detailed survey of the stowage.
58. Later in this judgment, I shall say something more about the position of the five pallets I have identified as being damaged. I shall also comment on the cause of any damage to the remaining 33 pallets, if (contrary to my finding) they were indeed damaged.

(2) Is the defendant liable for the damage?

59. I have set out the relevant parts of the CMR Convention in paragraph 30 above. The logic of the (somewhat convoluted) provisions is, on the facts of this case, as follows:

The defendant is prima facie liable for the damage to the cheese: Article 17, paragraph 1.

The defendant will be relieved of liability if it proves that the damage was caused by inherent vice of the goods: Article 17, paragraph 2; Article 18, paragraph 1.

If the defendant proves (i) that the damage could be attributed to special risks inherent in the nature of the goods, which particularly exposed them to damage through decay, (ii) that all steps incumbent on it in the circumstances with respect to the maintenance and use of the trailer's equipment to protect the goods from the effects of heat were taken, and (iii) that it complied with any special instructions issued to it, a rebuttable presumption will arise that the damage was so caused: Article 17, paragraph 4(d); Article 18, paragraphs 2 and 4.

If the defendant proves that the damage could be attributed to special risks inherent in the stowage of the goods, a rebuttable presumption will arise that the damage was so caused: Article 17, paragraph 4(c); Article 18, paragraph 2.

60. The defendant cannot rely directly on Article 17, paragraph 4(d). It has not proved that it took all steps incumbent on it with respect to the maintenance of the trailer: the fact that the trailer did not have a valid ATP Certificate tends, indeed, to show to the contrary that the defendant did not take all steps incumbent on it in that regard. Further, as the defendant concedes that it did not run the refrigeration unit in continuous mode in accordance with the instructions in the Transport Confirmation, it cannot show that it complied with special instructions issued to it. Although the point strictly arises later in the analysis, I say now that I reject the ingenious suggestion that the defendant can rely indirectly on Article 17, paragraph 4(d) because the claimants could have relied on that provision to defeat a claim by LAÏTA for compensation. Whatever other objections may exist to defeat that suggestion, it seems to me that the claimants would have failed to prove the matters required by Article 18, paragraph 4: in particular, they could not have proved that they had taken all steps incumbent on them in the circumstances with respect to the maintenance of the trailer or indeed, on their own case, the use of it.
61. Therefore, in respect of this second question, I am concerned only with relief from liability on the ground of inherent vice (Article 17, paragraph 2) or on the ground of a special risk inherent in the stowage of the goods (Article 17, paragraph 4(c)). It is therefore convenient to turn to the factual question: why did the cheese overheat?
62. In broad terms, on the facts of this case there are two kinds of possible cause: first, there was something wrong with the cheese when it was loaded; second, something went amiss with the cooling of the cheese in the trailer. As to the second kind of possible cause, the problem could be of two kinds: incorrect use of the "Start/Stop" mode; and poor circulation of cold air within the trailer. And poor circulation of cold air could the result of either or both of two matters: incorrect configuration of the trailer; and inappropriate stowage of the load.
63. As regards the cheese, there is no direct evidence either that the cheese was in an advanced state of maturity or that it was too warm when it was loaded. The evidence relating to the cheese and its storage tends to indicate, to the contrary, that it was young cheese, received (in the case of the Depot 441 Cheese) and stored at appropriate temperatures. A finding that, nevertheless, the damage was caused by the condition of the cheese on loading would be an inference from a number of matters: Mr Fraser's evidence was that this trailer had carried cheese from Ancenis to Evercreech on numerous previous occasions, always operating the refrigeration unit in the Start/Stop mode, always without damage to the cheese; the data logger shows that the Start/Stop mode operated correctly and that the refrigeration unit maintained appropriate temperatures; if poor circulation of cold air is not a credible explanation for the temperature of the cheese when it reached Evercreech, the explanation must lie in the condition of the cheese on loading.
64. For reasons set out below, I do not consider that it is necessary or appropriate to have recourse to any explanation that involves the condition of the cheese on loading.
65. As regards the operation of the refrigeration unit, Mr Lawton's primary argument concerned the use of the Start/Stop mode instead of the Continuous mode. For the reasons set out below, I am not persuaded that the use of that operating mode is a sufficient or the primary explanation for the damage to the cheese, though I consider that it was a contributory factor.

1) I accept Mr Fraser's evidence that the same delivery journey was made on a weekly basis for several years and that the unit always ran in Start/Stop mode. There was nothing unusual about the way in which the unit was operated on this occasion. Yet some of the cheese was damaged.

2) Mr Lawton was unable to point to any guidance, instructions or literature to support his contention that the Start/Stop mode is generally unsuitable for cheese. The manufacturer's guide did not support the contention.

