England and Wales High Court (Chancery Division) Decisions

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION

		Royal Courts of Justice Strand, London, WC2A 2LL
		3 April 2009

MR JUSTICE ARNOLD :

Contents

Topic	Paragraphs
Introduction	1-4
Secrecy	5-6
The parties	7-18
The Claimants	7-9
The Defendants	10-18
Dr Skovmand and IIC	19-27
The databases	28-33
The Danish proceedings	34-46
Copyright claims	36-38
Trade secrets claims	39-41
Contract claim	42-43
Criminal investigation	44
IIC claim	45-46
Other proceedings	47-50
France	48
India	49-50
Procedural history of this claim	51-79
The witnesses	80-102
VF's witnesses	81-94
Mr Vestergaard Frandsen	81
Finn Kjaergaard	82
Phan Quy Chi (Chi Phan)	83
Dr Helen Jamet	84
Sicco Roorda	85
Dr Robin Medenwaldt	86
Bruno Ducoulombier, Nicholas Rose and Jørn Vestergaard-Jensen	87
Mr Howe	88-94
The Defendants' witnesses	95-102
Dr Skovmand	95-96
Mr Larsen	97-98
Mrs Sig	99
Other witnesses	100-102
VF's motivation for bringing the claim	103-111
My approach to finding the facts	112-115
The facts	116-621
Introduction to the technology	116-134
Polymer technology	117-124
Mosquito control	125-134
Linear regression analysis	135-140
VF's development of ZeroFly, Fence and PE bed nets	141-304
The terms of Mr Larsen's employment by VF	305-307
The relationship between VF and Dr Skovmand IIC	308-335
Contract?	309-311
Express term?	312-327
Other matters	328-335
The information contained in the Fence database	336-355
Samples 7-20	339-345
65F.0503.5 and 65F.0304.5 recipes	346
Other information	347-355
The [REDACTED] formulation	356-357
The Netprotect database	358-360
The development of Netprotect	361-533
The source of the first Netprotect recipes	537-621
Dr Skovmand's story	541-543
The "preliminary formulations"	544-546
The Ciba recommendations	547-553
Other sources	554
The odd numbers	555-573
Dr Skovmand's "mathematical protocol"	574-599
The seven improbabilities.	600-609
Dr Skovmand's "search pattern"	610-611
Other points	612-617
Conclusion	618-621
Applicable law	622
Legal analysis	623-670
The necessary quality of confidence	626-627
Circumstances importing an obligation of confidence	628-642
Implied term	629-641
Equitable obligation	642
The scope of the obligation of confidence: the Ciba recommendations	643-647
The scope of the obligation of confidence: trade secrets	648-660
The nature of the work	653
The nature of the information itself	654
The attitude of the employer at the time	655
Steps taken to protect the information	656
The separability of the information	657
The commercial value of the information	658
The usage and practices of the trade	659
Conclusion	660
Misuse	669-670
Remedies	671-672
Conclusion	673
Appendix 1: The Extract DB documents	674-711
Appendix 2: Dr Skovmand's evidence to the Danish court	712-718

Introduction

1. This is a claim for misuse of confidential information consisting of technical trade secrets. The Claimants manufacture and sell long-lasting insecticidal mosquito nets ("LLINs"), in particular a polyester ("PES") net marketed under the trade mark PermaNet which has been commercially very successful. The Defendants are two ex-employees of the Claimants, and companies associated with them, who have set up in competition with the Claimants. The Defendants developed a polyethylene ("PE") LLIN called Netprotect, which is marketed by Syngenta under the trade mark Icon Life.
2. The Claimants allege that in developing Netprotect the Defendants misused information contained in a database which records development work undertaken for the Claimants and thus, it is said, is confidential to them. This database is referred to by the Claimants as the PE Net database and by the Defendants as the Hegn or Fence database (hegn being Danish for fence). I shall refer to it as the Fence database since that is what it is mainly called in the contemporaneous documents.
3. The Defendants' primary defence is a denial that any information from the Fence database was used to develop their product. They also dispute that the information in the Fence database was subject to an obligation of confidence in favour of the Claimants anyway. At one stage the Defendants appeared to be contending that the information in the Fence database which is alleged to have been misused did not have the necessary quality of confidence, but no such contention was advanced at trial. It will nevertheless be necessary for me to consider the quality of the information, since the Defendants do contend that, to the extent that any information was used, it did not amount to a protectable trade secret as opposed to knowledge and experience that an employee or consultant is entitled to use after the termination of his employment or consultancy.
4. Mosquito nets for human use were frequently referred by the parties, both in contemporaneous documents and at trial, as "bed nets". I shall do likewise.

Secrecy

5. As I have already said, this claim concerns confidential information. Both sides were concerned to protect their confidential information from the other side and from third parties. As a result, a confidentiality regime was put in place at an early stage and a number of the interim hearings in these proceedings were held in secret. At trial, counsel for the Defendants very properly pointed out that the hearing should be in public save insofar as was necessary to preserve alleged trade secrets. In the result, however, it was agreed that large parts of the trial should be heard in secret. In addition, by consent I made an interim order under CPR rr. 31.22 and 32.13 restricting use of confidential disclosure documents and witness statements. I shall deliver this judgment in the first instance in secret, and then invite submissions as to what parts, if any, should be redacted before a public version is released.
6. The Defendants have periodically complained that the pre-trial confidentiality regime hampered them in the preparation of their defence, in particular because they were only able to inspect documents claimed to be confidential by the Claimants at the Defendants' solicitors' offices in London. I accept that this made preparation of the case more difficult than it would otherwise have been, but my sympathy for the Defendants on this score is limited for three reasons. First, the confidentiality regime was bilateral, so that the Claimants were only able to inspect the Defendants' confidential documents at the Claimants' solicitors' offices in London. Secondly, the Defendants did not apply for the regime to be relaxed in this respect. Thirdly and most importantly, the Defendants had independent access to the key documents in the case, and in particular the databases.

The parties

The Claimants

7. The Claimants, who I shall refer to collectively as "VF", are part of a group of companies which was founded in Denmark in 1957. In 2005 the group moved its centre of operations from Denmark to Switzerland. The Third Claimant, Disease Control Textiles SA, is a Swiss company which is now the holding company of the group. The Second Claimant, Vestergaard Frandsen SA, is a Swiss company which is a wholly-owned subsidiary of the Third Claimant. Between them the Second and Third Claimants own all of the intellectual property rights relating to the group's products. The First Claimant, previously known as Vestergaard Frandsen A/S and now called MVF 3 Aps, is a Danish company which was the parent company of the group prior to 2005.
8. The First Claimant was founded by Kai Vestergaard Frandsen. His son, Torben Vestergaard Frandsen, started working for the company in 1967 and took over the running of it in 1970 or 1971. Until the late 1980s, VF were primarily engaged in selling corporate work wear, such as uniforms for hotel and supermarket staff, and fabric for such work wear. In about 1986, VF started to diversify into emergency relief products, initially selling refuge blankets. In about 1993-94 VF began to manufacture and sell PE tarpaulins for use in refugee camps. At about the same time, VF began to diversify into disease control products, beginning with a tse-tse fly trap which was made from a woven PES fabric impregnated with deltamethrin insecticide. The clothing division was sold in about 1997.
9. In 1998 Torben Vestergaard Frandsen handed over the running of VF to his son Mikkel Vestergaard Frandsen, with whom he entered into an agreement giving the latter the option to purchase his shares. Mikkel Vestergaard Frandsen exercised this option in 2005. Since 1998 Torben Vestergaard Frandsen has been VF's Director of Development. Although he is not a scientist, he has enough knowledge and experience to conceive new ideas for products. In the remainder of this judgment references to "Mr Vestergaard Frandsen" are references to Torben.

The Defendants

10. The Fifth Defendant, Torben Holm Larsen, has a degree in Chemical Engineering. Subsequently he had a number of jobs in the textile processing field, including one with Ciba Speciality Chemicals ("Ciba"). From 1 November 2000 he was employed as VF's Head of Production pursuant to a contract dated 28 September 2000 which was later replaced by a contract dated 3 November 2002.
11. Mr Larsen gave notice of his resignation on 18 May 2004. He worked out the first half of his notice period. In mid July 2004 he asked to be relieved of his duties, and VF agreed to this. His employment terminated on 31 August 2004. His contract of employment contained a post-termination covenant against competing with VF lasting 12 months.
12. The Sixth Defendant, Trine Sig, was employed by VF from 1 November 2000, initially as a sales and marketing assistant and later as Regional Sales Manager for Europe and Latin America. She handed in her notice on 17 June 2004, and it was agreed that her employment should terminate on the same day.
13. Shortly before Mrs Sig left VF, she and Mr Larsen set up a company called Pure Culture I/S to import handicrafts from the Far East. On 3 August 2004 Mr Larsen and Mrs Sig incorporated a Danish company called Intection A/S, later Intection ApS ("Intection"), to compete with VF in the field of LLINs, although as I shall relate below the project was already under way by that date. The principal shareholders were Hans Johnsen, Mrs Sig's father, and Ole Ringberg-Larsen, a professional investor. Mrs Sig was the sole director. Intection ceased to trade in about October 2005 after proceedings had been brought against it by VF in Denmark as related below.
14. The corporate Defendants are all English companies which were set up at about the time that Intection ceased trading. The First Defendant, Bestnet Europe Ltd ("Bestnet"), is the principal corporate defendant. Its shareholders are Mr Johnsen's and Mr Ringberg-Larsen's corporate vehicles together with Intelligent Insect Control SARL ("IIC"). Mr Larsen and Mrs Sig were appointed as directors of Bestnet in 2008.
15. The Second Defendant, Intection Ltd, was originally set up to perform the role now performed by Bestnet. Its activities, if any, were very short-lived. It was joined in the Danish proceedings. It is now in liquidation and the claim against it is stayed except for the purpose of giving disclosure.
16. The Third Defendant, 3T Europe Ltd ("3T"), is the corporate vehicle through which Mr Larsen and Mrs Sig provide their services to Bestnet.
17. The Fourth Defendant, Intelligent Insect Control Ltd, is a company owned by Dr Ole Skovmand, his wife and two relatives of theirs. It does not appear to have traded.
18. In this judgment I shall refer generally to "the Defendants" without distinguishing between them unless it is necessary to do so. For the reasons that I shall now explain, when I am discussing the development of Netprotect, references to "the Defendants" include Dr Skovmand and IIC unless the context indicates to the contrary.

Dr Skovmand and IIC

19. Although not formally parties to the claim, Dr Skovmand and IIC are central to it. Through IIC, Dr Skovmand is a 20% shareholder in Bestnet. Dr Skovmand was primarily responsible for the development of both VF's PE products and the Defendants' Netprotect product, although Mr Larsen was also involved in both developments
20. Dr Skovmand has an MSc in neurobiology and a PhD in neuro-acoustics. After his PhD, he worked for six years for the Danish Pest Infestation Laboratory evaluating and testing insecticides on flies and mosquitoes. His work included testing insecticides incorporated in polymers, such as plastic ear tags. From 1988 to 1995 he was employed by Novo Nordisk's bio-control laboratory running bioassays and developing pesticide products.
21. In 1995 Dr Skovmand moved to Montpellier, where he has been based ever since. From 1995 to 1998 he was employed by the Institut de Recherche pour le Développement ("IRD") in Montpellier. During this period he worked on formulating a sustained release product for mosquito larval control. This project was financed by the Danish International Development Agency (DANIDA) and others.
22. In 1998 Dr Skovmand left IRD and started his own consultancy. At around the same time, he started to advise VF on an occasional, unpaid basis. In late 1998 VF engaged Dr Skovmand as a paid consultant. From about April 1999 onwards he was engaged by VF to develop its PES bed net, and subsequently other products. At some point, he set up IIC as the corporate vehicle through which his services were provided. From about July 2001 Dr Skovmand provided his services to VF through IIC. To begin with IIC consisted of just Dr Skovmand and his secretary, Ms Natera, but in 2002 it also employed Rob Fleuren, an agronomist. (Subsequently, in 2003, Mr Fleuren became a consultant directly to VF.) IIC became an SAS in 2003 and reverted to being an SARL in 2006.
23. It is common ground that from April 1999 until he resigned on 7 December 2004 Dr Skovmand spent a significant proportion of his time working for VF, although what proportion is disputed. The relationship was not governed by any written agreement and there is a considerable dispute about the nature of the relationship.
24. On 28 January 2005 Dr Skovmand through IIC entered into a written contract with VF for the provision of consultancy services relating to two specific projects, PermaNet 2.5 and ZeroFly. This contract was terminated by Dr Skovmand on 28 April 2005, although Dr Skovmand continued to do some work for VF until the autumn of 2005. This contract, and the work done under it, has little or no bearing on the issues in the present case.
25. During the course of these proceedings VF have periodically complained that Dr Skovmand and IIC have exploited the fact that, as explained below, Dr Skovmand and IIC were not joined as defendants by giving selective disclosure of documents in Dr Skovmand's possession relating to the development of Netprotect which are not in the possession, custody or control of any of the Defendants. The Defendants riposted that it was VF's decision not to join Dr Skovmand and IIC as defendants and they must live with the consequences so far as disclosure is concerned. I agree with this. Moreover, as I pointed out in an earlier judgment in these proceedings, VF did not apply for an order for third party disclosure against either Dr Skovmand or IIC, nor did they serve a witness summons on Dr Skovmand requiring him to produce documents.
26. Nevertheless, it is fair to say that, as I shall explain in more detail below, Dr Skovmand did produce documents piecemeal. Certain documents were produced by him during the pre-trial stages of the case. He produced additional documents not only on the eve of trial, but also during the course of and immediately after the adjournment hearings discussed below and even during the course of giving evidence. Dr Skovmand explained that he was very reluctant to produce any more documents (or parts of documents) than were necessary to defend the claim because he was concerned at the possibility of VF misusing information confidential to him or IIC which was contained in them. I understand that concern. Given the confidentiality regime that was in place, however, I consider that Dr Skovmand was unduly sensitive to the risk of misuse of confidential information. The information he was most concerned about could have been dealt with by redaction, as it ultimately was. Furthermore, I do not accept that this concern justified his piecemeal approach to disclosure. This undoubtedly complicated and prolonged the trial.
27. The upshot, for which blame is attributable to both sides, is that even now I am not satisfied that Dr Skovmand has disclosed all documents in his possession relevant to the history of the development of Netprotect. This has made the task of fact finding considerably more difficult than it would have been had fuller disclosure been given at an earlier stage.

The databases

28. Also central to the claim are two databases compiled by Dr Skovmand with the assistance at least in the first case of his personal assistant Valerie Natera. The first is the Fence database already referred to. This contains formulation data, test results and other information relating to PE net products on which Dr Skovmand worked for VF. The second is referred to as the Netprotect database. It contains formulation data, test results and other information relating to products on which Dr Skovmand worked for the Defendants, and in particular Netprotect. It was and remains Dr Skovmand's practice to use such databases as working tools.
29. Both databases were created by Dr Skovmand in the form of Lotus 123 spreadsheets. For the purposes of this case they have been converted into Microsoft Excel spreadsheets. Generally, each database consists of a number of tabs, usually three or four. The spreadsheets contain a considerable number of columns and rows. While it is possible to print them out on paper, the paper copies are unwieldy. They are best appreciated in electronic form.
30. In terms of their format, the two databases are very similar. In itself, there is nothing untoward in this.
31. Because the databases were created and maintained electronically, it was a relatively simple matter for Dr Skovmand to alter their structures periodically as well as adding to their contents. Unsurprisingly, he was not always consistent in the way he did this. This can be seen, for example, where formulae within the spreadsheets link to incorrect cells. In addition there are sometimes inconsistencies in the data, for example in the units used.
32. It was also a simple matter for Dr Skovmand to make complete or partial copies of the databases and to transmit such copies to others by email or memory stick. Dr Skovmand's evidence was that it was his practice not to provide others with complete copies, that is to say, copies which included a sheet containing his comments; but it appears that he did provide copies which were complete apart from his comments from time to time as well as more limited partial copies. As explained in more detail below, Dr Skovmand sent or gave copies of the Fence database to VF employees such as Mr Larsen on a number of occasions.
33. As I have said, Dr Skovmand was assisted in maintaining the Fence database by Ms Natera. Once the database was established, it was mainly Ms Natera who entered data into the database. Dr Skovmand accepted that VF paid for Ms Natera's time spent in doing this.

The Danish proceedings

34. The present claim is a sequel to proceedings between some of the same parties in Denmark.
35. On 16 December 2004 VF's Danish lawyers sent letters before action to Mr Larsen, Mrs Sig and the directors of Intection. This led to a meeting to discuss settlement in January 2005, but no agreement was reached.

Copyright claims

36. On 17 February 2005 VF applied to the Enforcement Courts in Kolding and Esbjerg respectively for search and seizure orders against Intection and Mr Larsen personally alleging that during the period February to August 2004 Mr Larsen had made copies of VF's documents for use in Intection's business. The orders were granted on 17 March 2005 and resulted in the copying by Court-appointed experts from KPMG of some 56,000 files copies of which were found on CDs in Mr Larsen's apartment, including a copy of the Fence database. The Defendants say that Mr Larsen had legitimately made back-up copies of all the files on his laptop computer, which included personal files as well as VF files, immediately before returning the laptop to VF at the end of July 2004. (It appears that one of the CDs is dated 30 August 2004, which is around the time that Mr Larsen returned a second laptop to VF. I presume that the Defendants would advance a similar explanation in respect of this.)
37. On 31 May 2005 the Enforcement Court made an order that part of the recovered material should be disclosed to VF. On 6 September 2005 the Enforcement Court made an order that the remainder of the material should be disclosed to VF. Mr Larsen appealed against the order dated 6 September 2005 to the Western Division of the Danish High Court. On 2 December 2005 IIC intervened in this appeal, but subsequently it withdrew. On 5 July 2006 the Western Division of the Danish High Court allowed the appeal.
38. In the meantime VF commenced proceedings on the merits for copyright infringement against Intection and Mr Larsen before the Civil Courts in Kolding and Esbjerg respectively. On 15 March 2006 VF obtained judgment on its claim against Intection, which was undefended. The claim against Mr Larsen is still pending.

Trade secrets claims

39. On 15 June 2005 VF applied to the Enforcement Court in Kolding for an interim injunction against Intection to restrain misuse of trade secrets. On 31 October 2005, the day before the hearing of that application, Mrs Sig resigned as a director of Intection, leaving it with no director. On 1 November 2005 the injunction was granted by default.
40. On 15 November 2005 VF commenced proceedings on the merits for misuse of trade secrets against Intection before the Maritime and Commercial Court in Copenhagen. The Third Defendant was subsequently joined to the proceedings.
41. On 30 January 2006 the Bankruptcy Court in Kolding issued a bankruptcy order against Intection. The trustee in bankruptcy did not defend the Copenhagen proceedings, and the Court entered judgment against Intection on 1 September 2006. The Court granted judgment in default against the Third Defendant on 11 October 2006. The money part of the judgment against the Third Defendant was registered with the High Court on 30 April 2007. VF subsequently served a statutory demand and winding up petition against the Third Defendant which was wound up on 17 October 2007.

Contract claim

42. On 15 June 2005 VF applied to the Enforcement Court in Esbjerg for an interim injunction against Mr Larsen for breach of his non-competition covenant. Mr Larsen did not defend this application and so an injunction was granted on 5 August 2005 until 31 August 2005.
43. On 19 August 2005 VF commenced proceedings on the merits for breach of his non-competition covenant against Mr Larsen before the Civil Court in Esbjerg. In a judgment dated 22 February 2007, the Court held that Mr Larsen had made one persistent act of competition in breach of the covenant but dismissed other claims by VF.

Criminal investigation

44. In September 2005 VF made a complaint to the Danish police regarding Mr Larsen's alleged misuse of trade secrets and copyright infringement. The police investigated but dropped the matter due to insufficient evidence. VF complained to the District Attorney, as a result of which the investigation was resumed.

IIC claim

45. On 6 March 2006 IIC brought proceedings against VF in respect of unpaid invoices before the Western High Court in Viborg. VF disputes that the sums claimed are due and has counterclaimed against both IIC and Dr Skovmand for misuse of trade secrets.
46. On 7 June 2006 IIC applied for a search and seizure order against VF before the Enforcement Court in Copenhagen. The application was later withdrawn.

Other proceedings

47. In addition to this claim and the Danish proceedings, VF have brought two other sets of proceedings.

France

48. On 8 March 2007 VF obtained a search and seizure order against IIC and Dr Skovmand from the President of the Tribunal de Grande Instance in Montpellier. On 17 April 2007 IIC and Dr Skovmand applied to set the order aside, but the application was dismissed on 14 June 2007. IIC and Dr Skovmand appealed against the order of 14 June 2007, but the appeal was dismissed by the Cour d'Appel of Montpellier on 13 March 2008. In the meantime VF applied on 3 July 2007 for the appointment of a Court expert, and a panel of two experts was appointed on 27 September 2007. These experts are reviewing the materials seized on the execution of the search and seizure order. On 19 August 2008 VF commenced proceedings on the merits for unfair competition against IIC and Dr Skovmand before the Tribunal de Grande Instance.

India

49. On 30 March 2007 VF commenced proceedings in the Delhi High Court against all the present Defendants, IIC, Dr Skovmand, Mr M. Sivasamy trading as Siva Enterprises/Sivalakshmi Plastics (Bestnet's net manufacturer), Alok Masterbatches Ltd (Bestnet's masterbatch producer, which appears also to trade as Alok Industries, "Alok") and Biotech International. On 2 April 2007 VF obtained a search and seizure order against the three Indian defendants. VF claim that Mr Sivasamy and Alok failed to comply with this order and have commenced proceedings for contempt of court against them. On 15 May 2007 VF obtained an interim injunction against all the defendants in the Indian proceedings. VF claim that the Defendants have breached the injunction and have commenced proceedings for contempt of court against them as well. The Indian defendants have applied to discharge the injunction and that application is pending, although VF have settled with Biotech International.
50. On 17 October 2008 VF commenced proceedings against Syngenta, Syngenta India Ltd, its managing director and ATM Knits in the Delhi High Court. On 22 October 2008 VF obtained an interim injunction against those parties. VF claim that ATM Knits is another trading name of, or a company owned by, Mr Sivasamy.

Procedural history of this claim

51. For reasons that will appear, I must describe certain aspects of the procedural history of this claim in some detail.
52. The Claim Form was issued on 19 December 2006. In addition to the six extant Defendants, Dr Skovmand and IIC were named as Defendants. On the same day VF filed a without notice application seeking early specific disclosure of certain categories of documents.
53. On 8 January 2007 VF filed Amended Particulars of Claim confining their claim to the present Defendants. I was told by counsel for VF that the claim against Dr Skovmand and IIC was not pursued because of a concern about duplication of the proceedings which were intended to be commenced by VF in France against Dr Skovmand and IIC. The Amended Particulars of Claim alleged that four categories of information were confidential to VF, namely (i) the Fence database as at about July 2004, (ii) another database, (iii) three masterbatch recipes and (iv) two Fence recipes, namely those referred to as 65.F.0304.5 and 65.F.0503.5.
54. On 11 January 2007 Kitchin J made an order on VF's without notice application requiring the Defendants by 1 March 2007 to provide to VF copies of the following categories of documents in so far as they were in their possession, custody or control:

"(i)     instructions to all parties involved in the manufacture of the master batches and/or the product in respect of each of the following products:

(a)     the product submitted to [IRD] for testing, resulting in the report entitled 'Phase I evaluation of LLINs from In-tection' dated May 2005…

(b)     the product launched as 'Netprotect' at the conference in Cameroon in November 2005…

(c)     the products sent to Insight Pharm [PanAsia] in Thailand in March 2006…; and

(d)     the product licensed to Syngenta Crop Protection and sold as IconLife…;

     and

(ii)     instructions to any institutions or individuals involved in testing each such product."

55. The Order dated 11 January 2007 also established a confidentiality regime in respect of the confidential schedules to the Particulars of Claim and confidential exhibits to VF's witness statements and in respect of documents disclosed by the Defendants pursuant to the Order. This was subsequently replaced by the confidentiality regime outlined above.
56. On 2 March 2007 the Defendants' solicitors ("GMR") wrote to the Claimants' solicitors ("FFW") in response to the Order of 11 January 2007 in the following terms:

"We understand that in practice most of the instructions to parties involved in the manufacture of the master batches and/or the products were given orally, for reasons of confidentiality. However, our clients have located the following documents from which the instructions would have been drawn and which we list using the numbering in the Order:

1.(i)(a) We enclose the recipe for this product which is entitled 'HDPE net forsøgsoversicht, Karur January 2005'.

1.(i)(b) We enclose a document entitled 'Trial, Karur 20 -, June 2005' and a second page containing time, test nr, denier and temperature. These relate to samples produced in testing using a series of different formulae.

1.(i)(c) and (d). Our clients have not yet located the instructions sent to the manufacturer in each of the these two cases but our clients confirm that the two descriptions relate to the same net and they will continue to search for the actual instructions provided. In the meantime, we enclose a detailed form of the description of the process used to produce these two products and confirm that they are the same as our clients' current product. This document is called 'Netprotect Production Manual' and contains the raw material specifications and recipes."

57. On 27 March 2007 the Defendants filed an application that the claim be struck out on the ground that it was vexatious and an abuse of process. The application was supported by the first, second and third witness statements of each of Dr Skovmand and Mr Larsen and by a report from an expert witness, Professor Gary Stevens. In due course Dr Skovmand and Mr Larsen each made fourth witness statements in reply to the evidence served by VF.
58. On 13 April 2007 FFW wrote to GMR requesting copies of documents referred to by Dr Skovmand in his first witness statement. One of the requests was for copies of the "preliminary formulations" referred to by Dr Skovmand in paragraph 31 of the statement (as to which, see below).
59. On 4 May 2007 the Defendants served their Defence. In the Defence the Defendants denied misuse of confidential information and pleaded that "the Defendants rely upon the design process set out in the Defendants' Confidential Schedule 2 hereto". Confidential Schedule 2 is entitled "The development history of Netprotect".
60. Annexed to Confidential Schedule 2 were copies of a number of documents referred to in it, including three documents entitled "Extract DB oct04.xls", "extract DB Apr05.xls" and "Extract DB July05.xls". I shall refer to these collectively as "the Extract DB documents". As their names imply, these are Excel spreadsheets, electronic copies of which were disclosed by the Defendants on 16 May 2007. Also annexed to Confidential Schedule 2 was a document entitled "NPT details.doc", which was another copy of the Netprotect Production Manual.
61. On 17 May 2007 GMR wrote to FFW explaining certain things about the documents annexed to Confidential Schedule 2. This letter stated:

"As you will see from the electronic version of this database …, the document in tab 13 [i.e. 'extract DB Apr05.xls'] is made up of 3 different documents.

Our clients created a database (which we will refer to as the 'Netprotect Database') which contained all the tests and results in the development of Netprotect.

In the development history of Netprotect … , we deliberately chose to exhibit relevant extracts from that database to illustrate that history. The overall database (and what is contained in Tab 13 and called 'A' in the electronic version) was included in error in Tab 13…

The Karur 0705 results were also annexed in error to Tab 13…"

62. Thus the Defendants had not intended to disclose the A sheet of extract DB Apr05.xls. As explained in more detail below, this is a fairly complete copy of the A sheet of the Netprotect database down to about March 2006. Nor had the Defendants intended to disclose the Karur 0705 sheet of extract DB Apr05.xls.
63. Also on 17 May 2007 GMR wrote to FFW responding substantively to FFW's letter of 17 April 2007 and disclosing a number of documents. In response to the request made in relation to paragraph 31 of Dr Skovmand's first statement, GMR said:

"A copy of the information referred to is enclosed with this letter. (See Tab 4 – Intelligent Insect Control – Ciba Speciality Chemicals Joint Project – objective – investigate the influence of Ciba additives on the migration and UV stability of insecticides in polyolefin films for tarpaulin coverings for the African market.)"

64. On 21 May 2007 VF served a document setting out Further Information in respect of the Particulars of Claim. This document identified a number of rows in the Fence database, identified by sample number, as containing the information relied upon by VF as being confidential and having been used by the Defendants in the development of Netprotect. The samples so identified began with numbers 8, 9 and 13, but included a considerable number of others. The document also stated that VF would rely on the very close similarity between the concentrations of [Additive A], [Additive B] and [Additive C] relative to deltamethrin between the Fence samples and various Netprotect recipes and samples.
65. The Defendants' application to strike out the claim came before Roger Wyand QC sitting as a Deputy High Court Judge on 25 June 2007. In a judgment dated 27 October 2007 he dismissed the application. During the course of the hearing, counsel for VF relied on the fact that, as I describe below, the Fence database contains certain odd numbers which also appear in the Netprotect database. During the hearing counsel for the Defendants offered an explanation for one of these numbers on instructions from Mr Larsen. After the hearing but before judgment, Dr Skovmand made a fifth statement giving a different explanation.
66. On 25 July 2008 VF filed an application seeking inter alia specific disclosure of certain documents. On 18 August 2008 VF filed a further application seeking an order that the Defendants serve a supplemental list of documents and provide standard or specific disclosure of certain categories of documents, in particular documents in the possession of Dr Skovmand which the Defendants had a right to possession of.
67. In late August 2008, the parties exchanged witness statements for trial. The statements served by the Defendants included Dr Skovmand's sixth witness statement and Mr Larsen's fifth witness statement.
68. VF's disclosure applications came before His Honour Judge Fysh QC sitting as a High Court Judge on 4 September 2008. He made an order that Mr Larsen make and serve a witness statement setting out inter alia the steps he had taken to request Dr Skovmand (both personally and as a director of IIC) to confirm whether Dr Skovmand had any of the documents and items of information identified in Schedules 1 and 2 to the Order and, if Dr Skovmand had them, the steps he had taken to request Dr Skovmand to provide copies of documents to the Defendants for use in these proceedings. The Schedules to the Order included the following items:

"SCHEDULE 1

1.     Unedited and original copies, in electronic form, of the databases (in their entirety and including the comments of Mr Skovmand) from which come the extracts described as 'Extract DB Oct04.xls', 'Extract DB Apr05.xls' and 'Extract DB July05.xls' attached as Confidential Schedule 2 to the Defence.

…

5.     The complete and unredacted copy of the 'Fence Database' identified in paragraphs 115-116 of the witness statement for trial of Mr Skovmand.

…

SCHEDULE 2

1.     To the extent that the information is not known to the Defendants, the recipe of (a) the net or nets submitted to the WHO as a result of the application in September 2005 by Intection ApS for WHOPES I evaluation…"

69. In compliance with this Order Mr Larsen made his eighth witness statement. Mr Larsen exhibited to this statement a letter dated 5 September 2008 which he had written to Dr Skovmand asking him to disclose documents falling within Schedules 1 and 2 and a reply from Dr Skovmand dated 12 September 2008.
70. In late October 2008 the parties exchanged experts' reports. VF served an expert report from Gary Howe dated 24 October 2008.
71. The case had been fixed for trial in a trial window commencing on Monday 24 November 2008. At 14:23 on Friday 21 November 2008 the Defendants served a seventh witness statement of Dr Skovmand responding almost paragraph by paragraph to the first 495 paragraphs (out of 613, not including appendices) of Mr Howe's first report. This led to an application by the Claimants on Tuesday 25 November 2008 for an adjournment of the trial. For the reasons given in my judgment of that date [2008] EWHC 2928 (Ch), I declined to adjourn the trial substantially, but ordered a short postponement to Friday 28 November 2008. I express concerned about two matters, one of which was that the nature of the Extract DB documents, and the basis upon which they had been prepared, was unclear. Accordingly, I admitted Dr Skovmand's seventh witness statement on two conditions, the second of which was as follows:

"Secondly, the defendants should serve a short statement from the appropriate witness, and when I say 'the appropriate witness' I do not yet know who that is because it is not clear to me whether the appropriate witness is Dr Skovmand or Mr Larsen, which will contain the following information and no more, that is to say, an explanation as to what the three sets of extracts are, how precisely they were derived from the database from which they were derived, and when."

72. On Wednesday 26 November 2008 GMR replied to a list of queries which had been raised in a letter from FFW arising out of Dr Skovmand's seventh witness statement, together with additional queries raised by counsel for VF in his submissions to me the day before. GMR's letter acknowledged a number of mistakes and typographical errors in Dr Skovmand's seventh witness statement and notified a number of corrections to it.
73. On Thursday 27 November 2008 the Defendants served an eighth witness statement of Dr Skovmand, which was also verified by a statement of truth signed by Mr Larsen, in apparent compliance with the second condition I had imposed. In this statement Dr Skovmand said:

"1.      As I was the one who made the 3 'Extracts' of databases, I have been directed to explain what they are, how they were derived from the Netprotect database and when this was done.

2.      The three extracts are the databases (or copies of the databases) which were made by me and provided to Bestnet in or about October 2004, July 2005 and April 2006 respectively. They were disclosed by Bestnet as relevant documents. To the best of my knowledge and belief I provided Bestnet with no later extracts."

74. On Friday 28 November 2008 VF again applied for an adjournment of the trial. In support of that application VF contended inter alia, for reasons set out in their counsel's skeleton argument and a witness statement made by Beatriz San Martin of FFW, that paragraph 2 of Dr Skovmand's eighth witness statement was untrue. The Defendants requested time to file evidence in response and accordingly the application was adjourned to Monday 1 December 2008.
75. On Sunday 30 November 2008 the Defendants served a ninth witness statement of Dr Skovmand in which he gave a different explanation as to the provenance of the "Extract DB" documents to that which was given in his eighth statement.
76. On Monday 1 December 2008 I heard argument on the application. For the reasons given in my judgment of that date [2008] EHWC 2988 (Ch), I decided to adjourn the trial to 12 January 2009.
77. At trial, the evidence of the Claimants' witnesses was completed on Monday 19 January 2009. The Court did not sit on 20 and 21 January 2009 as I had other commitments. On Wednesday 21 January 2009, that is to say the day before Dr Skovmand started giving evidence, the Defendants notified VF of a considerable number of corrections to all nine of Dr Skovmand's witness statements. These included corrections to passages in his sixth witness statement which had already been the subject of corrections. Many of the corrections were of a minor nature, but others were not. The corrections were accompanied by a document which purported to explain them. It should be noted that when Dr Skovmand gave evidence he was not asked to verify the accompanying document (referred to below as "the explanatory sheet", although it consists of a number of sheets of paper).
78. When VF's counsel received the corrections, they pointed out that, although Dr Skovmand had corrected his eighth statement, that statement had also been signed by Mr Larsen and the Defendants had not stated whether or not Mr Larsen agreed to the corrections. They also pointed out that the corrections made by Dr Skovmand to his ninth statement indicated that there was a difference in recollection between him and Mr Larsen as to the provenance of the Extract DB documents. They therefore requested that Mr Larsen should serve a statement setting out his position on these matters. This led to the service of a tenth witness statement of Mr Larsen shortly before the hearing on 22 January 2009.
79. At that hearing counsel for VF sought a short adjournment to consider the new evidence. In support of that application he relied upon the discrepancies between the statements of Dr Skovmand and Mr Larsen. This was resisted by counsel for the Defendants, who submitted that there was only one difference between the recollections of the two men. For the reasons given in my ruling that day, I declined to adjourn the hearing. Dr Skovmand was then called, and made an additional couple of small corrections to his sixth statement in his evidence in chief.

The witnesses

80. All of the Danish witnesses gave evidence in English and all spoke excellent English.

VF's witnesses

81. Mr Vestergaard Frandsen. I have introduced Mr Vestergaard Frandsen above. Much of his evidence was unchallenged in cross-examination. Other parts were formally challenged, but not tested in detail. He gave evidence in his witness statement suggesting that, after the comparison testing of VF's PE bed net against PermaNet and Olyset in Burkina Faso and Benin (as to which see below) had been completed, in late 2004 and in 2005, he had chased Dr Skovmand for the results but Dr Skovmand only gave him vague and disencouraging feedback. It was put to him in cross-examination that his account was untrue on the basis that he had said in his statement that "The lady organising the trials in Burkina Faso apparently only spoke French, which meant that all of her emails to me … had to go through OS", whereas Mr Vestergaard Frandsen was shown to have sent the lady in question, Léa Toyé, an email in English. Mr Vestergaard Frandsen's response was that Ms Toyé had broken English but it was very difficult to communicate with her. This is supported by an email from Mrs Sig to Dr Skovmand in which she said "You were assigned for this task because I'm having such a hard time communicating with Lea, as she only speaks French". Accordingly, I am satisfied that Mr Vestergaard Frandsen was an honest witness. Whether he was correct in thinking that Dr Skovmand had withheld information from him about the Burkina Faso tests is unclear and unnecessary for me to decide.
82. Finn Kjaergaard. In October 1998 Mr Kjaergaard became an advisor to VF, in October 1999 he joined the Board of the First Claimant and in 2000 he became the Chairman of the Board. Since 2001 he has been an executive director and Chairman of a number of companies in the group. One of his functions has been to oversee VF's intellectual property protection. Mr Kjaergaard gave evidence while recovering from an attack of influenza and while still suffering from quite a bad cough, which made giving evidence quite an ordeal for him. His credit was attacked in cross-examination on a ground which was later withdrawn. He was cross-examined extensively on VF's motives for bringing these proceedings. For the reasons explained below, I have found this of little assistance in determining the issues between the parties. He was also cross-examined quite extensively on the relationship between VF and Dr Skovmand, but since it is clear that Mr Vestergaard Frandsen has more direct knowledge on this subject I have primarily relied on the latter's evidence.
83. Phan Quy Chi (known as Chi Phan). Ms Phan has a Bachelor's degree in organic chemistry and a Master's degree in materials science. She has also studied organic analytical chemistry. She has been the Laboratory Manager of VF's laboratories in Hanoi in Vietnam since September 2002. She gave evidence in English via video-link, and thus there were some difficulties in communication. Counsel for the Defendants submitted that she had failed to distinguish in her evidence between matters she had firsthand knowledge of and matters she had inferred without having direct knowledge. In my judgment there is some force in this, but I do not think it significantly undermines the value of her evidence. I would add that some of her evidence was really in the nature of expert evidence rather than evidence of fact, but no objection was taken to it on that ground.
84. Dr Helen Jamet. Dr Jamet obtained a BSc in Applied Biology in 1994, an MSc in Medical Parasitology in 1996 and PhD in Infectious Diseases in 2002. She then worked for two years as a Research Fellow at the London School of Hygiene and Tropical Medicine, followed by two years employed by an American company, working on mosquito protection. She joined VF as a Project Manager in May 2007 and since July 2008 has been VF's Head of Entomology. Her evidence was really in the nature of expert evidence, but no objection was taken to it on that ground. It was of little assistance, however, since she was not able to speak to the position prior to her employment by VF.
85. Sicco Roorda. Mr Roorda is a textile engineer of some experience. He joined VF as Production Manager on 1 February 2005. He was the permanent replacement for Mr Larsen, Lars Fensbo having temporarily filled the post in the meantime. Apart from providing some general background about the production and testing of PE nets, his evidence was of little assistance on the issues between the parties.
86. Dr Robin Medenwaldt. Dr Medenwaldt is a patent attorney instructed by VF, to begin with as a partner in Patrade A/S and more recently as a freelance working for Patrade A/S. He was a straightforward witness whose evidence concerned only one small point, as discussed below.
87. Bruno Ducoulombier, Nicholas Rose and Jørn Vestergaard-Jensen. Mr Ducoulombier is a partner in FFW's Paris office who acts for VF in the French proceedings. He gave evidence about those proceedings. Mr Rose is a partner in FFW's London who has the conduct of these proceedings for VF. He gave evidence about the Indian proceedings brought by VF. Mr Vestergaard-Jensen is a partner in the firm of Kromann Reumert who acts for VF in the Danish proceedings. He gave evidence about those proceedings (his statement also contained expert evidence about Danish law, but this was not admitted). None of these witnesses' evidence was challenged.
88. Mr Howe. Mr Howe graduated from the University of Bath with a BSc in materials science in 1990. From 1990-1992 he was employed by British Aerospace Military Aircraft Ltd. From 1992-2001 he was employed by Rapra Technology Ltd ("Rapra"), initially as a Consultant in Plastics Technology and from 1999 as Head of Plastics Technology. From 2001-2002 he was employed by New Transducers Ltd as Manager Components and Technologies Ltd. Since 2003 he has provided consultancy services through his own company Independent Polymer Technology Ltd.
89. Mr Howe's field of expertise is that of polymer technology with particular emphasis on the properties, long term performance and failure of plastic components. He has had considerable experience of consulting for clients in industry and also of work as an expert witness in legal disputes. Most of his work as an expert witness has been in connection with product failures of one kind or another, but he has also been involved in patent disputes.
90. Three points should be noted about Mr Howe's expertise. First, although he had some knowledge of the science of experimental design, he did not profess to be an expert in that field.
91. Secondly, Mr Howe did not purport to be an expert in statistics. Nevertheless, he was familiar with the statistical technique of linear regression analysis and able to use the Statistix 9 software package to carry out such analyses. This is the same software as Dr Skovmand used, except that Dr Skovmand had version 8. No one suggested that there were any differences between the two versions that were relevant to the present case.
92. Thirdly and most importantly, this was the first occasion on which Mr Howe had been asked to carry out a forensic investigation which primarily involved consideration of a paper trail rather than technical analysis of physical articles. To a considerable extent Mr Howe's reports involved the skills of a detective rather than those of a scientist. For that reason I entertained some doubts as to the extent to which Mr Howe's reports properly qualified as expert evidence even though counsel for the Defendants raised no such objection. I am satisfied, however, that they are admissible as expert evidence. Although much of Mr Howe's work involved trawling through the disclosure documents and the Defendants' statements and analysing the information contained in them, he was able to call upon his knowledge of polymer technology and his experience in forensic investigation of a more conventional kind to assist him, and thus the Court, to understand those materials.
93. Two further points are important with regard to Mr Howe's evidence. The first is that, in his seventh witness statement, Dr Skovmand fiercely criticised Mr Howe's first report in a number of respects. For example, Dr Skovmand said at paragraph 18:

"As to the basis of the analysis in the IPT report, it is flawed and invalid. I use statistical methods a great deal and I am shocked that IPT can put up this report as a serious analysis, when it does not employ proper statistical methods."

Mr Howe replied to Dr Skovmand's seventh statement in his third report. While he accepted some points made by Dr Skovmand, he rejected others. Overall, he essentially stood by the analysis in his first report. Few of Dr Skovmand's criticisms were put to Mr Howe in cross-examination. Moreover, the cross-examination of Dr Skovmand demonstrated that at least some of them were unfounded.

