Granitigard Pty Ltd v Termicide Pest Control Pty Ltd

The issues on the appeal 7 While Granitgard's notice of appeal to this Court contains some fifteen grounds, in its written submissions in this appeal, its counsel grouped those grounds under four headings, as follows: 1. The representation issue - grounds 1, 2, 11, 12 and 15 2. The size and shape issue - grounds 3, 5, 7, 11, 12 and 15 3. The "testing" issue - grounds 6, 7, 8, 11, 12, 14 and 15 4. The performance issue - grounds 9, 10, 11,12 and 14. 8 A number of points need to be made about these groupings. First, there is some obvious overlap between the grounds relied on for these four issue headings. Secondly, ground 4 of the notice of appeal was not placed within any particular group. I will assume that was deliberate. Thirdly, ground 13, which is also not mentioned above, separately raises a further issue relating to the subsequent costs orders made by the learned trial judge: see [2010] FCA 381. 9 Finally, at the hearing of the appeal, the Court was told that the second, third and fourth issues (above) all broadly involve the question of the falsity of the representations concerned. Thus, those issues essentially go to the performance of Termicide's Termiglass Termite Barrier System: did it provide an effective barrier to subterranean termites as Termicide claimed it did? Viewed in this way, the label Granitgard has given to the fourth issue (above) may create confusion. In reality that issue is quite narrow, in that it is directed to the performance of Termiglass at one location where it was installed, known as the Lagoona Court installation. I will therefore refer to that issue as the Lagoona Court installation issue and deal with all these three sub-issues as part of the broader performance issue. Because of these considerations, I consider the following three broad issues arise for consideration in this appeal: 1. The representation issue - did Termicide make the three representations the learned trial judge found it did not make? 2. The performance issue - were Termicide's claims that the Termiglass Termite Barrier System provided an effective barrier to subterranean termites false? This includes the three sub-issues mentioned above: the size and shape issue; the testing issue; and the Lagoona Court installation issue. 3. The costs issue - did the learned trial judge err in making the indemnity costs order against Granitgard? 10 I will now turn to address these three issues in this order.

The pleaded representations 11 The representation issue raises the question whether Termicide made the three representations the learned trial judge found it did not make. In its second further amended statement of claim, Granitgard alleged (in para 8) that: For the purpose of promoting "termiglass", Termicide has and continues to represent that the installation of termiglass: (a) forms "an impenetrable barrier" to termites; (b) provides a barrier that is "too hard to chew, too heavy to move and too small to crawl through"; (c) will satisfy the requirements of Australian Standard 3660.1-2000 'Termite Management - New building work' (the "Australian Standard"); and (d) provides a suitable physical barrier against subterranean termite entry. (e) termiglass and the termiglass termite barrier system have been tested by the CSIRO and appraised as satisfying the Australian Standard 3660.1 - 2000 'Termite management - New building work' (f) has a life expectancy of 50 years (collectively referred to as the "Representations") (Emphasis added)