3) Mr Lawton produced no objective support for his assertion that cheese is a "temperature sensitive" cargo that requires the constant temperatures provided by the Continuous mode, and the assertion is in my view implausible. When pressed to justify the assertion, Mr Lawton said that the temperature-sensitivity of cheese was demonstrated by the damage that occurred in the present case; that is a circular argument. I accept that some goods, such as some pharmaceutical products and some foodstuffs, have to be carried and stored within very narrow temperature ranges. However, the evidence shows that the acceptable temperature range for the storage and carriage of cheese is at least 4°C and maybe as much as 6°C. The evidence does not support the contention that such a range may not be maintained by the Start/Stop mode.

4) However, it seems to me that there is a significant difference between the adequacy of the Start/Stop mode to maintain an acceptable range of temperatures in conditions of good air circulation from its adequacy for that purpose where the air is not circulating properly. I come back to this point below.

66. All of the other explanations in terms of inadequate cooling posit that the cool air introduced by the unit into the trailer did not circulate sufficiently to cool the load properly. Three particular matters might support some such explanation: first, the fact that poor air circulation is a recognised cause of inadequate cooling of loads; second, the fact that there is a no direct evidence either that there was anything wrong with the cheese or (as I think) that there was anything wrong with the operation of the refrigeration unit; third, the evidence of the temperatures of the cheese. It is probable that the high temperature recorded upon delivery at Evercreech was taken at the rear of the trailer, near the doors. The readings taken by CL Surveys on the following day (paragraph 18 above) indicate that the higher temperatures were towards the rear of the trailer, and that the cheese near the forward bulkhead was cooler. The pattern suggests, though it does not conclusively demonstrate, that the temperatures were determined by the position of the various pallets within the trailer.
67. Mr Lawton pointed to the lack of baffle plates over the return air vents as a cause of circulation bypass or "short-circuiting" of the airflow, which could have been but was not mitigated by the installation of an air sock. He did not, however, suggest that this was the principal cause of the damage; that, he thought, was the inappropriate use of the Start/Stop mode. Rather, it was a contributory factor, and he was unable to give an opinion as to the extent of the contribution. In my judgment, it is improbable that any contribution made by these matters was significant.

1) As Mr Lawton could not give an opinion as to the extent of any contribution made by the configuration to the overheating of the cheese, his evidence does not suffice to persuade me that it was significant.

2) This trailer had been making deliveries on a weekly basis for several years, yet this was the only recorded problem with overheating. There is no evidence to suggest that the configuration of the trailer was in any way different on this occasion; it was not suggested that it was. This makes it relatively unlikely that the overheating of the cheese was caused or materially contributed to by the configuration of the trailer.

These two observations suffice to justify my conclusion. There are additional points, however.

3) Mr Lawton did not positively contend that an air sock ought to have been used; his point, rather, was that it would have assisted air circulation: in particular, that it would have mitigated the bad effects of the missing baffles. (I shall say more about air socks in the context of stowage.)

4) I share Mr Wilkie's scepticism regarding the complaint that baffles were "missing". Mr Lawton could point to nothing in the trailer's operating manual to support his opinion that there should have been one or two additional rows of baffles to cover the outlet duct. He observed that the manufacturer of the trailer is not the manufacturer of the refrigeration unit, but that does not seem to me to assist his argument: one might still expect some generic instruction in the trailer's operating manual regarding the way in which the baffles might be configured to achieve their purposes; and Mr Lawton did not refer to any relevant provision in the operating manual for the refrigeration unit either. The experts were in agreement that the primary purpose of a false bulkhead—which had not been compromised in the present case—is to prevent the cargo pressing up against the forward bulkhead and thereby blocking the air return; it thereby allows a clear route for the warmed air in the trailer to return to the evaporator. (The joint memorandum referred to the return of "the cooling air", but that seems to be inaccurate. The cooling air is what is introduced into the trailer at the inlet duct. The air that returns ought in principle to be warmer than the air introduced, because it is will have removed heat entering the trailer from outside and heat produced by the load.)