94. The second point is that the cross-examination of Mr Howe took a somewhat unusual course. Although counsel for the Defendants had estimated that his cross-examination of Mr Howe would take 1½ -2 days, in the event it lasted little over an hour. Having secured the key answer he wanted from Mr Howe, counsel for the Defendants sat down. I shall consider this evidence in detail below. At this stage, the point to be noted is that large parts of Mr Howe's reports were not challenged at all.

The Defendants' witnesses

95. Dr Skovmand. I have introduced Dr Skovmand and briefly outlined his role in the events with which this claim is concerned above. I have also described Dr Skovmand's written evidence and the circumstances in which most of it came into the case. Counsel for VF originally estimated that his cross-examination of Dr Skovmand would take up to two days. In the event the cross-examination extended over three days, although half a day was lost because Dr Skovmand produced some further documents during the course of his evidence which led to a short adjournment. In the end I was forced to impose a guillotine on the cross-examination because, as Dr Skovmand had explained when it became evident that there was risk of the estimate being exceeded, Dr Skovmand had a business trip to Brazil arranged which he was able to postpone by 24 hours but no more. Counsel for VF told me that ideally he would have wished to cross-examine Dr Skovmand for a further half day, but I am satisfied that counsel had a more than reasonable opportunity to question Dr Skovmand about all matters of real significance. Nevertheless, I have borne in mind that the cross-examination was curtailed.
96. Dr Skovmand's demeanour in the witness box was superficially impressive. He was calm, he was articulate, his answers were direct and to the point and he made a number of concessions, in particular as to a number of inaccuracies in his statements in addition to the corrections which had previously been notified and as to inaccuracies in the Defendants' accounts of the development of Netprotect. As a result, I consider that his oral evidence was more accurate than his previous witness statements. Nevertheless, I was driven to the conclusion that he was not a truthful witness, in particular with regard to the development of Netprotect. The principal reason for this was that Dr Skovmand was unable to give a consistent or coherent account of what he had done, when and why. On the contrary, during the course of the proceedings his account changed repeatedly in a number of important respects. The reason it kept changing was that VF were repeatedly able to demonstrate that his account was not correct. I will give a series of examples of incorrect statements by Dr Skovmand in his witness statements and oral evidence below, but these are by no means exhaustive. I was also concerned by his evidence regarding the provenance of the Extract DB documents, which is dealt with in Appendix 1 to this judgment and by his evidence to the Danish court, which is dealt with in Appendix 2 to this judgment. I was not satisfied by his attempts to explain himself in cross-examination in relation to either of these matters. Counsel for the Defendants submitted that, if and to the extent that Dr Skovmand had made inaccurate statements, they were merely mistakes and errors of recollection. I do not accept this, but even if I were to accept it I would still be bound to conclude that Dr Skovmand was not a reliable witness.
97. Mr Larsen. Again, I have already introduced Mr Larsen. Although not to the same degree as Dr Skovmand, and with one exception, Mr Larsen's demeanour in the witness box was also superficially impressive. He too was calm, though not as dispassionate as Dr Skovmand, articulate, direct and made concessions. The exception is that he became very defensive about the chain of emails dated 10 and 11 August 2004 (as to which see below), which considerably prolonged the cross-examination. Again, however, I was driven to the conclusion that he was not a truthful witness, in particular with regard to the development of Netprotect. My reasons are similar to those I have given in relation to Dr Skovmand. Again, I shall give non-exhaustive examples of incorrect statements by Mr Larsen below. Mr Larsen was also central to the question over the provenance of the Extract DB documents. Again, I was not satisfied with his explanations in cross-examination with regard to this. As with Dr Skovmand, counsel for the Defendants submitted that any inaccuracies were innocent mistakes. Again I do not accept this, but again I would in any event be bound to conclude that Mr Larsen was not a reliable witness.
98. I would add two further points about Mr Larsen's credibility. First, his letter dated 7 December 2004 (as to which, see below) was an outright lie. Secondly, as he was forced to accept in cross-examination, his behaviour in setting up Intection while still employed by VF, and using contacts made on behalf of VF for that purpose, was a clear betrayal of his employer.
99. Mrs Sig. Again, I have already introduced Mrs Sig. On her account, supported by Mr Larsen, she had very little involvement in, or knowledge about, the technical aspects of the development of Netprotect. She played an important role on the commercial side. Her oral evidence about the Defendants' commercial position was candid. I therefore accept that evidence, although I believe she knew more about the technical side than she was prepared to admit.
100. Other witnesses. The Defendants served a witness statement for trial from Yves Lambert of INEOS Polyolefins, formerly Innovene, formerly BP Solvay ("Innovene"). During the course of the trial the Defendants applied for permission to serve a hearsay notice out of time in respect of this statement. The application was supported by evidence that Mr Lambert's employer had declined to permit him to give evidence. Accordingly I granted the permission sought. In the absence of cross-examination I was not able to give Mr Lambert's statement much weight.
101. The Defendants also served a witness statement for trial from Sivasamy Marappan, who it appears is the same person as Mr M. Sivasamy. He was not called and no hearsay notice was served in respect of his evidence. I have therefore disregarded this evidence.
102. Although the Defendants served an expert report from Professor Stevens for trial, he was not called. I was told by counsel for the Defendants that this was due to problems with his availability. I have no reason to question that explanation, although it was unsupported by evidence. Counsel for VF relied upon a passage in an earlier report by Professor Stevens which had been adduced by the Defendants in support of their strike-out application. He thereby put the whole of that report in evidence, but counsel for the Defendants sensibly did not seek to rely upon it in the absence of cross-examination.

VF's motivation for bringing the claim

103. The Defendants contend that this claim is without foundation and has been brought by VF to harass and oppress the Defendants in order to protect a very profitable market for VF's PermaNet. Moreover, the Defendants contend that it forms part of a pattern of harassing litigation brought by VF against the Defendants, their suppliers and their distributor.
104. I entirely accept that, as Laddie J said in Ocular Sciences Ltd v Aspect Vision Care Ltd & Ors [1997] RPC 289 at 359, "it is well recognised that breach of confidence actions can be used to oppress and harass competitors and ex-employees". It follows that the courts must scrutinise such claims with care to see if they are well-founded or not, which is what I intend to do in the present case. It does not follow that all such claims are unfounded and harassing claims.
105. I am in no position to decide whether, as the Defendants contend, VF have succeeded for a number of years in maintaining an inflated price for PermaNets by restricting competition. I do accept that VF have a profitable business in marketing PermaNets to organisations funded by governments, non-governmental organisations and private philanthropists. Contrary to the suggestion advanced by the Defendants, I see nothing wrong in that. The advances made by the pharmaceutical industry are based on harnessing the profit motive to fund and reward research and development for the greater benefit of humanity. There is no reason why the production of insecticidal products to combat malaria should have to be conducted upon some more altruistic principle. Moreover, as I shall discuss below, it is clear that the Defendants' development of Netprotect was not motivated by altruism, but by their desire to exploit a perceived commercial opportunity.
106. I also accept that VF's purpose in bringing these proceedings is to defend their PermaNet market against what they contend is unlawful competition. Again, I see nothing wrong in that provided that the complaint is well-founded. In principle, VF's behaviour in bringing this claim is no different to that of a research-based pharmaceutical company suing a competitor for patent infringement.
107. In saying that, I do not overlook the fact that experience has regrettably shown that claimants in confidential information cases all too frequently bring claims that are abusive because they are speculative or brought without any real belief in the allegations made. I am not satisfied that this is such a case. The highest it can be put is that VF's claim has changed somewhat since it was first launched at the time of the without notice application. The claim presented by VF at trial, however, was within the scope of the case pleaded in VF's Further Information of 21 May 2007, although it was more narrowly focussed. VF's witnesses were adamant that VF had brought the claim to protect VF's confidential information, and it was not put to them in cross-examination that they had no genuine belief in the allegations made.
108. Nor am I satisfied that VF have conducted this claim in an abusive manner. Indeed, the Defendants did not seriously contend that VF had done so.
109. As for the foreign litigation, I can quite see that it must have been distressing and difficult, particularly from the perspective of time and resources, for the Defendants, Dr Skovmand and IIC to be sued by VF in a number of different countries. But given that the Defendants shifted their corporate base from Denmark to England with the express intention of trying to avoid the consequences of the Danish litigation, that Dr Skovmand and IIC are based in France and that the Defendants' manufacturers are located in India, it is difficult to see that it was illegitimate for VF to bring proceedings in those jurisdictions. Mr Wyand QC rejected the submission that it was abusive in his judgment on the striking-out application and the submission was not in terms renewed before me.
110. Finally, the Defendants pointed out that on 25 August 2005 VF wrote to the Director of the Roll Back Malaria Partnership ("RBM", as to which, see below) at the World Health Organisation ("WHO") informing him of VF's proceedings against Intection and Mr Larsen in the Danish courts and asking WHO to notify VF if Intection, Mr Larsen or Mrs Sig approached WHO. As I shall discuss below, upon receipt of this letter WHO stopped evaluating Netprotect and it took the Defendants some time to persuade WHO to resume evaluation. I have to say that I am unimpressed by VF's conduct in sending this letter, but other than confirming that VF have been determined to stamp on what they perceived as a competitive threat, I cannot see that it is particularly relevant to the merits of this claim. Indeed, Dr Skovmand candidly said that the delay proved advantageous since it enabled him to improve the formulation of Netprotect.
111. In short, I do not consider that VF's motives for bringing the claim shed any light on whether it is well founded or not. To decide that, it is necessary for me to find the facts and then apply the law to those facts.

My approach to finding the facts

112. There are considerable factual disputes between the parties both as to the development of VF's PE products and as to the development of Netprotect. In resolving these factual disputes, I have faced a number of difficulties. A general problem is that, as I have said, I have concluded that the two key witnesses, Dr Skovmand and Mr Larsen, are not reliable. There are also specific problems with regard to both VF's development of its products and the Defendants' development of Netprotect which I shall explain below. I have therefore adopted the approach which I articulated in BRUTT Trade Marks [2007] RPC 19 at [24] as follows:

"In a well-known passage in The Ocean Frost [1985] 1 Lloyd's Rep 1 at 57 Robert Goff L.J. (as he then was) said:

'Speaking from my own experience, I have found it essential in cases of fraud, when considering the credibility of witnesses, always to test their veracity by reference to the independent facts proved independently of their testimony, in particular by reference to the documents in the case, and also to pay particular regard to their motives and to the overall probabilities. It is frequently very difficult to tell whether a witness is telling the truth or not; and where there is a conflict of evidence such as there was in the present case, reference to the objective facts and documents, to the witnesses' motives, and to the overall probabilities, can be of very great assistance to a judge in ascertaining the truth.'

Thus before addressing the conflicts of evidence the fact-finding tribunal should first consider what is common ground, what is asserted by one party and not contested by the other and vice versa and what is shown by the documentary evidence. Once these matters have been ascertained and put into chronological order, it is frequently much easier to resolve the remaining conflicts. In resolving such conflicts it is of assistance to consider which version of events is the more likely."

113. In the present case, I have found it necessary to construct a chronological account of both developments. I have constructed this account primarily by reference to the disclosure documents in the case and secondarily by reference to unchallenged evidence. I have then used this chronological account to test the disputed evidence.
114. I have also borne in mind that the allegations made by VF are serious ones and that, the more serious the allegation sought to be proved is, the more cogent the evidence relied upon to support it must be: see Re H (Minors) [1996] AC 563.
115. Some of the disclosure documents were written in Danish. Where I quote from a document written in Danish, I shall quote from the certified translation into English without further acknowledgement. I have corrected typographical errors, except errors in the spelling of proper names, but not grammatical or syntactical errors.

The facts

Introduction to the technology

116. This case involves two main areas of technology. The first is polymer technology. The second is the use of insecticides to control mosquitoes. Before outlining the technology it is necessary to say a few words about terminology. VF's PES net, PermaNet, is "impregnated" with insecticide. The term "impregnated" is used to refer to a process whereby knitted PES net is coated with a material which includes the insecticide. VF's PE net products discussed below and the Defendant's PE net, Netprotect, "incorporate" the insecticide. The term "incorporate", and cognate expressions, are used to refer to a process whereby the insecticide is incorporated into the polymer before extrusion of the fibre. The insecticide migrates from the body of the net to its surface, thereby enabling the net to "regenerate" its insecticidal effect after washing. Although impregnation and incorporation are technically distinct, sometimes impregnation is referred to as "incorporation" and vice-versa.
117. Polymer technology. The polymer with which this dispute is principally concerned is PE. PE consists of chains of polymerised ethylene molecules. PE is not a purely linear material, but has side branches attached to the main backbone. The catalysts and conditions used during polymerisation control the number and length of the side branches. The number and length of the side branches affect the density, and other properties, of the PE There are a number of different types of PE, including the following (although strictly speaking LLDPE is a copolymer with a higher olefin):

118. PE fibres are made by a melt spinning or extrusion process. This involves melting the polymer and then forcing it through a spinneret under pressure. The polymer is stored, usually in the form of granules, in a hopper which discharges into the barrel of an extruder. The polymer is transported along the barrel, mixed and compressed by a rotating screw. The polymer is heated both by heating elements round the barrel and by shearing of the polymer as it moves along the barrel. A single stage extruder is shown schematically below:

119. It is common practice to introduce additives into PE during processing by means of a masterbatch. A masterbatch is a concentrated granule of one or more additives incorporated into a polymer binder. The purpose of the masterbatch is to ensure dispersion of the additive through the polymer during extrusion.
120. It is also common practice for other chemicals to be added to the PE to influence the properties of the PE fibres. A considerable variety of additives are marketed by suppliers such as Ciba. The selection of the additive or additives to use, and the determination of the concentration level of the selected additive(s), is a key aspect of the development of PE products such as those with which this case is concerned. It requires skill and experience, but the evidence in this case makes it clear that it is something of a black art dependent on trial and error.
121. The additives which principally feature in this case are as follows:

i) [Additive A] is [REDACTED]. The same chemical is sold by other suppliers under the names [REDACTED].

ii) [Additive B] is [REDACTED]. The same chemical is sold by another supplier under the name [REDACTED].

iii) [Additive C] is a [REDACTED].

iv) [Additive D] is [REDACTED].

v) [Additive E] is [REDACTED].

vi) [Additive F] is [REDACTED].

vii) [Additive G] is [REDACTED].

viii) [Additive H] is [REDACTED].

ix) [Additive I] is [REDACTED].

x) [Additive J] is [REDACTED].

xi) [Additive K] and [Additive L] are [REDACTED].

xii) [Additive M] is [REDACTED].

122. An additive will be held "in solution" within the "free volume" of the polymer. The amount of additive which can be incorporated into a polymer is controlled by the solubility of parameters of both the polymer and the additive. If the solubility limit is exceeded, the additive will have a tendency to migrate from the bulk of the polymer to the surface where it will be lost through a combination of evaporation, surface abrasion, UV degradation, surface washing and so on. The rate of diffusion from the bulk will be controlled by a number of factors. With a mixed additive system, the migration rate of each additive will be affected by the other additives present. Even if the additive concentration is below its solubility limit, there will be a tendency for it to migrate to the surface to replace material removed from the surface.
123. The use to which a particular product is to be put is likely to give rise to a number of specific properties that are required. Thus in general a product intended for use as an agricultural fence will require a higher tensile strength and stronger UV protection than one intended for use as a bed net. Conversely a bed net is more likely to be subject to detergents while being washed. These requirements will influence the selection of additives.
124. PE nets, whether for use as bed nets or as agricultural fence, are knitted from extruded fibres. VF's ZeroFly PE tarpaulin product discussed below is a laminate of three layers, of which the central layer is woven from PE fibres.
125. Mosquito control. It is estimated that each year around 1 million children under the age of 5 die from malaria-related diseases, mainly in Africa, Latin and South America and South-East Asia. To provide a coordinated global approach to fighting malaria, in 1998 RBM was launched by WHO, the United Nations Children's Fund (UNICEF), the United National Development Programme and the World Bank. RBM now embraces a wide range of partners, including malaria-endemic countries, non-governmental and community-based organisations, the private sector, philanthropic foundations and academic institutions. One of the key methods that has been employed to combat malaria by WHO and RBM is the provision of insecticidal mosquito nets.
126. A number of insecticides are used to kill malarial mosquitoes. The ones which principally feature in this case are deltamethrin, permethrin, alpha-cypermethrin and bifenthrin. These are all well-known pyrethroid insecticides.
127. Only insecticide present on the surface of the net fibre is effective at killing mosquitoes. Like any other additive, the insecticide becomes depleted from the surface, particularly by washing. Accordingly, for a PE net to retain its insecticidal properties over time, it is necessary for insecticide contained in the bulk of the fibre to migrate to the surface. This is referred to as "regeneration". Furthermore, migration must be controlled so that it is not too quick, resulting in the net not lasting long enough, or too slow, resulting in the net not killing enough mosquitoes to be effective. Ideally, the surface depletion rate and the bulk to surface migration rate would be balanced to provide a constant level of insecticidal performance.
128. WHO has a specification for LLINs. To be a LLIN, a net must achieve a minimum insecticidal effect after 20 "washes" and 3 years of recommended use under field conditions.
129. WHO operates the WHO Pesticide Evaluation Scheme (WHOPES). The major customers for LLINs require nets to have at least an interim, and preferably a full, recommendation, from WHOPES. WHOPES involves three phases of testing. Phase I is a laboratory test by a WHO-approved laboratory such as the Laboratoire de Lutte Contre les Insectes Nuisibles ("LIN") at IRD, Phase II is a three month small scale field test at WHO-approved testing station such as the Centre de Recherches Entomologiques de Cotonou in Benin and Phase III is a three year large scale field test.
130. The standard laboratory bioassay used by WHO-approved laboratories is the "cone" test. A plastic cone is held over the material being tested. Five mosquitoes are introduced into the cone which is then held over the sample for three minutes. The mosquitoes are then transferred to plastic cups where they are held with access to sucrose solution for 24 hours. A total of ten replicates is used for each sample tested, giving a total of 50 mosquitoes per sample. Knockdown (i.e. the total number of paralysed insects) is measured 60 minutes after exposure and mortality 24 hours after exposure. Efficacy may also be measured by a "tunnel test".
131. Regeneration is tested by carrying out by performing bioassays at 24 hour intervals on samples that have been washed using a standard washing procedure until initial biological activity is restored. Wash resistance is tested by washing the sample using the standard procedure at intervals equivalent to the time required for regeneration. Nets must cause at least 80% mortality and 95% knockdown after at least 20 washes to pass WHOPES Phase I.
132. Surprisingly, there was a lively dispute at trial as to whether the criterion used by WHO was effective or actual mortality (dead insects) or functional mortality (dead insects plus insects that have three legs or fewer). VF's witnesses said it was functional mortality, whereas Dr Skovmand said that it was effective mortality. The first report evaluating Netprotect issued by LIN in May 2005 expressly uses the criterion of functional mortality. The second report evaluating Netprotect issued by LIN in September 2006 is not explicit about this, but appears to use the same criterion since it states "Mosquitoes able to fly were counted as alive". On the other hand, during the trial Dr Skovmand produced an email from one of the authors of these two reports dated 16 January 2009 stating that all evaluations by LIN used the criterion of effective mortality. Mr Larsen admitted, however, that until then he had thought that the criterion used was functional mortality. Dr Jamet said that both criteria were used and Ms Phan said that the term "mortality" was not used consistently or with precision. My conclusion is that practice in this respect has recently been tightened up. Nowadays WHO and its collaborating centres use effective mortality as the criterion, but in the past they used functionality mortality or were unclear.
133. In any event, I think it is clear that during the period 2001-2004 the criterion which Dr Skovmand was using was functional mortality. Furthermore, Mr Howe demonstrated that functional mortality is closely correlated with effective mortality.
134. Bioassay results depend on the strain of mosquito used. Mosquito strains in some countries are more resistant to pyrethroid insecticides such as deltamethrin. An inter-laboratory study led by Dr Skovmand in 2004 found that wash resistance measured in Salatiga, Indonesia was 2.4 times higher, and wash resistance measured in Hanoi, Vietnam was 1.7 times higher, than in Montpellier, France, apparently due to differences in the susceptibility of the local mosquito strains.

Linear regression analysis

135. This case also involves the statistical technique of linear regression analysis. Linear regression is a form of regression analysis in which the relationship between one or more independent variables and another variable, called the dependent variable, is modelled by a linear function. This function depends on one or more model parameters, called regression coefficients. The values of the regression coefficients are chosen to minimise a least squares function. A linear regression equation with one independent variable represents a straight line when the predicted value (i.e. the dependant variable from the regression equation) is plotted against the independent variable: this is called a simple linear regression.
136. Where there is more than one independent variable, the multiple linear regression equation takes the form:

Y = A + B₁X₁ + B₂X₂ + B₃X₃ ….

where Y is the dependent variable, A is a constant, X_i is the value of the independent variable i and B_i is the coefficient. A larger coefficient indicates a greater influence of the independent variable on the dependent variable, a positive coefficient indicating a positive effect and a negative coefficient a negative one. If a linear relationship between independent and dependent variables does not exist, then alternative non-linear techniques are required.

137. The significance of each of the coefficients can be assessed using the P value. This value measures the probability that the same result would be obtained again if the "null hypothesis" (i.e. that the variable has no effect on the measured parameter) were true. Hence a P value of 0.01 indicates that, if the variable had no effect on the dependent variable, then there would be a 1% probability that the same result could be obtained randomly.
138. The overall fit of data to the calculated linear equation can be measured using the R² value. This value is between 0 and 1 with R²=1 indicating a perfect linear fit.
139. Linear regression analysis can be carried out in a stepwise manner, either forwards or backwards, but I do not think it is necessary to go into these details for the purposes of this judgment.
140. Within linear regression analysis, there can be a problem of correlation between the independent variables. This is referred to as co-linearity. It occurs when two variables change in tandem with each other. It can arise either (a) when the two variables are not independent or (b) when they have been varied in concert so that it is impossible to separate the effect of one from the other. Problems associated with co-linearity can be identified using the variance inflation factor (VIF), which is an index that measures how much the variance of a coefficient is increased because of co-linearity. The VIF indicates the co-linearity of the independent variables (i.e. the parameters chosen for the experiment) and is not affected by the dependent variable (i.e. the set of results of the experiment). Ideally, VIF will be as close to 1 as possible. There are various rules of thumb as to what value of VIF is taken to indicate a problem.

VF's development of Zerofly, Fence and PE bed nets

141. Although in this section I shall primarily concentrate on the development of VF's PE products, it will be necessary to place this work in context by reference to the development of VF's PES bed net.
142. In explaining the development of VF's PE products, VF faced the difficulty that the two people mostly closely involved in the work were Dr Skovmand and Mr Larsen. The only factual witnesses called by VF who were involved in the work were Mr Vestergaard Frandsen and Ms Phan, and their involvement was somewhat limited. VF did not adduce evidence from other employees who had been involved in the work, such as Ritu Ahuja, but again they would only have been able to give a limited account.
143. As for Dr Skovmand and Mr Larsen, they did not give a systematic account in their evidence. Furthermore, the accuracy of their account was challenged by VF. Accordingly, as explained above, I have constructed the following account primarily by reference to the disclosure documents and secondarily by reference to unchallenged evidence from certain witnesses, particularly Mr Vestergaard Frandsen.
144. Mr Vestergaard Frandsen had the idea of developing a PE tarpaulin incorporating insecticide as a result of seeing samples of polyvinyl chloride ("PVC") film incorporating deltamethrin produced by Matthew Holmes and John Carroll of the American company Green Shield at a conference in Africa in September 1995. By this time VF had already had some experience with manufacturing and selling their own PE tarpaulins without insecticide. Mr Vestergaard Frandsen thought that a PVC tarpaulin incorporating insecticide would be too expensive and therefore decided to develop a PE tarpaulin incorporating insecticide. This later became VF's ZeroFly product.
145. In July 1996 Mr Vestergaard Frandsen commissioned Messrs Holmes and Carroll to undertake pilot trials at an Indian manufacturer, Duck Tarpaulins ("Duck"), to produce deltamethrin-incorporating PE tarpaulins. Mr Vestergaard Frandsen chose to use deltamethrin for these trials since this was a WHO-approved insecticide which VF had used for tse-tse fly traps since 1993. A number of samples were produced, most of which were of a sandwich construction consisting of a central weave of PE yarns coated with deltamethrin and then laminated with a PE film.
146. At this time Mr Vestergaard Frandsen also discussed with Messrs Holmes and Carroll the idea of developing a long-lasting insecticide-impregnated PES net.
147. In late 1996 or early 1997 VF parted company from Messrs Holmes and Carroll since they could not agree on the contractual basis for Messrs Holmes and Carroll's future involvement in the development of the product. Mr Vestergaard Frandsen gave unchallenged evidence that he orally agreed with Messrs Holmes and Carroll the basis of their work for VF, including the fee to be paid to them, the scope of the project and the fact that any information arising from the collaboration should be kept confidential.
148. After the collaboration with Messrs Holmes and Carroll on the PE tarpaulin broke down, VF continued to work on the project with Duck until some time in 1998. Mr Vestergaard Frandsen gave unchallenged evidence that VF had a written agreement with Duck which included a provision that Duck would not supply other companies with the same products as they manufactured for VF.
149. In around 1997/1998 Mr Vestergaard Frandsen asked Messrs Holmes and Carroll to start developing an insecticide-impregnated PES net. This became VF's PermaNet product. Again, Mr Vestergaard Frandsen gave unchallenged evidence that he orally agreed with Messrs Holmes and Carroll the basis of their work for VF, including the fee to be paid to them, the scope of the project and the fact that any information arising from the collaboration should be kept confidential. In addition, it was agreed that VF would own the rights in the product resulting from the development.
150. In 1998 Mr Vestergaard Frandsen was given Dr Skovmand's name by a contact at the Danish Bilharzia Laboratory. In the spring of 1998 Mr Vestergaard Frandsen met Dr Skovmand in Montpellier, where Mr Vestergaard Frandsen was attending a meeting with Dr Pierre Guillet, the Director of LIN. It was shortly after this meeting that Dr Skovmand left IRD to set up his own consultancy. (Dr Guillet also left IRD and went to work for WHO/RBM in Geneva.)
151. For some time after this, Dr Skovmand advised Mr Vestergaard Frandsen on an occasional, unpaid basis on matters relating to the development of disease control textiles.
152. In late 1998 Mr Vestergaard Frandsen asked Dr Skovmand to supervise Messrs Holmes and Carroll's work on the PES net as he was not satisfied with their progress. Mr Vestergaard Frandsen orally agreed with Dr Skovmand that VF would pay him an hourly rate for the time spent working for VF, together with expenses. It appears that Dr Skovmand's first invoice to VF was dated 12 January 1999, covering work done during the preceding month or so. Thereafter Dr Skovmand (or from about July 2001 IIC) invoiced VF approximately monthly.
153. At some point in early 1999 Dr Skovmand informed Mr Vestergaard Frandsen that Messrs Holmes and Carroll were not doing a very good job, and that he could better.
154. In March 1999 Mr Vestergaard Frandsen was invited to a meeting of the global mosquito net industry by Dr Guillet. Dr Guillet's purpose in arranging the meeting was to get someone to take up the challenge of creating a long-lasting insecticide-impregnated bed net for a reasonable price. Although there were about 40 participants at the meeting, only VF expressed interest in this proposal. Mr Vestergaard Frandsen wanted Dr Skovmand to attend the meeting, but Dr Skovmand was in Africa at that time. At the meeting, Mr Vestergaard Frandsen became aware that Sumitomo was selling a PE net incorporating permethrin under the name Olyset. At that time this net was quite expensive at around $30 per net.
155. In April 1999 VF engaged Dr Skovmand to work on the development of the PES bed net in place of Messrs Holmes and Carroll. Mr Vestergaard Frandsen passed onto Dr Skovmand some information he had received from Messrs Holmes and Carroll.
156. Dr Skovmand carried out some tests on PES nets impregnated with deltamethrin in the spring of 1999 which achieved a wash resistance of about 3-6 washes. By December 1999 Dr Skovmand had identified a formulation which achieved a wash resistance of about 10-12 washes. VF sent bed nets made from these formulations for field testing in Burkina Faso.
157. On 25 November 1999 VF filed Danish Patent Application No. PA 1999 01702 to protect PermaNet. This was used as the priority document for International Patent Application No. WO 01/37662 filed on 24 November 2000 and published on 31 May 2001. This names Dr Skovmand as the sole inventor. Dr Skovmand executed an assignment of his rights in this invention to VF on 29 December 1999.
158. In December 1999 VF launched PermaNet (later called PermaNet 1.0 when PermaNet 2.0 was introduced) at an average price of around $4.50 per net (the price varies according to the size of the net). The product was manufactured by 10-10 Joint Stock Textile Company ("10-10") in Hanoi. VF received funding from DANIDA to defray the costs of setting up, training, technical assistance and technology transfer to 10-10. To begin with, VF experienced production difficulties with ensuring that each net contained the correct level of deltamethrin, but subsequently this problem was resolved.
159. On 5 December 2000 Mr Larsen, who by then had been employed by VF for just over a month, sent an email to Bruno Terrier of Ciba. After saying that Mr Terrier might remember him from Mr Larsen's time at Ciba and explaining his new position at VF, Mr Larsen said:

"We are in a situation where the capacity of our present manufacturing facility is highly overbooked. I am therefore forced to look for subcontractors and this gives us a number of problems. One of these is securing the secrecy of our formulation."

He went on to ask whether Ciba could mix the chemicals it supplied to VF into a single product delivered under "a pseudo-name like PermaChem or whatever".

160. After Mr Terrier replied asking about VF's formulation, Mr Larsen said in an email of 21 December 2000:

"I will give you the exact formulation, but not over e-mail and not before we have a secrecy agreement with you. We are at the moment filing a patent on the PermaNet(tm)."

161. This led to Nicholas Kolesch of Ciba sending Mr Larsen a letter dated 11 January 2001 expressed to constitute a secrecy agreement between Ciba and VF in the following terms:

"With regard to the specific formulation requested by CUSTOMER to be developed using CIBA products for the finishing treatment of impregnated mosquito nets:

CIBA agrees to treat as strictly confidential all written or oral information provided by CUSTOMER with regard to the specific formulation to be developed.

Information related to the specific formulation will not be disclosed to any third party without the prior consent or agreement of the customer."

162. Meanwhile, in late 2000 VF re-commenced work on an insecticide-incorporating PE tarpaulin, and also decided to develop an insecticide-incorporating PE net. Mr Vestergaard Frandsen asked Dr Skovmand to work on this project, as well as the ongoing PermaNet development.
163. At an early point in the PE tarpaulin work, Dr Skovmand consulted Ciba about the use of additives sold by Ciba to control the migration and UV stability of insecticides in PE film. It appears that Dr Skovmand met David De Corte of Ciba's Basel office in late 2000 or early 2001.
164. Dr Skovmand has disclosed six pages of manuscript notes, three of which are on Ciba notepaper, which appear to have been made during this meeting. Understandably, the notes are not very legible. Moreover, although mostly they are in English, it appears that some passages may be in Danish. Dr Skovmand has purported to provide a transcription of the notes. Unfortunately, the transcription suffers from two defects. First, particularly in the case of pages 2 and 3, Dr Skovmand has not so much transcribed as explained what he contends the note means. Secondly, even where Dr Skovmand has refrained from exposition, his transcription is not entirely accurate. For example, a reference on the first page to "[Additive J] [illegible word]" is not included in the transcript, although the rest of the page has been transcribed with reasonable accuracy as far as I can tell.
165. It appears from the notes, however, that during the meeting Dr Skovmand told Mr De Corte about the proposed application and the performance he was looking for. For example, it appears that he said that he wanted the migration of the insecticide out of the material to be a first order process rather than a second order one. It also appears that Mr De Corte discussed the following additives with Dr Skovmand and described their properties in general terms: [Additive A], [Additive B], [Additive H], [Additive I], [Additive F], [Additive C], [Additive D], [Additive J] and [Additive G]. According to Dr Skovmand, one of the passages on page 4 reads or means (words in square brackets admittedly being interpolated by him):

"The blooming is a function of solubility [of the deltamethrin in the polymer] and of migration. Migration is a kinetic process, but solubility is a thermodynamic process and the combination of these two parameters is temperature dependent. Therefore it will be impossible to predict [how to combine] and instead you had to try."

"Blooming" refers to deltamethrin rising to the surface of the yarn in the net.

166. Following this meeting, either Dr Skovmand or Mr De Corte prepared an undated document which appears to date from before mid February 2001 headed "Intelligent Insect Control – Ciba Speciality Chemicals joint project". I shall refer to this as the Joint Project memo. It appears that whoever wrote the first version of the Joint Memo sent an electronic copy to the other, and that the second person subsequently annotated and sent it back to the original author who made further annotations. As a result, there are three different versions of the document. If the different versions are compared, it appears likely that Mr De Corte wrote the first draft, Dr Skovmand annotated it and Mr De Corte annotated it again.
167. The first version of the Joint Project memo states:

"Objective: investigate the influence of Ciba additives on the migration and UV stability of insecticides in polyolefin films for tarpaulin coverings for African market.

Application/problem description: tarpaulin coverings are used to protect food bags from sun and insects in Africa. Insecticides are incorporated during the processing of the tarpaulin, and will migrate to the surface where there activity is needed. However, due to poor UV stability and uncontrolled migration of the insecticide its effect is not retained during the 2yr exposure requirement.

Program:

1) Evaluation of the Safety regulations for the handling and processing of the insecticide 'Deltamethrin' at Pontecchio.

2. If there are no safety issues the following formulations will be extruded in [REDACTED].

1) w/o [REDACTED]

     2. w [REDACTED]% [Additive A]

     3. w [REDACTED]% [Additive B]

     4. w [REDACTED]% [Additive A] / [REDACTED]% [Additive B]

     5. w [REDACTED]% [Additive A] / [REDACTED]% [Additive B] / [REDACTED]% [Additive H]

     6. w [REDACTED]% [Additive A] / [REDACTED]% [Additive B] / [REDACTED]% [Additive J]

     7. w [REDACTED]% [Additive A] / [REDACTED]% [Additive B] / [REDACTED]% [Additive I]

     8. w [REDACTED]% [Additive F]

These samples will be weathered in a WOM [Weatherometer] (% Retention of tensile strength) for a period corresponding to 1yr exposure at 180kLys.

Before and after WOM exposure samples will be sent to IIC to measure the insecticide effectiveness."

168. Dr Skovmand's annotations include the following:

"Suggestion: we will just test in a single lamellar film of LD (since this will be the outer film) to ease the process.

…

We will probably use [REDACTED] deltamethrin.

…

The first trials will be in India and the trial sample will be 25kg…"

169. Dr Skovmand made the point in paragraph 70 of his sixth witness statement that the use of additives such as [Additive A], [Additive B] and [Additive C] for UV protection was recommended in Ciba's literature, and that Ciba had obtained a number of patents dealing with the problem of halogen-containing insecticides like deltamethrin and polyethylene films exposed to sun. Ciba's literature contains little useful information about the levels or combinations of additives, however. For example, one of the documents exhibited by Dr Skovmand refers to use of [REDACTED]% [Additive A] with [REDACTED]% [Additive N] and a combination of [REDACTED]% [Additive C] and [REDACTED]% [Additive A] with [REDACTED]% [Additive N]. There is no mention of the use of [Additive A] and [Additive B] in combination.
170. Mr De Corte's recommendations both in the Joint Project memo and in the 23 August 2001 email discussed below differ from the information contained in his company's published literature. It appears that the explanation for this is that his recommendations were directed to a specific application and specific requirements explained to him by Dr Skovmand. Furthermore, it can be seen that Mr De Corte was not able simply to recommend a single recipe which he knew would work. In particular, he was not able to predict the effect of the additives on deltamethrin migration. Instead, either he suggested, or he and Dr Skovmand together devised, a number of formulations for trial.
171. The development of a PE net started in earnest in spring 2001. It appears that initially VF decided to develop the net for use as an agricultural fence to protect animals and crops against mosquitoes, tse-tse flies and other insects. Mr Vestergaard Frandsen had got the idea of using an insecticidal net as a fence from Dr Burkhard Bauer, a vet who had been working in Kenya to reduce trypanosomiasis, in early 2000. Dr Bauer's idea was for a PES net. VF tested the idea by fencing an area planted with yellow cabbages near Kolding in Denmark, and found that the netting had an immediate impact in reducing the number of flies and other insects. VF decided that a PES net would not be strong enough for this purpose, however, and decided to develop a PE net instead. According to Mr Vestergaard Frandsen, this was Mikkel Vestergaard Frandsen's idea. According to Dr Skovmand, it was his idea. It is unnecessary for me to decide who is right about this. Either way, this eventually led to the development of VF's Fence product.
172. At around the same time, VF decided to develop a PE bed net to compete with Olyset. VF thought it would be cheap to produce, and perhaps cheaper than PermaNet. Later VF found out that this was incorrect. There is a considerable dispute between the parties as to how much work was done on developing a PE bed net before the end of 2004. VF's case is that a reasonable amount of work was done, and although the development was put on the back burner, it was never abandoned. The Defendants' case is that very little work was ever done on developing a PE bed net and even development of Fence was abandoned in 2003.
173. At some point in 2001 Olyset became the first LLIN to obtain a WHOPES recommendation, which at that time was not split into interim and full.
174. On 5 July 2001 VF filed Danish Patent Application No. PA 2001 01060 to protect the concept behind Fence. This was used as the priority document for International Patent Application No. WO 03/003827 filed on 5 July 2002 and published on 16 January 2003. This names Dr Bauer and Dr Skovmand as joint inventors. The inventors executed an assignment of all rights, title and interest in and to this invention to VF on 24 July 2001.
175. On 8 July 2001 Dr Skovmand sent Mr De Corte an email asking for information about Ciba's additives. On 9 July 2001 Mr De Corte sent Dr Skovmand datasheets for the additives he was using. On 25 July 2001 Dr Skovmand sent Mr De Corte an email asking two questions, of which the second was:

"[W]hat is the way that [REDACTED] works?"

176. On 26 July 2001 Mr De Corte replied, and in answer to the second question said:

"Concerning the [Additive C], due to its composition it has a better resistance to acids and/or ag-chemicals. I can not tell you at this stage how much it will influence the blooming of the insecticide, but it looks important to investigate!"

177. After another query from Dr Skovmand, Mr De Corte replied at the end of July or beginning of August 2001:

"It matter how fast and how much the deltamethrin goes to the surface and therefore the plastic, which acts as container, should be able to retain or make the deltamethrin bloom (depended on the required conditions) without loosing mechanical properties. In this point of view, I think we have set-up the correct testing formulations and I am very interested to see the results."

178. After a further email from Dr Skovmand commenting on some initial results from the trials planned before mid-February 2001, Mr De Corte replied on 6 August 2001:

"Good to see that your pesticide effect remains after a day of QUV-B exposure. The question is then how this correlated to reality.

In these tests, did you see any differentiation in blooming speed of the deltamethrine for the 8 proposed formulations? [REDACTED].

Depending on you latest experience, tell me which formulations and quantity you will include in the real life Pakistanian trials."

179. On 22 August 2001 Dr Skovmand sent an email to Mr De Corte and Mr Vestergaard Frandsen in which he reported data apparently from the tests which had been planned:

"We have – finally – received data from the chemical analysis of the samples UV treated with UV B for 16 and 24 hours. Knowing that Gembloux found a level of [REDACTED] g deltamethrin/kg polyethylene, the results are very good

[Table setting out results for samples with no UV protection, with [Additive A], with [Additive B] and with both]

It is evident that [Additive B] by itself gives very little protection of the deltamethrin, whereas [Additive A] [giv]es a very good protection. Further, it may be speculated that [Additive B] retards the blooming of deltamethrin and that may explain the additive effect obtained by the combination of the two."

180. The email goes on to report that samples made with [Additive C] or [Additive F] instead of [Additive B] have been tested in bioassays before and after heating.
181. It is evident that there was a telephone conversation between Dr Skovmand and Mr De Corte following this email. On 23 August 2001, Dr Skovmand sent Mr De Corte an email asking for clarification of what he had said during the conversation. On the same day Mr De Corte replied:

"I referred to the active ingredients. The concentration suggested yesterday are the correct ones.

For the LDPE part of the terpauline [sic]

[REDACTED]% [Additive F]

[REDACTED]% [Additive C]

[REDACTED]% [Additive A]

[REDACTED]% [Additive B]

…

For the HDPE middle layer, you can use [Additive D]. However if the deltamethrin is present also in this part, we might consider another [REDACTED] next trial. Let's first see what happens with this formulation."

182. It is clear from exchanges between Dr Skovmand and Mr De Corte in July and August 2001 that, although Mr De Corte had been able to suggest potential additives and concentrations for testing prior to mid February 2001, Mr De Corte did not know if they would have the desired effect and that testing was required. It is also clear that Mr De Corte refined his recommendations as he received more information from Dr Skovmand, and in particular the results of the initial tests, but again he was not able to predict whether success would be achieved. Finally, it is clear that an aspect Mr De Corte was particularly uncertain about was the effect of the different additives on migration of the insecticide to the surface of the PE.
183. By September 2001, VF had engaged Elastic Enterprises ("EE") of Bangalore in India to produce PE net. On 25 September 2001 Dr Skovmand sent an email to Srinivas ("Shri") Shetty of EE an email introducing himself as follows:

"I work with Torben Vestergaard on the development of bednets, and I would like to add a few comments to the discussion on adding LD polyethylene to HD polyethylene fibers to use these for bednets."

184. On 26-27 November 2001 Dr Skovmand, Mr Larsen and Ritu Ahuja (an employee of VF's Indian subsidiary) visited EE to conduct trials in which six formulations containing bifenthrin, [Additive A], [Additive B] and [Additive C] in either HDPE or a mixture of HDPE and LDPE were extruded into yarn at [Temperature 1]. These are first six samples about which information is recorded in the Fence database.
185. The Defendants pleaded that VF's formulations had come into the public domain as a result of these trials, and the trials carried out in April 2002 (as to which, see below), but this contention was not pursued at trial.
186. On 31 January 2002 VF filed Danish Patent Application No. PA 2002 00148 to protect ZeroFly. This was used as the priority document for International Patent Application No. WO 03/063587 filed on 30 January 2003 and published on 7 August 2003. This names Mikkel Vestergaard Frandsen and Dr Skovmand as joint inventors. The inventors executed an assignment of all rights, title and interest in and to this invention to VF on 19 March 2002.
187. The specification of the published international application states [REDACTED].
188. Although the specification clearly discloses [REDACTED], it does not disclose any of the recipes in the Fence database.
189. The first version of ZeroFly tarpaulin was marketed some time in 2001-2002.
190. In early February 2002 Mr Larsen asked Anders Anden of Ciba Sweden to recommend one or two masterbatch producers close to Ahmedabad where VF was planning to produce PE film for Fence. Mr Anden put Mr Larsen in touch with Sudhir Kulkarni of Ciba India. On 8 February 2002 Mr Larsen sent Mr Kulkarni an email in which he said:

"We are in the process of developing a tarpaulin with incorporated insecticide. In order to prevent problems with the extruder at the factories and also to keep our formulations a secret we want to supply the products from Ciba as masterbatches."

In a subsequent email dated 19 February 2002 Mr Larsen said:

"I demand the following from the master batch producer:

*     He should be able to keep the identity of the chemicals a deep secret."

Mr Kulkarni responded by recommending two masterbatch producers in Mumbai.

191. Later in February 2002 VF received the results from chemical analyses of a number of the bifenthrin-incorporating PE nets carried out by FMC Corporation in the USA.
192. On 1 March 2002 Mr Vestergaard Frandsen sent Mr Kjaergaard, Mr Larsen, Dr Skovmand and others an email concerning the scheduled timing for key developments in preparation for a Product Development Committee meeting on 13 March 2002. In this email he wrote:

"3: Polyethylene net with incorporated insecticide

3 final products to be developed here:

A. Fencing.

B: Bednet

C: Netting for application to blanket.

Production trial 1^st week of April at Shri Shetty.

A: Fencing trials with UV, deltamethrin and alpha-cypermethrin in heavy yarns.

B: Bed net trials with deltamethrin and alpha-cypermethrin in light weight yarns.

C: Blanket netting trials with etofenprox and permethrin. Yarns and mesh to correspond with insecticide demands."

193. Dr Skovmand replied on 4 March 2002 saying:

"ad B: for bed net it probably has to be deltamethrin which is cheaper and more UV stable."

194. It is clear from this exchange, as well as other documents, that development of PE Fence and PE bed nets was proceeding in parallel at this date.
195. Also on 4 March 2002 Mr Larsen sent Miss Ahuja an email with regard to a planned visit to the masterbatch producers, in which he referred to the four Ciba additives which VF was using and said:

"Please do not discuss our recipe with anyone!!! I know you will not, but this is so important that I thought I better remind you anyhow.

You may also prepare the companies with which you speak, that they have to sign a secrecy agreement with us."

196. Miss Ahuja duly visited the masterbatch producers with Navneet Garg, another employee of VF's Indian subsidiary, on 6-7 March 2002. On Mr Garg's recommendation Rajiv Plastic Industries ("Rajiv") was selected. On 23 March 2002 Mr Larsen emailed Miss Ahuja a draft secrecy agreement with Rajiv. (Later, in July 2002 Mr Larsen sent Mr Shetty of EE a draft secrecy agreement in the same form which Mr Larsen and Mr Shetty subsequently signed on 5 June 2003.) It is not clear whether Rajiv signed the secrecy agreement, but it seems probable that it did.
197. In late March 2002 VF purchased additives from Ciba which were made into masterbatches by Rajiv and sent to EE for PE net manufacture and to Gujarat Raffia Industries Ltd ("GRIL") in Ahmedabad for PE tarpaulin manufacture. Because the [Additive F] had to come from Europe it was sent by Ciba direct to GRIL at the request of Mr Larsen in email dated 27 March 2002. In this email Mr Larsen said:

"… could you please change the labels on the drum and call it [REDACTED] instead? We would like to keep our recipe a secret."

198. On 2 April 2002 EE produced 9 samples, and on 3 April 2002 EE produced 5 samples, of PE net containing varying concentrations of deltamethrin, [Additive A], [Additive B], [Additive C] and [Additive G] but no [Additive F]. These are samples 7-20 recorded in the Fence database. I shall set out the full recipes for each of these samples below, but two points are important to note at this stage.
199. The first concerns the deltamethrin content. Each of the recipes specified an intended deltamethrin concentration of either [Figure 1] or [Figure 2] g/kg (stated to two decimal places). As Dr Skovmand himself pointed out for the first time during the course of cross-examination, it can be seen from the "receptor" sheet of the version of the Fence database disclosed by Dr Skovmand on 28 November 2008 that [Figure 2] g/kg was a calculated figure. It was calculated from a target concentration of [REDACTED] mg/m² in a net of [REDACTED] g/m². Similarly [Figure 1] g/kg was calculated from a target concentration of [REDACTED] mg/m² in a net of [REDACTED] g/m².
200. The second concerns the [Additive A], [Additive B] and [Additive C] contents. The recipes specified a [Additive A] concentration of either [Figure 3] or [Figure 4] g/kg, a [Additive B] concentration of [Figure 5] or [Figure 6] g/kg and a [Additive C] concentration of [Figure 7], [Figure 3] or [Figure 4] g/kg. Dr Skovmand's evidence was that these concentrations were based on Mr De Corte's recommendations, in particular in the latter's email dated 23 August 2001.
201. I accept that, when Dr Skovmand planned these experiments, he was to some extent working from Mr De Corte's recommendations in the 23 August 2001 email. It can be seen by comparing the recipes for samples 7-20 with the email, however, that Dr Skovmand did not simply follow those recommendations. Rather, he devised a set of experiments which was in part based on what Mr De Corte had said and in part on his own judgment in the light of the experience he had gained up to that point. As I have pointed out, he omitted [Additive F] even though it had been recommended by Mr De Corte and even though it had been purchased by VF from Ciba for the purpose. So far as [Additive A], [Additive B] and [Additive C] were concerned, Dr Skovmand constructed what he called a "grid" of concentrations roughly around those recommended by Mr De Corte except that the grid included no [Additive B].
202. It is not apparent either from the Fence database or from any other contemporaneous documents why Dr Skovmand chose to use the figure of [Figure 3] g/kg as part of this grid rather than, say, [Figure 8] g/kg. During the course of cross-examination Dr Skovmand came up with an explanation for his use of the same number in the Netprotect recipes (as to which, see below), which I consider is the explanation for its use in samples 7-20 in the Fence database. I shall explain this in more detail below, but in short Dr Skovmand chose [Figure 3] with the intention of avoiding problems of co-linearity in linear regression analysis of the experiments. As it happens, Mr Howe demonstrated that in the case of samples 7-20 this was an unnecessary precaution on the part of Dr Skovmand since the VIF values would have been equally low using [Figure 8] g/kg; but I do not think Dr Skovmand appreciated that at the time.
203. EE also went on to produce six samples containing etofenprox and four samples containing permethrin on 3 April 2002. These samples are samples 21-26 and 27-30 in the Fence database.
204. On 3 April 2002 Mr Larsen emailed Miss Ahuja a spreadsheet entitled "PE net trial data.xls" tabulating the various PE net samples which were the subject of these trials. Samples 1-6 in this spreadsheet are the six bifenthrin formulations which were the subject of the November 2001 trials. Samples 7-15 are headed "New trials 2, april 2002". Samples 8-20 are headed "New trials 3, April 2002". For each of these samples the entry in the second column headed "insecticide" is "deltamethrin [REDACTED]%", but a later column gives the deltamethrin concentration in g/kg. The latter figures, and the figures for [REDACTED] (i.e. [Additive A]), [REDACTED] (i.e. [Additive B]) and [REDACTED] (i.e. [Additive C]) correspond to the figures contained in the Fence database. Samples 21-26 in the spreadsheet contain etofenprox, while samples 27-30 contain permethrin. It is not clear who was the author of the spreadsheet; it may have been Dr Skovmand. Either way, the information contained in this spreadsheet was also entered by Dr Skovmand into the Fence database.
205. Samples 7-20 were subsequently chemically analysed for deltamethrin content at the Walloon Agricultural Research Centre ("WARC") in Gembloux, Belgium. The results of these analyses were entered in the Fence database, and I shall set them out below. In short, however, the analyses only found between 50 and 66% of the nominal amount of deltamethrin.
206. There are a number of documents from July and August 2002 which refer to these analyses. Dr Skovmand spent some time investigating this issue and concluded that [REDACTED]. It appears that deltamethrin has a boiling point of 190^oC, while most PE extrusion is carried out at temperatures in the range 250-300^oC. Although there are eight possible stereoisomers of the deltamethrin molecule, there are two main isomers which are referred to (somewhat inaccurately) as the S isomer and the R isomer. Of these two, the S isomer is active while the R isomer is inactive. The S isomer is converted to the R isomer by the action of heat and UV light.
207. In early July 2002 samples 7-20 (and samples 21-30) were sent to the Vector and Reservoir Control Research Unit ("VRCRU") at Salatiga for wash resistance bioassays. The results were again recorded in the Fence database and I shall set them out below. There was a dispute at trial as to the reason why these samples were subjected to wash resistance bioassays. In paragraph 19 of his first witness statement Dr Skovmand said:

"This polyethylene net [viz. Fence] was in the beginning challenged in the same way as mosquito nets by wash-dry-and-bioassay cycles. The reason was that this was the only challenge method VF had available as standard test. These challenge tests stopped in 2002 after it was decided in VF that a polyethylene mosquito net should not be developed."

The first and second sentences were repeated almost word for word in paragraph 59 of Dr Skovmand's sixth witness statement.

208. In my judgment it is clear from the contemporary documents that the reason why the sample were subjected to wash resistance bioassays was that VF was developing the PE net for use as a bed net as well as Fence. Furthermore, in a memorandum reviewing the status of the HDPE net dated 21 August 2003 to which Dr Skovmand himself referred in paragraph 55 of his first statement (and as to which, see also below), Dr Skovmand stated:

"Our bioassay has primarily been mosquito tests, but the net aims for a multi-purpose application, with mosquito control being only a small part."

209. There was also a dispute at trial as to the value of the results given that they were carried out at Salatiga. The Defendants rely upon the 2004 inter-laboratory study as showing that the results were more favourable than would have been obtained by tests carried out by LIN in Montpellier. Although this is later in time, I accept that this is so. The relevance of this is another matter, which I shall deal with below.
210. On 15 July 2002 Dr Skovmand emailed Dr Alberto Fereres of Centro di Ciencias Medioambientales ("CCMA") of the Spanish National Council for Scientific Research, copied to Mr Kjaergaard, a draft secrecy agreement for CCMA to carry out field trials of Fence. On 19 July 2002 Dr Fereres sent Dr Skovmand proposed amendments to the secrecy agreement and a draft contract. Dr Skovmand replied that he would "leave the juridical questions to Finn Kjaergaard".
211. On 24 July 2002 Ms Ahuja sent Dr Skovmand an email saying that she was producing "version no 17" but did not have [Additive G] to put in according to the recipe, and asking him to advise whether she could proceed without [Additive G]. Dr Skovmand replied, "Drop it, it apparently has very little importance".
212. On 15 August 2002 Dr Skovmand emailed Mr Larsen a copy of the Fence database as it then stood. It is clear that Dr Skovmand periodically sent copies of the database to Mr Larsen by email or transferred them to him by means of a memory stick, and that both men used them as working tools to collate and interpret results, and to plan future tests.
213. Also on 15 August 2002 Dr Skovmand sent Miss Ahuja an email, copied to Mr Larsen, in which he said:

"We have now been analysing the bioassay data on the net, and they are indeed very interesting. It seems likely that we can beat Olyset by far! But we have no data on effect of UV treatment.

So, could you please cut out 2 samples (30cm x 30cm) of the following:

7 – black, 8 –blue, 11 orange, 12 red, 17 light yellow and 19 light green.

These should be sent to Teknologik Institut in Copenhagen for UV B treatment 30 and 60 hours."

214. A similar message is conveyed by a minute dated 16 August 2002 of a meeting between Dr Skovmand and Mr Larsen which probably took place shortly before that. The minute states:

"Results on PE net with deltamethrin are very encouraging. Samples that only lasted for 4-5 washes was heated for 10 min at 80^oC (WHO std.) and have now been washed for 21 washes and still holding on.

Samples No. 7 black, 8 blue, 11 orange, 12 red, 17 l. yellow and 19 l. green to be sent to Technological Institute for 30 and 60 hours UV radiation. After radiation to be sent to bio assay and subsequently to chemical analysis. Action: OS"

215. Dr Skovmand gave evidence that, when he made the observations about the encouraging results, he was not aware that, as the 2004 inter-laboratory study showed, the results were over-optimistic because the wash tests had been carried out at Salatiga. I am prepared to accept this. Even with the benefit of the hindsight of 2004, however, I consider that the results indicated whether any, and if so which, of the samples were promising for use as a bed net. I shall return to this point below.
216. The minute dated 16 August 2002 also records discussions between Dr Skovmand and Mr Larsen about laboratory facilities in Vietnam. In about June or July 2002 Dr Skovmand had set up a VF Bioassay Laboratory at the premises of 10-10 in Hanoi. It had two entomologists on part-time secondment from the Vietnamese National Laboratory for Malaria Prevention. At the time of the meeting, VF had recruited Ms Phan as its Laboratory Manager but she had not yet started work. According to the minute, the Laboratory was then capable of carrying out 20 bioassays a day. Dr Skovmand wanted 40 a day within a month, and it was proposed to bring the capacity up to 120 a day by employing another three laboratory assistants. When Ms Phan started work in September 2002, she set up a new Chemical Laboratory at the premises of 10-10 as well as managing the Bioassay Laboratory. At that time she was the only VF employee and laboratory assistants were employed by 10-10. A second VF employee joined in 2003. Since then VF has further expanded its laboratories in Hanoi and now also has a laboratory in Bangkok in Thailand.
217. On about 22 August 2002 Mr Vestergaard Frandsen sent Dr Skovmand an email containing the agenda for the Product Development Committee on 9 September 2002. Items 5, 9 and 9a on the agenda were as follows:

"5. PE bed net. Sum up from discussion under point 4. (OS, THL)

WHOPES test on PE bed nets?

…

9: Fencing: Including update from point 4. (OS, THL)

9a: Latest development in field test and cooperation with Bukhard Bauer (FK, OS)"

218. On 29 August 2002 Mr Larsen sent Mak Onn Low of Ciba Singapore an email saying that VF was about to launch a new tarpaulin containing [Additive B], [Additive C], [Additive A] and [Additive F]. He continued:

"We need to supply these products as MB to both India and Vietnam and possibly other countries in the future. In order to avoid duty problems in India and to create secrecy around our recipe, we want to produce the MB centrally and mix all 4 product in the right ratio into the same MB.

Since you have your main store in the region for these products in Singapore, it would be natural for us to also produce the MB there.

Do you have some good and reliable MB-producers in Singapore, that could perform the manufacture, marking and shipment of the MB to our producers. It is crucial that the MB-producer we work with is reliable, because our recipe will be revealed to him. He will of course have to sign a secrecy agreement."

219. On 3 September 2002 there was a meeting between Dr Skovmand, Mr Larsen and Mr De Corte in Basel. There are minutes of the meeting prepared by both Mr De Corte and by Mr Larsen. Mr De Corte's minute is marked "CONFIDENTIAL". During the meeting there were discussions about both tarpaulins and PE net. It is clear from the minutes that the discussions were interactive, with Dr Skovmand and Mr Larsen reporting test results from the existing formulations and Mr De Corte explaining the properties of Ciba's additives and making suggestions for improvements. Both minutes record a proposal for an improved tarpaulin formulation comprising [REDACTED]% [Additive C], [REDACTED]% [Additive O] and [REDACTED]% deltamethrin in the external LDPE layers and [REDACTED]% [Additive F], [REDACTED]% [Additive O] and [REDACTED]% deltamethrin in the internal HDPE layer. The part of Mr Larsen's minute dealing with PE net includes the following:

"[Additive G] might make the fibre brittle.

[Additive A] will in such a thin fibre be gone very quickly.

UV protection: combination of [REDACTED]% [Additive O] and [REDACTED]% [Additive C] 3 years in Africa (guess)"

220. Although not entirely clear from this minute, it is clear from subsequent documents referred to below that, as Dr Skovmand acknowledged, what Mr De Corte suggested was to omit [Additive A] from the PE net formulation.
221. The minutes of the Product Development Committee meeting on 9 September 2002 include the following under the heading "PE bed net":

"Besides what has been substantially covered in the points above, this product has been finalised in Deltamethrin.

As this is a price product it is highly essential that we design protect the pyramid shape. Action FK.

This will be sent for WHOPES evaluation ASAP. WHO will make the claims themselves as they test phase one. This will be sent latest end October, and will send from either Spain or Italy."

222. On 12 September 2002 Mr De Corte emailed Dr Skovmand and Mr Larsen his minute of the meeting on 3 September and added:

"In order to understand better the effect of our additives on the deltamethrin performance, could you please send me a summary of the results you obtained in the bio-assays and other tests? This will help us to recommend a more optimal LD package for the mosquito nets."

223. Mr Larsen replied the same day, saying:

"I would like to have the actual recipes deleted from the minutes. There is no reason why this should be distributed around the world with the risk of a third party picking it up."

224. On 17 September 2002 Dr Skovmand replied, saying:

"The most convenient method for giving a fast data review is the attached file from STATISTIX in which I have included 20 analysis of 20 different formulations of tarpaulins. They include 5 ingredients, and the dependent variable is percentage remaining insecticide after 60hr UV radiation.

The data clearly says that the only significant parameter for the survival of deltamethrin is [Additive A] and the probability that [Additive B] contributes is nil. [Additive G] has no impact. Other ingredients have little impact, and surely not significant.

In a trial with impregnated fibres where 6 formulations were tested, [Additive F] and [Additive C] were varied at constant rates, whereas [Additive B] and [Additive A] were varied independently. Analysis of these showed again, that [Additive B] had no impact on deltamethrin survival, whereas [Additive A] had high impact and [Additive F]/[Additive C] also had significant impact. This was again confirmed in bioassays.

Therefore, I cannot agree totally on Davids conclusion. My suggestion will be to test 2 rates of [Additive A] in central core and laminate, and to separate the effect of [Additive C] and [Additive F]. Tests separating effect of [Additive C] and [Additive F] have already been made with fiber net, but we lack feed back from chemical analysis. It would be nice to have absorbtions spectra for all Ciba UVG filters, also for future product developments perhaps with other insecticides. Could these be forwarded in print, David?

Besides the role of these chemicals for the insecticide, UV protection of the PE should also be considered. I understand that [Additive B] or [Additive O] can play such a role and suggest that [Additive O] is chosen for an amount that only secure the PE, but for a period of 2 years. That level should be chosen as a non variable. David should suggest a level.

I further understand that [Additive F] can be replaced by [Additive P] (from my memory, the number could be another), which should be tried also.

These samples can then be tested in accelerated UV test and ALSO sent to Thomas in Kenya or a laboratory closer to Sahara to get an max UV load and then cut off pieces should be analysed after 6 and 12 months.

Meanwhile, the [Additive B] level can be reduced to the level suggested by David also in coming productions."

225. On 14-15 November 2002 Mr Larsen visited EE to conduct trials on HDPE netting and to produce HDPE netting for jarlids. The trials were conducted with etofenprox and permethrin as insecticides and [Additive C], [Additive A] and [Additive G] as additives. Samples 33.1-33.10 in the Fence database were made from "formulation no 11" with deltamethrin, [Additive A] and [Additive C].
226. On 11 February 2003 Dr Skovmand sent Miss Ahuja an email about PE nets in which he said:

"Remember to warn Valerie when you send samples somewhere, so she can add a code number and enter in a database. So far, we use one database for all PE nets without caring whether they are used for jarlids, sewed on blankets, use for tse tse nets or whatever. We may split it up later, when specific formulations are developed for each area."

227. On 3 April 2003 Dr Chang Moh Seng of WHO's Western Pacific Regional Office sent an email to Mr Garg in connection with samples of PE net which had been sent for bioassay testing, asking questions about the specifications of the nets. This was referred by Miss Ahuja to Mr Larsen, who replied to her the following day:

"We have several UV protectants build in the netting material. The complete concentration is not very informative, since the UV protectants serve different purposes. The exact recipe we can obviously not disclose, since it is a business secret of VF and also under patent application.

The complete concentration of UV protectants is approximately [REDACTED] g/kg."

228. On 11 April 2003 Dr Skovmand emailed Mr Vestergaard Frandsen, Mr Larsen and Miss Ahuja instructions for the manufacture of samples of PE net for comparison testing against Olyset and PermaNet nets. The recipe was to be [REDACTED]% [Additive B], [REDACTED]% [Additive C] and [REDACTED]% deltamethrin. Dr Skovmand also suggesting replacing the HDPE with a mixture of HDPE and LDPE, saying "that should ease extrusion at the low temperature [Temperature 1] and make the net softer".
229. On the same day there was a meeting of VF's Management Committee. The minutes record the product development status of HDPE netting as follows:

"We are impregnating HDPE for different purposes with different insecticides:

a.     For bed nets and Fencing - Deltamethrin has been used. Formulation no. 11 has been the most effective so far.

In the laboratory in Salatiga, after passing 9 washes it failed, then the same samples has been re-heated. Now the samples are still continuing and passed 25 washes.

b.     For Blanket – We are using formulation no. 28 with Permethrin …"

230. On 22 April 2003 Miss Ahuja sent Mr Larsen an email asking three questions, of which the second was:

"We also discussed that so far HDPE net, formulation no 11 has passed 25 washes and still working – Please advise why are we washing HDPE net when the washing is not required in HDPE fabric."

231. Mr Larsen replied on 26 April 2003:

"I will try to speed up the new databases, that can be distributed in a wider forum, so that you can find these results yourself in the future.

…

2.     Why do you say washing is not required for HDPE netting? The netting material is among others meant for bed nets. For fence we do perform UV tests, but these tests are very cumbersome and are therefore only performed on a very few promising versions."

232. In paragraph 19 of his first witness statement Dr Skovmand said:

"These challenge tests stopped in 2002 after it was decided in VF that a polyethylene based mosquito net should not be developed."

As the last three documents I have quoted demonstrate, this is wrong. In April 2003 the development of PE bed nets and Fence were still proceeding in parallel.

233. Also on 26 April 2003 Mr Larsen sent an email to Dr Skovmand and his secretary saying:

"We agreed sometime back to make a database, where we have reduced the information amount, in order to use this for wider distribution. I remember having seen some preliminary work on this. What is the time frame for this database? We need to make these databases for both PermaNet®, hegn and tarpaulins. Why don't we change the name of the 'hegn' database to 'fence'? It seems a little awkward to write 'hegn' in English."

234. In May 2003 there was some email correspondence between Dr Skovmand and Miss Ahuja regarding which formulations of PE net were being used for which tests. This led to Mr Larsen emailing to Miss Ahuja on 22 May 2003 a document entitled "Fence recipe No. 65.F.0503.5". (This is one of the two specific recipes claimed to be confidential in Confidential Schedule 1 to the Amended Particulars of Claim.) The document indicates that 65 refers to [REDACTED]. The list of ingredients includes "ZF1" (i.e. deltamethrin), [REDACTED] (i.e. [Additive B]) and [REDACTED] (i.e. [Additive C]). The document states at the bottom:

"This recipe replaces recipe No 11

Approved, 22 May 2003

Shri Shetty      Torben Holm Larsen

Elastic Enterprises      Ole Skovmand

     Vestergaard Frandsen A/S".

235. The Fence database indicates that this recipe was used to make samples 63-75 dated 23 and 24 May 2004, 81-93.2f dated 12 May 2003 to 17 June 2003 and 100-150 and 152-156 dated 20 June 2003 to 21 July 2003 (many of these samples being net samples rather than yarn samples).
236. On 5 June 2003 Mr Larsen and Miss Ahuja attended a meeting with Mr Shetty at EE. The minutes of the meetings include the following entries:

"… It was however made clear that for future orders VF will have the right to claim if the Deltamethrin content are low to errors in the production (high temperature in extrusion, not confirming to agreed recipe etc.)

Stitching of bed nets with Deltamethrin incorporate net fabric is at present a problem for EE…

New secrecy agreement signed during the meeting."

237. On 18 June 2003 Dr Skovmand emailed Ms Phan, with copies to Mr Larsen and Miss Ahuja, copies of two databases saying:

"To help you, I have made two files with data from sun exposure of tarpaulins and fence, first fence (all data, all bioassay, strictly confidential)

…

I have checked I can open them with Xcel, but that does not mean that Ritu also can. Let me know, Ritu, when you have Lotus on your computer, that will safe a lot of my time for updating you."

238. In June 2003 EE produced a quantity of PE bed nets for VF to send to Burkina Faso and Benin for the comparison testing against Olyset and PermaNet. VF experienced certain problems with EE, which drew the following comment from Dr Skovmand in an email dated 1 July 2003 to Miss Ahuja, copied to Mr Larsen and others:

"… I feel this project went far too fast from the first trial we made together to production, and would prefer that it goes back to lab status to solve the various problems we know we have: apparently slow KD (bauers test), high losses of insecticide in production even at the lowest temp of extrusion possible in Shris factory, leakage of insecticide in water (probably that on the surface) especially in flush test."

239. The comparison testing was a socioeconomic study paid for by WHO to test the acceptability of the nets to consumers. VF provided 1,200 nets free of charge for the test. In addition, VF paid for the condition of the nets over time to be monitored. There is a dispute between the parties regarding the results of these tests, which I shall deal with below.
240. In the summer of 2003 VF launched PermaNet 2.0. This lasted for 21 washes.
241. On 16 July 2003 Dr Skovmand emailed a recipe for Fence production to Miss Ahuja, with a copy to Mr Larsen, saying "Do not show the total receipt to anyone outside VF and IIC…". Mr Larsen replied on 4 August 2003:

"I am not happy about having both MB and net recipe on the same piece of paper, it gives away all information, please make it on separate papers in future."

242. In July and early August 2003 there was some email correspondence between Mr Vestergaard Frandsen and Dr Chang concerning some test results on jar covers made from PermaNet, Olyset and VF's PE net in which the PE net did not perform satisfactorily. This led to an email from Mr Vestergaard Frandsen to Dr Chang dated 4 August 2003 in which he said:

"We are constantly working on the improvement of the PE netting, and we have just decided to put more resources into this product. In our opinion there are so many aspects where you can use a properly working insecticide incorporated PE netting, especially with UV protection incorporated, that this product is a must for us. We shall not rest until it is fully developed."

243. When the results were forwarded to Dr Skovmand, he sent an email dated 7 August 2003 to Ms Phan, copied to Miss Ahuja and Mr Larsen, in which he suggested:

"We can add [Additive A] to the formulation, but we [k]now more than well, that when all ingredients are added as separate masterbatches, [Additive A] will increase the migration speed of deltamethrin and hence the loss in the extrusion process.

But what will happen if we use a mixed masterbatch for deltamethrin and all ingredients EXCEPT [Additive A] that is added separately as a separate masterbatch?

That may reduce the loss in the extrusion process and on the other hand help a later migration and perhaps give some UV protection and – now I am wildly optimistic – some reduced wash off … "

244. Mr Larsen suggested that Ms Phan try heating the net, which she did and found that it increased the surface concentration of deltamethrin. On 8 August 2003 Dr Skovmand commented:

"But it does show the problem I talked about yesterday. We need to find a way to increase the blooming of deltamethrin AFTER production, without having it during production, where it just leads to great losses. This is the dilemma I call the Olyset dilemma.

Since we are more intelligent, we should solve it.

The suggestion made yesterday could be a first trial: two types of premixed masterbatches.

We should also play with the conc of migration inhibiting chemicals, we are perhaps overdoing."

245. Mr Larsen replied on the same day:

"We also have the possibility to work with the somewhat lower extrusion temperatures they are using at the PE yarn manufacturer in West Bengal that Ritu visited last week."

246. Dr Skovmand responded later that day:

"Thanks, let us remember that, that could move the compromise for [Additive A]."

247. On 13 August 2003 Mr Larsen emailed Ms Phan two recipes for masterbatches, saying:

"Please be aware that these are top secret and for your eyes only. I have password protected the documents."

248. The Fence database records that sample 151 dated 18 August 2003 was "PE net" made from "formulation 13".
249. On 21 August 2003 Dr Skovmand wrote a memorandum to Mr Kjaergaard and Mr Fleuren, with copies to Mikkel Vestergaard Frandsen and Mr Larsen, containing a status report on the development of HDPE net. The report discussed bioassay data on mosquitoes, chemical analysis, tests on JarLids, tests on blankets, tests on tsetse fence, field tests on Fence and laboratory tests on Fence. Dr Skovmand's conclusion was that HDPE net was an underdeveloped product and that further work on a number of fronts was required for a successful product.
250. In paragraph 55 of his first witness statement Dr Skovmand said that he wrote this memorandum "after the company had transferred the development of this product [viz. Fence] to their Indian polymer chemist, Gopi Nath". As Dr Skovmand more accurately stated in paragraph 47 of his sixth witness statement, Mr Nath was hired after this memorandum, not before. Although it is not clear precisely when Mr Nath started work on PE nets, it was probably in 2004: the first reference to him in the documents that I can find is in an email dated 6 May 2004, although that email suggests that he was already working on the project by then. Furthermore, Dr Skovmand continued to be closely involved in the work after the 21 August 2003 memorandum.
251. On 25 August 2003 there was a Product Development Committee meeting. The minutes of this meeting state under the heading "PE – Netting":

"It [Additive A] has been put back into PE Yarn, as its removal slowed/stopped the migration of active ingredient to surface.

This has impact on all the trials we are doing on PE nets, and would have to be considered/amended in the following programmes: PE bed nets, Fencing, ZeroFly Blankets, ZeroFly Tents and Jar lids."

It also states under the heading "Matters arising":

"2) PE Yarn

TVF and THL visited one Chinese and two Taiwanese Companies in Week 36.

The Bengali yarn manufacturer is being evaluated by NG and RA.

A complete overview will be made by THL by end September.

OS has found a lab extrusion company in Germany and is taking a quotation on this."

252. On 29 August 2003 Dr Skovmand sent Mr Larsen a new recipe for yarn for ZeroFly tents incorporating [Additive A], saying that it was the only ingredient of which they had experience which could increase the migration rate of deltamethrin after extrusion.
253. In early September 2003 there was email correspondence between VF, a Vietnamese masterbatch producer called Hoang Phuong Production ("HPP") and Ciba about the production of masterbatches for ZeroFly. During the course of this correspondence Mr Larsen wrote to Tran Van Thanh of HPP on 5 September 2003:

"I am sure it is not necessary to point this out to you, but I will do it anyway: Please be aware that the recipe you are using for these 3 new MB is bound by our mutual secrecy agreement and I will not hesitate to take action, should the recipe become known to any third party."

254. On 12 September 2003, Dr Skovmand sent Miss Ahuja an email, copied to Mr Larsen and Ms Phan, saying:

"… PE yarns are now used for many things in VF, but they are basically the same, and it is optimal for me that I can calculate on any PE yarn whatever it is used for.

Therefore, the database fence will be rebaptised 'PE yarn' and all kind of yarns including those for central weave of tarpaulins will be entered in that (also old data). This database is especially made for data analytical purposes

The fence database will be continued under that name (to avoid confusion of old numbers, but will be for test se fence and agro only. I do not know where we will put mosquito net PE, but for the moment it can rest in 'fence'."

255. Shortly after this there was email correspondence between Mr Larsen, Dr Skovmand and Ms Natera concerning the databases being maintained by Dr Skovmand and Ms Natera. During the course of this correspondence Ms Natera wrote to Mr Larsen between 20 and 24 September 2003:

"… for the moment 10 databases are running: bednet, fence, pressening, blanket, tent, competitor, masterbatch, Pe yarn, THL db, tse tse (hope I have not forgotten one!)"

256. On 24 September 2003 Dr Skovmand wrote to Mr Larsen and Ms Natera, saying:

"I have been thinking some time on this problem. There are two considerations of importance in these databases, and perhaps they can be treated differently.

For me, it is important to be able to analyse data together that are related. I do not care if a PE net is used for agricultural fence, backside of carpet, curtain or whatever, as long as I know the application and thus can see that the right type of tests are applied and can draw data from any tests of PE net into analysis. This is why I have asked for this combined database."

He went on to say that data entry in the big bednet (i.e. PermaNet) database was taking a lot of Ms Natera's time and could perhaps be done in Vietnam, that he did not like the idea of having the database on the internet for data entrance because of the risk that it could be hacked into and that putting an effective firewall into 10-10 could easily stop the system.

257. Mr Larsen replied on the same day:

"I could not agree more. The solution is to move the data entry to India and give Ritu the responsibility for this. This is the kind of job she would be able to handle well. By doing that the database stays inside VF and India has quite OK data transfer speeds."

258. Dr Skovmand responded on 25 September 2003:

"no problem with that with the various PE databases, since Ritu is involved in the trials, the database form will force her to enter all data.

For the bednet data, where data are generated in Hanoi, Salatiga and Montpellier, I am not convinced it is a good idea. It would be better to have someone in Hanoi to enter data under the eyes of Chi.

…

For me, it is of course no argument that the database is on a VF computer…"

259. In a report addressed to Mr Larsen dated 31 October 2003 Dr Skovmand reported on laboratory extrusion tests he had carried out at the premises of a company called Brabender in Duisburg, Germany. The report begins:

"The extrusion test in Brabender was for me very useful. I learned a lot about extrusion being among real experts, about various types of extrusion screws, of measurements that can be carried out to see impact of various ingredients on viscosity at various temperatures (they have a special instrument for that), types of HDPE and our formulation.

These are experiences that I gained partly by listening to all their explanations and partly by working with our own formulations….

What I learned by doing was:

1.     HDPE quality – there are so many qualities available that we need to test a series of these to find one that is optimal for the temperature regime we prefer to use…

…

3. [Additive A] has a profound impact on [REDACTED] in the extruder. Without, [REDACTED]e increase to the double, but there is little difference between [REDACTED] and [REDACTED]%. If [Additive A] is removed, we should perhaps think of a [REDACTED] to decrease [REDACTED] and enabling us to run at [REDACTED]."

260. The Fence project was discussed at a Product Committee Meeting on 15 or 16 December 2003. The minutes from this meeting are not in evidence, but according to a subsequent email dated 5 March 2004 from Dr Skovmand to Mr Kjaergaard and Mikkel Vestergaard Frandsen, copied to Mr Larsen, Dr Skovmand said at the meeting:

"a realistic time schedule for the fence project is found in our proposal. We have time to make 2 tests before the summer to attain a production in late March early April. These will be no more than an interim guidance, and it will not be an optimal solution but just an improvement compared to the existing (which is poor)."

261. As Dr Skovmand put it in his 5 March 2004 email:

"As mentioned, a reasonable time schedule can be found in the EU proposal, everything else would be a gamble. VF decided to gamble…"

262. It appears that what was decided at the meeting was to try producing PE net in China for a trial in Denmark.
263. In paragraph 27 of his first witness statement Dr Skovmand said:

"VF had a specific decision in a combined marketing and product and development meeting that polyethylene nets for mosquito control would not be developed."

In that witness statement this decision is dated to 2002 (see the quotation from paragraph 19 set put above). Similarly, in paragraph 21 of his second witness statement Dr Skovmand said:

"The so-called VF PE bed net does not exist. FK refers to a product made for a sociological project in Burkina where OLYSET and PermaNet were compared. This net was specifically produced for that study to interpret preferences, not to introduce a new VF product. On the contrary, VF had the year before decided NOT to develop a PE mosquito net. No entomological data was ever collected for this net, so VF does not even know if it is worked."

Again, this dates the decision to 2002 since the Burkina study was in 2003. (I note in passing that Dr Skovmand's statement that the PE bed net was specifically produced for that study is wrong, as is his statement that no entomological data were collected for the PE bed net.)

264. The statement that VF decided not to develop a PE bed net in 2002 is clearly wrong. In paragraph 53 of his sixth witness statement, Dr Skovmand dated this decision to late 2003. By a correction to this statement, he dated it to late 2003 or early 2004. Similarly, Mr Larsen and Mrs Sig said that a decision to stop work on PE bed nets was taken at a VF Product Development Committee meeting in late 2003. The only possible candidate for this meeting is the one in December 2003.
265. Both Mr Vestergaard Frandsen and Mr Kjaergaard accepted that there came a point at which the general opinion of the Product Development Committee was to stop testing PE bed nets and concentrate on Fence, although they said that work did continue. The Fence database indicates that in 2004 VF did not pursue PE bed net development, but instead concentrated on Fence. I do not accept that VF abandoned the idea of producing a PE bed net, however. It was simply not a priority at that time.
266. On 7 January 2004 Dr Skovmand sent Miss Ahuja an email, copied to Mr Larsen and Ms Phan, in response to earlier emails commenting on some unsatisfactory test results on samples from EE, saying:

"To my mind, it is not worth the time to work on data from Shri Shetty; just move the production away from him and let us get things under control. Rob is starting rain tests now with various formulations made in Brabender, Germany, and we will see how much we loose from rain treatment with various formulations. Then we will choose in consequence of that and produce perhaps in China where we hope to find a better factory (Torben has identified). Then hopefully the nightmare of Plastic [sic] Enterprises is over."

267. In January 2004 PermaNet 2.0 passed WHOPES Phase II and thereby achieved an interim recommendation from WHOPES. PermaNet thus became the second LLIN to achieve a WHOPES recommendation.
268. On 19 January 2004 Mr Lambert of Innovene sent Mr Larsen an email, copied to another VF employee, stating:

"We are a worldwide HDPE supplier and we are involved in a project concerning Insecticide Impregnated Nets. We will have a trip to Denmark the week 5 and we would like if possible to meet you or your colleagues the 27^th of January 2004 to discuss a possible collaboration."

This email appears to have come out of the blue, but it is evident that Innovene was aware of VF's interest in this field. The email led to Mr Larsen and Mr Lambert arranging to meet at a trade fair in Geneva on 29 January 2004, although Mr Larsen's evidence was that the meeting did not materialise.

269. On 17 February 2004 Mr Larsen sent Dr Skovmand, Mr Kjaergaard, Mikkel Vestergaard Frandsen and others a time plan for the production of approximately 5,000 m² of PE net by Shangdongtex in Shanghai, China for the Fence trial in Denmark. This plan involved films being submitted to a company called Atlas for accelerated sun exposure in the second half of February, a decision on formulation at the end February, extrusion of yarn on 4-6 March and knitting of net on 6-20 March.
270. It appears from email correspondence in early March that Atlas did not carry out the UV tests as planned, but Dr Skovmand and Mr Larsen proceeded with production of PE yarn in China anyway.
271. On 6 March 2004 Mr Larsen (and Dr Skovmand) sent Mr Kjaergaard and Mikkel Vestergaard Frandsen an email reporting on the production of PE yarn in China as follows:

"As already told Mikkel verbally, our test in China was no success. We didn't succeed in forming a yarn by extrusion at [Temperature 1]. The yarns kept breaking.

There are probably several reason for these problems.…

Subsequently we've had an emergency meeting with the people from Shandongtex, and it turns out that the tests they've been making on extruding yarn with deltamethrin have been conducted in another and smaller extrusion plant. During these tests the work has been successful at a temperature of [Temperature 1].

We may therefore still have a small chance of getting some yarn produced for the fence test in Denmark. The technician from Shandongtex who made the tests at the low temperature has been involved all the time during our test, and he is competent. He has agreed to run, at the small plant, the production of the 300kg of yarn needed for the DK test.

However in order to do this we need to get some more MB from India and send it Shanghai….

It should be noted that we'll probably still have to decide on the formulation to be used in the DK test without knowing the result of the UV exposure. Quite a lot of guesswork will therefore be involved in this.

So gentlemen it is decision time. Shall we:

1.     do what we can to make the 300 kg for the DK test, under the conditions outlined above and well aware that the formulation we'll use will probably not be the correct/final formulation.

Or

2.     cancel the test in DK and thereby give Ole and me and our new man in India 1 year to find the correct formulation, the correct PE grade, the correct extrusion, net and MB supplier, etc., well aware that this will postpone the fence project for 1 year but also save us a lot of test costs on products that we won't need anyway."

272. It appears that VF decided to proceed.
273. Shortly before 20 March 2004 Dr Skovmand met Mr De Corte. It appears from a subsequent email exchange on 20 and 22 March 2004 that they discussed ways in which HDPE fibres could be extruded at lower temperatures without adversely affecting the mechanical properties of the end product and that Mr De Corte recommended use of processing additives. It is not clear whether these included [Additive M], but it seems likely. In the email dated 22 March 2004 Mr De Corte sent Dr Skovmand the contact details of Michel Grangrette, a French extrusion specialist.
274. On 24 March 2004 Mr Larsen emailed Miss Ahuja (by now Mrs Gulati) and Dr Skovmand a recipe for Fence entitled 65.F.0304.5. (This is the second of the two specific recipes claimed to be confidential in Confidential Schedule 1 to the Amended Particulars of Claim.) The format of this document is very similar to that of the document for Fence recipe 65.F.0503.5 described above. It differs because, as Mr Larsen explained in his covering email, VF were unable to get [Additive F] in time and so had to use some masterbatches that were in stock.
275. The Fence database indicates that this recipe was used for samples 260.1-260.6 dated 1 April 2004, samples 263-264, 269-277 and 281.1-288 dated 23 May 2004 to 4 June 2004 and samples 292-308.2 dated either 26 May 2004 or undated.
276. On 7 April 2004 Dr Skovmand sent Mr Larsen, Ms Phan, Mr Fleuren, Mrs Gulati and others an email with an update on Fence:

"The fence formulation that Torben and I made based on reflections and the preliminary analysis of the Brabender sheets has now been compared to the experiences gained on these Brabender samples after having the analysis of Chi and Huong on samples returned from Kenya and the UV exposure in Germany (report will follow, I still miss a few data from Huong).

Actually, the suggestion we made and already realised by Ritu in India is not too bad, and should give better results than the existing formulation used last year. Especially, the UV protection and ratio of insecticide on the surface versus inside the yarns should be much better."

277. On 20 April 2004 Dr Skovmand sent Mikkel Vestergaard Frandsen an email, copied to Mr Fleuren, Mr Kjaergaard and Ms Phan, saying:

"1. In the production of yarns, we loose from 10 to 60% of the deltamethrin, and it is not consistently related to anything known (except when we know they do not follow temperature demand in extruding).

2.     [Additive A] was originally put into yarn to protect against UV as in tarpaulin. Ciba then told us that this absolutely waste, since [Additive A] would migrate fast and then disappear. We then took it out. The result was a slower migration and loss of bio-activity. We then put it back in the last formulations, but it seems that whatever it did to migration during the extruding, the blooming deltamethrin was gone. We now await new analysis of yarns that will be presented as an accelerated migration technology. We will then see if surface extractable deltamethrin is at a higher level. As written previously to-day, I called the Indian office Saturday and arranged that such samples are forwarded to Chi.

If such a heating will increase the deltamethrin level, this fence is our best choice.

3. In the formulation tested in the field last year, there was no [Additive A]. In the new formulation, the level is [Figure 6] g/kg. "

278. On 22 April 2004 Ms Phan sent Dr Skovmand and Mr Larsen the results of some tests she had done on the effect of heating on deltamethrin in PE samples with and without [Additive A]. Her conclusions were:

"1. Heating or not, [Additive A] migrates itself very fast to the surface….