5) Mr Lawton contended that a secondary purpose of a false bulkhead was to prevent the cool air short-circuiting and returning directly to the outlet duct. I accept that contention, but I do not accept that it has been demonstrated to have relevance to the circumstances of the present case. That a secondary purpose of a false bulkhead is the prevention of short-circuiting of cool air is supported by the one piece of learned literature cited by either expert: "Air Circulation Inside Refrigerated Semi-Trailers Transporting Fresh Produce", a 2001 M.Sc. thesis by Ka Po Catherine Hui. Ms Hui refers to various designs of false bulkhead and, at page 20, remarks that "solid/pressure bulkheads" were better for temperature management than "frame bulkheads" (both illustrated in her figure 2.11), because the latter might allow some inlet air to by-pass the load, whereas: "Solid or pressure bulkhead generates a pressure difference across the outlet and inlet of the fan". For reasons already indicated, this remark does not materially assist Mr Lawton's theory in practical terms. However, I do not think it lends it much theoretical support either. First, Mr Lawton makes no mention of pressure differentials; his argument is that the covering of the outlet duct itself bars what he called the path of least resistance. However, unless there is a pressure differential the baffles could make no difference, because air that short-circuited back to the baffles would not be circulating around the load at the rear of the trailer, regardless of whether it went straight out of the outlet duct. Second, Mr Lawton made no attempt to show that an extra row or two of baffles would create a pressure differential or what the practical significance of the differential might be. Third, confirmation that Ms Hui's thesis and Mr Lawton's evidence proceeded on different premises came when he was cross-examined about section 2.3.3.3 (page 20) of the thesis. Of the sentence, "The bulkhead can cover the full width and half the height of the front wall", Mr Lawton referred to what he took to be Ms Hui's assumption that an air sock would be used. If that means that false bulkheads are designed not to cover the outlet duct because the use of an air sock is presumed, it seems to undermine Mr Lawton's argument from the secondary purpose of the bulkhead. In fact, Ms Hui clearly understood that many trailers were not equipped with air socks, which are widely considered to have practical disadvantages. It seems more likely that her thesis indicates that the secondary purpose of the false bulkhead can be achieved by a half-height bulkhead and that the pressure differential is created by the air between the forward bulkhead and the false bulkhead that makes its way to the outlet duct. This is why the focus is not on the advantage of a false bulkhead that is high enough to cover the outlet duct but rather on the advantage of a solid over a frame (that is, lattice) design. This is very different from Mr Lawton's insistence that the false bulkhead should extend nearly to the top of the trailer. Fourth, in agreement with Mr Wilkie, I find implausible in the present case the short-circuiting mechanism posited by Mr Lawton, regardless of the position in other cases. The refrigeration unit worked by introducing a jet of air from the front of the trailer, directed towards the back. It is hard to see why, in normal circumstances, air thus introduced should short-circuit to the bulkhead in significant quantities (of course, it may well do so to some extent) rather than continuing towards the rear of the trailer, unless it met an obstruction, which might be either a solid obstruction such as a high load or a build-up of air pressure. The former kind of obstruction was not present in this case, and no one has suggested that the latter kind was present.

68. I turn to consider stowage and Article 17, paragraph 4(c). This was an alternative argument raised by Mr Wilkie in his first report, though not mentioned in his supplementary report, and it was defended by him in cross-examination. I have set out the reasoning as it appears in paragraphs 7.4 and 11.14 of the first report in paragraph 47 above. Mr Lawton disagreed with Mr Wilkie. He opined that the space towards the trailer doors would have aided airflow, not hindered it: I refer to sections 9.1 to 9.3 of his first report. Neither expert did much more than assert his opinion; no literature was referred to on this issue.
69. In these circumstances, and subject to the further remarks concerning the Start/Stop operating mode set out below, it would suffice to say that the defendant had established that the damage could be attributed to stowage, that the claimants had not proved that it was not in fact attributable to the stowage, and that therefore and by reason of Article 17, paragraph 4(c), and Article 18, paragraph 2, the defendant is relieved from liability. However, I would go further and find that a substantial cause of the damage was the manner of the stowage of the cheese by the sender, LAÏTA. My reasons are as follows:

1) There is no doubt that inappropriate stowage methods may lead to poor air circulation and thereby to poor temperature distribution. This fact does not establish that there was poor stowage in the present case, but it renders the possibility inherently plausible.

2) The explanation given by Mr Wilkie seems to me to make good sense. The cold air introduced at the front of the trailer had to make its way to the back, proceeding over the top of the load, before returning under and to some extent through the load to the front of the trailer. It was common ground between the experts that the air would tend to follow the line of least resistance. It seems to me that the air that reached the end of the pallets on the right of the trailer would be likely to return to the front before approaching the doors. This would be likely to result in the area adjacent to the rearmost single pallet on the left being relatively warm, because the cool air had not circulated through it. Moreover, the single pallet on the left was a low pallet compared to the higher ones that were forward of it. The cool air that went over the top of the pallets in front of it on the left would tend to some extent to return to the front as it came to the lower level of the final pallet, though because that air had not circulated around the single final pallet more of it might be expected to return through the load (as distinct from under the load) than would otherwise be the case.