2. Thanks to its mobility, [Additive A] supports actively to the immigration of Deltamethrin to the surface…

3.     [REDACTED] plays important key on the rate of immigration of DELTA in PE…"

279. The next day Dr Skovmand sent an email to Mr Fleuren and Mr Kjaergaard, copied to Mikkel Vestergaard Frandsen and Ms Phan, commenting on these tests:

"Chi has now made a small series of test with nets from last year and nets from this year.

The conclusion is very clear. Blooming rate is 8 times in the new net than in the old net both with heating and without heating.

That means that the new net has a much higher capacity of replacing deltamethrin at the surface after UV destruction.

UV light can also destroy the deltamethrin inside the yarns. Last year, the fence material lost 50% of its deltamethrin due to weathering, and since the blooming rate was slow, the surface concentration must have been very small.

With the new net, the surface concentration will be higher since the blooming rate is higher. The loss inside the yarns will probably be smaller because of the UV filter, but this UV filter migrates very fast, so the protection may be gone before the end of the season.

In conclusion, the new net will probably work better for the first year, but none of the nets are likely to be good to use again the year after. Since this is a one year study, you should use the new nets, NOT the old."

280. On 26 April 2004 Mr Lambert telephoned Mr Larsen to follow up on the communications in January 2004. It appears that during the conversation Mr Larsen asked Mr Lambert about grades of PE supplied by Innovene. Later the same day, Mr Larsen sent Mr Lambert an email saying:

"Please be assured that we would be highly interested in cooperating with you on the incorporation of insecticides into PE.

As agreed I would encourage you to substantiate your ideas on cooperation between our two companies.

In the mean time I have a question you might help me with.

In order to scale up production we are looking for a good, stable and qualitative good manufacturing facility for PE nets in India. I have been informed that one of the largest and most modern PE net manufacturers (extrusion and knitting) is located some 2 hours drive from Bangalore in India. They are only producing for the domestic market as far as I have been informed. I have been unable to locate this manufacturing unit so far. Could you help me?

Alternatively could you help me located a good manufacturing unit for PE net in either China or India.

We agreed that a possible date for a meeting at our office here in Demark could be June 18^th in the morning."

281. On 28 April 2004 Mr Lambert replied:

"The 18^th of June is ok for me. Please find hereafter the requested information. We have in our product mix the well defined Eltex HDPE [Grade I] grade for such application."

282. As related below, Mr Lambert subsequently recommended the same grade of HDPE to Mr Larsen and Dr Skovmand for use in Netprotect, and the Defendants started to use it in March 2006.
283. On 6 May 2004 Dr Skovmand sent Gopi Nath, an employee of VF's Indian subsidiary, an email, copied to Mr Larsen, Ms Phan and a Mrs Huong, about "new fence", saying:

"We should prepare the next field tests very soon. You need to make 3 different formulations for me that we send to Kenya and Benin for field exposure and to Chi for surface extraction and surface migration rate and to Huong for bioassay and perhaps also to Dr Tring for bioassay in Plutella (diamond back moth).

The receipt will be variations of that you made last time, with lower of [Additive A] and other levels of [Additive C]."

284. On 18 May 2004 Ms Phan emailed Mr Larsen and Dr Skovmand the results of tests on HDPE yarns made in India and China and commented:

"The data from China is so good compared to the same production in India…

Delta+Risomer are found at almost 100% compared to target data. It shows very little loss in production. Please be noted that R-isomer is semiquantified only, but we can conclude that very little loss of Delta except the loss due to R-isomerisation."

285. Dr Skovmand said in paragraph 18 of his first witness statement that:

"No new formulations [of Fence] were tested after 2003, and most development work was done in 2002."

As the two emails I have just quoted show, this is not correct.

286. On 28 May 2004 Mr Larsen sent faxes to Mahaveer Fabrics, R. Kandasamy, Life Time Associates and VKA Polymers in India, each of which said:

"We are a Danish group working with among others mosquito nets for international aid.

We are looking for Indian companies that can produce HDPE nets for us. We have a technology for incorporation of insecticide into the HDPE fiber and we need a manufacturing unit to produce the material for us. "

Mr Larsen also sent emails to Nackul Plastics and SS AG International ("SSAG") in the same terms. He received a reply from Dr B.J. Fernando of SSAG and entered into email correspondence with Dr Fernando.

287. On 2 June 2004 Mr Larsen sent Rajiv an email saying:

"We are getting close to the point where we start producing PE net in large scale. Fortunately for you this will mean that our need for MB will increase accordingly.

To get this process running I am looking for new HDPE net suppliers, the one we have do not fulfil our needs for capacity and speed.

I know that there are many HDPE net producers in the Karur area in Tamil Nadu. I also know from WHO that there is at least 1 that is big and have good equipment. I need to find this supplier or any other as good for that matter.

Can you help me with this?"

288. On the same day Mr Larsen sent an email with a similar enquiry to a company called Lohia Group. When a representative of that company asked for more information, Mr Larsen replied on 11 June 2004:

"The product we are producing is a HDPE netting fabric, warp knitted. Square meter weight from 40-70g yarn diameter 0.1 – 0.5 mm. Mesh size 25 holes per square cm.

The net is used in agriculture, construction, mosquito bed nets etc."

289. It is unclear whether Mr Larsen met Mr Lambert on 18 June 2004 or not.
290. On 1 July 2004 Dr Skovmand sent an email to Ms Phan, Mr Vestergaard Frandsen, Mr Fleuren and another VF employee an email saying:

"In my last comment to UV exposures, I forgot an important point. Chi took samples made previously without [Additive A] and compared with the last productions made with [Additive A]. At same conditions, deltamethrin survive 60% with [Additive A] and 30% without, this is a clear indication that [Additive A] helps.

Our data across samples for polyethylene products (excluding tarpaulins) shows that [Additive A] survives better than deltamethrin.

Therefore, Ciba was not right stating that [Additive A] would not help and would be burned off very fast and not protect deltamethrin.

At least we can conclude, that [Additive A] is there to stay until a better UV filter is found."

291. On 27 August 2004 Dr Skovmand sent Mr Nath an email saying:

"The formation of isomers and the loss of deltamethrin can be reduced by the addition of [Additive C], but this also reduce the blooming later, so a proper choice of concentration is a delicate balance."

292. In September 2004 Mr Kjaergaard told Dr Skovmand to send the Fence database to Ms Phan in Hanoi with a copy to Mr Nath in India.
293. On 19 October 2004 Dr Skovmand sent Mr Fensbo and Ms Phan an email saying that Ms Natera would stop working for IIC at the latest at the end of December 2004. He continued:

"The transition of databases to Hanoi goes slower than I thought. And you only have the bednet database, but besides that fence and tarpaulin databases are updated regularly and then there are the smaller ones for shadenet, blankets.

Rather sooner than later, you will have to take that over also, or you decide to hire Valerie yourself to do so.

Beside the databases, she also follows samples sent for UV exposure and a few other special treatments, and survey that they return at the right time. Finally, she checks the production plans and see that she gets data from all productions.

She is doing a very good job, she is a smart and intelligent assistant, but since she left Montpellier, she mostly worked for you."

294. On 20 October 2004 Mr Fensbo replied:

"I have had a meeting with Mrs Oanh today – together with Chi.

It is still to my interest that the total database is moved to Hanoi.

I will however, need to be assured that the database is well operating before I let go of your kind support Valerie, wherefore I kindly ask you for support for another month – can you do that?"

295. After some intervening correspondence, Dr Skovmand sent an email dated 2 November 2004 to Ms Phan, copied to Mr Fensbo and other VF employees, saying:

"I have a bit the impression that the existing database – made by me – is now only used by me?! I do no more analyse production and test samples together, too many other factors are different.

So, the two databases can just as well be separated.

But I could be hit by a car tomorrow – God forbidden - and then the rest of these databases are more or less useless for a period until someone found the time to sit down and get them organised so he or she can understand.

So, I would strongly recommend that your database group speed up their work. Valerie will stop working for me January first….

I would therefore strongly recommend that you find out what info people use (you, Gopi, Lars and Benny), get an idea on when the databases are transferred and working, and if this is not before January first, find out what to do."

296. It is not clear from the documents precisely when it occurred, but Dr Skovmand's evidence was that sometime in 2004 he sent a copy of the Fence database (not including the "recepter" and "comments" tab) to Mrs Oanh, a Vietnamese employee of VF, for her to convert into a relational database.
297. On 19 November 2004 Dr Skovmand sent Mr Fleuren and Ms Phan an email saying:

"As you know, [Additive A] accelerate deltamethrin migration to the surface and also provide some UV protection. But in a meeting with Ciba, we were told that the migration was very fast (weeks) and it was the destroyed (in yarns), so it was a waste of money. We (being THL and me) then took it out, but that gave nets with lower effect despite high concentrations of deltamethrin. At the same time we found out that [Additive A] was much more stable in the product than Ciba told, and I (THL had left) then put it back.

To know what products have [Additive A] and which have not, you need to look into the fence database. In the future, [Additive A] will be a component.

Never listen to experts, always try, see and measure."

298. On 20 November 2004 Dr Skovmand responded by email to some earlier correspondence from Ms Phan and Mr Fleuren about deltamethrin levels in Fence, saying:

"I personally think that this product is not ready for registration, though I well know that this will be badly accepted in VF. We will have to prove that we have a method of production that gives a constant product, and we are not there."

299. On 2 December 2004, following earlier correspondence involving Dr Skovmand and others, Mr Fleuren gave instructions for 100 metres of Fence made in China to be sent to Kenya for exposure tests for a minimum of 145 days with monthly analysis of the surface deltamethrin level.
300. On 7 December 2004 Dr Skovmand sent Mikkel Vestergaard Frandsen and Mr Kjaergaard a "good-bye letter" resigning his consultancy, although VF did not receive this letter until 13 December 2004. Most of this letter is devoted to his reasons for resigning, but it includes the following passages:

"This move raises a lot of practical problems. The database work is already transferred to Hanoi, so this is not a problem. Nobody but my self can probably figure out in the old databases, so I will keep a copy of them and you can always ask on historical data. They will not be opened to other people and will not be used for other purposes, except perhaps to finish this article on wash methods and implications.

…

Let me take the products one by one and give you an update:

…

Fence: There has been no new developing in fence this year on the formulation. All productions during the last 2 years have been carried out by your Indian staff (lately Gopi), and I have as you only reports to refer to. The basic problem is the loss of deltamethrin and [Additive A] in production, which makes it difficult to relate formulations to bio-efficacy. Field tests are in the hands of Finn and Rob, so I have been out of this for so long."

301. Although Mr Vestergaard Frandsen replied by email on 13 December 2004, trying to persuaded Dr Skovmand to change his mind, Dr Skovmand could not be persuaded. Dr Skovmand did agree, however, to enter into a short term contract to work on PermaNet 2.5 and ZeroFly.
302. In June 2004 VF employed a full time agrochemical specialist, Niels Borre, to market Fence. VF have so far only succeeded in selling Fence in Norway in 2006, although they have also made sales of a PES version of the product in a number of other countries. VF hope that sales will grow in the future.
303. VF's evidence is that they are currently using PE net technology in a number of products. Apart from Fence, they have made small sales of JarLids, a product for protecting containers used to collect water, and PE net tse-tse fly traps. Apart from bed nets, they have field tested other three products, namely blankets incorporating permethrin, an internal wall-lining and shade nets. VF have not yet commercialised a PE bed net, although it appears that PE net is used in the roof of PermaNet 3.0 nets.
304. In December 2008 PermaNet 2.0 achieved a full WHOPES recommendation, while PermaNet 2.5 and PermaNet 3.0 achieved interim recommendations.

The terms of Mr Larsen's employment by VF

305. Clause 9 of Mr Larsen's contract of employment dated 3 November 2002 provides:

"The Production Manager shall keep all internal information about the Company and its activities confidential. He has been informed that the Company's business goals are based on methods which have been developed in the Company and are therefore to be considered business secrets."

306. Clause 10 of the contract contains a 12 month post-termination non-competition covenant.
307. A list of Mr Larsen's responsibilities in a document dated 8 February 2002 includes "Implementation of secrecy and no-competition agreements where relevant". It can be seen from the foregoing narrative that Mr Larsen diligently discharged this responsibility.

The relationship between VF and Dr Skovmand/IIC

308. I have already touched on this topic above, but since it formed one of the principal areas of dispute at trial, I must now consider it more specifically. It is common ground that it is not necessary to distinguish between Dr Skovmand and IIC, and accordingly I shall concentrate on the position of Dr Skovmand.
309. Contract? As I have already found, in late 1998 Mr Vestergaard Frandsen orally agreed with Dr Skovmand that VF would engage Dr Skovmand as a consultant on the basis that VF would pay him an hourly rate for the time spent working for VF, together with expenses.
310. It is expressly admitted in paragraph 14(A) of the Defence that:

"before or during the early phase of the co-operation with the First Claimant, Mr Skovmand entered into an oral contract with the First Claimant to develop, on a consultancy basis, disease control textiles for the First Claimants to have produced for them for sale."

In paragraph 15 of the Defence, this contract is referred to as a "contract for services".

311. Despite this, counsel for the Defendants submitted in his closing submissions that there was no contract between Dr Skovmand and VF, and that, if VF had not paid Dr Skovmand, his only remedy under English law (as to which, see below) would have been to make a quantum meruit claim. In my judgment this submission is not open to the Defendants. In any event, I reject it as being contrary to the evidence. In addition to the evidence of Mr Vestergaard Frandsen concerning the oral agreement, there is contemporaneous documentary evidence that Dr Skovmand considered that there was a contract. Thus in a report addressed to Mikkel Vestergaard Frandsen and Mr Kjaergaard dated 25 August 2001, Dr Skovmand wrote under the heading "History of VF-IIC cooperation":

"In the beginning, VF had sale of Permanet and IIC had product and production development. Using a factory in Vietnam, VF could not go on without IIC, IIC was contract (oral) bound to work with VF on nets and tarpaulins."

An email from Dr Skovmand to Mr Vestergaard Frandsen, copied to Mr Kjaergaard, dated 24 January 2002 contains a passage to similar effect.

312. Express term? In paragraph 13 of the Defence the Defendants expressly admitted the fifth sentence of the Amended Particulars of Claim, in which VF pleaded that:

"… during this period [sc. late 1998 to late 2000] the confidentiality and commercial sensitivity of the product development carried out by Mr Skovmand was well understood by him and was an issue which he had explicitly discussed with Mr Torben Vestergaard Frandsen."

313. It was Mr Vestergaard Frandsen's evidence that, from the outset of the relationship between Dr Skovmand and VF:

"OS understood that the information arising out of the work we gave him was confidential to VF, and that VF owned the rights in this information, and indeed we orally discussed both of these issues a number of times."

Consistently with the Defendants' pleaded admission, this evidence was not challenged in cross-examination. Furthermore, when Mr Vestergaard Frandsen was cross-examined about the December 2000 agreement (as to which, see below), he repeated it:

"… we already had discussed the confidentiality and the rights to the parts [semble products] that would come out of it. We did that already when Ole was taking over from Matt Holmes and John Carroll."

314. This evidence is supported by the fact that Dr Skovmand assigned all his rights to the PermaNet, ZeroFly and Fence inventions to VF. As noted above, the first of these assignments was executed in December 1999 and concerned an invention in relation to which Dr Skovmand was the sole named inventor.
315. Furthermore, Dr Skovmand's evidence in cross-examination was as follows:

"Q. You see, what I suggest is that there was never any intention

or suggestion that my clients could not manufacture nets in

accordance with recipes that you supplied in the course of

this relationship. My clients were entitled and free to make

nets in accordance with any of these recipes they felt like?

A. Sure.

Q. And they would not have to pay you any royalty in respect of

that?

A. No, there was not (inaudible).

Q. And it was not contemplated that you would be free to use any

of those recipes for your own manufacture of nets or to permit

third parties to manufacture nets in accordance with those

recipes? Correct?

A. Right.

Q. Was that yes, you agreed with me?

A. Yes; this was the general concept, yes.

Q. It would make no conceivable commercial sense that you could

take any of these recipes and simply licence Olyset, Sumitomo?

A. Of course not. That would be unethical.

Q. It would be inconsistent with your relationship?

A. Yes.

Q. So that if someone had come up to you from Vestergaard and

said, 'Dear Dr. Skovmand, As between the two of us,

Vestergaard and IIC' -- Skovmand – 'the information that is

developed in relation to the products is confidential.' You

would have said, 'Well, of course it is.'

A. Yes, probably."

316. I therefore conclude that from the outset it was an express term of Dr Skovmand's contract with VF that (i) information arising out of the work that VF gave him would be confidential to VF and (ii) the rights to the products developed during the course of that work would belong to VF.
317. VF also rely upon an oral agreement between Mr Vestergaard Frandsen and Dr Skovmand during a visit to the United Nations High Commission for Refugees (UNHCR) in Geneva in December 2000, as to which there was conflict of evidence.
318. Mr Vestergaard Frandsen's evidence was as follows. The agreement was concluded during a brief conversation between himself and Dr Skovmand outside the UNHCR building. Although Mikkel Vestergaard Frandsen was accompanying them on the visit, he was not present for most of the conversation in question since he was trying to get hold of someone inside the building. Dr Skovmand expressed concern that he was becoming dependent on the work he was doing for VF. The two men went over what had already been agreed with regard to confidentiality and the rights to the products being developed. In addition it was agreed that (i) Dr Skovmand would have a good consultancy fee for the work he was doing, (ii) VF would not employ anyone who could do the kind of work Dr Skovmand was supposed to do and (iii) Dr Skovmand would make sure he was not competing with VF in any way. The agreement was a "gentleman" agreement (it appears that the English word is used in Danish), meaning an agreement which is intended to be respected but binding in honour only. The agreement was not recorded in writing at that time.
319. Dr Skovmand's evidence was that the agreement was accurately set out in an email he sent Mr Kjaergaard on 16 September 2002. Dr Skovmand was replying to an email from Mr Kjaergaard of the same date in which Mr Kjaergaard said:

"I am now back in the office and MVK has given me your proposal for a confidentiality agreement with BF/IIC/OS and VF A/S.

…

I have looked for the present agreement between OS/IIC and VF A/S but cannot find it in the relevant folder.

I know that such agreement exists – you have mentioned this to me, and both MVF and TVF have confirmed this."

320. Dr Skovmand replied:

"There is no wr. agreement between VF and IIC so you don't have to look for it. There is only a gentleman agreement between Torben, Mikkel and myself which was made simply as follows:

vestergaard Frandsen will not use any other consultants or companies than IIC in its product development without a prior OK from me (us), and VF is the first priority and has one month to decide on products developed by or plans to develop within the business area of Vestergaard.

This obviously provides a very intimate connection of the two companies and based on this I have rejected companies which through Guillet or directly to me have proposed cooperation.

But you are very welcome to put it on paper, especially as VF's business area has become a dynamic entity."

321. This exchange appears to have led to protracted negotiations between Dr Skovmand and Mr Kjaergaard over the terms of a written agreement which lasted into 2004. No agreement was concluded, however.
322. On 22 October 2002 Gerhard Hesse of Bayer Environment Science sent Dr Skovmand an email asking a number of questions, including "does the secrecy agreement we have with Vestergaard on mosquito nets cover your activities?" Dr Skovmand replied the next day:

"… the secrecy agreement you have with VF covers my activities. The deal with VF is quite simple: the products I work on within their range of interest have a first refusal with them. If they accept, we work with them (we are actually 3 persons I IIC), if not, we can go wherever we want. bednet is clearly a VF project and so is tarpaulin."

323. On 13 December 2002 Mr Kjaergaard sent Dr Skovmand an email about the Fence patent and about the relationship between Mr Fleuren and IIC, in which he said:

"In principle, this is not a concern between IIC and VF, but in this special situation it is important that we have full clarity about the employment situation of Rob in IIC, especially because VF A/S has no formal agreement with IIC but just a gentleman agreement."

324. On 23 December 2002 Dr Skovmand emailed Mr Kjaergaard and Mikkel Vestergaard Frandsen a document setting out his thoughts on the relationship between IIC and VF. Under the heading "Present situation", this states:

"IIC has to-day an oral agreement originally made with Torben Vestergaard-Frandsen, which simply says: When IIC develops products and patents within the business area of VF, then VF has a first right of rejection. The other way round, when VF has a product idea within the expertise of IIC, they first ask IIC to help with the development.

Examples of a products well defined inside the agreement and made within the agreement: Permanet, Zero Fly.

…

Example of a product that to-day is outside the VF business area, but nevertheless developed inside this co-operation: Fence for agriculture. It is expected that VF will develop this market."

325. When Mr Vestergaard Frandsen replied to Dr Skovmand's "goodbye letter" on 13 December 2004, he said:

"So far we have had 5 years of cooperation even without a written agreement, where the main points were that we would not cooperate with someone with your skills and you would not cooperate with any competitor within textiles or plastic…."

326. In a further email dated 15 December 2004 Mr Vestergaard Frandsen said:

"Finally a very personal matter: As you have not signed a new agreement with the company there is still only one agreement between us, and that is the one you and I made verbally many years ago, where we agreed that you would never work for a competing company, that is within textiles and plastic for disease control and agriculture. If you break that promise you will set me in a very bad position here."

327. My conclusions on this point are as follows. First, there was an oral agreement between Mr Vestergaard Frandsen and Dr Skovmand in December 2000. Secondly, it was understood by both men to be a gentleman's agreement i.e. binding in honour only. Thirdly, the main focus of the agreement was the exclusivity which each party wanted. So far as that is concerned, I find that it was agreed that VF would not engage any other consultant to assist in product development without Dr Skovmand's consent and that Dr Skovmand would not compete with VF unless VF rejected a proposal of Dr Skovmand's i.e. VF had a right of first refusal. Fourthly, the question of confidentiality was taken as read between the two men.
328. Other matters. Mr Vestergaard Frandsen's evidence was that from April 1999 onwards Dr Skovmand spent about 90% of his time working for VF. Dr Skovmand's evidence was prior to the end of 2000 it was only about 20%, but that after that VF became his biggest client. Ms Phan's evidence was Dr Skovmand had told her in around 2003-2004 that about 75% of IIC's time was dedicated to VF. It was put to her in cross-examination that a more accurate figure was about 50%, based on the fact that in an email from Dr Skovmand to Mr Kjaergaard dated 10 January 2003 he said that the time he spent on VF projects annually was approximately 1,000 hours. I note, however, that in an email from Dr Skovmand to Mikkel Vestergaard Frandsen and others dated 3 July 2003, he said that for the last three years cooperation with VF had covered 50-70% of IIC's activities. In any event, I do not think the precise percentage matters. What is clear is that, particularly from December 2000 onwards, Dr Skovmand devoted a substantial proportion of his time to working for VF.
329. Mr Kjaergaard gave unchallenged evidence that VF paid Dr Skovmand approximately $1.8 million between 1999 and 2005, of which nearly $493,000 was paid in 2003. These sums include expenses, but it is clear that Dr Skovmand was well rewarded for his work for VF.
330. Mr Vestergaard Frandsen, Mr Kjaergaard and Ms Phan gave evidence that, from December 2000, Dr Skovmand was effectively VF's head of development. Counsel for the Defendants poured scorn on this suggestion, pointing out that (a) Dr Skovmand was not employed by VF and (b) Mr Vestergaard Frandsen was VF's Director of Development. As to (a), VF have never suggested otherwise. As to (b), this is true but in my judgment beside the point. The point that I understood VF's witnesses to be making was that Dr Skovmand was the person who made the technical decisions with regard to VF's development work on a day to day basis. As Mr Vestergaard Frandsen put it in his witness statement:

"OS's role was to oversee the product development of the ideas originated by me, although other VF employees … were also very much involved in the development work."

Ms Chan's evidence in cross-examination was to similar effect:

"What I meant is Ole is not a director of research and development. I mean he is head – he is really working – he is key person working for research and development. He is head of that work. I am not talking about the position."

331. So far as the development of VF's PermaNet, ZeroFly and PE net products (in particular Fence and PE bed net) during the period from about December 2000 to about December 2004 is concerned, I accept this evidence, which in my judgment is supported by the documentary evidence and, indeed, to a considerable extent by Dr Skovmand's own evidence. Moreover, it is common ground that Dr Skovmand attended VF's Product Development Committee meetings, where he presented the results of product development. VF's witnesses say that he was a member of the Committee, together with Mr Kjaergaard, Mr Vestergaard Frandsen, Mikkel Vestergaard Frandsen and Mr Larsen.
332. Mr Vestergaard Frandsen and Ms Phan also gave evidence that Dr Skovmand presented himself as a representative of VF in dealing with VF's suppliers and manufacturers and at conferences. Again, I did not understand them to mean that Dr Skovmand presented himself as an employee, but rather as someone with authority to represent VF. As Ms Phan put it, giving Ciba as an example:

"He would speak to [additive suppliers such as Ciba] regarding what additives could potentially be used in VF products… As far as I know, OS always made it clear to Ciba that he was working on behalf of VF, even though he was working through IIC."

Again I accept this evidence which in my judgment is supported by the documentary evidence and to a considerable extent by Dr Skovmand's own evidence.

333. Indeed, it is clear that some people with whom Dr Skovmand dealt thought that he was employed by VF. Thus Ned Walker of Michigan State University, who it appears was involved with some field tests in Kenya, sent an email to Dr Skovmand dated 26 March 2004 in which he said (ellipsis in the original):

"Regarding your previous emails about the involvement of the Vestergaard Frandsen group: yes of course they are involved … through you. I fully understand their interests and yours in obtaining new chemicals and trying new formulations and combinations, but Kate Aulman has clearly steered me away from that avenue of work; and in addition I have no more direct access to these chemistries than you do apparently. Actually I had thought all along that you and V-F were the same entity and that you were regularly examining new chemistries in your work but that is not so I see."

334. Similarly, Dr Medenwaldt's evidence was that, during the first couple of years of working with Dr Skovmand, Dr Medenwaldt was under the impression that Dr Skovmand was an employee of VF with an influential position.
335. This is not surprising, because occasionally Dr Skovmand was held out in VF documents as if he were an employee. For example, the Fence recipe 65.F.0503.5 and 65.F.0304.5 documents (which were drawn up for use by EE) hold out Dr Skovmand as being an employee of VF alongside Mr Larsen.

The information contained in the Fence database

336. I have described the Fence database in general terms above. I must describe the information contained in the Fence database in more detail. In doing so I shall draw upon the analysis of Mr Howe, which as noted above was not challenged in cross-examination.
337. In the form in which it was disclosed by Dr Skovmand to VF, the Fence database is a spreadsheet with four tabs, namely (i) "hegn database", (ii) "UV analysis", (iii) "analysis result" and (iv) "materbatch [sic]". In compliance with one of the two conditions for admission of Dr Skovmand's seventh witness statement which I imposed on 25 November 2008, Dr Skovmand disclosed a more complete version on 28 November 2008 which includes two additional tabs, "recepter" and "comments". This version was reconstructed by Dr Skovmand from the version he sent to Mrs Oanh in 2004.
338. Attention was primarily focussed on the "hegn database" tab, and hence references to the Fence database are to this tab unless otherwise specified. This has 76 columns A-BX and 706 rows (of which four are taken up by the headings to the columns). Of particular interest are the following columns:

i) A – "Sample NR".

ii) B – "Date". This generally appears to refer to the date of production of the yarn.

iii) C – "Type". This generally provides an indication of the purpose of the sample.

iv) D – "recipe". From row 40/sample 31.1 onwards this is generally used to record formulation or recipes numbers for repeat samples manufactured using formulations set out earlier.

v) M – "insecticide". This records the type of insecticide.

vi) Q – "ins. i g/kg". This records the amount of insecticide introduced into the formulation in g/kg.

vii) Y – "[REDACTED] [Additive A]". This records the nominal [Additive A] content in g/kg.

viii) Z – "[REDACTED] [Additive B]". This records the nominal [Additive B] content in g/kg.

ix) AA – "[REDACTED] [Additive C]". This records the nominal [Additive C] content in g/kg.

x) AB – "[Additive G]". This records the nominal [Additive G] content in g/kg.

xi) AC – "LD pure". This records the LDPE content in g/kg.

xii) AD – "[Additive D]". This records the nominal [Additive D] content in g/kg.

xiii) AE – "[Additive F]". This records the nominal [Additive F] content in g/kg.

xiv) AF – "[Additive E]". This records the nominal [Additive E] content in g/kg.

xv) AI – "analyse where". This identifies the laboratory or laboratories which analysed or bioassayed the sample. These are typically identified as "VCH" (VF Chemical Laboratory, Hanoi), "VBH" (VF Bio Laboratory, Hanoi), "Sal" (the VRCRU in Salatiga) and "Gembloux" or "Gem" (the WARC in Gembloux).

xvi) AK – "analysed deltamethrin content g/kg". This records the deltamethrin content found by chemical analysis of the sample.

xvii) AO – "% af tilsat". This records the analysed deltamethrin content as a percentage of the nominal amount.

xviii) BH-BK – "after UB B radiation 30 Hr mg/kg", "60 hr", "30 HR%", "60 HR%". These columns contain the results of UV exposure tests after 30 hours and 60 hours in terms of deltamethrin content.

xix) BN and BO – "90%" and "%". These columns refer to columns I and J in the "analysis result" spreadsheet. They set out the number of washes at which performance of the net reduces to a 90% and 80% mortality value.

339. Samples 7-20. The information which is primarily relied upon by VF in this case is the information relating to samples 7-20, and in particular samples 8, 9 and 13. The recipes for samples 7-20 as recorded in the Fence database are set out below, with samples 8, 9 and 13 highlighted:

[REDACTED]

340. I have set out the deltamethrin concentrations in this table as they appear in column Q, but if one clicks on the cells in question, in most cases the spreadsheet displays the concentration to 14 decimal places.
341. The nominal and analysed deltamethrin contents for these samples recorded in the Fence database are as follows:

[REDACTED]

342. In the Defendants' closing submissions it was pointed out that the figures for analysed deltamethrin for samples 9 and 11 recorded in the database are calculated from two other figures. It was suggested that this was because the samples had been analysed twice, once at Gembloux and once at VF's Hanoi laboratory. No evidence was cited in support of this submission, and it is difficult to reconcile with the fact that the calculation takes the ratio of the two figures and not the average. In any event, I do not consider that this point is of any particular significance.
343. The database also records the average of the analysed/nominal figures as [REDACTED].
344. The wash test data recorded in the Fence database for samples 7-20 are as follows:

Sample no.	90%	80%
7	4	5
8	7	18
9	8	18
10	6	12
11	8 and 31	21 and 39
12	6	9
13	7	10
14	5	7
15	0	0
16	4, 24 and 42	6, 33 and 41
17	5 and 24	16
18	4	13
19	5	13
20	3, 24 and 36	4, 31 and 38

345. It can be seen from the "analysis result" tab that the reason why there is more than one result in the case of some of the samples is that the sample has been heated after falling below the 80% or 90% level and the additional numbers represent how many washes (inclusive of earlier washes) it took to reduce to that level again.
346. 65.F.0503.5 and 65.F.0304.5 recipes. The Fence database also contains details of the 65.F.0503.5 and 65.F.0304.5 recipes for Fence. These two recipes are the ones which the database records as having been most frequently tested during 2003 and 2004. The 65.F.0503.5 recipe was [Figure 4] g/kg deltamethrin, [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C]. The 65.F.0304.5 recipe was [Figure 9] g/kg deltamethrin, [Figure 6] g/kg [Additive A], [Figure 10] g/kg [Additive B], [Figure 11] g/kg [Additive C] and [Figure 12] g/kg [Additive F].
347. Other information. According to Mr Howe's analysis, there are some 69 unique formulations resulting from differences in insecticide type, insecticide content, [Additive A] content, [Additive B] content, [Additive C] content, [Additive D] content, [Additive F] content, [Additive E] content, [Additive G] content, HDPE content, LDPE content and "LD pure" content. "LD pure" appears to mean LDPE added as a separate ingredient rather than as a separate masterbatch binder.
348. Deltamethrin is used in 57% of the unique formulations, [Additive A] in 73% of the unique formulations, [Additive B] in 46% of the unique formulations and [Additive C] in 97% of the unique formulations. Deltamethrin is used at concentrations ranging from [Figure 1] g/kg to [Figure 9] g/kg with the main concentrations being [Figure 1], [Figure 2], [Figure 13], [Figure 14] and [Figure 4] g/kg. [Additive A] is mainly used at concentrations of [Figure 6] g/kg or greater, most commonly at [Figure 15] g/kg or [Figure 4] g/kg. [Additive B] is almost always used at a concentration of [Figure 6] g/kg. [Additive C] is predominantly used at concentrations of [Figure 15] g/kg or [Figure 4] g/kg.
349. The Fence database contains some 674 samples of which a number have missing or incomplete data. Of the 665 samples for which there is data, 95.5% incorporate deltamethrin, 79.4% incorporate [Additive A], 55.3% incorporate [Additive B] and 98.9% incorporate [Additive C]. The only other additive used with any real frequency is [Additive F], used in 28.3% of samples.
350. Many of the deltamethrin-containing samples are based on a deltamethrin content of either [Figure 2] g/kg (38.9%) or [Figure 4] g/kg (26.1%). The [Figure 4] g/kg level is particularly associated with samples for use as Fence.
351. Most of the deltamethrin-containing samples are based on a [Additive A] content of [Figure 5], [Figure 6] or [Figure 15] g/kg. Most of the deltamethrin-containing samples are based on a [Additive B] content of [Figure 5], [Figure 6] or [Figure 10] g/kg. In 59% of all samples [Additive C] is used at a concentration of [Figure 4] g/kg.
352. The data contained in the Fence database demonstrates that [Additive A] acts as a "high migration" additive which increases the migration of deltamethrin from the bulk to the surface of the fibres, producing an increase in the number of washes before the mortality rate drops below 80%. It also demonstrates that [Additive C] acts as a "low migration" additive which reduces the migration of deltamethrin from bulk to surface.
353. 39 samples were subjected to wash tests. Of these, 24 used deltamethrin in 14 unique formulations. All these samples are based on a deltamethrin content of [Figure 1] or [Figure 2] g/kg. All of the samples which were washed more than 21 times are based on deltamethrin. As noted above, three samples reached or got close to 20 washes before mortality dropped below the 80% mortality threshold, namely samples 8, 9 and 11. The best performing sample was sample 11.
354. The best performing sample in four and eight month outdoor exposure tests and in one month Kenyan exposure tests was sample 9. The best performing samples in UV-B exposure tests were samples 7 and 8.
355. The "Comments" sheet records certain conclusions drawn by Dr Skovmand from his analyses of the data, in particular that [Additive A] reduces the pressure in the extruder and that [Additive C] reduces the formation of the R-isomer of deltamethrin.

The [REDACTED] formulation

356. A formulation which is not included in the Fence database, but does feature in the Netprotect database, is one referred to by VF as "[REDACTED]". This is a recipe containing [Figure 2] g/kg deltamethrin, [Figure 6] g/kg [Additive A], [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C].
357. Mr Howe gave unchallenged evidence that the wash test data in the Fence database showed that an additive ratio of [Figure 3]/[Figure 6]/[Figure 3] was better than [Figure 4]/[Figure 6]/[Figure 3], that this could be explained on the basis that reducing the [Additive A] concentration was beneficial in reducing the rate of migration, and that another way to achieve the same effect would be to increase the [Additive C] concentration. This reasoning would indicate that [REDACTED] was a formulation which was well worth trying.

The Netprotect database

358. As noted above, Dr Skovmand created and compiled a database recording the development of Netprotect which is very similar in structure to the Fence database. As also noted above, a number of different versions of, or extracts from this database, were disclosed by Dr Skovmand during the course of the proceedings. For the reasons discussed in Appendix 1 to this judgment, two of the three extracts originally produced, Extract DB oct04.xls and Extract DB Jul05.xls are of doubtful provenance and cannot be relied on. Extract DB Apr05.xls consists of three tabs, "A", "bioassay" and "Karur 0705". VF have not questioned the reliability of the A tab so far as it goes.
359. The final version of this database produced by Dr Skovmand on 30 November 2008 consists of three tabs, "A", "bioassay" and "comments and recepts". The A tab is similar to the A tab of Extract DB Apr05.xls, but there are differences in both presentation and content. The principal difference in terms of content is that it includes samples 123-139 which are not included in the A tab of Extract DB Apr05.xls. Furthermore, there is slightly more information about some of the later samples, such as the date of samples 114, 116-117 and 120-122. The bioassay tab contains considerably more bioassay data than anything which had been disclosed prior to that point. The "comments and recepts" tab turns out to have been the source of two documents which had previously been disclosed before which are discussed below.
360. Unless otherwise stated, references in this judgment to the Netprotect database are to the A sheet of the 30 November 2008 version.

The development of Netprotect

361. In determining how Netprotect was developed, I have been faced with three difficulties. The first is that I have been convinced by VF that the account of the development given by the Defendants, particularly in Confidential Schedule 2 to the Defence, but also in the evidence of Dr Skovmand and Mr Larsen, is inaccurate.
362. The second is that I have been convinced by VF that a number of the documents disclosed by the Defendants to support their account are of doubtful provenance and cannot be relied on. This is true not only of Extract DB oct04.xls and Extract DB July05.xls, but also at least one other document. My reasons for reaching this conclusion are set out below.
363. The third is that on 27 January 2009, towards the end of the evidence, I invited both counsel during their closing submissions to provide me with their side's respective proposed findings of fact. I made it clear to counsel for the Defendants that I considered the most important factual question in the case was how the Defendants had developed their product, and that so far as that was concerned what I was expecting was a historical recitation. In the event, counsel for the Defendants did not provide me with such a history in his closing submissions. I infer that the reason why he did not do so was because the state of his clients' evidence did not enable him to.
364. I have therefore had to reconstruct the story myself as best I can. In doing so I have placed primary reliance upon the documentary evidence, but excluding the documents which I have concluded cannot be relied on. In this regard, it is important to note that VF did not suggest that the Netprotect database itself (as distinct from the doubtful Extract DB documents i.e. other than tab A of Extract DB Apr05.xls) was not an accurate record as far as it went. I have been assisted in understanding the documentary evidence, and in particular the Netprotect database, by the evidence of Mr Howe.
365. The story starts in about September 2003 when, on their own account, Mr Larsen and Mrs Sig discussed leaving VF during a visit to Vietnam. This led to a discussion about what to do afterwards. They noticed that many visitors were taking lacquer goods home from Vietnam and decided that there was a commercial opportunity to be explored.
366. Mr Larsen and Mrs Sig appear to have started serious planning for their new business during a trip to South America in March 2004. Their initial plans are set out in a memorandum headed "Lacquer products – thoughts from a bar in Caracas" dating from 10 March 2004. This memorandum contains an entry headed "Mosquito nets" which includes the following passage:

"Our aim is to rope in 1 customer for starters: IKEA. This is to create sales and provide capital.

- price of IKEA net. Trine to buy 2 nets in Aarhus. Torben to negotiate with Shandongtex and take 1 net with him to VN and talk to MK/Tan Quang/10-10/a.o. perhaps a cautious chat with C.K./Akrungaroon."

367. It appears that at this stage Mr Larsen and Mrs Sig were thinking about non-insecticidal mosquito nets. Nevertheless, it may be noted that the suppliers referred to were all actual or potential VF suppliers or subcontractors. Shangdongtex was the Chinese producer which Mr Larsen and Dr Skovmand visited to carry out PE net production trials in March 2000. MK (Minh Khai) was a sub-contractor to 10-10, VF's Vietnamese PermaNet manufacturer. Tan Quang was a supplier of materials to 10-10. CF Trading was a Thai manufacturer which made PermaNet for VF until 2004. Akrungaroon was another Thai supplier which had supplied fabrics and nets to VF until around 2001.
368. Some time in the spring of 2004 Mr Larsen told Dr Skovmand of his plans to leave VF and set up in business with Mrs Sig. According to Mr Larsen, Dr Skovmand said that he was no longer satisfied with his cooperation with VF, and that he had a new idea for an impregnated PES net using polyacrylate as a coating agent. On 23 March 2004 Mr Larsen sent Dr Skovmand an email telling Dr Skovmand to use his Yahoo email address for "private stuff".
369. On 15-16 April 2004 Mr Larsen and Dr Skovmand met in Montpellier. Mr Larsen's minute of the meeting states:

"Purpose of the meeting: To discuss the development of a LLIN based on alternative technology and how to speed up this process. To discuss the set up of a company to handle this business and how to handle the situation that THL has a competition clause in his contract.

We agreed that it is necessary for us to have our own laboratory set up. We need to separate from VF now in order to avoid future claims that the technology development was paid by VF. There are the following needs:

Physical textile laboratory:

Within a time frame of 2-3 month a fully operational textile lab will be installed at Ole's new house in Montpelier (MP)

…

Until the lab is established in MP we will be using the facilities at TEKO in Herning.

Chemical analysis laboratory:

It is not practical to install an HPLC at the lab in MP, since we will have to install this equipment at the production site in the near future. We have the following alternative possibilities:

-      MP University …

-      Rudolf Chemie …

-      We ask Chi to look for external analysis possibilities in Hanoi at institutes, universities etc. This will require that we let Chi in on our project. We both agree that we should do so anyway. We might need her to remove some records from the laboratory and we could easily use her for setting up lab facilities in the region. OS and THL will meet with Chi on the 15^th May in Hanoi….

-      Hanoi Chemical Lab, Dr Lan. We used this laboratory in VF before we had our own HPLC. This is last possibility since we were not too happy with the results.

Bioassay laboratory:

The future solution is to install an insectarium at the lab in MP. ..

For now we will use OS contacts at bioassay lab in Burkina Faso. We need to invest in a washing machine for the lap in MP immediately, since the lab in BF cannot do washing. … We could run wash test at LIN for a limited period of time.

For the laboratory impregnations that we are doing now, we will use the following tests: Chemical analysis at 0 and 6 washes, Bioassay at 0 and 6 washes. We will use bioassay at 0 washes as criteria for go/no go decision on the individual samples in order to save laboratory time and cost.

THL/TS will run a large series of impregnations in the lab in Herning. We will work with the following parameters: Deltamethrin/Alfacypermethrin, +/- [Additive A], 80/120/150°C, Rucoplast PES/Rhodia DS910/Rhodia duramul/Dicrylan AC at different concentrations. Detailed test plan has been discussed and agreed. Dicrylan AS has been ordered from Ciba and Rucoplast PES from Rudolf Chemie. Trials will be run when chemicals have been received, probably in week 18. Concentrates containing insecticide, solvents, dispersant and UV filter were produced. Action THL/TS

...

One of the most critical success factors for our new company is how we overcome the competition clause in THL's employment contract with VF. There are basically 2 scenarios:

A.      THL gets fired from VF. In this case there are no problems

B.      THL has to resign from VF. In this case he will be formally restricted from working in areas that are in competition with VF.

…

Other actions:

TS/THL will visit Tan Quan in HCMC in week 19 to investigate the equipment at the factory and the interest of the company in working with us on the LLIN project.

THL will visit Shandongtex in Shanghai on the 10^th May, to inspect a net factory with a capacity of 300.000 pcs. Per month."

370. Dr Skovmand explained in cross-examination that the reason why he and Mr Larsen had contemplated asking Ms Phan to remove records from VF's chemical laboratory in Hanoi was to remove evidence that VF had carried out certain tests on PES nets. He accepted that it would have been a dishonest act if it had been carried out.
371. In paragraphs 24 and 25 of his first witness statement Dr Skovmand said:

"24.     I wrote a letter (which I have been informed by the Defendants' solicitors VF have exhibited at ('TVF1') outlining my frustrations. The response (also I am informed exhibited at 'TVF1'), of course, came from Torben VF.

25.     I later became aware that two of VF's employees, Torben Holm Larsen ('Torben HL') and Trine Sig, were setting up on their own as 'Intection' to produce mosquito netting and wanted me to help… "

372. The correspondence referred to in paragraph 24 is Dr Skovmand's letter dated 7 December 2004 and Mr Vestergaard Frandsen's reply dated 13 December 2004. In context, therefore, Dr Skovmand's statement in paragraph 25 "I later became aware" means later than 13 December 2004. Consistently with this, the impression given by the first witness statement is that Netprotect was developed during 2005. In fact, as the minute of 15-16 April 2004 shows, Dr Skovmand was already collaborating with Mr Larsen and Mrs Sig by then.
373. Dr Skovmand's sixth witness statement is closer to the truth in this respect, but still not accurate:

"29.     … in the early summer of 2004, Torben approached me to see if I would like to assist a new company that he planned to create. He had thought about something I had said to him some time earlier, which is that I had an idea to develop a polyacrylide insecticide treated mosquito net…

30.     … I went to a meeting of potential investors in Kolding. I thought the time planning which was presented to the potential investors was very optimistic and totally unrealistic, which I told them. I suggested that the first product should be a polyethylene net…

155.     I had a first meeting with the group behind Intection in August, 2004. … The recipes given to Torben before the meeting were for polyester nets, not polyethylene nets…."

In fact, as I shall relate below, the meeting at which Dr Skovmand suggested that a PE net should be developed took place in May 2004.

374. On 26-28 April 2004 Mr Larsen and Mrs Sig carried out PES impregnation tests at TEKO in Herning, Denmark. Mr Larsen said that this work was done "in our spare time and not as part of our work for VF". At that date neither Mr Larsen nor Mrs Sig had handed in their notice, and Mr Larsen was employed to develop and produce PES nets for VF.
375. Sometime in May 2004 Mr Larsen, Mrs Sig and Dr Skovmand met to discuss the business plan for the new LLIN business. During this meeting Dr Skovmand suggested that the new business should develop a PE bed net, and estimated that Intection would be ready to start selling such a net by the beginning of December 2004. Mr Larsen, Mrs Sig and Dr Skovmand agreed that Intection would launch a PE net first due to uncertainty over how long it would take to develop a PES net. It is clear that they were optimistic that they could develop a PE net very quickly.
376. A budget for Intection dating from 9 June 2004 provides for US$15,400 worth of sales of nets in December that year. Mr Larsen and Mrs Sig sent this to Dr Skovmand on 12 June 2004. This budget was, then or subsequently, used as part of Intection's business plan to raise finance both by way of equity investment from Mr Johnsen and Mr Ringberg-Larsen and also by way of bank loans. Mrs Sig accepted that, for that reason, it was quite important that it should be as accurate as possible. She said that the sales projection was based on information she had received from Dr Skovmand and Mr Larsen as to when Intection would be able to start selling the nets.
377. On 15 July 2004 Dr Skovmand sent Mrs Sig an email saying "I have sent Torben a suggestion in principle to PE net formulations". The document containing the suggested formulations has not been disclosed by either the Defendants or Dr Skovmand.
378. On 3 August 2004 Mrs Sig sent Dr Skovmand an email saying that a trip to India to visit potential PE net producers was planned for week 35. In the event it appears that Mr Larsen and Mrs Sig visited India from 24 to 29 August 2004.
379. On about 3 August 2004 Mr Larsen sent an email under the pseudonym Hans Johnson of Intection to Dr Fernando of SSAG (with whom he had previously been in correspondence under his own name on behalf of VF) enquiring about PE net production. On 3 August 2004 Dr Fernando replied, asking from where Mr Larsen had found SSAG's name and address? Mr Larsen replied (under the pseudonym):

"Your company were recommended to me by WHO and I found your details on the Internet."

380. At about the same time Mr Larsen sent emails under the pseudonym to other producers in Bangalore and Karur, including Nackul, enquiring about PE net production.
381. As noted above, VF required Mr Larsen to return his laptop computer at the end of July 2004. He asked for and was given a replacement, which he returned at the end of August 2004. When the second laptop was returned, it appeared that it had been wiped clean. Despite this, in late 2005 a data recovery company engaged by VF was able to recover a number of deleted documents from this laptop. One of these was a chain of emails dated 10 and 11 August 2004.
382. The first email in the chain is an email dated 10 August 2004 from Rune Bosselmann, Dr Skovmand's son, to Dr Skovmand, in which he comments on Intection's business plan (a document which does not appear to have been disclosed by the Defendants):

"I share your point of view as regards its time complications. There is no product yet, but nevertheless sales are expected from October. You say that it can easily take one year before you have a finished and registered product. That is inconsistent with the fact that Trine is expecting to receive salary already from August. I cannot see what kind of sales and marketing she should do in the time. So you have to ask how sensitive you are (Trine in particular) to a postponement of income. I do not know how far you have come with the development of the product…"

383. On 11 August 2004 Dr Skovmand replied (emphasis added):

"We completely agree. The problem is that Torben and Trine will then try to start with a Polyethylene net (as opposed to Permanet, which is a polyester net). Due to the fact that Sumitomo's Olyset net (approved by WHO, but is a rather inferior product based on PE but tested with acetone extraction and not washes as stress factors) is racing ahead, they believe we can get in that market. The problem is then that PE nets I cannot make in my lab. They can only be made in a factory, i.e. someone must make a trip out there, make tests and have them evaluated in bio and chemical assay. The travel costs in that connection must be paid for, and the test lab must be paid. Torben and Trine were in South America – a market VF has not really touched – and have identified some very interested customers at government level. Based on our Fence tests we can make some good guesses as to how a formulation should be, but we had huge problems with the then Indian manufacturer, who could not control his production (last time he delivered unimpregnated nets, which in bioassay and later in chemical analysis turned out to be full of deltamethrine). Torben has during a trip to India identified some major and hopefully better manufacturers. But it is still uncertain when such a product can be in place, i.e. developed and tested for washability and continuous activity, so I have suggested I join with capital on 1 January but do not stop earning money with VF until Intection has a product that has been tested and considered OK. The WHO approval may come later, precisely on the South American market it is not so important – as opposed to the African, donor-dominated market. As regards the rest of your comments, I can only agree to them, and I will forward them to Torben."

384. Dr Skovmand forwarded this email to Mr Larsen, who replied later the same day (emphases added):

"A fact that you seem to ignore is that we are pressed for time because there is a window of opportunity in the market at present as everybody is screaming for a competitor to Permanet. 1 December is the time when we expect sales of PE nets in the budget – that is what Ole estimate at our meeting in May. If this is to be realised Trine can not just sit back and do nothing until then; a proper preparatory sales effort has to be made. Furthermore she will lose her hold on the market and her contacts if she does not start contacting the market already in September/October. … That we have chosen to launch a PE net first, and Ole, do not say that it is Trine and I who want that, at the meeting in May you agreed to make that part of our plan, is precisely due to the uncertainty about the time schedule for the development of the PES net. At the same time we know that the product we can make will be better than the Olyset net in several regards (mesh, weight, efficiency, regeneration and possibly fire resistance)."

385. Mr Larsen accepted that the last sentence I have quoted from his email was based on his knowledge of VF's PE net. I would go further: the only possible basis both for Dr Skovmand's view that the Fence tests enabled them to make some good guesses as to a formulation and for Mr Larsen's certainty that the product would be better than Olyset is the information in the Fence database, and in particular the bioassay data in respect of samples 7-20.
386. On about 27-28 August 2004 Mr Larsen and Mrs Sig visited Karur to find a PE net manufacturer. One of the manufacturers they visited was Mr Sivasamy, who they selected as Intection's manufacturer. On 28 August 2004 Mrs Sig signed a secrecy agreement on behalf of Intection with Mr Sivasamy trading as Sivalakshmi Plastics & Shobika Industries in respect of "the collective technologies of [Intection] for long lasting impregnation/incorporation of insecticide on/into different materials".
387. In the first week of September 2004 Mr Larsen and Mrs Sig visited Dr Skovmand in Montpellier. While they were there, they visited Sophie Guillot of the University of Montpellier ("UMP") to discuss chemical analysis of PE net samples.
388. On 14 September 2004 Mr Larsen under the pseudonym Hans Johnsen sent Vikram Bhadauria of Alok saying:

"We are a European company working with incorporating insecticide into HDPE and have developed a technology that can be used in mosquito nets….

At the moment we are setting up a production facility in India hence therefore we are looking for suppliers of master batches within India."

389. Amit Puri of Alok replied on 15 September 2004, saying that Alok was one of India's largest producers of masterbatches and had plants in New Delhi and in Silvassa near Mumbai. Mr Larsen responded on 16 September 2004 saying that he was flying into Mumbai late on 22 September 2004 to visit masterbatch producers in Mumbai and then flying onto Ahmedabad.
390. On 20 September 2004 Mr Larsen (under the Hans Johnsen pseudonym) established email contact with Parikshith Jhaver of Tagros Chemicals India Ltd ("Tagros"), a pesticide manufacturer. On 22 September 2004 he asked Tagros to quote for the supply for deltamethrin and alpha-cypermethrin.
391. On 23-24 September 2004 Dr Skovmand attended a meeting in Johannesburg where he met Mr Lambert.
392. According to Dr Skovmand's account in paragraph 33 of his first witness statement:

"I had two years of chemistry at University and 29 years experience in formulations, but I decided I knew too little about the plastics themselves. Therefore, in 2005, I contacted Yves Lambert of Innoven, a leading producer of polyethylene and partly owned at that time by Solvay who develop new polyethylene types, for help. I met Mr Lambert at the meeting in Johannesberg where I had the final discussion with Mikkel VF."

393. Although it was not subsequently corrected by Dr Skovmand, on any view the date of 2005 is inconsistent with the date of "second half of 2004" given for the discussion with Mikkel Vestergaard Frandsen by Dr Skovmand in paragraph 23 of the same statement. This appears to have been something of a Freudian slip, since the chronological context of paragraph 33 in Dr Skovmand's first statement is that (i) Dr Skovmand did not start working with Intection until after December 2004 (which as discussed above is wrong) and (ii) he contacted Mr Lambert after the initial results from tests on preliminary formulations (which are not dated in the first statement, but in fact took place in October 2004 as discussed below).
394. Dr Skovmand's account in paragraph 63 of his sixth witness statement is chronologically more accurate:

"During a malaria meeting in South Africa in September 2004, where I had participated for IIC (not VF), I was lucky to meet Mr Yves [i.e. Mr Lambert] from Innoven, with whom I spoke. I got his business card. Innovene is one of the biggest polyethylene producers in Europe and they also make new polyethylene types, specially developing them for customers. I contacted Mr Yves in October 2004 and went with Torben to meet him in November 2004 to discuss with him and his colleagues what sort of polythene mixes would be appropriate for making insecticide-filled blankets."

Similarly, Mr Larsen says in paragraph 32 of his fifth witness statement:

"Ole had discovered from talking to Innoven that the crucial factor was the plastics mix and had Netprotect's plastics mix specially formulated for the product by Innoven. I met Yves Lambert of Innoven in November 2004 with Ole."

395. While I accept that the first time that Dr Skovmand met Mr Lambert was at the meeting in Johannesburg in September 2004, I do not accept Dr Skovmand's story that this was a chance encounter. Nor do I accept that Mr Larsen met Mr Lambert for the first time in November 2004. As related above, but not mentioned by Mr Larsen in his statement, Mr Larsen was in email contact with Mr Lambert in January and April 2004 even assuming that they did not actually meet in either January or June 2004. During the course of Mr Larsen's contact with Mr Lambert on behalf of VF, Innovene had recommended a particular grade of HDPE as being suitable for PE nets which is the grade of HDPE that Bestnet subsequently used for Netprotect. I infer that Mr Larsen put Dr Skovmand in touch with Mr Lambert.
396. Mr Larsen and Mrs Sig visited India again from 23-27 September 2004. During this trip they visited Alok and selected it as Intection's masterbatch producer.
397. On 28 September 2004 Mrs Sig sent an email to LIN to book time for testing a new LLIN at the beginning of December 2004.
398. On 8 October 2004 Intection and Alok entered into a secrecy agreement in respect of "the collective technologies of [Intection] for long lasting impregnation/incorporation of insecticide on/into different materials". On the same day Mr Larsen (still under the pseudonym Hans Johnsen) sent Mr Puri an email saying:

"The products we need for the MB are the following:

[Additive C]

[Additive B]

[Additive A]

Please check availability and price with Ciba…

Please be aware that this is confidential information and this should not be communicated to any third parties."

399. Mr Puri replied on 9 October 2004:

"All information shared regarding formulation will be kept strictly confidential."

400. On 12 October 2004 Dr Skovmand forwarded to Mr Larsen Mr De Corte's email dated 22 March 2004 saying "another visit we need to make after November".
401. On 25-27 October 2004 Mr Larsen and Dr Skovmand carried out PE extrusion and net manufacture trials at Mr Sivasamy's factory in Karur. Samples 1-12 in the Netprotect database are yarn extrusion samples from these trials, while samples 13.1-24 are net samples made from the yarns.
402. Confidential Schedule 2 to the Defence states:

"Oct. 2004:     Development of recipe and production method for insecticide containing master batch.

Oct. 2004:     Initial production trials run at Shobika Industries, only involving insecticide with polyethylene, to evaluate the production equipment of Shobika Industries. See 'Shobika Initial'. Also trials with different MB concentrations."

The first of these entries goes on to say "See 'insect-control ech 7 et 10 janv(1).xls'" and the second goes on to say "See 'Extract DB oct04.xls'" in the column headed "Tests performed".

403. The document called 'insect-control ech 7 et 10 janv(1).xls' contains three columns headed "Echantillons" (French for sample), "deltaméthrine (g/Kg)" and "alphacyperméthrine (g/Kg)". It appears to show three samples containing deltamethrin and three containing alpha-cypermethrin. Extract DB oct04.xls lists "alfa", i.e. alpha-cypermethrin, as the insecticide in samples 6 and 7. Extract DB oct04.xls has no entries in the columns headed "[Additive A]", "[Additive B]" and "[Additive C]".
404. Similarly, Mr Larsen states in his fifth witness statement at paragraph 31:

"This led to the first masterbatches being produced in September/October 2004, for test purposes i.e. for testing the machinery and process, rather than establishing a formulation for the product."

405. In fact, the October 2004 trials were not of insecticide and PE only, nor were they simply for testing the machinery and process rather than establishing a formulation, nor did any of the samples contain alpha-cypermethrin. As Dr Skovmand acknowledged in paragraph 72 of his sixth witness statement, the samples all contained [Additive A], [Additive B] and [Additive C] as well as deltamethrin. The samples constituted the Defendants' first shots at a PE net formulation, and the Defendants had arranged in advance to have some or all of them chemically analysed at UMP and bioassayed at LIN.
406. So far as insecticide, additives and LDPE are concerned, the recipes for samples 1-12 are as follows:

[REDACTED]

407. Mr Howe's assessment was that the October 2004 trials appeared to be aimed at investigating the effect of (i) reducing the extruder temperature, (ii) changing the amount of LDPE in the masterbatch, (iii) doubling the deltamethrin content and (iv) changing the amounts of [Additive A] (which increases insecticide migration) and [Additive C] (which reduces insecticide migration) within the formulation to control migration of the increased deltamethrin content.
408. Two of the formulations tested by the Defendants in the October 2004 trials were the same as two of the recipes in the Fence database with respect to deltamethrin and additive concentrations, if [Figure 3] is rounded to [Figure 15]:

Netprotect yarn samples	VF sample	Deltamethrin g/kg	[Additive A] g/kg	[Additive B] g/kg	[Additive C] g/kg
9-10	13	[Figure 2]	[Figure 4]	[Figure 6]	[Figure 15]
11-12	8 & 9	[Figure 2]	[Figure 14]	[Figure 6]	[Figure 15]

409. Dr Skovmand said in paragraphs 72-73 of his sixth witness statement:

"From the results [of the October 2004 tests], it was clear to me that there was too much sample variation. The 6 formulations made could not be evaluated….

In my view, the problem was not in the extrusion process, but simply because Torben, in an attempt to be economical, had asked for concentrated (40%) masterbatches, which makes homogenous mixing much more difficult."

Mr Larsen said in paragraph 33 of his fifth witness statement:

"In January 2005, the results of the trials were submitted to LIN/IRD and for the chemical analysis at Gembloux. At this time Ole and I discussed the way in which we would deal with the results although we realised they would not arrive for some time. In fact, the results were not good"

410. Leaving aside the inconsistency of saying that the formulations could not be evaluated and that the results were not good, what actually happened, as I shall relate below, is that net sample 19 from the October 2004 trials was submitted to LIN for bioassay, and the results were good. Sample 19 was a net made from the recipe of yarn sample 3, namely [Figure 2] g/kg deltamethrin, [Figure 4] g/kg [Additive A], [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C]. This is the [REDACTED] formulation, which was referred to by Dr Skovmand in a number of the contemporaneous documents as the "standard" or "reference" formula or recipe.
411. On 28 November 2004 Dr Skovmand and Mr Larsen (as Hans Johnsen) met Mr Lambert in Brussels. According to Mr Lambert's statement, Dr Skovmand told him that the high melting point, and hence the high extrusion temperature of HDPE, caused additives, and in particular the insecticide, to vaporise. Mr Lambert therefore suggested a blend of HDPE (specifically Eltex [Grade I], the same grade previously recommended to VF) and MDPE (specifically Eltex [Grade II]). Mr Larsen's minute of the meeting says:

"Intection should run a series of trial in Karur with ELTEX, to see how it performs. This would probably be in January. Action: HJ to send address of Sivalakshmi to Yves. Yves will supply 300kg sample of ELTEX to Karur."

412. Confidential Schedule 2 to the Defence states:

"Nov. 2004     Initial contact and discussions with Innovene on development of special mixtures of PE to enhance insecticide migration, facilitate low temperature extrusion and secure strength of the final net material. See 'Innovene 1'."

'Innovene 1' consists of two emails from Mr Larsen (as Hans Johnsen) and Dr Skovmand to Mr Lambert and a colleague dated 29 November 2004. Given the previous contacts recorded above, I do not accept that the "initial contact and discussions with Innovene" occurred in November 2004.

413. On 7 December 2004 Mr Larsen sent Mr Kjaergaard an email on the subject of his non-compete clause stating:

"As you probably know, my partner in Pure Culture I/S has set up a new business. This business may be regarded as a competitor to Vestergaard Frandsen A/S. In this respect, I have therefore, after consultation with my trade union concerning my non-compete clause, decided to take the following steps:

…

It should be emphasised that in the period in which the non-compete clause applies, I have neither been, am being nor will be involved in my partner's other business."

This statement was quite simply untrue.

414. On around 17 December 2004 Mrs Sig sent eight samples (19.1-19.8) of a PE net to LIN for bioassay testing. The Defendants pleaded in paragraph 30 of the Defence this was "one of the products produced under IIC advice and inspection in Kahrur, January 2005", but in fact the net was sample number 19 from the October 2004 trials.
415. Mr Larsen's evidence was that the purpose for which these samples were submitted to LIN was to establish a correlation between LIN and the laboratory which the Defendants were using in Burkina Faso. Even if that was a secondary objective, it is clear that the primary purpose of the tests was to test the efficacy of the prototype net according to a protocol approximating to the WHOPES Phase I protocol. The protocol was not precisely in accordance with WHOPES I, since it involved heating to 40^oC; but even on Mr Larsen's account, this would have made the test regime more realistic not less.
416. On 30 December 2004 Mrs Sig sent an email to her father attaching Dr Skovmand's proposal for a contract between IIC and Intection. In the email Mrs Sig said that certain items had been agreed with Dr Skovmand by telephone that day, including:

"- IIC transfers money to Intection next week and is subsequently a shareholder of Intection.

…

- The net technology (netprotect) or other technologies developed in under the auspices of Intection, IIC is a shareholder of Intection, belongs to Intection."

417. On 31 December 2004 IIC paid Intection 40,200 euros or DKr 300,000. The purpose of the payment was recorded on the bank transfer note as "acquisition parts sociales" (i.e. acquisition of shares).
418. In paragraph 31 of his sixth witness statement Dr Skovmand stated:

"Neither IIC nor I ever became a shareholder of Intection. IIC became a shareholder of Besnet Europe Ltd. IIC paid Intection DKr 300,000 as a help for the start up company. This was paid on 31 December 2004… "

In cross-examination Dr Skovmand maintained his denial that IIC became a shareholder in Intection, and claimed that the payment was a loan, but was unable to explain the terms of the bank transfer note. By contrast, Mrs Sig accepted that Dr Skovmand had purchased shares in Intection and that the shares were issued.

419. On 11 January 2005 Ms Guillot sent Dr Skovmand the results of UMP's chemical analyses of both PE and PES net samples. The attachment to this email, which is entitled "insect-control ech 7 et 10 janv.xls" (not "insect-control ech 7 et 10 janv(1).xls"), does not appear to have been disclosed by the Defendants.
420. On 14 January 2005 Dr Vincent Corbel of LIN sent Mrs Sig preliminary results of the wash resistance testing after 10 washes. On 18 January 2005 Mr Corbel sent Mrs Sig the final results, showing that the net achieved 78% effective mortality and 96% functional mortality after 20 washes. Dr Corbel commented:

"Results are particularly good and I encourage you to continue further testing on these LLINs."

421. On 18-21 January 2005 Dr Skovmand went to Karur to carry out further production trials for the Defendants' net. Samples 25-93 in the Netprotect database are dated Jan-05 and were produced during these trials. Samples 25-41 are yarn extrusion samples identified as being made from recipes 1-8 (two samples from each recipe). These recipes do use insecticide masterbatches containing [REDACTED] g/kg ([REDACTED]%) rather than [REDACTED] g/kg ([REDACTED]%). Five contain deltamethrin as the insecticide and three alphacypermethrin. So far as insecticide, additives and LDPE are concerned, recipes 1-5 are as follows:

[REDACTED]

422. Recipe 8 is as follows:

[REDACTED]

423. The first document disclosed by the Defendants under cover of GMR's letter dated 2 March 2007 was a document entitled "HDPE net forsøgsoversigt [Danish for trial summary], Karur January 2005". GMR described this as "the recipe for this product", that is to say, the product submitted to LIN for testing which resulted in the report dated May 2005 discussed below. This is incorrect. In fact, as I have said, the product submitted to LIN for testing which resulted in the May 2005 report was sample 19 made in October 2004, and not a product made in January 2005.
424. Curiously, this document is not referred to in Confidential Schedule 2 to the Defence. It can now be seen that it substantially reproduces a section of "comments and recepts" tab of the Netprotect database, although for no apparent reason some entries on the right hand edge of that section have been omitted in the document as originally disclosed.
425. The document lists a set of recipes numbered from 1 to 8 by reference to masterbatches which are identified by codes. "[REDACTED]" is the [Additive A] masterbatch, "[REDACTED]" is the [Additive C] masterbatch, "[REDACTED]" is the [Additive B] masterbatch, "NPTD1" is the deltamethrin masterbatch and "NPTA1" is the alpha-cypermethrin masterbatch. Recipes 1-5 and 8 correspond to recipes 1-5 and 8 in the Netprotect database. Recipes 6 and 7, which are alpha-cypermethrin recipes, differ slightly.
426. Confidential Schedule 2 to the Defence states:

"Jan. 2005:      Initial trials with Innovene polymer and different UV additives. Nets from this trial sent to LIN/IRD and to Walloon Agricultural Research Centre (Gembloux)."

427. This is wrong in a number of respects. First, there were no trials with Innovene polymer at this date: although Mr Lambert arranged for a batch of Eltex [Grade I] to be sent to Bangalore for the January 2005 trials, the polymer did not clear Indian customs until after these trials were carried out. Secondly, there were no different UV additives used: the UV additives used were the same as those used in the October 2004 trials. Thirdly, nets from these trials were not sent to LIN, the nets sent to LIN were from the October 2004. As for the two WARC analysis reports referred to, one of these (2005/20985) again relates to the October 2004 trials: the samples were sent from LIN to WARC for chemical analysis after the wash tests had been carried out. The other report (2005/20928) relates to some samples received by WARC on 21 January 2005. The identity of these samples is mysterious, but at all events they cannot emanate from the January 2005 trials.
428. On 24 January 2005 Mr Larsen sent Dr Skovmand an email saying:

"We know already that LIN achieves good results with PE nets, and at the time of writing this email we're about to find out how far our samples get at LIN – if they reach 30 washes, I guess we must conclude that the PE concept is working even though we are uncertain about the exact concentration of insecticide on the samples at LIN."

429. On 17 February 2005 Dr Corbel sent Mrs Sig further test results, showing that the samples achieved 60.5% effective mortality and 80.5% functional mortality after 30 washes.
430. On 18 February 2005 Dr Skovmand sent Mr Sivasamy an email, copied to Mrs Sig, saying:

"We only have the chemical results I told you already. These show that the variation in deltamethrin is reduced, but it is still very big. This can be further improved, when a final masterbatch is made. It does mean that currently all bioassays must be based on average results of 5 samples, and not of single samples per formulation!! These 5 samples must of course be taken a bit apart to give a true average of the net. Chemical analysis is the same thing."

431. On 16 March 2005 Mr Larsen (as Hans Johnsen) sent Mr Sivasamy an email requesting him to make 10 nets for field testing in Africa. Mr Larsen instructed Mr Sivasamy to use the net material from "recipe 1". Mrs Sig arranged for these nets to be field tested by Abdoulaye Diabaté of Centre Muraz in Burkina Faso.
432. On 16-21 April 2005 Mr Larsen visited Delhi, Karur and Chennai. During the course of his trip he visited Alok, Mr Sivasamy and Tagros among others. Mr Larsen's visit report records that Mr Sivasamy was to produce nets for field trials as follows:

"Net for field trials:

Indonesia:

60 nets recipe 1 (white)

50 nets recipe 3 (violet)

50 nets recipe 8 (dark blue)

BF

50 nets recipe 1 (white)"

433. Mr Howe points out that there are three undated samples in the Netprotect database, numbered 94-96, which appear to fit these descriptions. The utilisation of all three is stated to be "sumba field test". Sumba is an island in Indonesia. Sample 96 is white and is in accordance with recipe 1. Sample 95 is coloured "lilla" (Danish for purple) and is in accordance with recipe 3. Sample 96 is dark blue and is in accordance with recipe 8.
434. Confidential Schedule 2 to the Defence states:

"April 2005:      Further trials with different PE mixtures and UV additives. 3 formulations from this trial sent to Indonesia for field trial. Results from this field trial received in August 2005 and published at the MIM [Multilateral Initiative on Malaria] conference in Cameroon October 2005. See 'Recipes for field trial 0405'."

435. The document referred as "Recipes for field trial 0405" begins by reproducing the passage I have quoted from Mr Larsen's visit report, except that it gives the colour of the 50 nets for BF (i.e. Burkina Faso) as "blue". It then sets out three recipes, identified as "Recipe 1", "Recipe 3" and "Recipe 8", with recipe 1 being repeated in blue as well as white. The first six ingredients are masterbatches identified as "NPT1", "NPT2", "[REDACTED]", "[REDACTED]", "NPTD1" and "NPTA1". Each recipe also contains [REDACTED]kg "[Additive M]".
436. It is clear that, as set out above, "NPTD1" and "NPTA1" are deltamethrin and alpha-cypermethrin masterbatches. It seems likely that "[REDACTED]" is again a [Additive A] masterbatch. As to the masterbatches called "NPT1", "NPT2" and "[REDACTED]", the Defendants have variously identified these as follows:

i) In paragraph 56(ii) of his fourth witness statement dated 20 June 2007 Mr Larsen identified NPT1 as a [Additive A] masterbatch, NPT2 as [Additive C] and [REDACTED] as an [REDACTED].

ii) In a response to a request for further information which was verified by a statement of truth signed by Mr Larsen on 24 January 2008, the Defendants said that NPT1 and NPT2 were "believed to be single ingredient Masterbatches", yet in the next sentence said "NPT1 and NPT2 are probably mixed Masterbatches", but without specifying the contents of these masterbatches in either case.

iii) In the same response, [REDACTED] was identified as a mis-print for [REDACTED], a [Additive C] masterbatch.

iv) On pressing, the Defendants said in a letter dated 30 April 2008 "NPT1 and NPT2 are Deltamethrin".

437. It is plausible that [REDACTED] is a typo for [REDACTED], a [Additive C] masterbatch, but whether this is right I am not sure. It is clear that neither NPT1 nor NPT2 can be a deltamethrin masterbatch, but it is not clear what they are.
438. More importantly, perhaps, samples 94-96 in the Netprotect database do not contain [Additive M]. The first recorded use of [Additive M] in the database is in either sample 114 dated 23 March 2006 or sample 118 dated 8-11 January 2006. Moreover, the correspondence considered below suggests that it was first discussed in around December 2005. Accordingly, I do not accept that the Defendants used [Additive M] in April 2005. Moreover, Mr Howe points out that the recipes in the "Recipes for field trial 0405" document add up to a total mass of 50.05kg rather than 50kg as specified. As he says, the discrepancy appears to be due to the mass of [Additive M] (0.05kg) in the recipes.
439. All in all, "Recipes for field trial 0405" is not a reliable document.
440. On 22 April 2005 Dr Olivier Pigeon of WARC sent Mrs Sig the 20985/2005 test report on net samples received from LIN. She forwarded the report to Mr Larsen, who forwarded it to Dr Skovmand. Dr Skovmand commented on the report in an email dated 25 April 2005:

"So the positive side is that they are extremely wash resistant, not even half have been washed off after 35 washings, and here is probably a product that can't be exhausted by washing in practice.

The negative side is the big spot to spot variation, however, if one used WHO's test technology, you'd cut out 5 samples of a net, cut them into small pieces and look at the average for 1 net and then it's the value of that net. A number of whole nets extracted from a production or order then all have to (if a large order, then except for one or two) be within the declared average."

441. On 29 April 2005 Dr Skovmand sent Mr Larsen, Mr Bosselmann and others an email saying "have the polyethylene formations proper tested, they really are very promising".
442. Shortly before 3 May 2005 LIN sent Intection its final report by J. Bonnet, S. Duchon and V. Corbel on the eight Netprotect samples entitled "Phase 1 evaluation of LLINs from In-tection" ("the LIN Report"). Under the heading "Bioassay procedure" the LIN Report states:

"The WHO cone test measures knock down and mortality of susceptible mosquitoes exposed in a small chamber to a piece of treated netting (25 x 25 cm) for a 3 min exposure time. … KD effect is recorded 60 min post exposure and percentage mortality after 24 hours. Mosquitoes able to fly are counted as alive. However, it is commonly observed, when testing pyrethroids, surviving mosquitoes able to fly with several legs missing, sometimes up to 5. This phenoment of leg auto-section related to pyrethroids is likely to occur in the field as well as the lab. For this reason, we also record mosquitoes surviving with 3 legs or less, since these mosquitoes are unlikely to survive in the field (not able to bite and/or to lay eggs). We count them as 'functional mortality (dead + alive with 3 legs or less) by opposition to effective mortality. Difference between effective and functional mortality is sometime important and this double count gives a better estimation of the overall killing effect of the insecticide."

443. Under the heading "Drying procedure", the LIN Report states:

"Once washed, net samples were laid vertically in clotheslines at room temperature during 2 hours. After drying, nets were stocked in a clamtic chamber at constant 27^oC and 80% RH until next washing. According to In-tection recommendations, net samples were placed in alufoil and stored at 40^oC during 48 hours after the last washes and before the next bioassays (weekend period)"

444. Under the heading "Testing protocol", the LIN Report states:

"Two samples of LLIN were tested after different wash regimen. Samples 19.1 and 19.2 were tested at 0 wash, samples 19.3 and 19.4 after 5 washes, and samples 19.5 and 19.6 after 10 washes. Finally, samples 19.7 and 19.8 were tested after 20, 25, 30 and 35 washes. This cycle of wash-dry-wash was repeated until KD effect after 60 min falls down below 95% and/or mortality after 24 h falls down below 80% (WHO thresholds). In this case, the samples were tested after a 5 min exposure time. If efficacy was still below WHO thresholds after a 5 min exposure, regeneration procedure was carried out by storing the samples at 40^oC during 48 hours in alufoil ('heat assisted' reactivation). Samples were then tested after a classical 3 min exposure time."

445. Under the heading "Results" and the sub-heading "Mortality of samples from series 19", the LIN Report states:

"Figure 2 showed that effective mortality remained above the cut-off point (WHO threshold) until 25 washes (we did not consider the decrease of mortality to 78% after 20 washes as significant). However, after 30 washes, effective mortality decreased under the cut off point (61%). With a 5 minutes exposure (or after 'heat assisted' reactivation) efficacy of the net samples was restored (mortality was 94%). The same trend was noted after 35 washes, where a longer contact period on nets allows mortality to increased above the WHO threshold (82%).

Figure 2 showed that functional mortality was above 80% (WHO threshold) until 35 washes. Exposure of 5 min and/or 'heat assisted' reactivation allowed efficacy to recover 100% for both samples at 30 washes."

446. The LIN Report incorporates the deltamethrin analyses conducted by WARC. Under the heading "Conclusion", the last sentence of the LIN Report reads:

"I would now recommend to Intection to test these mosquito nets following standardized procedures as recommended by WHOPES for laboratory testing of LLINs (WHO/CDS/WHOPES/GCDPP/2005.11)."

447. Looking ahead slightly, Bestnet subsequently posted a copy of the LIN Report on its website as part of the promotional materials for Netprotect. It remained there until after the commencement of the trial. Shortly after VF had drawn this to my attention, it was removed. In addition, Figures 1 and 3 from the LIN Report were reproduced by Syngenta in a brochure for Icon Life together with the statement:

"Icon Life has been evaluated by the WHO collaborating centre Laboratoire de Lutte Contre les Insectes Nuisibles (LIN) at the Institut de Recherche pour le Développement (IRD) using the methods established for WHOPES Phase I (Guidelines for laboratory and field testing of long lasting insecticidal mosquito nets – WHO/CDS/WHOPES/GCDPP/2005.11). These studies have demonstrated that Icon Life has the ability to control malaria vectors for at least 35 washes."

448. On 3 May 2005 Mrs Sig sent Dr Morteza Zaim of WHOPES an email saying:

"We were in contact one month ago, discussing Intection's possibilities for submitting our product for WHOPES evaluation. At the time you kindly requested us to wait for the new guidelines to be published (in 3-4 weeks time).

Please let me know if these guidelines are no available? And if yes, where I can find them.

For your information, we have just received the final report from our testings at LIN, France and also we have started field trials in 2 different countries. And of course we are eager to get the WHOPES evaluation started as soon as we know the importance of this evaluation."

449. Dr Zaim responded on 9 May 2005 by sending Mrs Sig a pre-publication copy of WHO's Guidelines for Laboratory and Field Testing of Long-Lasting Insecticidal Mosquito Nets.
450. On 24 May 2005 Mrs Sig sent Dr Zaim a copy of the LIN report together with some information about Netprotect.
451. On 15 June 2005 Dr Skovmand sent Mr Lambert an email confirming a request to send HDPE to Mr Sivasamy and to an unidentified masterbatch producer (presumably Alok). He also enquired about a new, softer HDPE that Mr Lambert had mentioned. Mr Lambert replied the same day saying that he had developed a new grade, Eltex [Grade II], for artificial grass, which could be run with a lower temperature profile on the extruder. Confusingly, Eltex [Grade II] is described by Innovene as both a HPDE and a MDPE. Likewise, it appears that the Defendants sometimes refer to it as HDPE and sometimes as MDPE. I think it is correctly described as a MDPE, which is how Mr Howe describes it.
452. In paragraph 83 of his sixth witness statement Dr Skovmand said that "IIC used a new type of polymer that VF neither had nor had access to". In context, it is clear that he was referring to an Innovene polymer. He did not identify the polymer in question, but it seems clear that he meant Eltex [Grade II] since (a) he was referring to the final recipe for Netprotect and (b) Eltex [Grade II] is the only Innovene polymer listed in the Netproduct Production Manual. It is true that this was a new polymer developed by Innovene which VF did not have at the time of the work recorded in the Fence database, but given that Innovene recommended that same grade of HDPE successively to both VF and the Defendants, there is no reason to think that it would not have recommended Eltex [Grade II] to VF in the summer of 2005 if asked. Eltex [Grade II] was not, and is not, supplied by Innovene exclusively to the Defendants. Furthermore, as I shall discuss below, the Defendants have not always used this polymer in their production of Netprotect.
453. On 17 June 2005 Claus Bogh, a Danish consultant known to Dr Skovmand who had been hired to carry out the Indonesian field tests, sent Dr Skovmand, Mr Larsen and Mrs Sig an email in which he commented "First and foremost I have to say that you have made a good product".
454. On 21 June 2005 Dr Skovmand sent Mr Lambert an email requesting that a sample of Eltex [Grade II] be sent to Karur for trials on 17 July 2005. Mr Lambert replied on 22 June 2005 that this would be done, and added:

"In order to have a good relationship in the future, I would like to propose you a mutual Secrecy Agreement."

455. On 22 June 2005 Mr Larsen sent Mr Sivasamy an email requesting him to make more nets and send them to Indonesia urgently.
456. On about 1 July 2005 Mr Larsen sent Mr Puri an email saying that he would be in Karur from 19-23 July to supervise the start of production as well as running some trials, and requesting Alok to deliver some masterbatches there.
457. On 7 July 2005 Dr Skovmand sent Mrs Sig an email suggesting:

"We go ahead with the HDPE net while at the same time talking and making plans with Innoven, i.e. we start by submitting to WHOpes (still no reply from Zaim) and at the same time commence sales to get some experience."

458. Mrs Sig replied the same day:

"I … think your plan about going ahead with the present HDPE net while at the same time talking and making plans with Innoven is the only way to go right now."

459. On 18-24 July 2005 Mr Larsen visited Karur. His visit report states:

"Siva Enterprises (SE)

Mr. Sivasamy

Mr. Mahindra

Mr. Marappan

9 trials were run, basically the trials were aiming at trying out some alternative chemicals in the recipe and 2 trials were designed to try a new polyethylene polymer from Innovene.

The trials went well except 1 trial that were designated with the colour Dark Blue….

The trials with Innovene polymer went extremely well. We ran the second trial at only [Temperature 2] (normally we have problems keeping below our max temperature of [Temperature 3]) and we could have went further down in temperature….

We started bulk production of Netprotect® late Friday. There was a great sense of excitement surrounding the event. Unfortunately the first 50 kg did not run well and we created some 30 mg of waste. I decided to raise the temperature in the extruder to the limit ([Temperature 3]), reduce the speed of the screw and switch from blue colour to white colour. This improved the running of the yarns and even before the switch of colour was effected, the production was running smoothly. CONGRATULATION INTECTION – WE NOW HAVE A PRODUCT!!!!!!!

…

The following will be sent to SGS, Chennai for chemical analysis:

…

2 samples of each colour from bulk production (blue and white)

Siva will send samples of net from all 9 trials and 2 production run to DK. THL will send to BF, together with samples of net from field trial in Indonesia, for bioassay.

…

Mr. Vikram Bhadouria

Vikram visited us while we were at Siva. We discussed the future MB situation.

…

Vikram will immediately make a trial a single MB – I gave him the recipe – to see if it is possible. I underlined that it is of outmost importance that only very few people at ALOK knows this recipe. He promised me that he, Amit and one other trusted employee would be the only ones to know the recipe."

460. As is recorded in the visit report, on 23 July 2005 Intection and N. Marappan (who it appears is Mr Sivasamy's father) trading as Siva Enterprises entered a Licence and Co-operation Agreement under which Intection licensed Siva to use the Know-How exclusively for the manufacture of Products for Intection. Product is defined as meaning "long lasting insecticide treated/incorporated Nets or LLIN's", while Know-How is defined as:

"the recipe of the chemicals for the purpose of adding to polyethylene for giving it a long lasting insecticidal effect and the process by which it is incorporated into HDPE yarns for creating the Products."