3) As Mr Wilkie pointed out, this explanation accounts for the uneven temperature distribution in the trailer. The evidence indicates that the single final pallet was the warmest and that the load towards the front of the trailer was the coolest. But the evidence also seems to indicate that the ¼ trailer temperatures were higher on the left side (i.e. facing towards the front), which is what one would expect if the single final pallet caused a problem with air circulation, resulting in a warm area on the right and a warm pallet on the left. The ¼ trailer temperatures on the right, by contrast, had the benefit of proper circulation of cool air, though they were adjacent to a warm area.

4) I do not attach significance to the fact that Mr Wilkie did not propose this as the most likely cause of the warming and did not mention it in his second report. It clearly represented his continuing opinion, as he made clear in cross-examination. In my view, the reason why he did not make more of it in his evidence was that he became side-tracked by unprofitable speculations about missing temperature records and supposed breakdowns of system at Ancenis.

70. Nevertheless, I do not regard the stowage as the sole explanation of the damage to the cheese. For reasons I have indicated, I am satisfied that in conditions of adequate air circulation the Start/Stop mode would have been adequate for the storage of the cheese in the trailer. But it does not at all follow that it would be adequate in conditions of interrupted air flow. The intended effect of the Start/Stop mode is to ensure that, when the temperature in the trailer rises to the upper end of the permissible range, heat is removed by cool air until the temperature returns to the lower end of the range; thus there is oscillation but within acceptable parameters. However, if the air is not circulating properly, the efficacy of this control will be compromised. It is true that use of the Continuous mode does nothing to improve air circulation. However, it will tend to arrest the increases of temperature and thereby to mitigate, though it could not eliminate, the adverse effects of poor circulation.
71. In the present case, the temperature readings from the inlet probe show that the refrigeration unit itself was working properly. However, there were no readings from the load itself. The probe at the outlet duct would have given an indication of the adequacy of the air flow, because the temperature of the return air should be higher by reason of the ambient and cargo-generated heat it had removed; but the measurements from this probe are not available. Other evidence shows that insufficient heat was being removed from the rear of the trailer; this is indicative of air flow. However, the evidence from the CL Survey report shows that it was possible to remove more heat from the load at the rear, which indicates that although air flow was compromised it was still capable of cooling the load. Therefore the choice of operating mode cannot be dismissed as irrelevant on account of poor air flow. The actual reductions in temperature achieved in the period 5 – 6 May will doubtless have owed something to the initial period of running at 0°C. But, as explained above, it is reasonable to suppose that they also owed something to the use of the Continuous mode. No doubt that is why the defendant adopted it and maintained it after the set point had been returned to 2°C.
72. Accordingly, I find that the damage was caused by two factors: the manner of stowage, and the mode of operation of the refrigeration unit. Doing the best I can on the basis of the evidence before me, for the purpose of Article 17, paragraph 5, I assess that the contribution of these two factors to the damage was equal.
73. Finally, I should say something further about the greater part of the goods, which I have found were not damaged. If, contrary to that finding, the remainder of the goods was damaged, I consider that the damage was caused by a risk inherent in the stowage of the goods (and a fortiori that the claimants have not proved that it is not attributable to such a risk). This conclusion does not reflect any express opinion of the experts or submission of counsel, but it seems to me to follow from the evidence and submissions and I shall explain it briefly.
74. When discussing the merits of air socks, the experts disagreed as to the relevance of the height of the load. Mr Lawton said that an air sock was suitable with a low load, whereas a high load acted as its own directional tunnel to take air to the rear of the trailer. Mr Wilkie said, to the contrary, that a low load had the same effect as an air sock and that one would only use an air sock with a high load. I accept Mr Lawton's opinion on that point, because it seems to me to be clear that the air will be more channelled to the rear of the trailer if it passes through a small space between the top of the load and the top of the trailer. However, Mr Lawton's own argument about air circulation would, in this light, tend to point to stowage as the problem, rather than anything for which the defendant was liable. Mr Lawton did not opine or prove that the defendant ought to have used an air sock; he said that it would have improved air circulation, particularly in the context of the incorrect configuration of the false bulkhead. (I have dealt with these matters.) In fact, this trailer (like many others, as Ms Hui's thesis made clear) was not fitted with an air sock. The photographs in the CL Survey report show that the load was nevertheless stowed to no more than half of the permissible height. In those circumstances, any damage resulting from poor circulation of the cooling air, other than in respect of the arrangement of pallets at the rear of the trailer, would properly be considered as the result of the stowage, which ex hypothesi compromised the flow of air to the rear of the trailer. Accordingly, the defendant would be relieved from liability under Article 17, paragraph 4(c).

Conclusion

75. In the light of my findings, the defendant is liable to the extent of one half of the modest amount of damage that I have found proved but is not liable for the other half.
76. This judgment is being handed down at a hearing in the absence of the parties. As they have been unable as yet to agree what are the appropriate terms of order, I shall adjourn the hearing part-heard to a later date so that counsel may be heard.