461. The Netprotect database identifies samples 97-107 as having been produced on 20 July 2005 using recipes 9-17.
462. The second document disclosed by the Defendants under cover of GMR's letter dated 2 March 2007 was a document entitled "Trial, Karur 20-, June 2005". This document was disclosed in response to the Order of Kitchin J for disclosure of instructions relating to the product launched at the conference in Cameroon in November 2005.
463. Again, this document is not referred to in Confidential Schedule 2 to the Defence. The first section of the document does, however, reproduce the "Karur 0705 tab" of Extract DB Apr05.xls, except that it omits the information contained in columns R-Z of the latter. It now appears to be an extract from the "comments and recepts tab" of the Netprotect database, but with a few differences (particularly in the case of recipe 15) as well as the omission of the same information.
464. This document gives a set of recipes 9-17 in terms of masterbatches identified by the codes "NPTD1" (i.e. deltamethrin), "[REDACTED]" (i.e. [Additive A]), "[REDACTED]" (i.e. [Additive B]), "[REDACTED]" (i.e. [Additive C]), "[REDACTED]" ([Additive Q], see below), "[REDACTED]" ([Additive R], see below), "[REDACTED]" ([Additive S], see below) and "[REDACTED]" ([Additive F]). Recipes 9-15 are headed "Normal Polymere" while recipes 16 and 17 are headed "Innovene Polymere". Recipes 9-15 have appended to them the comment (in Danish) "Mistake in weighing, there is 1 kg HDPE more than there should be".
465. As Mr Howe explains, the mistake is due to an error in this spreadsheet, which fails to include the [Figure 1] kg of deltamethrin masterbatch in the calculation. The result was that the concentrations of the additives were 96% of those intended. If this is corrected for, recipes 9-15 were as follows:

[REDACTED]

466. As recorded in Mr Larsen's visit report, these trials appear to have been designed to try alternative additives. [Additive Q] is BASF's trade name for the chemical supplied by Ciba as [Additive A]. [Additive S] is another BASF additive which is similar to [Additive B]. [Additive R] is a BASF blend of [REDACTED]. In addition, samples 97-105 made from recipes 9-15 had a net weight of 30 g/m².
467. Recipes 16-17 contain Innovene HDPE with no LDPE. It is not entirely clear whether Innovene HDPE here means Eltex [Grade I] or Eltex [Grade II], but I consider it is probably the latter. I think it was only later that Dr Skovmand appreciated that [Grade II] was a MDPE. Recipe 16 contains [Figure 2] g/kg deltamethrin, [Figure 4] g/kg [Additive A], [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C]. Recipe 17 contains [Figure 2] g/kg deltamethrin, [Figure 4] g/kg [Additive A], [Figure 16] g/kg [Additive B], [Figure 14] g/kg [Additive C] and [Figure 17] g/kg [Additive F]. These samples had a net weight of 28 g/m².
468. Neither the Netprotect database nor any other documents disclosed by the Defendants and Dr Skovmand disclose the formulation used in the July 2005 production run. In the case of the "Trial, Karur 20-, June 2005" document enclosed with GMR's letter dated 2 March 2007, there is a heading "Production" but nothing underneath it. The same is true of the "Karur 0705" tab of Extract DB Apr05.xls and the "comments and recepts" tab of the Netprotect database.
469. It can be seen from Mr Larsen's visit report that the yarn made in the production run was originally intended to be blue, but the colour was switched to white during the course of the run. Of the nine formulations discussed above, recipe 9 was white and recipe 11 was blue according to the Netprotect database.
470. In an email dated 21 June 2005 Dr Skovmand said that "In Torben's trial in India the standard formula will … be used". As I have mentioned above, the "standard formula" was the [REDACTED] formulation of [Figure 2] g/kg deltamethrin, [Figure 4] g/kg [Additive A], [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C] i.e. the recipe used in yarn samples 1-5 and net sample 19. This corresponds to recipe 9 if the weighing mistake is corrected for as set out above.
471. Given this and given that the product launched by Bestnet in October 2005 (as to which, see below) was made in accordance with the [REDACTED] formulation, I conclude that the standard formula was used for the July 2005 production run. It appears from an email dated 31 August 2005 that the weight of the net was 30 g/m².
472. Nets from the July 2005 production run were subsequently marketed by Bestnet. The "Thai Sample" obtained by VF in March 2006 (as to which, see below) bears a label stating that it was made in July 2005.
473. Confidential Schedule 2 to the Defence states:

"July 2005     Further trial with Innovene polymer. See 'visit report India 0705'."

"Visit report India 0705" is Mr Larsen's visit report quoted above. The entry goes on to refer to Extract DB July05.xls.

474. The statement that there was a "further" trial with Innovene polymer in July 2005 is wrong: it was the first trial.
475. Nets from the July 2005 trials were subjected to physical and chemical testing by SGS India Pvt Ltd in Chennai during August 2005. The deltamethrin analyses showed insecticide retention rates of around 98% for most samples. In a subsequent email to Mr Larsen dated 20 November 2006, however, Dr Skovmand expressed scepticism about the Chennai analyses since analyses by CDC found lower deltamethrin concentrations. He commented:

"It seems certain we have a loss somewhere, which is OK, I guess, as long as we can keep it constant (of course. I'd rather be without it)."

476. On 9 August 2005 Mrs Sig sent Mr Sivasamy an email stating that 65,000 nets had to be delivered before 31 September 2005.
477. On 23 August 2005 Mrs Sig wrote on behalf of Intection to WHO in the following terms:

"Intection hereby kindly request WHOPES to evaluate a new LLIN, Netprotect®.

WHOPES I & II:

We kindly request WHOPES to run tests, WHOPES I & II of Netprotect at the same time.

Selection of net:

For your information we have taken a random selection of 20 nets from a running production of 50,000 nets. Hereof we have send 15 nets to LIN, France and 5 nets we have kept for your own reference.

Protocol:

The tests are to follow the protocol for PermaNet as this has been the protocol for Netprotect under the product development.

Info:

A copy of this letter has been sent to LIN, France together with the nets. Nets left India … 15^th August 2005.

Nets are being produced in India at Shobika Enterprises …"

478. On 24 August 2005 Mrs Sig wrote to LIN saying:

"Samples of Netprotect for WHOPES evaluation

…

As per agreement with Ole Skovmand I hereby forward you 15 samples for testing."

479. It is clear from these letters that the nets submitted by the Defendants for WHOPES I evaluation in August 2005 came from the July 2005 production run. It is also clear that the Defendants originally requested WHOPES I and II evaluation to be carried out in parallel.
480. In paragraph 36 of his first witness statement Dr Skovmand said:

"At the end of 2005, IIC had a preliminary recipe that was used to make nets for WHO. It was slightly modified during 2006 to become more economically viable without compromising quality and the new formulation was sent to WHOPES II for simulated field evaluation including simulated field washing and is now used in production."

481. IIC did not arrive at the preliminary recipe that was used to make nets for WHO at the end of 2005. The nets which were submitted to WHO in August 2005 were made in July 2005 according to a formulation which IIC had first tested in October 2004.
482. On 2 September 2005 IIC and Innovene signed a secrecy agreement. This document does not appear to have been disclosed by the Defendants or Dr Skovmand.
483. On 7 September 2005 Dr Skovmand sent Dr Duchon and Dr Corbel of LIN an email saying:

"the 15 samples have been sent directly to you and and at the same time, a letter has been sent by Trine to WHOPES.

The purpose is to start WHOPES I and II evaluation of these nets according to the WHO protocol."

484. On 9 September 2005 Dr Zaim of WHO sent an email to Dr Skovmand, Mrs Sig and others saying that WHO had received a letter from VF (i.e. VF's letter dated 25 August 2005) informing it of the legal action against Intection and that WHOPES evaluation of the Intection product required clarification of the issue by the two companies and the advice of WHO legal counsel.
485. On 12 September 2005 Mrs Sig wrote on behalf of Intection to WHO enclosing a comparison sheet showing the differences between Netprotect and PermaNet.
486. On 19 September 2005 Mrs Sig sent Dr Corbel at LIN an email saying that she was sending three PE and two PES samples for testing the next day.
487. In Dr Skovmand's letter dated 12 September 2008 (as to which, see above), he stated in relation to the request for the recipe of "the net or nets submitted to the WHO as a result of the application in September 2005 by Intection ApS for WHOPES I evaluation":

"there was never sent any nets to WHO in September to my knowledge, the processs of sending the nets would have been expected to start some months later after WHO had accepted the request and Intection had provided a tox evaluation. It never came so far before VF had stopped the process with a letter to WHO claiming that Netprotect was a copy of Permanet and thus patent protected."

It can be seen that this is not an accurate account.

488. Confidential Schedule 2 to the Defence states:

"Sept. 2005     Refinement trials with UV and other supporting ingredients."

No supporting documents are cited in the Schedule, none have been disclosed, there are no entries in the Netprotect database datable to September 2005 and no work during this period is mentioned in the Defendants' evidence.

489. Netprotect was commercially launched at a Netmark conference in Kenya in October 2005 and the MIM conference in Cameroon in November 2005. One of VF's employees obtained a blue sample of Netprotect at the conference in Cameroon. This sample was subsequently analysed by the Danish Technology Institute on behalf of VF. A white sample was also obtained in India. Dr Skovmand and Mr Larsen accepted in cross-examination that this product was made in accordance with the [REDACTED] formulation.
490. On 24 October 2005 Mrs Sig wrote to WHO to say that Intection would hand the licence for Netprotect back to IIC so that IIC could get it evaluated. On 27 October 2005 Mrs Sig wrote to WHO to say that Intection would go into liquidation. Also on 27 October 2005 Dr Skovmand wrote on behalf of IIC to Dr Zaim requesting WHOPES to start evaluating Netprotect in accordance with the protocols for WHOPES I and II in parallel. In this letter Dr Skovmand said:

"Protocol:

I have asked LIN to run tests of the product for my own knowledge of product performance. These tests showed that the net is of a regenerating type if the net is first washed 3 times and then tested repeatedly to see if and when a level is reached. I know well that this is NOT a WHOPES I test, but with this information I now know that it is behaving like a regenerating net as the prototype Olyset although with a faster regeneration (5 days in the test carried out). I therefore kindly request WHO to consider using the Olyset test model and the fast track model as discussed on the Informal Consultation on the development of guidelines for testing/evaluation of long-lasting insecticidal mosquito nets held in WHO headquarters, Geneve 4-7 this year.

Where to send nets:

Please inform where to send nets and how many."

491. It can be seen from the first paragraph that I have quoted that Dr Skovmand was aware that the tests carried out by LIN and reported in the LIN Report were not in accordance with the WHOPES I protocol, but that nevertheless he was hopeful that these results would persuade WHO to fast-track the evaluation of Netprotect.
492. On 3 November 2005 WHO wrote to Mrs Sig to say that it would agree to evaluate Netprotect on request by IIC if IIC undertook that (i) it had the legal right to Netprotect and (ii) it was not involved in any dispute with VF regarding Netprotect. On 13 November 2005 Dr Skovmand wrote to Dr Zaim saying:

"You have asked for 3 sets of information and confirmation, which I shall happily give you.

One is the ownership of the Net Protect technology. This product is developed by my company Intelligent Insect Control and we own the technology. It will be licensed out for production, but we will control production…

Another is about relations ship to Vestergaard Frandsen. … We are not involved in any legal dispute with Vestergaard Frandsen.

The third is the origin of the insecticide used for the production of Net Protect. IIC has used deltamethrin and alfa cypermrthrin from Tagros during the product development phase and will continue to use this for production. In connection with license agreement(s) it will be a condition that license receiver(s) use a WHOPES recommended insecticide producer of one of these two chemicals."

493. I note that Dr Skovmand's letter was written around three weeks before IIC intervened in the copyright proceedings between VF and Mr Larsen. I consider it unlikely that Dr Skovmand only decided to take this action after he wrote this letter.
494. It is not clear what happened between November 2005 and March 2006 so far as WHO is concerned, but for the reasons given below it appears that IIC submitted further samples of Netprotect to LIN for WHOPES I evaluation in March 2006.
495. In paragraph 79 of his sixth witness statement Dr Skovmand said:

"In October or November [2005], Torben was in Kahrur to produce 6 formulations. Two of these were used for bioassay and chemical analysis. High levels of [Additive C] were tested together with [Additive L]."

This is not mentioned by Mr Larsen in his fifth witness statement.

496. The only samples in the Netprotect database between those dated 20 July 2005 and those dated 8-12 January 2006 are two undated deltamethrin samples recipes B and C (samples 108 and 109) and four undated alpha-cypermethrin samples (110-113). The recipe for samples 108 and 109 contains [Figure 18] g/kg of [Additive C]. Sample 108 is the first sample in the Netprotect database which contains [Additive L].
497. On 22 November 2005 Dr Skovmand sent Mr Larsen and Mr Ringberg-Larsen an email saying:

"Net Protect must be as different as possible from any VF product whatsoever, so I'm happy about the last recipe I sent. It's really different and expectedly also a lot better."

It is not clear which recipe is the "last recipe" referred to, but it may be the [REDACTED] formulation discussed below.

498. On 23 November 2005 Dr Skovmand sent Mrs Sig, Mr Larsen and others an email in which he said:

"I hope we all agree that until we have sent a product to WHOPES and made official recipes for use in production, Netprotect is a working title of an experimental product, even though we've had a lot of them packed in real packages to be given as samples and possible for sale to minor customers, who will of course receive the same message.

Tests performed at LIN of the first recipe are not the same as tested in the field in Burkina and Sumba or the second time round on LIN. The tests on Sumba clearly indicated a future direction for a final recipe and away from the previous recipes, which look more like Fence recipes. Why on earth would I test 3 products if I felt the recipe was finished? Besides, we still need results from our polymer chemist's friend parameter test of extrusion to be able to optimise net strength, which must also be in place for the final recipe and production guide.

Lastly, we need to select insecticide, having run parallel with alpha-cypermethrine and deltamethrine. Out of price concerns and other concerns I've said we should choose alpha-cypermethrine. That way we can avoid a market profile as 'yet another deltamethrine' product.

I've also worked with new types of additives to decrease the loss of insecticide in the process and in use, a promising area I have no intentions of letting go.

I cannot accept that the work on a preliminary recipe for concept testing of small-mesh polyethylene nets suddently becomes a final product…"

499. Confidential Schedule 2 to the Defence states:

"Nov. 2005     Laboratory evaluation of different mixtures of Innovene polymers in relation to strength of yarn a.o. factors. Performed at Innovene pilot plant."

The entry goes on to refer to two documents referred to as "Lara 13760 II". These do indeed appear to show tests on Innovene polymers (namely Eltex [Grade I] (HDPE), Eltex [Grade II] (MDPE) and a mixture of Eltex [Grade I] with [Grade III] (LDPE)) carried out on 29 November 2005. These documents were emailed by Mr Lambert to Dr Skovmand and Mr Larsen on 23 December 2005.

500. On about 30 November 2005 Dr Skovmand sent Mr Puri of Alok an email saying:

"As you may know one of the main point in producing insecticide impregnated nets is process control and low temperature. This is an area where we have done too little with additives.

I send you attached a page that you can check on the internet about how to make masterbatches with [Additive M]. Very good blending (preferably from powder is needed); these ingredients are a bit expensive, but so is the insecticide that we currently loose due to process temperature problems, so it is worth doing.

The final concentration should be max [Figure 12] g/kg, it would be interesting to try [REDACTED] g/kg, [REDACTED] g/kg and [Figure 12] g/kg for the final product. The masterbatch could be made as a [REDACTED]% concentrate.

…

You may already have good experience with [Additive M] masterbatch production, if so, so much the better.

Let me know what you think. We better prepare this work for the next trials"

501. It appears from this and a number of emails which follow it that this was the first time that use of [Additive M] was proposed.
502. On 14 December 2005 Mrs Sig sent Dr Skovmand an email saying that Bestnet had received an order for 110,000 nets. She added:

"… everyone is saying that WHOPES will succeed because it is a formality more than anything else."

503. Dr Skovmand stated in paragraph 79 of his sixth witness statement that:

"In January 2006, two larger scale test productions were made, one the standard formulation and the other with a lower level of insecticide and [Additive A] (to see if the starting surface concentrations would then be the same)."

504. The Netprotect database records that on 8-11 and 12 January 2006 nets were made for production, samples 115 and 118. The formulation for sample 115 is [Figure 19] g/kg deltamethrin, [Figure 14] g/kg [Additive A], [Figure 20] g/kg [Additive B] and [Figure 21] g/kg [Additive C]. The recipe for sample 118 is [Figure 4] g/kg [Additive A], [Figure 6] g/kg [Additive B] and [Figure 4] g/kg [Additive C], but with deltamethrin slightly reduced to [Figure 16] g/kg. Sample 118 also includes [REDACTED] g/kg MDPE and is the first sample in the Netprotect database recorded as including MDPE (although as noted above this may be because of an earlier misunderstanding about Eltex [Grade II]). The database also indicates that sample 118 may have contained [REDACTED] g/kg [Additive M], but since the entry is followed by "?" this is unclear. In both cases the net weight is 35 g/m², which is the weight of most of the subsequent samples.
505. In March 2006 the well-known investigation company Carratu acting on behalf of VF and using the alias Insight Pharm purchased a sample of Netprotect in Thailand ("the Thai Sample"). The Thai Sample was subsequently analysed by Rapra on behalf of VF.
506. Dr Skovmand stated in paragraph 80 of his sixth witness statement:

"In March 2006, I went to Karur to test 9 formulations, this time testing alternative producers of [Additive A] (called [Additive A] when from Ciba), different mixes of polymers and two levels of [Additive C]. Samples were tested in 3 round by the bioassay laboratory in Burkina together with two nets from a competitor of PE nets (from China, the product did not survive the WHOPES evaluation). The tests also included the first tests with mixed masterbatches: until this time, all masterbatches have been one ingredient per masterbatch. The results showed that it was possible to mix everything in just one masterbatch, but for better economy, we chose to have the insecticide alone and at low concentration in one and the rest of the additives together in a more concentrated masterbatch."

507. The document enclosed with GMR's letter dated 2 March 2007 headed "Trial, Karur 20-, June 2005" contains a section headed "test i marts 2006" (i.e. test in March 2006). It can now be seen that this is another extract from the "comments and recepts" tab of the Netprotect database. The layout of this document is quite confusing. As analysed by Mr Howe, there are three groups of formulations:

i) The first group comprises recipes 1 and 4-9. These are based on a 100kg batch size made from masterbatches NPTD1, [REDACTED], [REDACTED], [REDACTED], "NPTS3" and "NPTS2" together with [Additive M] and [Additive T]. NPTS3 contained [REDACTED] and NPTS2 contained [REDACTED], which are the same chemicals as [Additive A] and [Additive B] respectively, but supplied by Sunny Chemical Corp Ltd of China. Recipe 5 includes "[Grade I]" and recipe 6 includes "[Grade II]" in place of "hdpe".

ii) The second group comprises recipes 2 and 3. Recipe 2 contains NPTTOT, while recipe 3 contains NPTMIX1 and NPTMIX2. These are mixed masterbatches. In recipe 2 NPTTOT contains all the additives, while in recipe 3 the additives are split between NPTMIX1 and NPTMIX2.

iii) The third group comprises recipes 8 and 9, but this time for a 25kg batch size.

508. Recipes 1 and 4-7 correspond with samples 114 and 119-122 in the Netprotect database, each of which is dated 23 March 2006. Recipes 2 and 3 correspond with samples 116 and 117, which again are dated 23 March 2006. Recipes 8 and 9 correspond with samples 123 and 124, which are undated.
509. Sample 114 is recorded in the Netprotect database as being white and as being made to a recipe containing [Figure 16] g/kg deltamethrin, [Figure 14] g/kg [REDACTED] (equivalent to [Additive A]), [Figure 6] g/kg [REDACTED] (equivalent to [Additive B]) and [Figure 21] g/kg [Additive C]. This was referred to at trial as the [REDACTED] formulation. It also contained [REDACTED] g/kg [Additive M]. As noted above, either sample 114 or sample 118 was the first sample to include [Additive M].
510. Confidential Schedule 2 to the Defence does not state when samples of Netprotect were made for submission to WHOPES I. In paragraph 62 of his fourth witness statement, paragraph 98 of his sixth witness statement and paragraph 155 of his seventh witness statement, and in his letter dated 12 September 2008, Dr Skovmand said that the formulation for WHOPES I was made in January 2006. On the day before he gave evidence, this was corrected to March 2006. The explanatory sheet stated that this correction was made having regard to an email string. VF asked for disclosure of this email string, and it was disclosed by the Defendants shortly before Dr Skovmand gave evidence.
511. This shows that on 26 March 2006 Dr Skovmand sent Dr Zaim an email saying that he had received nets from production to send to WHO and asking where to send them. Dr Zaim replied that the samples should be sent to LIN. There is no documentation which shows that Dr Skovmand did samples to LIN after this exchange of emails, let alone what samples, but it seems likely that he did.
512. During the course of cross-examination Dr Skovmand said that sample 114 in the Netprotect database was what was submitted for WHOPES I evaluation. As noted above, sample 114 is a [REDACTED] formulation. The accuracy of this evidence is challenged by VF, who suggest that the nets submitted for WHOPES I were made from the [REDACTED] formulation.
513. Although none of the March 2006 samples in the Netprotect database are described as being for production, as noted above sample 114 is recorded as having been white. No other sample dated March 2006 is recorded as having been white. Nor are either of the production runs in January 2006, which are recorded as having been light blue. Given that the LIN report issued in September 2006 (as to which, see below) records that the nets received and tested by LIN were white, I am prepared to accept that Dr Skovmand was correct to say that these nets came from sample 114.
514. Dr Skovmand said in cross-examination that the use of the [REDACTED] formulation was a courageous guess, but not a very daring one since it was based on the January 2005 tests. He agreed that the reason for decreasing the concentration of [Additive A] and increasing the concentration of [Additive C] was to reduce the rate of migration of the deltamethrin.
515. Sample 120 is described in the Netprotect database as being for "polymer test, HD" while sample 121 is described as being for "polymer test, MD". The entry for sample 120 includes the comment "Innoven HDPE, stronger polymer, thinner yarn" while the entry for sample 121 includes the comment "Innoven MDPE, lower extrusion temp". This is consistent with the indication in "test i marts 2006" that these contained Eltex [Grade I] and Eltex [Grade II] respectively. It appears the latter was used in place of all the HDPE since there is no entry in the MDPE column for sample 121.
516. On 13 September 2006 WHO wrote to Dr Skovmand enclosing a WHOPES Phase I final report by J. Bonnet, S. Duchon and V. Corbel of LIN and a risk assessment issued by the Finnish Institute of Occupational Health. The report records that IIC provided LIN with four white LLINs. Under the heading "WHO cone test method" the report contains a passage which is very similar to the passage under the heading "Bioassay procedure" which I have quoted from the May 2005 report, except that it stops with the sentence "Mosquitoes able to fly are counted as alive".
517. Under the heading "Discussion", the report states:

"Netprotect® from IIC showed good performance against susceptible mosquitoes of Anopheles gambiae through laboratory assays and met the WHO criteria to undergo Phase II testing (at least 80% mortality and/or 95% KD after 20 washes). One should note that after 20 washed, the KD effect remained maximal (100%) whereas mortality fell slightly under the cut-off point (76%)."

518. Confidential Schedule 2 to the Defence states:

"Sept. 2005-

Nov. 2006     Several refinement production trials at Shobika Industries involving alternative UV chemical and other supporting chemicals (Antioxidants, Silicone etc.) Arriving at the present recipe as described in 'NPT details.doc' as earlier served."

519. There is no evidence of any "refinement production trials" during the period September 2005 to October 2006 other than those discussed above.
520. The Defendants have disclosed an undated document, or part of a document, described in their supplemental disclosure list as "Test recept November 2006 dedacted". This appears to be part of a communication from Dr Skovmand to another person in which he requests the other person:

"Please run this production test as soon as possible.

Then send me 5 m² of each sample.

Please use the sample numbers attached (126 to 130)."

521. The document sets out five formulations made from NPTD1, [REDACTED], [REDACTED], [REDACTED], NPTA1, [Additive M], LDPE and LLDPE. The formulations appear to correspond with samples 126-130 in the Netprotect database, the first of which is dated Dec 06 and each of which is described as being for "polymer test". Each of these samples is made from a recipe containing [Figure 22] g/kg deltamethrin. In terms of [Additive A], [Additive B] and [Additive C], samples 126-128 are [REDACTED] formulations while samples 129 and 130 are [REDACTED] formulations. "Polymer test" refers to the fact that they contain either [REDACTED], [REDACTED] or [REDACTED] g/kg of LDPE and either [REDACTED] or [REDACTED] g/kg of LLDPE. They also contain [REDACTED] g/kg [Additive M].
522. Sample 131 in the Netprotect database is dated Nov-Dec 06 and is described as "production" for "sale". The recipe for this sample is as follows: [REDACTED] g/kg deltamethrin, [Figure 6] g/kg [Additive B], [Figure 16] g/kg [REDACTED], [Figure 18] g/kg [Additive C], [Figure 12] g/kg [Additive L], [REDACTED] g/kg [Additive M], [REDACTED] g/kg LDPE, [REDACTED] g/kg of MDPE. The database includes the comment "error in mixing", which presumably refers to the low deltamethrin concentration.
523. The third document enclosed with GMR's letter dated 2 March 2007 was entitled "Netprotect Production Manual". This was described as "a detailed form of the process used to produce these products [i.e. the Thai Sample and Icon Life] … [which] are the same as our clients' current product". It was said to contain "the raw material specifications and recipes". The same document was annexed to Confidential Schedule 2 to the Defence, where is referred to as "NPT details.doc". I shall refer to it as "the Manual".
524. The statement that the recipe in the Manual was used to produce the Thai Sample (which was repeated by Michael Tackley of GMR in paragraph 3(iii) of his fourth witness statement made on 22 August 2008 on instructions from Mr Larsen) is inconsistent with the chronology in Confidential Schedule 2 to the Defence. The Thai Sample was purchased in March 2006, but Confidential Schedule 2 to the Defence indicates that the recipe in the Manual was not arrived at until November 2006. In fact, the Thai Sample was produced in July 2005 as discussed above.
525. The recipe shown in the Manual is [Figure 16] g/kg deltamethrin, [Figure 14] g/kg [REDACTED] (equivalent to [Additive A]), [Figure 6] g/kg [REDACTED] (equivalent to [Additive B]), [Figure 21] g/kg [Additive C], [Figure 12] g/kg [Additive L], [REDACTED] g/kg [Additive M], [REDACTED] g/kg MDPE and [REDACTED] g/kg LDPE. The Manual identifies the MDPE as being [Grade II] from Innovene. It also indicates that the masterbatches are LLDPE masterbatches.
526. Mr Howe observed in paragraphs 530-539 in his first report that this recipe seemed questionable for a number of reasons, and in particular that it was not clear how the blend of polymers had been arrived at. In Table 3-7 of Professor Stevens' second report served by the Defendants on 24 October 2008, Professor Stevens set out a comparison between the polymer composition in VF's 65.F.0304.5 formulation and that in Netprotect, stating in paragraph 116 that "The polymer blend formulation has been corrected from that shown, with errors, in 'NPT details.doc'". This was the first time that it had been suggested by or behalf of the Defendants that the Manual contained errors. FFW wrote to GMR asking for an explanation of this on 30 October 2008. On 9 November 2008 GMR replied:

"[The Manual] indicates in error that the HDPE content was only [REDACTED]% where in fact it was about [REDACTED]%. Mr Skovmand has confirmed that [REDACTED] of LDPE was used per 28 kilogram total batch. Professor Stevens has used these corrected proportions in Table 3-7 of his report."

527. This explanation was queried by Mr Howe in paragraphs 69-72 of his second report and by FFW in a letter to GMR dated 12 November 2008. I cannot trace a response to that letter, but there may have been one which has escaped my attention.
528. Confidential Schedule 2 to the Defence states:

"Nov.2006     Final recipe delivered to WHOPES phase II testing."

529. Sample 133 in the Netprotect database is dated Jan-June (presumably 2007) and is described as "production" for "sale". The recipe for this sample is [REDACTED] g/kg deltamethrin, [Figure 14] g/kg [Additive A], [Figure 6] g/kg [Additive B], [Figure 21] g/kg [Additive C], [Figure 12] g/kg [Additive L] and [REDACTED] g/kg [Additive M]. The entry includes the comment "error in mixing", which I assume explains the low deltamethrin concentration. I note that this recipe contains no LDPE or MDPE and does not correspond to the recipe set out in the Manual in a number of respects.
530. Samples 134-134.3 in the Netprotect database are undated and described as "WHOII". This suggests that these samples were made for submission to WHOPES II some time in 2007. They are all made from the same recipe, namely [Figure 16] g/kg deltamethrin, [Figure 14] g/kg [Additive A], [Figure 6] g/kg [Additive B], [Figure 21] g/kg [Additive C], [Figure 12] g/kg [Additive L], [REDACTED] g/kg [Additive M], [REDACTED] g/kg LDPE and [REDACTED] g/kg MDPE. The database also indicates that they contain [REDACTED] g/kg LLDPE from masterbatches.
531. In paragraph 35 of his first witness statement Dr Skovmand stated (the underlining and strikethrough shows the changes made immediately before he gave evidence):

"The formulation which we submitted for the WHOPES II recommendation test is No. 131 134 i.e. 130 compositions were tested before it, combined with different product conditions and test conditions."

He went on to say in paragraph 69 that Netprotect contained [Additive K] and [Additive M], although this statement was not tied to the formulation submitted for WHOPES II.

532. In paragraphs 81, 84 and 94 of his sixth witness statement Dr Skovmand stated (again as corrected):

"81.      In December 2006, there 4 tests planned in which LLDPE (low linear density polyethylene) was to be tested against low density polyethylene, in order to see the impact on release rate but only one I think was made. Before the test above were was analysed, we had to decide on the formulation to be sent to WHO for WHOPES II. We chose [REDACTED] Plus [Additive L] and with MDPE and LDPE added ([REDACTED]% of total polyethylene). This then became the recipe for Netprotect in January 2007. The current product is not identical to the test product of January 2005, and this one was different from the product sent to WHO in 2006 for the WHOPES I test. The one sent to WHOPES II is slightly different, but of course it cannot be very different or we would have had to have redone WHOPES I.

…

84.      All in all, over 130 samples from 6 series of experiments were made, tested and analysed to arrive at the formulation for Netprotect submitted for WHOPES II testing in (insert date) Spring 2007. …

…

98.     … As stated before, we did however change this [formulation] slightly in November 2006, especially in the aspect of the polyethylene composition. This is the product supplied for WHOPES II and sold now."

533. Notwithstanding my reservations about Dr Skovmand's evidence, I accept that the recipe submitted for WHOPES II evaluation was sample 134, that is to say, a [REDACTED] formulation plus [Additive L] (not [Additive K]), [Additive M], LDPE and MDPE. As he said, this is different to the recipe submitted for WHOPES I, but not very different. I think that Dr Skovmand was closer to the mark when he said that the submission to WHOPES II was in Spring 2007 than when he said that it was in November 2006.
534. The formulation of sample 134 is similar to, but not quite the same as, that set out in the Manual. The main differences are that (1) sample 134 contains [REDACTED] g/kg [Additive M] while the Manual gives [REDACTED] g/kg [Additive M] and (2) sample 134 contains [REDACTED] g/kg LDPE while the Manual gives [REDACTED] g/kg.
535. According to the Netprotect database sample 135 dated "June till mid July" (presumably 2007) was for "sale". This recipe for this sample was [Figure 22] g/kg deltamethrin, [REDACTED] g/kg [Additive A], [REDACTED] g/kg [Additive B], [REDACTED] g/kg [Additive C] and [REDACTED] g/kg [Additive M]. The entry includes the comment "error in mixing", which I assume explains the low concentrations of deltamethrin and the additives other than [Additive M]. I note that this recipe contains no LDPE or MDPE. Nor does it appear to contain LLDPE.
536. In December 2007 Netprotect passed WHOPES II.

The source of the first Netprotect recipes

537. The single most important, and most disputed, factual issue in this case is the source of the Netprotect formulations, and in particular the recipes used in the October 2004 trials. VF contend that these recipes were derived from the Fence database. The Defendants dispute this. Both sides have tended to treat this as an allegation of copying. In my view this is unhelpful and obscures the real issue. It is not VF's case that Dr Skovmand simply copied a particular recipe in the Fence database without further ado. Nor is it the Defendants' case that Dr Skovmand created the recipes that were tested without any reference to knowledge and experience he gained while acting as VF's consultant. The real issue, as I shall consider in more detail below, is whether Dr Skovmand crossed the line between, on the one hand, using knowledge and experience which, although gained during the course of acting for VF, he was entitled to use thereafter and, on the other hand, using confidential information amounting to a trade secret of VF in a manner which he was not entitled to do. In order to address that issue, however, it is first necessary to ascertain what Dr Skovmand did as a question of fact.
538. It is common ground that Dr Skovmand had a copy of the Fence database in his possession at all material times, as VF were aware. VF make no complaint about that in itself. It is also common ground that Mr Larsen retained a copy of the Fence database in his possession after the termination of his employment with VF. Again, in these proceedings, VF do not complain about that in itself.
539. I have set out above my general findings as to the history of the development of Netprotect. These demonstrate that in numerous respects the account of the development given in Confidential Schedule 2 to the Defence is not accurate. The same is true of the differing accounts given by Dr Skovmand and Mr Larsen in their various witness statements and in their oral evidence, although the account given in Dr Skovmand's sixth statement is more accurate than that given in his first statement and the account they gave under cross-examination is somewhat closer to the truth than the accounts given previously.
540. At this stage it is necessary to focus more specifically on how Dr Skovmand came up with the recipes which were tested in October 2004, and in particular the concentrations of deltamethrin, [Additive A], [Additive B] and [Additive C] in those recipes.
541. Dr Skovmand's story. It is convenient to begin with Dr Skovmand's account in his witness statements. In his first witness statement Dr Skovmand said this (emphasis in the original, and a subsequent correction underlined):

"30.      IIC subsequently developed 'NETPROTECT' from scratch without any reference to the way in which any VF product had been made (beside what everyone must do to make such a product: make masterbatches, mix, extrude, knit the product etc). The development process was very different: firstly we identified a capable producer (which VF never did for FENCE during the time IIC consulted on Fence (or PE net)), then we sought expert help from a company specializing in the design of plastics, getting a special combination of 3 different types of polyethylene. Then we looked to the process of extrusion, solving problems and ensuring the production of a net of sufficient strength for the purpose. The additives were selected from those recommended by Ciba and a number of others which, as far as I am aware (and I ought to know), VF had never used, which I had identified from the literature (including web sites). Then we had a testing program. We also tested additives from several producers to reduce costs without losing quality….

31.     A colleague told me that Sumitomo (the major competitor of VF) had been able to produce their PE mosquito net in some factories in India and obtained good results, although they never started any production there. He would not identify the manufacturer. However, later on, THL went to India and tracked the manufacturer down. It was not a manufacturer that VF had ever used, which was another attraction in using him. The first point was now to see if he could make a net of our design, which would mean thinner yarns and therefore greater difficulty in manufacture without losing large quantities of the insecticide in the process. Having researched the available additives, I suggested some preliminary formulations for Intection to test the company. The additives I suggested were those suggested and promoted by Ciba and I also tested a number of others I found by researching the subject, for instance, by using Google and also on specialchem4polymers.com.

…

60.     … It is simply a matter of trial and error to determine the minimum levels of DM which are effective in the final product, although, of course, this depends also on the degree of migration in the plastic. We tested more than 10 different dosages in many combinations to arrive at the low dose we can use."

542. In his sixth witness statement Dr Skovmand said:

"69.     The additives were selected from those recommended by David de Corte of Ciba when I inquired in 2001 (when I was considering what additives I should use in Zerofly) (see the document exhibited already in OS1 and OS5 and e-mail from Ciba (pages 36-38 and 147 of 'OS6')). Others were selected based on literature and company information. David had suggested that the UV filter [Additive A] should not be used in yarns, because it leaves the yarn due to migration. This was all information freely given to me and clearly could not be considered any sort of trade secret of Ciba or, especially, VF. I chose to add it (despite possible negative experience from the Fence production) because the molecule is very close to a pyrethroid and thus is also a process aid.

…

71.      As can be seen from the Ciba recommendation, the percentages which David suggested I should consider were [REDACTED]% [Additive D], [REDACTED]% [Additive B], [REDACTED]% [Additive A] and [REDACTED]% deltamethrin. He later changed his mind and recommended [Additive C] instead of [Additive D] (see mail from David de Corte (pages 36-38 of 'OS6')). I started there: I did not use any of the information from the results recorded in the Fence trials …"

543. In his seventh witness statement Dr Skovmand said:

"21.      It is also remarkable that the Fence formulations which IPT claim to have been copied were … the two specific formulations which I came up with by looking again at the Ciba recommendations. Surely this illustrates the point that I was the same developer, starting again from the same point (the Ciba recommendations) and therefore likely to come up with much the same formulations to test. …

…

27.      The Netprotect development started exactly the same way as the Fence development (except that the former did not start with FMC bifenthrin masterbatch) with the advice of CIBA on these additives, their concentrations and effect, a grid of formulations and trial and error testing by analysis of results. It is inevitable, therefore, that one can find similarities in the two 'grids'."

544. The "preliminary formulations". As set out above, after the Defendants had served Dr Skovmand's first witness statement, FFW wrote to GMR on 13 April 2007 requesting copies of documents referred to in it including the "preliminary formulations" referred to in paragraph 31. I have quoted GMR's response to this request on 17 May 2007 above.
545. The document enclosed in tab 4 of the bundle of documents enclosed with GMR's letter dated 17 May 2007 was the first version of Joint Project memo. Reading Dr Skovmand's first witness statement together with this response, the message which is conveyed is that Dr Skovmand suggested that Intection should begin by testing the formulations set out in the Joint Project memo i.e. those recommended by Ciba.
546. In cross-examination Dr Skovmand said that this document was not what he had given Intection, but rather the information upon which he had based his proposals. He was unable to explain why GMR had said that it was the information referred to in paragraph 31 of his first statement. Mr Larsen said that he assumed that this came from Dr Skovmand. In closing submissions, counsel for the Defendants suggested that the production of this document was a mistake on the part of GMR, without any evidence to support the suggestion. I do not think it was a mistake on the part of GMR, I think that the document was supplied to GMR by Dr Skovmand in order to enable them to respond to the request. In any event, the question remains as to what Dr Skovmand did provide to Intection. As I have noted above, this has not been disclosed even now.
547. The Ciba recommendations. In his sixth witness statement Dr Skovmand claims that (i) the additives were "selected from those recommended" by Mr De Corte of Ciba in the Joint Project memo and that (ii) he "started from" the formulation in the email dated 23 August 2001.
548. I have already found that the initial recommendations were made by Mr De Corte in response to a specific request for advice in relation to particular proposed application and arose out of discussions between Dr Skovmand and Ciba. I have also found that Mr De Corte refined his recommendations as he received more information from Dr Skovmand. I shall consider the legal implications of this below. The question I shall address at this stage is whether Dr Skovmand did "start from scratch" from those recommendations.
549. So far as the Joint Project memo is concerned, this proposed eight formulations for making tarpaulin. All contained [REDACTED]% [Additive D]. One contained [REDACTED]% [Additive A], one contained [REDACTED]% [Additive B], one contained a combination of [REDACTED]% [Additive A] and [REDACTED]% [Additive B], and three contained the same combination with another additive. None contained [Additive C]. As noted above, Mr De Corte appears to have suggested [REDACTED]% deltamethrin, but Dr Skovmand decided that [REDACTED]% would be better. None of these formulations matches any of the recipes tested by the Defendants in October 2004. On the contrary, there are a number of differences. Accordingly, Dr Skovmand cannot have started from this document.
550. As for the 23 August 2001 email, this recommended [REDACTED]% [Additive A], [REDACTED]% [Additive B], [REDACTED]% [Additive C] and [REDACTED]% [Additive F] in the LDPE outer layers of a 3 layer sandwich with unspecified amounts (but presumably meaning [REDACTED]% and [REDACTED]% respectively) of deltamethrin and [Additive D] in the HDPE middle layer. Again, this formulation does not match any of the recipes tested in October 2004 even disregarding the differences in polymer composition and type (LDPE film versus for HDPE yarn if one concentrates on the outer layers, as the Defendants do in order to sideline [Additive D]) and usage (tarpaulin versus bed net). Indeed, the Defendants themselves emphasised in their evidence that one of the differences between the Netprotect recipes and some of VF's was that they did not use [Additive F]. Again, therefore, Dr Skovmand cannot have started from this document.
551. Furthermore, as Dr Skovmand acknowledged in his sixth statement, after the 23 August 2001 email Ciba recommended that [Additive A] be removed from the formulation. This was during the meeting on 3 September 2002. Dr Skovmand questioned this recommendation after an analysis of tests of different formulations carried out by or on behalf of VF. Nevertheless, Dr Skovmand did remove [Additive A] from the formulation, only to reinstate it in August 2003 at a concentration of [Figure 4] g/kg having found that its omission was disadvantageous. This did not constitute "possible negative experience", but positive experience. As Dr Skovmand put it in his email dated 19 November 2004 referring to this recommendation, "Never just listen to experts, always try".
552. Still further, Mr De Corte's recommendation for PE bed nets during the September 2002 meeting was of a recipe containing [REDACTED] g/kg [Additive C] and [REDACTED] g/kg [Additive O], described as a "guess", which bears no relation to the initial Netprotect recipes.
553. In his email dated 11 August 2004 Dr Skovmand did not refer to starting from recommendations made by Ciba. On the contrary, what he said was:

"Based on our Fence tests we can make some good guesses as to a formulation should be."

554. Other sources. In his first witness statement Dr Skovmand suggested that, in addition to those recommended by Ciba, he identified other additives for testing by his own researches into the literature including websites. This suggestion was repeated in his sixth witness statement. Leaving aside the fact that this suggestion is inconsistent with his claim that he started from the Ciba recommendations, it is clearly incorrect. As Dr Skovmand accepted in cross-examination, no other additives were tried in October 2004 or, indeed, in January 2005. The Defendants did not try other additives for use in Netprotect until later.
555. The odd numbers. One of the matters relied upon by VF as showing that Dr Skovmand used the information in the Fence database in devising the Netprotect recipes is the use of odd and unexplained concentrations of some of the ingredients in both, in particular [Figure 2] g/kg of deltamethrin and [Figure 3] or [Figure 15] g/kg of [Additive A] and [Additive C].
556. I have set out the reason why a deltamethrin concentration of [Figure 2] g/kg is specified in several of the recipes for samples 7-20 in the Fence database, namely that it was a calculated figure. Mr Larsen and Dr Skovmand gave a series of different explanations for the use of this figure in the Netprotect recipes.
557. The first explanation was provided by counsel for the Defendants on instructions from Mr Larsen during the course of the hearing before Mr Wyand QC. This was that [Figure 2] g/kg times [REDACTED] g/m² (which was said to be the weight of the Defendants' net) gives [REDACTED] mg/m² (which was said to be what WHO wanted). This explanation is clearly wrong: the weight of Defendants' net started out at 25 g/m², and WHO did not recommend an insecticide concentration of [REDACTED] mg/m².
558. The explanation given by Dr Skovmand in paragraph 2 of his fifth witness statement was as follows:

"This comes about because until recently, the total weight of masterbatch and basic granules used when testing and in production, was 28 kg. Instead of recalculating to make the total weight of the net and the additives 25 kg, we simply calculated the dosages to be added to a 25 kg sack of polyethylene HDPE. The [Figure 2] therefore comes from adding [REDACTED]g of deltamethrin [REDACTED]% masterbatch into 28kg of polyethylene (25kg plus weight of all masterbatches). If you look at the database, you can see we started with a net of 25 g/m² and [Figure 2]g DM/kg net. Since 2007 masterbatches were made so the workers can weigh out even number of kilo masterbatches on top of the 25kg sack and get an even number of % DM, like [Figure 16] g/kg."

     This explanation was repeated almost word for word in paragraph 137 of his sixth witness statement.

559. As Mr Howe pointed out, there are a number of problems with this explanation. These include the following:

i) There is no record of the Defendants having used a total mass of 28 kg per batch. Instead the recipes indicate a total mass of 25 kg.

ii) If the intended polymer mass was 25 kg, that would imply that the intended deltamethrin concentration was [REDACTED] g/kg. This is itself an unexplained figure with no support in the documents, and inconsistent with the suggestion that the target was [Figure 16] g/kg.

iii) Adding masterbatches to a 25 kg sack of HDPE would not give a constant total mass of 28 kg. On the contrary, as the formulations changed, the total mass would change. Hence the concentration of deltamethrin would change. This is not what the Netprotect database shows, which is a constant figure of [Figure 2] g/kg deltamethrin with differing concentrations of additives and other materials such as LDPE.

iv) It is inconsistent with the documents relating to the July 2005 trials in Karur and the note about the weighing error of 1 kg in HDPE.

560. In cross-examination Dr Skovmand accepted that his first explanation was wrong. He said that it was a deduction he had made.
561. In paragraphs 10 and 145 of his seventh statement, Dr Skovmand acknowledged that his previous explanation might be incorrect and offered an alternative explanation:

"10.      … I believed that the real origin of the figure was the one I set out in [paragraph 137 of his sixth statement] (because the database referred to it being an error). However, whether it was one or the other or some other reason is not really the point, the point being that it was not a figure which the manufacturers could achieve: they could never be that accurate, weighing out the ingredients individually …. There was no point specifying [Figure 2], rather than [REDACTED] or [REDACTED] or even [Figure 16] therefore. This is particularly the case because the bioassays would not discriminate significantly between such products. One would never target [Figure 2] and, in the final formulation, one would aim for a rather rounded figure anyway. In trials you would normally try to take a convenient number – one which made the job of weighing out easier. Having once made a calculation and come up with a figure of [Figure 2], that figure just became a sort of 'pet number' for me. It may be remarkable, but it is of no significance at all.

145.     … Whether if it was [Figure 16] or [Figure 2] g/kg of deltamethrin is of little importance and it can be seen, that, from March 2006, it is [Figure 16]. [Figure 2] was a 'pet number', but of no real significance. It just happened to be a number which I had once calculated (after the event) to have been used at one stage and I just used from then on: it was a quirk I had, rather than a number of any consequence."

562. In GMR's letter dated 26 November 2008 responding to the queries about Dr Skovmand's seventh statement raised in FFW's letter dated 24 November 2008 and subsequently by VF's counsel, GMR responded to a question in relation to the parenthesis in paragraph 10 asking which database referred to it being the result of an error as follows:

"The database referred to is the one called 'extract DB Apr05.xls' which is the database as it was in April 2006. It is the tab 'kahrur 0705', column L, line 12-18."

563. This explanation does not begin to stand up. As discussed above, the weighing error referred to meant that the July 2005 concentrations were not [Figure 2] g/kg when they were meant to be.
564. In a subsequent correction to paragraph 115 of his seventh statement, Dr Skovmand offered a further (third) explanation:

"Likewise, the difference between [Figure 2] or [Figure 16] g/kg was of no concern. The IPT conclusion in paragraphs 388 and 390, that IIC was aiming at [Figure 2] g/kg does not seem to follow and is wrong in any event: it would be pointless to aim for such an exact concentration. IIC was actually aiming for having [REDACTED]% in the product at that time, adding a little more to allow for losses in production."

565. When he was cross-examined on this point, Dr Skovmand initially gave an explanation that was a combination of his second and third explanations:

"The correct explanation is that [Figure 2] is something between 5 and 10% more than [Figure 16], that would be [REDACTED] and then I used [Figure 2]. That is a 'pet number' you have got. You can see [REDACTED], so it is not so difficult to remember."

566. Ultimately, however, Dr Skovmand accepted that the figure in the Fence database was a calculated one and gave a rather simpler explanation for its use in the Netprotect recipes:

"Q. The true answer is that you simply copied it from the Fence

database?

A. I simply remember[ed] it from the Fence database."

567. As for the figure of [Figure 15], the explanation given by Dr Skovmand in paragraph 3 of his fifth witness statement and repeated in paragraph 138 of his sixth witness statement was as follows (ignoring the subsequent "correction" discussed below):

"The others are the result of my correcting errors by the workers in formulating the samples: the figures are taken straight off my calculator. One example is the figure [Figure 3], which is again the result of the workers simply adding the additives to a 25 kg bag of the plastic substrate, instead of making up the total to 25kg, thereby reducing the amount of the additives from the intended [REDACTED]%. "

568. As Mr Howe pointed out in his first report, the arithmetic here is wrong. If the intended concentration was [REDACTED] g/kg (which is actually [REDACTED]% rather than [REDACTED]% but is clearly what is meant), the result of adding the masterbatches to a 25 kg sack of HDPE to make a total mass of 28 kg would be a concentration of [REDACTED] g/kg, not [Figure 3] g/kg.
569. Even though Dr Skovmand commented on the relevant paragraphs of Mr Howe's first report in his seventh witness statement, he was silent on this point. He neither refuted what Mr Howe had said, nor offered an alternative explanation.
570. Paragraph 3 of the fifth statement and paragraph 138 of the sixth statement were among those which were corrected by Dr Skovmand on the day before he gave evidence. In this case the correction is a striking one. It consists of deleting the references to "[Figure 3]" and "[REDACTED]%" and substituting "[REDACTED]" and "percentage" respectively. The explanation provided for this correction was as follows:

"[A] bad example chosen – reconsidering the Fence database, OS finds that there is no indication that the figure [Figure 3] was ever [Figure 3] (which would indicate a recalculation on a calculator), thus concluding that he is likely to have erred when thinking the figure arose from a mis-weighing of ingredients, rather than simply a choice made by him to use [REDACTED] in accordance with his mathematical protocol."

571. The effect of this correction is that the witness statements as corrected provide no explanation for the figure of [Figure 15], but instead seek to explain a figure of [REDACTED] which was not the figure which VF had contended needed explaining. Indeed, it is not a figure which appears in the Fence database. The suggestion that this was a "bad example" is impossible to accept: it was not a bad example, it was a failed attempt to explain that figure. Although the witness statements as corrected (which Dr Skovmand confirmed in his evidence in chief) provide no explanation for the figure, the explanatory sheet (which Dr Skovmand did not confirm) offers the unelaborated suggestion that this was "in accordance with his mathematical protocol".
572. As I shall discuss below, it was only during cross-examination that Dr Skovmand elaborated on how he said that his "mathematical protocol" explained the use of [Figure 15].
573. On any view, Dr Skovmand's attempts to explain the use of the odd numbers in the Netprotect recipes are highly unsatisfactory. Before coming to a final conclusion on this point, however, it is necessary for me to consider the linked matter of his "mathematical protocol".
574. Dr Skovmand's "mathematical protocol". Although it was not elaborated in the explanatory sheet, I am prepared to accept that the phrase "his mathematical protocol" was intended to refer to what Dr Skovmand had said in paragraph 13 of his sixth witness statement:

"Over the years (before I started to do any work for VF) I developed my own method of testing and experimenting. I used general mathematical criteria and developed a deterministic method, with the exercise of my own judgment. I knew from this, never to synchronise the steps which I used to vary the additives: there had to be asynchronism otherwise the results would just be synchronised as well. I therefore used what might appear to be slightly random figures but they were in fact chosen deliberately using my gut feel and judgment. I would also grade the parameters in importance. I used this same approach once I started to work for VF and it is still my method today."

575. It is important to note that: (i) none of Dr Skovmand's first to fifth witness statements referred to this; (ii) in the sixth statement as originally served, Dr Skovmand did not state that this was the explanation for the figure of [Figure 15], but gave the different explanation discussed above; (iii) Dr Skovmand said nothing more about this in his seventh witness statement; (iv) no mention was made of this in the Defendants' skeleton argument for trial; (v) Mr Howe was cross-examined in the manner discussed below after service of Dr Skovmand's seventh statement and before the corrections to Dr Skovmand's statements; (vi) the explanation in the explanatory sheet was not confirmed by Dr Skovmand in his evidence in chief; and (vii) even in the explanatory sheet there was no reference to VIF values.
576. The case presented by the Defendants in their skeleton argument for trial was that Dr Skovmand had devised the recipes in the Fence database as follows (emphasis added):

"In connection with that project Ole Skovmand's company started from additive compositions suggested by Ciba and proceeded by trial and error. It tried out a large number of samples, varying the ingredients stepwise and observing the results. Thus it built up a sort of grid – referred to in this case as the database."

The Defendants contended that Dr Skovmand had followed the same course when devising the Netprotect recipes:

"It was the same man on both occasions and in each case he started with additives in proportions recommended by Ciba, and proceeded by trial and error."

577. The case put to Mr Howe in cross-examination some seven weeks later was subtly, but significantly, different. Counsel first asked Mr Howe about his approach:

"Q. … You have put forward a report which suggests that Dr. Skovmand

copied from the Fence database. What steps did you take to

investigate the contrary hypothesis before you accused your

scientific colleague of cheating?

A. I looked at the claims that the formulations had come from

published information, from information provided by Ciba

Geigy, from information provided by Innovene, and I could not

correlate that information to the formulations that occur in

the October production trials.

Q. So were you, therefore, as it were, investigating the

hypothesis that the explanations for this or that numerical

parameter where they came explicitly in that form from that

source? That is to say, if for example it says [Figure 3], that

specifically that number came from a specific source, such as

an additive that Ciba Geigy supplied?

A. Yes, I was looking for a source of (a) the typical values used

and (b) the specific values which were used in the

formulations.

Q. To what extent do you consider that, first of all, this was

admittedly a case of the same man doing essentially the same

job a second time?

A. I do not believe it was the same job done the second time. It

was, in my opinion, it was a continuation of the same job

conducted by the same man."

578. Counsel then asked Mr Howe to consider "hypothesis 2", namely "that there is no copying, no cheating". The cross-examination proceeded as follows:

"Q. By looking at hypothesis 2, why do you suppose the

investigator might choose a number like [Figure 3] to investigate in

the first place?

A. When your other two variables are [Figure 6] and [Figure 4], it would appear

logical to go for the [Figure 8].

Q. Yes, reinforcing my question, really, yet it is [Figure 3]?

A. Yes.

Q. A strange number. Talking about when you did Fence, as a

scientist did you ask yourself why should that number of all

numbers that could have been was chosen?

A. I asked myself and I could find no evidence to support why.

The most likely conclusion I drew was that it was probably

some formulation error.

Q. Some formulation error. We will come to that in a moment.

Apart from cases where [Additive B] is set to [Figure 5], is this essentially

the pattern that is investigated? The variables are varied

step wise but because of introducing a number like [Figure 3], which

obviously has no logical relation to anything in particular,

it means that it is irregular step wise, do you follow?

A. Yes.

Q. What now that you think about it, might the advantage of that

be?

A. It would throw an irregularity into the data which would

presumably better test a response.

Q. Yes, and why would it be good to throw an irregularity into

the data, thus to better test?

A. To form better confidence on trends that were observed.

Q. Why would using [Figure 3] rather than, say, [Figure 8] or [Figure 17] give

better confidence?

A. I do not know.

Q. Yet throughout the Fence database again and again we see this

number [Figure 3], do we not?

A.     Yes.

…

Q. … Have you any scientific experience of a third, and I

suggest better, method, which is, provided I take proper steps

to vary one of the parameters in irregular steps -- irregular

steps like [Figure 3] -- then I can vary them all together, no more

than one anyway, thus getting the combined advantages of

getting a result which is sufficiently reliable, and saving

time and money. Correct?

A. Yes.

Q. Is that right?

A. Yes.

Q. OK. I do not want to be unfair to you but before now had you

heard of that method?

A. Yes, I have heard of similar experimental design procedures.

Q. What steps did you take when writing your reports to ask

yourself whether what was going on in the Fence database, the

reason the space, the N-dimensional space was being searched

in that way was because that method was being used?

A. I considered the data and the way that things were being

adjusted and I tried to see if the data would fit some sort of

quadratic high/low, high/low interaction.

Q. Is the short answer to my question you did sort of look into

it but you will now fairly admit you did not think far enough?

A. If that's the way the system was being varied."

579. I note that the use of irregular steps was suggested to Mr Howe as explaining the choice of [Figure 3] g/kg of additives, but not the choice of [Figure 2] g/kg deltamethrin. It is hardly surprising that Mr Howe did not consider this possibility in his reports when it was not put forward by Dr Skovmand in any of his first to seventh witness statements as being the explanation.
580. Counsel then turned to the question of the significance of the odd numbers:

"Q. …. I am just interested for the moment in the

numbers, [Figure 3], [Figure 6], [Figure 3] and so on. Was the choice of those

numbers second time around what led you to suspect cheating?

I want an honest answer to this, please.

A. The combination of [Figure 2] for deltamethrin content, which is

only doubled, there are only two values whichever database you

look at. ..

[Figure 2], [Figure 3] were strong, I took as strong indicators that

similar technology had been moved forwards."

581. I note that Mr Howe placed reliance on [Figure 2] as well as [Figure 3].
582. Counsel then put it to Mr Howe, and Mr Howe accepted, that (i) the Ciba datasheet for [Additive A] suggested combining it with a [REDACTED] and that [Additive C] was a [REDACTED]; and (ii) the ZeroFly patent application, of which Dr Skovmand was a named inventor, indicated that it was advantageous to combine [Additive A], [Additive B] and [Additive C].
583. Counsel then proceeded:

"Q. In the light of these matters, the feeling of the search

pattern and what you have said about that, [Figure 3], [Figure 6], [Figure 3] and

so on, the same man aiming at the same goal twice over, using

additives which we have just agreed he must have investigated

before, is there any scientific reason to suppose there is

anything remarkable about the same numbers turning up in the

two databases?

A. No, I do not think so.

Q. If I have been unfair to you, that was wrong of me because

I think I am now revising my opinion. As a fair minded

scientist, with the light of what I have been putting to you

in cross-examination, you would have expressed your report

somewhat differently or with certain qualifications, to say

the least, would you not?

A. I think the same methods could have been applied by the same

person and assuming the same starting point, then it is

reasonable to assume the same numbers could be used.

Q. Without any impropriety that one would think as a scientist

was going on?

A. Yes, as long as that data is freely available to be used."

584. I note that Mr Howe's answer here is carefully qualified in three ways: (i) assuming the same starting point (ii) as long as that data is freely available to be used (iii) the same methods could have been applied by the same person, in which case the same numbers could be used without any impropriety. As to (i), it was not put to Mr Howe that Dr Skovmand had started from any specific recommended concentrations, such as those in Mr De Corte's email dated 23 August 2001. As to (ii), VF's case is that the data was not freely available to be used. As to (iii), the question and answer are both predicated on the earlier cross-examination about the use of an "irregular stepwise" experimental design having been employed by Dr Skovmand on both occasions. As noted above, it was not suggested to Mr Howe that this explained the choice of [Figure 2] g/kg deltamethrin.
585. Counsel then put it to Mr Howe, and Mr Howe accepted, that in both cases the concentrations stated in the recipes were planned rather than actual concentrations.
586. The cross-examination concluded as follows:

"Q. I am going to ask you this as a perfectly formal question. I

am going to, I submit, test your fairness as a scientist. Now

that we have been through this is there anything in the

databases that as a scientist makes you infer copying at all?

A. Based on those three additives being used, and your line of

questioning, I think I need to concede that they could be

achieved by the same person using the same approaches.

Q. Yes, that is not quite the point I put to you. It is not that

it could be achieved. Is there any scientific reason for

saying that there has been copying at all, based on scientific

reasons and what is in the databases?

A. I think there is a hypothesis that says that there was not.

Q. We know there is a hypothesis, there is always in science two

hypotheses at least, the question is, is there any, I put it

this way, is there any scientific support for hypothesis one;

namely, that there was copying? I think you are a man of

integrity. You know what the answer is, do you not?

A. I think there is a strong -- I am struggling to respond to the

direct nature of the question. I think there is a scientific

approach which could lead to hypothesis two being correct,

yes.

Q. Legally I do not have to do this but I believe, contrary to my

first impression, you are actually a very honest man. I am

going to ask you a last time. Is there any scientific

evidence that supports hypothesis one; namely, copying? For

your benefit I give you that opportunity to answer that.

A. I think the support for hypothesis one is in the use of

factors like [Figure 2] for a deltamethrin content, where that does

not form part of a series that is being randomly varied or

systematically varied. That is a value that has been

selected.

Q. Same chap second time around. Is there any scientific basis

for saying he would never be able to remember [Figure 2]?

A. I do not see why he would specify [Figure 2] and not [Figure 16]. For [Figure

16] if you are making up a 25kg batch, you add [REDACTED].

…

Q. Final, final, final, time. It is your reputation as a

scientist now. Do you still accuse your scientific colleague,

Dr. Skovmand, of fraud, cheating, lying, simply based on the

scientific evidence available to you from these databases?

A. I still believe that the work that was done in Fence was used,

if you like, to springboard the Net Protect development.

Q. Is that based on scientific reasons, because if you say yes I

am going to ask you what they are. You do not get off with the

question, if I can put it that way.

A. I think, no, I cannot put a scientific basis behind that."

587. I note that Mr Howe again referred to [Figure 2] as being support for "hypothesis one", and that the explanation put to Mr Howe for this figure was that Dr Skovmand had remembered it. As discussed above, this was ultimately the explanation that Dr Skovmand gave in cross-examination. I also note that the last two questions restricted Mr Howe to "the scientific evidence available to you from these databases", whereas in his reports Mr Howe had not merely taken into account the information in the databases, but also in the disclosure documents. Moreover, he had considered at length and in detail the history of the Netprotect development put forward by the Defendants in Confidential Schedule 2 to the Defence, and concluded that it did not stack up. That evidence was not challenged in cross-examination. His final two answers must be seen in that light.
588. When he was cross-examined about his explanation for the use of [Figure 15], Dr Skovmand said this:

"A. So the logic behind this that you see on this page is that if you

want to have a series of dilutions which is [Figure 6], [Figure 8] and [Figure 4],

then your statistical analysis will go wrong. You cannot have

the same steps in your recipes because if you do that, you get

I think it is called co-correlation; you get high VIP data,

that means your analysis will come out and say that some of

your independent variables are not independent. Therefore,

I put in irregular number to avoid that problem.

…. So instead of putting in [Figure 8] here, I put

in [Figure 15].

Q. Why did you do that?

A. Because if you put in [Figure 8] you risk that you get, I said before

high VIP values, it means that two of your parameters are not

independent. The whole point of making these screens of

testing is that you can -- that I can make linear regression

analysis after that. And to do that the demand is that the

parameters are totally independent. If they are not totally

independent, then your analysis will go wrong. So what you

have to do is you have two possibilities, you can be very,

very careful in your planning or you can put in some odd

numbers. If you put in some odd numbers, that will work.

Q. What is VIP?

A. VIP is a statistical parameter which indicates that two

parameters are not independent."

589. By VIP Dr Skovmand meant VIF, as he subsequently confirmed. He went on to explain that he used Statistix 8 software to carry out linear regression analysis. Later he said that his rule of thumb for a VIF value indicating a problem of co-linearity was about 2.
590. On the next day Dr Skovmand was cross-examined upon a series of linear regression analyses Mr Howe had carried out (using Statistix 9 rather than 8, but it was not suggested this made any difference). I subsequently gave VF permission to serve a short statement from Mr Howe verifying these analyses, but the Defendants chose not to cross-examine Mr Howe on that statement.
591. Mr Howe carried out two analyses on the recipes relating to samples 7-20 in the Fence database. The first analysis was on samples 7-15 (excluding 9 since it is a duplicate), namely the samples which did not include LDPE or [Additive G]. This analysis gave a VIF of either 1.1 or 1.2 i.e. close to ideal. The second was on all the samples. This gave a VIF ranging from 1.5 for deltamethrin to 2.2 for [Additive G]. This analysis confirms the impression one gets simply from inspecting the recipes that samples 7-15 constitute a fairly well-planned series whereas samples 16-20 appear to be more of an afterthought.
592. Mr Howe repeated both sets of analyses using an additive concentration of [Figure 8] instead of [Figure 3]. This gave VIF values that were the exactly the same for samples 7-15 and ranged from 1.5 to 2.3 for samples 7-20 i.e. little different.
593. Dr Skovmand's evidence was that he had not carried out a VIF analysis in advance of planning the experiment for samples 7-20 or 7-15, but instead had relied upon his experience of VIF in linear regression analyses in designing the experimental grid and selecting [Figure 3]. Given that it turns out that [Figure 8] would have been very nearly as good, I am prepared to accept that this is so. Nevertheless, the point remains that samples 7-20, and in particular samples 7-15, constitute a well-designed set of experiments. As I have said above, I consider that they were planned with a view to avoiding co-linearity problems.
594. Mr Howe also carried out an analysis on the six recipes tested by the Defendants in October 2004. This gave VIF values ranging from 3.6 for LDPE to 31.2 for [Additive C]. These values are well above Dr Skovmand's threshold of 2 and, in the case of two of them, high enough on any view to indicate a serious problem of co-linearity. Repeating the analysis using [Figure 8] instead of [Figure 15] gave VIF values ranging from 3.6 to 23.6 i.e. better but still high.
595. Dr Skovmand's response when this was put to him in cross-examination was that it was not correct to put deltamethrin in as an independent variable because "you cannot foresee that your outcome is independent of your insecticide concentration", and that he had put it in as weighted parameter. As noted above, this explanation was not put to Mr Howe. In any event, it does not hold water for three reasons.
596. First, deltamethrin is an independent variable. As discussed above, "independent" refers to independence of the input parameters (here the concentrations of insecticide and additives) with respect to each other. The outcome, the dependent variable, is not independent of the input parameters, nor is it expected to be. Even if one takes the outcome in the Netprotect trials to be mosquito death, the fact that this is indeed expected to depend on deltamethrin concentration does not mean that the latter is not an independent variable. Furthermore, this was not the only outcome Dr Skovmand was interested in, he was also interested in deltamethrin retention during production.
597. Secondly, it is not the case that Dr Skovmand only included deltamethrin as a weighted parameter in his own linear regression analyses. Sometimes he did and sometimes he did not.
598. Thirdly, including deltamethrin as a weighted parameter does not lead to lower VIF values. Indeed, comparing two of Dr Skovmand's own analyses shows that weighting using deltamethrin concentration increased the VIF values of all variables except one, which remained the same.
599. I conclude that the Netprotect October 2004 trials were not a well-designed set of experiments which were planned to avoid problems with co-linearity in the way that samples 7-15 in the Fence database were, and that the explanation for the choice of [Figure 15] was not Dr Skovmand's "mathematical protocol".
600. The seven improbabilities. In the Defendants' skeleton argument for trial, it was contended that the allegation that Dr Skovmand copied from the Fence database "rested upon" seven improbabilities. I shall consider these in turn.
601. (1) It was said that Fence was intended to have a gauge of 70 g/m² so as to be suitable for agricultural use, whereas as Netprotect was originally intended to have a gauge of 25 g/m² (later 35 g/m²), and therefore the migration rate for insecticide in the former would be hopelessly wrong for the latter.
602. Like many of the Defendants' points in this case, this appears persuasive at first blush, but does not stand up to analysis. First, it is simply incorrect that all the Fence samples were 70 g/m²: some were 42 g/m², including sample 13. Secondly, samples 7-20 in the Fence database were not intended just for agricultural use, they were intended for use as bed nets as well. It follows that the insecticide migration rate was intended to be suitable for bed nets. Thirdly, when tested by means of wash resistance bioassays, Dr Skovmand considered the results to be very promising for use in bed nets. Fourthly, Dr Skovmand himself made the point in his email dated 12 September 2003 that "PE yarns are now used for many things in VF, but they are basically the same, and it is optimal for me that I can calculate on any PE yarn whatever it is used for". Fifthly, Mr Howe gave unchallenged evidence that the concentrations of deltamethrin per unit surface area of net, as achieved in practice in the Fence samples and as intended to be achieved in the first Netprotect trials, were very similar. Sixthly, Mr Howe gave unchallenged evidence that linear regression analysis of the wash test data in respect of Fence samples with [Figure 2] g/kg nominal deltamethrin concentration indicated that a reduction in net weight would be beneficial.
603. (2) It was said that, when working on Fence, Dr Skovmand had not succeeded in obtaining levels of insecticide that were consistent even within the same batch, and accordingly the results were not useful as a guide to formulation of a successful product. Once again, this appears persuasive at first blush, but rather less so on further consideration. First, it is quite true that the analysed deltamethrin concentration in samples 7-20 ranged from 50 to 66% of the nominal deltamethrin concentration. Mr Howe gave unchallenged evidence, however, that the samples had enough consistency to enable useful conclusions to be drawn. Indeed, that is what Dr Skovmand did in fact do. Secondly, if a formulation gives good bioassay results despite loss of deltamethrin in production, one would expect it to give better results if that loss can be reduced. Thirdly, as Dr Skovmand accepted in cross-examination, he and Mr Larsen received the news that VF's Chinese manufacturer had successfully produced PE yarn with very little loss of deltamethrin in May 2004, that is to say, around the time when Dr Skovmand, Mr Larsen and Mrs Sig decided to develop a PE net. Fourthly, Dr Skovmand's own evidence was that one of the reasons for specifying [Figure 2] g/kg deltamethrin was that he was aiming for a concentration of [Figure 16] g/kg and wanted to allow for some loss in production. Fifthly, it is clear that Dr Skovmand did not regard some degree of deltamethrin loss as a serious problem. Thus it was his evidence that the product submitted to LIN in May 2005 had unpredictable levels of insecticide. Even as late as 20 November 2006, the Defendants were experiencing some loss of deltamethrin in production, but Dr Skovmand was not too concerned about this.
604. (3) It was said that the reason why Netprotect was successful was not due to its additives, but due its polymer composition and manufacturing process. In particular, the Defendants relied on the contention that Netprotect contained LLDPE and MDPE, but the Fence samples did not, enabling a lower extrusion temperature to be employed. As to this, the evidence not only of Mr Howe but also of the Defendants' expert Professor Stevens was that additives play an important role in polymer processing. Thus in his report made in support of the Defendants' application to strike out the claim Professor Stevens said:

"The benefits of plastics additives are not marginal – they make the difference between success and failure in plastics technology."

Furthermore, it is clear from the evidence in this case that the choice and concentration of the additives makes a difference to the performance of the product, in particular so far as migration of deltamethrin is concerned.

605. I accept that polymer composition and manufacturing process are also relevant factors, but as discussed above the Defendants' production runs of Netprotect have not always included LLDPE or MDPE, or even LDPE. I would also observe that the Netprotect database shows that, contrary to the Defendants' claim to employ an extrusion temperature of [Temperature 3], the more recent samples were extruded at [Temperature 4]. In any event, however, I consider that the relevance of these factors is not to the question of the source of the early Netprotect recipes, but to another issue which I will consider below.
606. (4) It was said that the migration rate was strongly affected by the polymer composition and so the Fence results were useless in principle. This submission was unsupported by evidence. To the contrary, Mr Howe gave unchallenged evidence that the available data does not enable any such conclusion to be drawn.
607. (5) It was said that the samples of Fence relied upon by VF were early, rejected samples. In my judgment this is not a correct characterisation of the facts. Samples 7-20 do occur early in the Fence database, but some of these recipes (in particular those for samples 11 and 13) were used repeatedly in the subsequent work. There is no evidence that these recipes were rejected. Rather the position is that, as VF concentrated more on developing Fence and less on developing a PE bed net, so attention was focussed on some of the samples and on recipes developed from them.
608. (6) It was said that the later two field trial samples were obviously different. This is a reference to the 65.F.0503.5 and 65.F.0304.5 recipes, which VF no longer rely upon as such. This is really the same point as point (5), and the answer is the same. It is true that the Netprotect recipes are not as similar to those recipies as they are to the samples which VF do rely upon, but those recipies were specifically Fence recipes. Moreover, the work that was done using the 65.F.0503.5 recipies confirmed the need to include [Additive A], contrary to Mr De Corte's suggestion.
609. (7) It was said that the same man started from the Ciba recommendations in both cases. I have already dealt with this above: in devising the Netprotect recipes Dr Skovmand did not go back to the Ciba recommendations.
610. Dr Skovmand's "search pattern". In the Defendants' closing submissions, counsel for the Defendants' principal argument was that Dr Skovmand had merely used a similar "search pattern" both times. Counsel argued that the recipes for samples 7-20 in the Fence database represented a search pattern for a suitable PE net formulation devised by Dr Skovmand using (i) published information such as the information in the ZeroFly patent application, (ii) Ciba's recommendations and (iii) a grid of concentrations which he thought would yield useful data (i.e. using his "mathematical protocol"). Counsel went on to argue that in devising the early Netprotect recipes, Dr Skovmand had simply done the same thing.
611. I shall deal with this argument so far as it relates to samples 7-20 in the Fence database below. So far as devising the Netprotect recipes is concerned, I do not accept that this is what Dr Skovmand did for the reasons I have already given. He did not start from published information, he did not go back to the Ciba recommendations and he did not devise a set of experiments which were designed to avoid problems with co-linearity.
612. Other points. Three further points were made by the Defendants in their closing submissions in support of the contention that Dr Skovmand had not copied from the Fence database.
613. (1)     It was said that there was no magic in any of the numbers such as [Figure 2] and [Figure 3]/[Figure 15] in that VF had not demonstrated that these particular concentrations were essential to success. It is true that there is no evidence that a deltamethrin of [Figure 2] g/kg rather than [Figure 16] g/kg is essential to success, and on the contrary the evidence suggests that it is not. In my view this is beside the point, however. If there is no particular magic in the figure of [Figure 2] g/kg, it is all the more telling that the same figure re-appears in Netprotect database. The same goes for [Figure 3]/[Figure 15].
614. (2)     It was said that, because the bioassay results obtained in respect of samples 7-20 of the Fence were over-optimistic since the tests were carried out in Salatiga, Dr Skovmand had no reason to copy those recipes when devising Netprotect. The manner in which the evidence about this point emerged suggests to me, however, that this was not something which Dr Skovmand was conscious of, or at least regarded as significant, at the time.
615. In paragraph 51 of his sixth witness statement, Dr Skovmand said that the test laboratory was VF's bioassay lab in Hanoi, and that this produced results on average 30% more optimistic than LIN. In paragraph 24 of his seventh witness statement, Dr Skovmand said that the bioassays were carried out in Salatiga and they had to be corrected by a factor of 0.5. Both paragraphs were corrected on the day before he gave evidence, to refer to Salatiga and a factor of 240% or 2.4.
616. In any event, I consider that the evidence demonstrates that the bioassay data in the Fence database were regarded by Dr Skovmand at the time as providing the basis for making good guesses as to a formulation.
617. (3) It was said that, if Dr Skovmand had intended to copy from the Fence database, the obvious recipe to copy was that of sample 11 since that gave the best wash test results and was used a number of times subsequently. This is true if the wash test data are considered in isolation. It is clear from the history of the development of both Fence and Netprotect, however, that that was not Dr Skovmand's approach. On the contrary, one of the conclusions Dr Skovmand drew from the trials recorded in the Fence database was that [Additive B] should be included for its UV protective effect. The Defendants say that UV protection is not needed for bed nets as opposed to an agricultural product, but again the evidence shows that UV resistance was regarded as a desirable property for bed nets as well. Indeed, Mr Larsen accepted in cross-examination that Netprotect was promoted as being able to be dried in the sun and that one of the reasons for this was that it contained [Additive B].
618. Conclusion. In my judgment, the true position is set out in Dr Skovmand's email dated 11 August 2004. Dr Skovmand based the initial Netprotect recipes on the information recorded in the Fence database.
619. As for the odd numbers, in my judgment what Dr Skovmand said about [Figure 2] g/kg in cross-examination is close to the truth, except that Dr Skovmand did not need to remember the figure because he had the Fence database on his computer. The reason why he used [Figure 2] g/kg of deltamethrin and [Figure 15] g/kg of [Additive A] and [Additive C] was that he had used those concentrations in the recipes recorded in the Fence database. This was also the reason why he used the figures of [Figure 4] and [Figure 6] g/kg.
620. I find that, when devising the initial recipes for Netprotect, Dr Skovmand started from the Fence formulations which had been bioassayed and found to give good results, namely samples 7-16 and in particular samples 8, 9 and 13. (He did not start from samples 17-20 since he had concluded as a result of his later work on PE yarns for VF that the inclusion of [Additive G] was not a good idea.) He expected success with one of these formulations, or a variant of those formulations. That was why Mr Larsen replied on 11 August 2004 "we know that the product we can make will be better than the Olyset net in several regards (mesh, weight, efficiency, regeneration and possibly fire resistance) [emphasis added]". Dr Skovmand believed that the principal problem to be overcome was that of finding a reliable manufacturer, thus reducing the level of deltamethrin losses in production, and that the Defendants were on the track of such a manufacturer. Dr Skovmand expected some loss of deltamethrin in production, but was hopeful that continuing to start with [Figure 2] g/kg would yield an actual concentration close to [Figure 16] g/kg. Furthermore, use of the same numbers would enable Dr Skovmand to compare the results of tests on the new recipes with the results obtained previously.
621. Whether that amounts to an actionable misuse of confidential information is a separate question, which I shall consider below.

Applicable law

622. Although there was a very ill-focussed dispute on the parties' statements of case as to whether the law which governed the relationship between Dr Skovmand and VF was Danish law or French law and as to the relevant content of whatever law was applicable, during the course of the first hearing before me in November 2008, an agreement was reached between the parties that they would proceed on the basis that, whatever the applicable law was, it would be assumed to be the same as English law. The same approach was adopted with regarded to Mr Larsen and Mrs Sig. In the result, the case has been argued entirely on the basis of English law.

Legal analysis

623. The clearest statement of the elements necessary to found an action for breach of confidence remains that of Megarry J in Coco v A.N. Clark (Engineers) Ltd [1969] RPC 41 at 47:

"First, the information itself ... must 'have the necessary quality of confidence about it'. Secondly, that information must have been communicated in circumstances importing an obligation of confidence. Thirdly, there must have been an unauthorised use of the information to the detriment of the party communicating it."

624. This statement of the law has repeatedly been cited with approved at the highest level: see Lord Griffiths in Attorney General v Guardian Newspapers Ltd (No 2) [1990] 1 AC 109 at 268, Lord Nicholls of Birkenhead in Campbell v MGN Ltd [2004] UKHL 22, [2004] AC 457 at [13] and Lord Hoffmann in Douglas v Hello! Ltd (No 3) [2007] UKHL 21, [2008] 1 AC 1 at [111].
625. In the present case, the argument has focussed upon whether Dr Skovmand committed an actionable breach of confidence in developing Netprotect. The Defendants did not dispute that, if he did, Mr Larsen, Mrs Sig, Bestnet, 3T and Intection Ltd were also liable on one basis or another. Nevertheless, it is worth noting that, to the extent that the work was done by Mr Larsen, there is no dispute that he was subject to an express contractual obligation of confidence as set out above.

The necessary quality of confidence

626. The expression "the necessary quality of confidence" was coined by Lord Greene MR in Saltman Engineering Co Ltd v Campbell Engineering Co Ltd (1948) 65 RPC 203 at 215. Lord Greene defined this quality by antithesis: "namely, it must not be something which is public property and public knowledge".
627. As noted above, the Defendants did not in the end dispute that the information in the Fence database possessed the necessary quality of confidence.

Circumstances importing an obligation of confidence

628. I have already held that it was an express term of Dr Skovmand's contract with VF that he would keep information arising out of his work for VF confidential to VF. Nevertheless, I will deal with two alternative contentions advanced by VF. First, VF contend that, even if there was no express term as to confidentiality, it was an implied term of Dr Skovmand's contract that he would keep the information arising out of his work for VF confidential to VF. Secondly, even if there was no contract between Dr Skovmand and VF at all, VF contend that Dr Skovmand was bound by an equitable obligation of confidence.
629. Implied term. Counsel for VF and counsel for the Defendants were agreed that guidance as to the correct approach to deciding whether such a term should be implied could be obtained from the statement of principles regarding the analogous question of ownership of copyright by Lightman J in Ray v Classic FM plc [1998] FSR 622 at 640-643, which was approved by the Court of Appeal in Griggs Group Ltd v Evans [2005] EWCA Civ 11, [2005] FSR 31 at [13]-[14]:

"The general principles governing the respective rights of the contractor and client in the copyright in a work commissioned by the client appear to me to be as follows:

(1)      the contractor is entitled to retain the copyright in default of some express or implied term to the contrary effect;

(2)      the contract itself may expressly provide as to who shall be entitled to the copyright in work produced pursuant to the contract. Thus under a standard form Royal Institute of British Architects ('RIBA') contract between an architect and his client, there is an express provision that the copyright shall remain vested in the architect;

(3)      the mere fact that the contractor has been commissioned is insufficient to entitle the client to the copyright. Where Parliament intended the act of commissioning alone to vest copyright in the Client e.g. in case of unregistered design rights and registered designs, the legislation expressly so provides (see section 215 of the 1988 Act and section 1(a) of the Registered Designs Act 1949 as amended by the 1988 Act). In all other cases the client has to establish the entitlement under some express or implied term of the contract;

(4)      the law governing the implication of terms in a contract has been firmly established (if not earlier) by the decision of the House of Lords in Liverpool City Council v Irwin [1977] AC 239 'Liverpool'). In the words of Lord Bingham MR in Philips Electronique v BSB [1995] EMLR 472 ('Philips') at 481, the essence of much learning on implied terms is distilled in the speech of Lord Simon of Glaisdale on behalf of the majority of the Judicial Committee of the Privy Council in BP Refinery (Westernport) Pty Ltd v The President, Councillors and Ratepayers of the Shire of Hastings (1978) 52 ALJR 20 at 26:

'Their Lordships do not think it necessary to review exhaustively the authorities on the implication of a term in a contract which the parties have not thought fit to express. In their view, for a term to be implied, the following conditions (which may overlap) must be satisfied: (1) it must be reasonable and equitable; (2) it must be necessary to give business efficacy to the contract, so that no term will be implied if the contract is effective without it; (3) it must be so obvious that "it goes without saying"; (4) it must be capable of clear expression; (5) it must not contradict any express term of the contract.'

Lord Bingham added an explanation and warning:

'The courts' usual role in contractual interpretation is, by resolving ambiguities or reconciling apparent inconsistencies, to attribute the true meaning to the language in which the parties themselves have expressed their contract. The implication of contract terms involves a different and altogether more ambitious undertaking: the interpolation of terms to deal with matters for which, ex hypothesi, the parties themselves have made no provision. It is because the implication of terms is so potentially intrusive that the law imposes strict constrains on the exercise of this extraordinary power.

...

The question of whether a term should be implied, and if so what, almost inevitably arises after a crisis has been reached in the performance of the contract. So the court comes to the task of implication with the benefit of hindsight, and it is tempting for the court then to fashion a term which will reflect the merits of the situation as they then appear. Tempting, but wrong.'

(5)      where (as in the present case) it is necessary to imply the grant of some right to fill a lacuna in the contract and the question arises how this lacuna is to be filled, guidance is again to be found in Liverpool. The principle is clearly stated that in deciding which of various alternatives should constitute the contents of the term to be implied, the choice must be that which does not exceed what is necessary in the circumstances (see Lord Wilberforce at p.245 F-G). In short a minimalist approach is called for. An implication may only be made if this is necessary, and then only of what is necessary and no more;

(6)      accordingly if it is necessary to imply some grant of rights in respect of a copyright work, and the need could be satisfied by the grant of a licence or an assignment of the copyright, the implication will be of the grant of a licence only;

(7)      circumstances may exist when the necessity for an assignment of copyright may be established. As Mr Howe has submitted, these circumstances are, however, only likely to arise if the client needs in addition to the right to use the copyright works the right to exclude the contractor from using the work and the ability to enforce the copyright against third parties. Examples of when this situation may arise include: (a) where the purpose in commissioning the work is for the client to multiply and sell copies on the market for which the work was created free from the sale of copies in competition with the client by the contractor or third parties; (b) where the contractor creates a work which is derivative from a pre-existing work of the client, e.g. when a draughtsman is engaged to turn designs of an article in sketch form by the client into formal manufacturing drawings, and the draughtsman could not use the drawings himself without infringing the underlying rights of the client; (c) where the contractor is engaged as part of a team with employees of the client to produce a composite or joint work and he is unable, or cannot have been intended to be able, to exploit for his own benefit the joint work or indeed any distinct contribution of his own created in the course of his engagement: see Nichols Advanced Vehicle Systems Inc v Rees [1979] RPC 127 at 139 and consider Sofia Bogrich v Shape Machines unreported, 4th November 1994 and in particular page 15 of the transcript of the judgment of Aldous J. In each case it is necessary to consider the price paid, the impact on the contractor of assignment of copyright and whether it can sensibly have been intended that the contractor should retain any copyright as a separate item of property;

(8)      if necessity requires only the grant of a licence, the ambit of the licence must be the minimum which is required to secure to the client the entitlement which the parties to the contract must have intended to confer upon him. The amount of the purchase price which the client under the contract has obliged himself to pay may be relevant to the ambit of the licence. Thus in Stovin-Bradford v Volpoint Properties Ltd [1971] 1 Ch 1007, where the client agreed to pay only a nominal fee to his architect for the preparation of plans, he was held to have a licence to use the plans for no purpose beyond the anticipated application for planning permission. By contrast in Blair v Osborne & Tompkins [1971] 21 QB 78, where the client was charged the full RIBA scale fee, his licence was held to extend to using the plans for the building itself. Guidance as to the approach to be adopted is provided in a passage in the judgment of Jacobs J in Beck v. Montana Construction Pty [1964-5] NSWR 229 at 235 cited with approval by Widgery LJ in Blair v Osborne & Tompkins supra at p.87:

'it seems to me that the principle involved is this; that the engagement for reward of a person to produce material of a nature which is capable of being the subject of copyright implies a permission, or consent, or licence in the person giving the engagement to use the material in the manner and for the purpose in which and for which it was contemplated between the parties that it would be used at the time of the engagement.'

(9)      the licence accordingly is to be limited to what is in the joint contemplation of the parties at the date of the contract, and does not extend to enable the client to take advantage of a new unexpected profitable opportunity (consider Meikle v. Maufe [1941] 3 All ER 144)."

630. In my judgment, even if there was no express term, it was an implied term of Dr Skovmand's contract with VF that he would keep the information arising out of the work that VF gave him confidential to VF, including in particular the information recorded in the Fence database. This term satisfies all five of the conditions identified by the Privy Council in BP v Hastings. It is necessary and obvious for the following cumulative reasons.
631. First, Dr Skovmand was acting as a consultant to develop new products for VF to manufacture and sell. He was paid an hourly rate and all his expenses.
632. Secondly, although Dr Skovmand was given considerable freedom by VF, he was subject to VF's direction through Mr Vestergaard Frandsen and the Product Development Committee. His status was close to that of a senior employee even though he was not employed by VF and did not work for VF full time.
633. Thirdly, although Dr Skovmand personally carried out some of the development work, much of it was done by VF's employees or by other contractors such as external laboratories which were paid by VF or both.
634. Fourthly, Dr Skovmand was not VF's only source of advice as to the development of the products, for example, advice was obtained from Ciba and Brabender which again was paid for by VF either directly or indirectly through payment for materials.
635. Fifthly, VF paid for the raw materials used in the development work.
636. Sixthly, most, if not all, of the information recorded in the Fence database was advice paid for by VF or results generated by tests paid for by VF or a combination of the two.
637. Seventhly, VF paid for the compilation of the Fence database itself.
638. Eighthly, Dr Skovmand handed over the Fence database to VF (albeit without his personal comments) and recognised in his goodbye letter that he should only use the information in the copy he retained for a limited purpose.
639. Ninthly, it was clearly understood between Dr Skovmand and VF that inventions made by Dr Skovmand during the course of the work belonged to VF.
640. Tenthly, it cannot have been intended that Dr Skovmand would be able to exploit the information created during the course of his work for VF, and in particular the information in the Fence database, for his own benefit either during or after the termination of his relationship with VF. Nor can it have been intended that Dr Skovmand would license third parties to do so. I think that Dr Skovmand essentially conceded this in the passage of cross-examination I have quoted in paragraph 315 above.
641. Finally, this is not a case in which something akin to an exclusive licence would suffice. VF needed to be in position both to exploit the information without fetter and to enforce their rights against others, including Dr Skovmand himself.
642. Equitable obligation. For the same reasons as I have given in relation to the implied term, even if there was no contract between Dr Skovmand and VF, I consider that he was subject to an equitable obligation to keep the information confidential to VF.

The scope of the obligation of confidence: the Ciba recommendations

643. VF contend that the scope of the obligation of confidence extended to Ciba's recommendations to Dr Skovmand as to which additives to use. VF did not dispute that Ciba was free to give advice about the use of additives to third parties who asked for it, but argued that Ciba's recommendations as to what proportions of what additives to use in the manufacture of PE yarn were confidential to VF in the hands of Dr Skovmand. I accept this, for the following reasons.
644. First, the relationship between VF and Ciba was confidential: as recorded above, there was an express confidentiality agreement in place and Mr De Corte marked his minute of the meeting on 3 September 2002 "Confidential".
645. Secondly, Ciba's initial recommendations were made in response to Dr Skovmand's confidential disclosure to Ciba of the proposed product and the purpose of the additives. Furthermore, Ciba's recommendations were refined in the light of feedback from Dr Skovmand.
646. Thirdly, at least the later recommendations were treated at the time as being confidential to VF. Thus Mr Larsen asked for the recipes to be deleted from the minutes of the meeting of 3 September 2002.
647. Fourthly, as noted above, Dr Skovmand accepted that when he visited Ciba in connection with his work for VF, VF paid for his time and all his expenses. He also accepted that he could not use Ciba's recommendations for the benefit of other clients:

"Q. Did you think when you went to see Ciba on the various

occasions that you did, in 2002, did you think that having

obtained their recommendations, you could then just go off and

use those recommendations for the purposes of some other

client?

A. I do not think I gave that a thought.

Q. But it would be pretty improper, would it not?

A. I think it would be a bit unethical, yes."

The scope of the obligation of confidence: trade secrets

648. It was common ground between counsel for VF and counsel for the Defendants that, whether it was express, implied or equitable in origin, the scope of any obligation of confidence owed by Dr Skovmand to VF did not extend beyond information that amounted to trade secrets of VF's once Dr Skovmand ceased to work for VF. It was agreed that, although Dr Skovmand was a consultant, in this respect his position was analogous to that of an employee. Accordingly, after the termination of the relationship, he was entitled to use for his own benefit or that of third parties information forming part of his own skill, knowledge and experience even if it was learnt during the course of the relationship, but he was not entitled to use any trade secrets. While I am not sure that this approach would apply to all consultants, in the particular circumstances of the present case I am prepared to adopt the agreed position of counsel. That being so, it is not necessary to explore the precise jurisprudential basis for it.
649. In Faccenda Chicken Ltd v Fowler [1987] Ch 177 at 135G-138H Neill LJ delivering the judgment of the Court of Appeal stated the principles to be applied in such cases as follows:

"(1)     Where the parties are, or have been, linked by a contract of employment, the obligations of the employee are to be determined by the contract between him and his employer: cf. Vokes Ltd v Heather (1945) 62 RPC 135, 141.

(2)     In the absence of any express term, the obligations of the employee in respect of the use and disclosure of information are the subject of implied terms.

(3)      While the employee remains in the employment of the employer the obligations are included in the implied term which imposes a duty of good faith or fidelity on the employee. For the purposes of the present appeal it is not necessary to consider the precise limits of this implied term, but it may be noted: (a) that the extent of the duty of good faith will vary according to the nature of the contract (see Vokes Ltd v Heather, 62 R.P.C. 135); (b) that the duty of good faith will be broken if an employee makes or copies a list of the customers of the employer for use after his employment ends or deliberately memorises such a list, even though, except in special circumstances, there is no general restriction on an ex-employee canvassing or doing business with customers of his former employer: see Robb v Green [1895] 2 QB 315 and Wessex Dairies Ltd v Smith [1935] 2 KB 80.

(4)      The implied term which imposes an obligation on the employee as to his conduct after the determination of the employment is more restricted in its scope than that which imposes a general duty of good faith. It is clear that the obligation not to use or disclose information may cover secret processes of manufacture such as chemical formulae (Amber Size and Chemical Co. Ltd v Menzel [1913] 2 Ch 239), or designs or special methods of construction (Reid & Sigrist Ltd v Moss and Mechanism Ltd (1932) 49 RPC 461), and other information which is of a sufficiently high degree of confidentiality as to amount to a trade secret. The obligation does not extend, however, to cover all information which is given to or acquired by the employee while in his employment, and in particular may not cover information which is only 'confidential' in the sense that an unauthorised disclosure of such information to a third party while the employment subsisted would be a clear breach of the duty of good faith. This distinction is clearly set out in the judgment of Cross J in Printers & Finishers Ltd v Holloway [1965] 1 WLR 1, [1965] RPC 239 where he had to consider whether an ex-employee should be restrained by injunction from making use of his recollection of the contents of certain written printing instructions which had been made available to him when he was working in his former employers' flock printing factory. In his judgment, delivered on 29 April 1964 (not reported on this point in [1965] 1 WLR 1), he said [1969] RPC 239, 253:

'In this connection one must bear in mind that not all information which is given to a servant in confidence and which it would be a breach of his duty for him to disclose to another person during his employment is a trade secret which he can be prevented from using for his own advantage after the employment is over, even though he has entered into no express covenant with regard to the matter in hand. For example, the printing instructions were handed to Holloway to be used by him during his employment exclusively for the plaintiffs' benefit. It would have been a breach of duty on his part to divulge any of the contents to a stranger while he was employed, but many of these instructions are not really "trade secrets" at all. Holloway was not, indeed, entitled to take a copy of the instructions away with him; but in so far as the instructions cannot be called "trade secrets" and he carried them in his head, he is entitled to use them for his own benefit or the benefit of any future employer.'

The same distinction is to be found in E. Worsley & Co. Ltd v Cooper [1939] 1 All ER 290 where it was held that the defendant was entitled, after he had ceased to be employed, to make use of his knowledge of the source of the paper supplied to his previous employer. In our view it is quite plain that this knowledge was nevertheless 'confidential' in the sense that it would have been a breach of the duty of good faith for the employee, while the employment subsisted, to have used it for his own purposes or to have disclosed it to a competitor of his employer.

(5)      In order to determine whether any particular item of information falls within the implied term so as to prevent its use or disclosure by an employee after his employment has ceased, it is necessary to consider all the circumstances of the case. We are satisfied that the following matters are among those to which attention must be paid:

(a)      The nature of the employment. Thus employment in a capacity where 'confidential' material is habitually handled may impose a high obligation of confidentiality because the employee can be expected to realise its sensitive nature to a greater extent than if he were employed in a capacity where such material reaches him only occasionally or incidentally.

(b)      The nature of the information itself. In our judgment the information will only be protected if it can properly be classed as a trade secret or as material which, while not properly to be described as a trade secret, is in all the circumstances of such a highly confidential nature as to require the same protection as a trade secret eo nomine. The restrictive covenant cases demonstrate that a covenant will not be upheld on the basis of the status of the information which might be disclosed by the former employee if he is not restrained, unless it can be regarded as a trade secret or the equivalent of a trade secret: see, for example, Herbert Morris Ltd v Saxelby [1916] 1 AC 688, 710 per Lord Parker of Waddington and Littlewoods Organisation Ltd v Harris [1977] 1 WLR 1472, 1484 per Megaw L.J.

We must therefore express our respectful disagreement with the passage in Goulding J's judgment at [1984] ICR 589, 599E, where he suggested that an employer can protect the use of information in his second category, even though it does not include either a trade secret or its equivalent, by means of a restrictive covenant. As Lord Parker of Waddington made clear in Herbert Morris Ltd v Saxelby [1916] 1 AC 688, 709, in a passage to which Mr. Dehn drew our attention, a restrictive covenant will not be enforced unless the protection sought is reasonably necessary to protect a trade secret or to prevent some personal influence over customers being abused in order to entice them away.

In our view the circumstances in which a restrictive covenant would be appropriate and could be successfully invoked emerge very clearly from the words used by Cross J in Printers & Finishers Ltd v Holloway [1965] 1 WLR 1, 6 (in a passage quoted later in his judgment by Goulding J [1984] ICR 589, 601):

'If the managing director is right in thinking that there are features in the plaintiffs' process which can fairly be regarded as trade secrets and which their employees will inevitably carry away with them in their heads, then the proper way for the plaintiffs to protect themselves would be by exacting covenants from their employees restricting their field of activity after they have left their employment, not by asking the court to extend the general equitable doctrine to prevent breaking confidence beyond all reasonable bounds.'

It is clearly impossible to provide a list of matters which will qualify as trade secrets or their equivalent. Secret processes of manufacture provide obvious examples, but innumerable other pieces of information are capable of being trade secrets, though the secrecy of some information may be only short-lived. In addition, the fact that the circulation of certain information is restricted to a limited number of individuals may throw light on the status of the information and its degree of confidentiality.

(c)      Whether the employer impressed on the employee the confidentiality of the information. Thus, though an employer cannot prevent the use or disclosure merely by telling the employee that certain information is confidential, the attitude of the employer towards the information provides evidence which may assist in determining whether or not the information can properly be regarded as a trade secret. It is to be observed that in E. Worsley & Co. Ltd v Cooper [1939] 1 All ER 290, 307D, Morton J attached significance to the fact that no warning had been given to the defendant that 'the source from which the paper came was to be treated as confidential.'

(d)      Whether the relevant information can be easily isolated from other information which the employee is free to use or disclose. In Printers & Finishers Ltd v Holloway [1965] RPC 239, Cross J considered the protection which might be afforded to information which had been memorised by an ex-employee. He put on one side the memorising of a formula or a list of customers or what had been said (obviously in confidence) at a particular meeting, and continued, at p. 256:

'The employee might well not realise that the feature or expedient in question was in fact peculiar to his late employer's process and factory; but even if he did, such knowledge is not readily separable from his general knowledge of the flock printing process and his acquired skill in manipulating a flock printing plant, and I do not think that any man of average intelligence and honesty would think that there was anything improper in his putting his memory of particular features of his late employer's plant at the disposal of his new employer.'

For our part we would not regard the separability of the information in question as being conclusive, but the fact that the alleged 'confidential' information is part of a package and that the remainder of the package is not confidential is likely to throw light on whether the information in question is really a trade secret."

650. When it comes to determining whether information constitutes a trade secret, two further authorities are of assistance. First, in Thomas Marshall Ltd v Guinle [1979] Ch 227 at 248E-H Sir Robert Megarry V-C said:

"If one turns from the authorities and looks at the matter as a question of principle, I think (and I say this very tentatively, because the principle has not been argued out) that four elements may be discerned which may be of some assistance in identifying confidential information or trade secrets which the court will protect. I speak of such information or secrets only in an industrial or trade setting. First, I think that the information must be information the release of which the owner believes would be injurious to him or of advantage to his rivals or others. Second, I think the owner must believe that the information is confidential or secret, i.e., that it is not already in the public domain. It may be that some or all of his rivals already have the information: but as long as the owner believes it to be confidential I think he is entitled to try and protect it. Third, I think that the owner's belief under the two previous heads must be reasonable. Fourth, I think that the information must be judged in the light of the usage and practices of the particular industry or trade concerned. It may be that information which does not satisfy all these requirements may be entitled to protection as confidential information or trade secrets: but I think that any information which does satisfy them must be of a type which is entitled to protection."

651. Secondly, in Lancashire Fires Ltd v S.A. Lyons & Co Ltd [1996] FSR 629 at 668-669 Sir Thomas Bingham MR (as he then was) delivering the judgment of the Court of Appeal said:

"In Faccenda Chicken (at page 137) the Court of Appeal drew attention to some of the matters which must be considered in determining whether any particular item of information falls within the implied term of a contract of employment so as to prevent its use or disclosure by an employee after his employment has ceased. Those matters included: the nature of the employment: the nature of the information itself: the steps (if any) taken by the employer to impress on the employee the confidentiality of the information: and the case or difficulty of isolating the information in question from other information which the employee is free to use or disclose. We have no doubt that these are all very relevant matters to consider. In the ordinary way, the nearer an employee is to the inner counsels of an employer, the more likely he is to gain access to truly confidential information. The nature of the information itself is also important: to be capable of protection, information must be defined with some degree of precision: and an employer will have great difficulty in obtaining protection for his business methods and practices. If an employer impresses the confidentiality of certain information on his employee, that is an indication of the employer's belief that the information is confidential, a fact which is not irrelevant: Thomas Marshall Ltd v Guinle [1979] Ch 227 at 248. But much will depend on the circumstances. These may be such as to show that information is or is being treated as, confidential; and it would be unrealistic to expect a small and informal organisation to adopt the same business disciplines as a larger and more bureaucratic concern. It is plain that if an employer is to succeed in protecting information as confidential, he must succeed in showing that it does not form part of an employee's own stock of knowledge, skill and experience. The distinction between information in Goulding J's class 2 and information in his class 3 may often on the facts be very hard to draw, but ultimately the court must judge whether an ex-employee has illegitimately used the confidential information which forms part of the stock-in-trade of his former employer either for his own benefit or to the detriment of the former employer, or whether he has simply used his own professional expertise, gained in whole or in part during his former employment."

652. Considering the various factors which have been identified in the authorities, the position is as follows.
653. The nature of the work. Dr Skovmand was engaged to develop new products for manufacture and sale by VF. The nature of his duties was such that inventions might reasonably be expected to result from the carrying out of those duties (compare section 39(1)(a) of the Patents Act 1977). Indeed, inventions did result and Dr Skovmand assigned them to VF without demur.
654. The nature of the information itself. The information in the Fence database primarily consists of two sorts of information: (i) recipes for PE yarns and nets and (ii) the results of tests on such yarns and nets. The specific information which VF contends has been misused consists of the recipes and results for samples 7-16, and in particular samples 8, 9 and 13. VF also relies upon the information which can be derived from the results recorded in the database, such as what they reveal about the effect of between [Additive A] and [Additive C] on the migration of deltamethrin. It has frequently been held that secret formulas and secret manufacturing processes are precisely the kind of information which can in appropriate circumstances constitute trade secrets. In my view experimental results and deductions drawn from experimental results, particularly concerning products under development, fall into the same category.
655. The attitude of the employer at the time. I have found that it was expressly agreed between VF and Dr Skovmand that information arising out of his development work would be kept confidential to VF. Furthermore, there is abundant evidence that both VF and Dr Skovmand regarded the information in the Fence database, and in particular the recipes, as confidential information.
656. Steps taken to protect the information, Again, there is abundant evidence both VF and Dr Skovmand took quite extensive steps to protect that information from misuse by third parties. These steps included obtaining secrecy agreements with third parties, limiting the dissemination of information and the use of codes. As I have noted above, Mr Larsen was particularly diligent in discharging his contractual responsibility to protect VF's confidential information. For what it is worth, the Defendants took the same precautions in respect of the Netprotect recipes.
657. The separability of the information. In my view the information in the Fence database was separable from Dr Skovmand's general skill, knowledge and experience including the general skill, knowledge and experience he gained during the course of his work for VF. This is particularly true of the recipes.
658. The commercial value of the information. It is clear that the information was regarded at the time as of commercial value. Moreover, it appears that it is actually of at least some commercial value.
659. The usage and practices of the trade. There is little evidence as to the usage and practices of the trade, but such evidence as there is lends support to the view that such information was regarded as being a trade secret.
660. Conclusion. All of these factors point more or less strongly towards the conclusion that the information contained in the Fence database, and in particular the recipes contained in that database, constituted VF's trade secrets.
661. Counsel for the Defendants submitted that the recipes for samples 7-16 (and samples 17-20) were merely Dr Skovmand's research proposals before a stroke of work had been done and accordingly constituted the product of his own skill, knowledge and experience rather than trade secrets of VF. Ingenious though this argument is, I do not accept it for the following reasons.
662. In the first place, the recipes were based on the Ciba recommendations. I have held that those recommendations were confidential to VF. They were not part of Dr Skovmand's stock of knowledge which he was free to use as he pleased. This is not affected by the fact that Dr Skovmand did not simply implement the Ciba recommendations, but also exercised his own judgment as to the additives and concentrations to be tested.
663. Secondly, Dr Skovmand formulated the recipes with the specific objective of developing suitable PE yarns for VF's proposed new products. This is not affected by the fact that, in devising his "grid" of recipes based on the Ciba recommendations, Dr Skovmand employed an approach to experimental design which I accept in itself did form part of his own skill and experience.
664. Thirdly and most importantly, I consider that the argument is based on a fallacy, which is that the recipes can be divorced from the experimental results. Whatever the position might have been if samples 7-16 had remained mere paper proposals, that is not this case. In fact they were bioassayed, and a number of them gave promising results, including samples 8, 9 and 13. That is why the recipes were of value and why they enabled Dr Skovmand to make good guesses as to a formulation for Netprotect.
665. Fourthly, it is not realistic to view the information relating to samples 7-16 in isolation from the information which Dr Skovmand gained subsequently and which is also recorded in the Fence database. In particular, Dr Skovmand subsequently tested the suggestion made by Ciba that [Additive A] could be dispensed with, and concluded that this was wrong advice. The information relating to samples 7-16, all of which contain [Additive A], has to be viewed in that light. Similarly, in relation to samples 17-20, Dr Skovmand subsequently concluded it was not a good idea to include [Additive G].
666. Counsel for the Defendants also submitted that the recipes constituted information which Dr Skovmand could carry around in his head, and that this supported the conclusion that they were not trade secrets. I do not agree. It is clear from the authorities that this does not matter if the information is separable. Furthermore, as I have already said, I do not accept that Dr Skovmand simply relied on his memory. He had the Fence database on his computer. It is clear from the evidence that his working method was to plan future experiments by analysing the results of past ones recorded in his databases. I have little doubt that that is what he did when devising the initial Netprotect recipes.
667. Accordingly, I conclude that the information contained in the Fence database, and in particular the recipes and bioassay results for samples 7-16, constituted VF's trade secrets. In the words of Cross J in Printers & Finishers at 255:

"… the information in question can fairly be regarded as a separate part of the employee's stock of knowledge which a man of ordinary honesty and intelligence would recognise to be the property of his employer, and not his own to do as likes with … "

668. I would add that it may be that at the time Dr Skovmand thought, or at least proceeded on the assumption, that he was free to use the information in the Fence database for his own purposes. That he did so would be consistent with it being the Defendants' case in these proceedings that he was under no obligation to VF not to do so. Even if that was what Dr Skovmand thought at the time, it does not affect my conclusion. I have found that he was subject to an obligation of confidence. Moreover, as I have pointed out, Dr Skovmand substantially accepted this in cross-examination. Whether the information in the Fence database constituted trade secrets of VF must be objectively assessed. On an objective assessment, I conclude that it did. It follows that Dr Skovmand could not in good conscience use that information for his own purposes or those of the Defendants.

Misuse

669. I turn therefore to the question of whether Dr Skovmand misused VF's trade secrets. In my judgment it follows from my previous conclusions that Dr Skovmand did misuse the information in the Fence database, in particular the information relating to samples 7-16 and more particularly the information relating to samples 8, 9 and 13, by using it to devise the initial Netprotect recipes which were tested in October 2004.
670. Counsel for the Defendants argued that the only information in the Fence database which could be protectable, if any, was the experimental results and that Dr Skovmand had not used that information. As I have said, I think that it is fallacious to divorce the recipes from the results. Furthermore, I consider that Dr Skovmand did use the results, in that he used them to identify which recipes were worth repeating and to decide what variations should be tested.

Remedies

671. VF claim an injunction to restrain continued misuse of their trade secrets. In his submissions counsel for VF made it clear that VF would want such an injunction to extend to preventing the manufacture and sale of the Defendants' current product. In support of this argument he referred me to Ocular Sciences (cited above) at 401-406. Understandably, counsel for the Defendants did not address me on this point. It would have been difficult for him to make any sensible submissions without knowing my findings of fact and my conclusion as to whether there had been an actionable breach of confidence. In these circumstances I have decided to follow the course adopted by Cross J in Printers & Finishers and invite further argument on the question of remedies, and in particular as to whether an injunction should be granted and, if so, in what terms.
672. Accordingly, at this stage I will merely observe that it is not obvious to me that an injunction is an appropriate remedy or, if it is, that it should extend to preventing the manufacture and sale of the Defendants' current product. Although I have concluded that Dr Skovmand misused VF's trade secrets, he did not simply copy any particular recipes. Moreover, the misuse of VF's trade secrets I have found was merely the starting point for a substantial program of further development which resulted in a formulation which is different from any of VF's recipes in a number of respects, and in particular (i) the polymer composition (at least in the case of the sample submitted for WHOPES II evaluation), (ii) the inclusion of [Additive L] and (iii) the inclusion of [Additive M]. In addition, a substantial period of time has elapsed since then.

Conclusion

673. I conclude that the Defendants are liable for breach of confidence, but that the question of what remedies VF are entitled to should be the subject of further argument.

Appendix 1: The Extract DB documents

674. VF contend that the Extract DB documents annexed to Confidential Schedule 2 to the Defence, and in particular Extract DB oct04.xls and Extract DB July05.xls, are forgeries concocted to support an untrue account of the development of Netprotect. Since this issue took up a lot of time at trial, and since the allegation of forgery is a serious one, I must deal with it in some detail. I shall do so as briefly as possible.
675. I have related above how the Extract DB documents came into the case, and what was said by GMR's letter dated 17 May 2007 about them.
676. I have also related the circumstances in which Dr Skovmand wrote his letter dated 12 September 2008. In this letter he said:

"1. I have found the database extracts that were sent, but they are all pass word protected and I have not the pass word anymore, they were sent april 2007. I do not know how useful they are, but they have been revealed in the court in DK, so I think VF can get them from there. I understand that I am asked to produce a full version of the database per July 05

It does no longer exist. … The closest I can get in the non deleted old versions is PE database extract from October 2005. It is however password protected and I have not written down the password. … So, I do not know how relevant it is, but you can have it."

This clearly indicates that the Extract DB documents were made by Dr Skovmand and sent by him (whether to Mr Larsen or to GMR) in April 2007.

677. In his first report dated 24 October 2008 Mr Howe considered the Extract DB documents in some detail. In short, he concluded that Extract DB oct04.xls and Extract DB July05.xls were questionable documents since they were consistent neither with the A tab of Extract Apr05.xls nor with other documentation and evidence. From this point onwards, the provenance of Extract DB oct04.xls and Extract DB July05.xls was a major focus of VF in the litigation. What subsequently transpired needs to be seen in that light. Particularly from the service of VF's skeleton argument on 20 November 2008, the explanations given by Dr Skovmand and Mr Larsen were given in the knowledge that VF were suggesting that these documents had been concocted to support a false account of the history of the development of Netprotect.
678. On 10 and 28 October 2008 FFW wrote to GMR raising queries about certain points, and in particular the names of the Extract DB documents given that Extract DB April05.xls contains samples dating from later than Extract DB April05.xls. On 2 November 2008 GMR replied:

"We have received the following response from Mr Skovmand to the point[s] raised in your letter. Following your numbering:

1. Extract DB October 04 is the first version of the database and only includes data from the test made in that month. Extract April 05 and 06 are the same and both contain data from products made up to January 2006, so the title 05 is an error. Extract July 05 is as it says an extract that includes data up until that time."

While this is unclear in certain respects, what is again clearly indicated is that the person who was responsible for making the Extract DB documents was Dr Skovmand.

679. In paragraph 109 of his seventh witness statement Dr Skovmand responded to Mr Howe's observation in paragraph 347 of his first report that both Extract DB oct 04.xls and Extract DB Apr05.xls appeared to derive from a database dating from July 2005 as follows:

"What IPT have are extracts of which one has been worked on by Torben, maybe both. They were picked from his computer during the search in his house. As explained before (witness statement 6), April 2005 is not from 2005 since it holds data from March 2006, so it most likely was from April 2006 and namely wrongly. The IPT conclusion that data were removed from the June [sic] 2005 version is therefore wrong."

This is very confused, not least because Dr Skovmand seems here to be proceeding on the basis that the Extract DB documents were obtained by VF from one of the searches of Mr Larsen's house, whereas in fact they were produced by the Defendants. Nevertheless, Dr Skovmand suggests that the extracts have been "worked on" by Mr Larsen.

680. In paragraph 110 of the same statement, Dr Skovmand responded to Mr Howe's reference in paragraph 351 of his first report to what Dr Skovmand had said in his letter dated 12 September 2008 (quoted above) as follows:

"In fact, I did not say I have found the database extracts that were sent, I do not have any databases with these names (I have run a search). What I said was I found one database – it could be something totally irrelevant – that I could not open since it was password protected."

The first sentence directly contradicts the first sentence from the 12 September 2008 letter quoted above. The second sentence, as is clear from what follows, refers to the version of the Netprotect database which he said in the same letter he also found at that time (and was disclosed by the Defendants on 21 November 2008 after GMR's IT manager was able to access it).

681. In paragraphs 138 and 139 of the same statement Dr Skovmand said (as subsequently corrected):

"138.      #444. IPT has not taken into account that the extract was in fact from a later date, almost certainly April 2006. Extracts from a database means exactly that, they are not the database. Data can be omitted or simplified, depending on the purpose. For example, when there were no additive information on the first formulations in a database extract sent to Torben, the reason was that I wanted to communicate the problem concerning the high variation in deltamethrin. He probably himself added data to that: the file called 'Extr July 2005', which I do not recognise. … Further, there is something wrong with the bioassay values presented in the table named 'July database' [taken by Mr Howe from Extract DB July05.xls]: they do not at all corresponde to those in the database…

139.      #445. As set out under #444, there is appears to be a typing error in the naming of Extract apr05.xls, which is apparent from its content. It covers up to the start of 2006 and is likely, therefore to be April 2006…."

682. In these paragraphs Dr Skovmand was saying that Extract DB July05.xls was made by Mr Larsen, that he did not recognise it and that it contained different information to the Netprotect database. On the other hand, he was giving an explanation (subsequently deleted) for what appeared in Extract oct04.xls.
683. I have related above the circumstances in which Dr Skovmand's eighth witness statement was made, and quoted from paragraphs 1 and 2 of it. In paragraph 3 he said:

"I did not name these databases and that is why I said in my seventh witness statement that I did not recognise the names. Torben Holm Larsen has told me (and I therefore believe) the following. He named them. The database entitled 'Extract db Apri05.xls' is so named because the data under the tab 'bioassay' relates to the 3 formulations sent to Indonesia for a field trial in April 2005… Other tabs include data added after April 2005, because the database was current at the date I sent the extract to him in about April 2006. … Torben will verify this by signing below."

684. Thus the story being told in this statement is that Dr Skovmand created the Extract DB documents and sent them to Mr Larsen, but Mr Larsen named them.
685. In Ms San Martin's first witness statement, she pointed out that paragraph 2 of Dr Skovmand's eighth witness statement did not appear to be correct for two reasons. First, Extract DB oct04.xls had column headings which relate to additives introduced in July 2005 (e.g. [Additive Q] and [Additive R]), which suggested that Extract DB oct04.xls was extracted from a version of the Netprotect database dating from July 2005 or later. Secondly, the electronic copy of Extract DB July05.xls included a link to a file which appeared to have been created or updated in 2006.
686. As related above, Dr Skovmand's ninth witness statement was made in reply to this. This is too long to quote, but in summary Dr Skovmand said that:

i) He had provided extracts from the Netprotect database to Mr Larsen until November 2005.

ii) When Mr Larsen's apartment was raided for the second time in November 2005, his files were removed and so the Defendants did not have any extracts. Dr Skovmand decided not to send him any more complete databases, but for a short time did send extracts for temporary use only.

iii) When the Defendants prepared Confidential Schedule 2 to the Defence, they did not have any databases to support it and so Mr Larsen asked for extracts for this purpose. Dr Skovmand found a database which was originally stored in February 2006 but updated to about mid March 2006. All three extracts were made from that database. Dr Skovmand did not recall how the three extracts were provided to Mr Larsen, but Mr Larsen had got them from him. Mr Larsen named them.

iv) When it had been pointed out by Ms San Martin that Extract DB oct04.xls and Extract DB July05.xls contained references to information later than the dates on which Dr Skovmand had assumed that they had been made, he had opened all three files. He was then able to see that they had all been extracted from the same file. He had then realised that Mr Larsen would have them from him at the same time, namely when Mr Larsen was producing the timeline in about April 2007.

687. In VF's supplementary skeleton argument for trial dated 9 January 2006, counsel for VF submitted that the account given by Dr Skovmand in his ninth witness statement must be incorrect for at least two reasons. First, because it was plain from a comparison of the Extract DB documents that they could not have been derived from the same source database. Secondly, because the data contained in Extract DB Oct04.xls and Extract DB Apr05.xls differed from that in the version of the Netprotect database disclosed by the Defendants on 28 November 2006, which was said by the Defendants to be complete and up to date to 13 June 2006.
688. Dr Skovmand corrected both the eighth and ninth witness statements before he gave evidence. In the explanatory sheet, these corrections were stated to be "[a]s a result of a further consideration of the history of the extracts, given that VF place importance on the changes made to the extracts once they had been extracted from the database".
689. Dr Skovmand corrected the first sentence of paragraph 2 of the eight statement to say:

"The three extracts are from the databases (or copies of the databases) which were made by me and provided to Bestnet in or about October 2004, July 2005 and April 2006 respectively."

The remainder of the statement was unchanged.

690. The corrections to the ninth statement were quite extensive. In particular Dr Skovmand changed paragraph 9 as follows:

"However, the Defendants did not have any databases. It is my recollection(but not Torben's he tells me) that Torben therefore asked if he could have extracts for this purpose [namely supporting Confidential Schedule 2]. I remember now that I was able to find a database but I do not know now when it was created. Certainly at some time, I had been able to provide Torben with a database (or part of a database) which was originally stored in February 2006… "

691. He changed paragraph 12 as follows:

"I certainly permitted Torben to have use the three extract in about April 2007, but that was certainly not the first time Torben had the information they contained… As for the 2006 database or partial database from which the 3 extracts have been taken, I believe that I would have provided Torben with it about April … 2006. Whether I did provide another copy later (when he was doing the time line) I cannot say. Since it seems he already had a copy, it would seem unlikely."

692. In his tenth witness statement, Mr Larsen said this about the Extract DB documents:

"2.      They were files in my possession, which had not been altered for the sake of these proceedings, except for the fact that they have been given new names to reflect the relevant parts f the Netprotect development history document.

…

4.      Ole did not send me them in connection with these proceedings. I do not remember asking him to send me any extracts for illustrating the timeline in the Defence. In fact, I think it is unlikely that I asked him to send me any extracts for that purpose, because I already had these extracts."

He went to say that Extract DB oct04.xls was a "strange document" which had features he could not explain.

693. During the course of argument on 22 January 2009, counsel for the Defendants stated (with both Dr Skovmand and Mr Larsen sitting behind him):

"There is only one real difference between the

recollections of Mr. Holm Larsen and Mr. Skovmand, and that is

that Mr. Skovmand says he remembers being asked to produce a

database to support the details in the timeline of the

protection of Net Protect. He does not remember when he

actually sent the database. It is common ground between the

two of them that the database is one which would have been

saved in March or April 2006 and that all three extracts came

from that database. That is common ground. It is common

ground that Mr. Larsen would have been sent that in 2006,

about April. The only difference is that Ole Skovmand does

not remember whether he sent the database again in 2007 or not

and Mr. Holm Larsen says he did not."

694. Dr Skovmand was cross-examined at some length on this topic. It is not feasible even to summarise the cross-examination, but I should record certain points.
695. First, Dr Skovmand had considerable difficulty in providing a coherent, still less a plausible, explanation of why his account of the provenance of the documents had repeatedly changed.
696. Secondly, Dr Skovmand said that the first time he had looked at the Extract DB documents in either printed or electronic form was after the service of Ms San Martin's statement. This in itself is quite incredible, given that: (i) he had discussed the documents in his letter dated 12 September 2008; (ii) he had discussed the contents of the documents in his seventh witness statement, for example in the paragraphs quoted above; (iii) the provenance of the documents was one of the issues debated in court in Dr Skovmand's presence on 25 November 2008; (iv) his eighth witness statement was made purely in order to explain the provenance of the documents; and (v) paragraph 3 of his eighth statement discussed the contents of one of the documents.
697. Thirdly, on the first day of his cross-examination Dr Skovmand said that after the service of Ms San Martin's first witness statement, he had opened the electronic files and he had then seen that they were all taken from one database. He testified that he was absolutely sure that all three extracts were made from a database which had been updated until shortly before 23 March 2006. He went on to say that he had supplied this to Mr Larsen on a USB memory stick during a meeting in March 2006. This is simply wrong: examination of the three Extracts shows that they cannot have been taken from one version of the database.
698. Fourthly, on the second day of his cross-examination Dr Skovmand pointed out for the first time that the "properties" data in the electronic copies of Extract DB oct04.xls and Extract DB July05.xls both state that they were created by Mr Larsen on 29 April 2007 whereas the "properties" data in the electronic copy of extract DB Apr05.xls states that it was created by an unidentified author on 8 March 2005. He went on to say that he definitely had created Extract DB Apr05.xls, but Mr Larsen had created Extract DB oct04.xls and Extract DB July05.xls. This, of course, contradicted the evidence he had given the previous day.
699. Fifthly, when pressed for explanations, Dr Skovmand repeatedly took refuge in semantics about what constituted an "extract".
700. After Dr Skovmand had given evidence, Mr Larsen was also cross-examined at some length on this topic. Again, I shall note certain points.
701. First, Mr Larsen's evidence was that not only was paragraph 2 of Dr Skovmand's eighth statement untrue, but also he (Mr Larsen) knew that it was untrue when he read and signed that statement.
702. Secondly, it was also Mr Larsen's evidence that aspects of Dr Skovmand's ninth witness statement were wrong, and that he knew they were wrong when he read that statement prior to it being served.
703. Thirdly, Mr Larsen said that he had had the three Extract DB documents on his computer, and that he had used them to support the timeline in Confidential Schedule 2 to the Defence which he had written, but that he could not remember how the documents had been created or by whom or when.
704. Fourthly, although he had signed the statement of truth on the Defence, Mr Larsen accepted that some of the entries in Confidential Schedule 2, including ones which the Extract DB documents were intended to support, were incorrect.

The discrepancies in the Extract DB documents

705. There are a number of discrepancies between what the Extract DB documents appear to show and what actually happened during the development of Netprotect. I have mentioned some of these points already, but not all of them.
706. Extract DB oct04.xls is suspicious in three respects: (i) it contains entries showing use of alpha-cypermethrin, when in fact no alpha-cypermethrin was used in the October 2004 trials; (ii) it appears to show no use of [Additive A], [Additive B] or [Additive C], when in fact they were used in the October 2004; and (iii) it contains no column heading for [Additive F], but does contain column headings for [Additive M] and [Additive K], when in fact use of [Additive M] was not contemplated at that time. VF contend that the first two represent changes made to support the story in Confidental Schedule 2 to the Defence that the initial trials were of insecticides and PE, and that the third represents a change made to support the claims made in Dr Skovmand's first witness statement and Mr Larsen's third witness statement as to the differences between the additives in Fence and those in Netprotect, which emphasised the absence of [Additive F] and the presence of [REDACTED].
707. The "bioassay" tab of Extract DB Apr05.xls is suspicious in that it purports to show that the [REDACTED] additives and [REDACTED] were used in samples 94-96 in April 2005 for the Indonesian field trial, when in fact the [REDACTED] additives were not tried until July 2005 and [REDACTED] was not tried until March 2006. VF contend that this was done to give the false impression that the Netprotect recipes were more different, sooner, from the Fence recipes than was the case.
708. Extract DB July05.xls is suspicious in that it purports to show that each of samples 97-107 contained [REDACTED] g/kg [REDACTED]. This information has been typed in rather than being looked up from elsewhere as is the case for information about the other additives. In fact, the Defendants do not appear to have used [REDACTED] until sample 108, which appears to date from October or November 2005. Again, VF contend that this is a change made to support the claims made in Dr Skovmand's first statement and Mr Larsen's third statement.
709. Neither Dr Skovmand nor Mr Larsen was able to explain all of these discrepancies. When pressed during cross-examination, Dr Skovmand said several times that only Mr Larsen could explain them. The best Mr Larsen could do was to suggest that the questionable Extract DB documents might have been made for the purpose of planning future trials, but even on this hypothesis he was not able to explain some of the discrepancies. I therefore do not consider this explanation at all plausible. Furthermore, even if it were correct that the documents had been made for the purpose of planning future trials, that would not excuse the Defendants' use of them as purporting to show what had actually been done at particular points in time.

Conclusion

710. If the discrepancies with regard to the "bioassay" tab of Extract DB Apr05.xls and Extract DB July05.xls had stood on their own, I might have concluded that the documents were unreliable as a historical record, but no more. In the case of Extract DB oct04.xls, however, the discrepancies are serious. I cannot see that there is any plausible explanation for them other than that suggested by VF, particularly when it is borne in mind that (i) the Defendants did not intend to disclose the A and Karur 0705 tabs of Extract DB Apr05.xls which first alerted VF to the discrepancies and (ii) the Defendants were applying to strike out the claim and thus avoid having to give disclosure. This conclusion is reinforced by the repeated twists and turns of both Mr Larsen and Dr Skovmand in trying to explain the provenance of these documents.
711. In my judgment the person who is primarily responsible for this is Mr Larsen since it appears likely that he was the person who created Extract DB oct04.xls and Extract DB July05.xls. Even assuming that Dr Skovmand was not responsible for creating the documents, however, he must have known that they were not genuine or least did not show what they purported to show.

Appendix 2: Dr Skovmand's evidence to the Danish court

712. On 2 December 2005 IIC's Danish lawyer submitted a short pleading to the Western Division of the Danish High Court contending that documents relating to IIC's work for Mr Larsen should not be disclosed to VF (i.e. pursuant to the search and seizure order) since this work was on new developments which had nothing to do with VF. In support of the pleading she submitted Exhibit J, a statement by Dr Skovmand on the technical development of the various products. It is clear from this, and confirmed by internal evidence, that this statement was written by Dr Skovmand in November 2005. It is also clear, both from the circumstances and from the fact that it begins "For use in a potential court action…", that the statement was written by Dr Skovmand for the purpose of submission to the Danish court.
713. In this statement Dr Skovmand gave a short account of his work on PermaNet, ZeroFly, Fence and Netprotect. This account is broadly similar to that which he subsequently gave in his first witness statement in these proceedings, but there are certain differences.
714. In this statement Dr Skovmand said:

"I then, in November 2005, made the final recipe for Net Protect in polyethylene nets. It uses components that are also found in Fence and Zero fly, but in different concentrations, and also several other chemicals; on the other hand, other of the chemicals in the two VF products are left out, among them the insecticide chosen by VF….

To summarise:

…

Net protect is made using an insecticide never used in any of VF's products. The insecticide in question has never been tested in VF's polyethylene products."

715. Dr Skovmand's explanation for this in paragraphs 145-146 of his sixth witness statement in these proceedings was as follows:

"145.      … at that time, I was intent on using alfacypermethrin. Samples had been made of each of these two insecticides and, whilst my earlier inclination had been to go with deltamethrin, principally for technical reasons, the fact that alfacypermethrin had done well in recent trials, together with the fact that it was 5 times cheaper than deltamethrin, changed my mind. I did not intend to suggest that we had not made samples with deltamethrin or that deltamethrin had never been the favoured insecticide.

146.      Since that time the decision on which insecticide to use has been reviewed prior to sending samples for approval and Bestnet (the client) preferred to go with deltamethrin after all, in spite of the fact that it is so much more expensive, based on the larger number of tests made with deltamethrin than with alfacypermethrin."

716. In cross-examination Dr Skovmand maintained that, at the time he wrote this statement, it was his "firm conviction" that Netprotect should be made with alpha-cypermethrin and that it was only later in December 2005 that the Defendants committed to using deltamethrin. He relied on his letter to Dr Zaim dated 13 November 2005 as showing that he had told WHO that alpha-cypermethrin might be used as the insecticide. He also said that on 27 December 2005 he had sent his lawyer an updated version of the document in which the relevant paragraphs had been deleted. No communication with his lawyer of that date and no updated version of the document have been disclosed, however. Nor is there any evidence that, if the document was updated, the revised version was sent to the Danish court.
717. I accept that Dr Skovmand told Dr Zaim in the letter dated 13 November 2005 that the insecticide would be either deltamethrin or alpha-cypermethrin. I also accept that, as shown by his email dated 23 November 2005, Dr Skovmand was suggesting to Bestnet at that time that alpha-cypermethrin should be chosen. As he was well aware at that time, however, most of the work that had been done prior to that point had involved deltamethrin, deltamethrin was the insecticide which was used in the July 2005 production run, deltamethrin was the insecticide in the nets which had been submitted for WHOPES I evaluation in August 2005, deltamethrin was the insecticide in the nets that IIC had asked WHO to evaluate (albeit not actually submitted for evaluation) in October 2005 and deltamethrin was the insecticide in the nets which had been commercially launched in October 2005. Dr Skovmand's suggestion that alpha-cypermethrin should be chosen as the insecticide was looking to the future, and it was not a suggestion that Bestnet had accepted. At most, Bestnet was undecided.
718. I therefore consider that Dr Skovmand's statement to the Danish court that the "final recipe" of Netprotect incorporated a different insecticide to that used by VF, i.e. not deltamethrin, was misleading. In my judgment Dr Skovmand must have intended to mislead the Danish court.