The representations in issue 12 The learned trial judge found that Termicide made the representations pleaded in paras 8(a), 8(b) and 8(f) - the former two having been expressly or implicitly admitted by Termicide - but not the representations pleaded in paras 8(c), 8(d) and 8(e). It follows that it is the latter three representations that are in issue in this representation issue. Granitgard provided the following particulars in support of the allegations in 8(c), 8(d) and 8(e): Termicide maintains a web site with the address www.termicide.com.au. That web site records the above Representations. Representations (c) and (d) are contained in a CSIRO Appraisal of termiglass contained on the web site. That appraisal records, inter alia: "In the opinion of CSIRO Appraisals, Termiglass Termite Barrier System will satisfy the requirements of Australian Standard 3660.1-2000 'Termite management - New building work' as a suitable physical barrier against subterranean termite entry by use of a screening system, in conjunction with Quikstrip strip shielding, under the following conditions:" Certain conditions are then set out. Representation (e) is made by including on the web site the CSIRO Appraisal including the words set out in these particulars. Under the heading "Basis of Appraisal", the CSIRO Appraisal records that a trial of the "Termiglass system" for exposure to termites was carried out by Ecospan Consulting Services in or about November 2003. 13 When these particulars are read together with the pleaded representations, it becomes apparent that the representations in para 8(c), 8(d) and 8(e) are solely based upon the contents of the CSIRO Appraisal of Termiglass that Termicide posted on its website: www.termicide.com.au. As the particulars state, the representations in paras 8(c) and 8(d) are taken from the statement in the CSIRO Appraisal document that: In the opinion of CSIRO Appraisals, Termiglass Termite Barrier System [as to 8(c)] will satisfy the requirements of Australian Standard 3660.1-2000 'Termite management - New building work' [as to 8(d)] as a suitable physical barrier against subterranean termite entry by use of a screenings system, in conjunction with Quikstrip strip shielding, under the following conditions: [Paragraph references and emphasis added] 14 However, it will be noted that there is some incongruity between the introductory words to para 8 (emphasised in [11] above), which solely refer to Termiglass, and that part of the CSIRO Appraisal document identified in the particulars (emphasised in [13] above) which refers to the Termiglass Termite Barrier System. This incongruity was identified by Termicide's counsel during oral submissions and it may be relevant to an aspect of the performance issue, however, for present purposes, ie the representation issue, the critical question is whether Termicide adopted the words identified from the CSIRO Appraisal document as its own words when it reproduced the CSIRO Appraisal document on its website. 15 As to para 8(e), the particulars state that the CSIRO Appraisal document includes "the words set out in these particulars". However, nowhere in the particulars is it stated (in terms of the representation pleaded in 8(e)) that: "Termiglass and the Termiglass Termite Barrier System have been tested by the CSIRO". Indeed, the only words that approach that form of representation are the words that follow the statement (above) in the particulars, viz: "Under the heading 'Basis of Appraisal', the CSIRO Appraisal records that a trial of the 'Termiglass system' for exposure to termites was carried out by Ecospan Consulting Services in or about November 2003". 16 Nonetheless, in oral submissions, Granitgard's counsel pointed to the following parts of the CSIRO Appraisal document as supporting this pleaded representation. First, note (ii) under the heading "General Conditions" near the beginning of the document as follows: At the time of the Appraisal testing is being undertaken of the effectiveness of the barrier systems against mastotermes darwiniensis. The results of this testing will be reviewed with the possibility of removing this restriction during the term of validity of this appraisal. Secondly, under the heading "Reports" near the end of the document, as follows: TEST OF THE SYSTEM. Ecospan Consulting Services, 10 Miranda Street, Coulandra 4551. (8th November 2003) This is a trial of the Termiglass system for exposure to coptotermes acinaciformis and mastotermes darwiniensis.. The trial has started and will be monitored during the term of validity of this Technical Assessment. (Errors in original) 17 Thus, for the purposes of this representation issue, so far as it relates to para 8(e), the critical issue is whether Termicide, or more particularly the CSIRO, in its Appraisal document, made a statement to the effect that the Termiglass Termite Barrier System (not Termiglass itself) had been "tested by the CSIRO".

The decision below on the representations in issue 18 In relation to the representations pleaded in paras 8(c) and 8(d), the learned trial judge concluded that Termicide did not merely act as "a postman" (a term taken from the decision of French J in Gardam v George Willis & Co (1988) 82 ALR 415 at 427) in reproducing the CSIRO Appraisal document on its website, but neither did it do anything to adopt the CSIRO Appraisal document as its own. Specifically, he said (at [57]-[58]): The continued reproduction of the CSIRO Appraisal and of the link to that document tells both in favour of the drawing of an inference of adoption and against that conclusion. To borrow from Gardam v George Willis & Co, I accept that Termicide is not just "a postman". A postman has an occupational role that allows the inference readily to be drawn that he or she is disinterested in the mail being conveyed and the contents of that mail. Termicide is clearly not disinterested in that sense. Yet the reproduction of and the link to the CSIRO Appraisal are also consistent with an inference that what Termicide is doing is not adopting the CSIRO Appraisal but rather inviting the reader to read that document and then form his or her own opinion as to the worth of claims that Termicide does itself state on its website. The presence of the "CSIRO Appraisals" logo is also, in my opinion, a neutral factor. It certainly alerts the reader to the fact that this product and system has been appraised by the CSIRO but, objectively, does not, without more, carry with it the necessary implication that the publisher is adopting all or any particular one of the CSIRO's opinions. This is not to deny that a publisher aware of the favourable contents of such a report and of the standing of a body such as the CSIRO is inherently likely to be aware of a strong probability that a reader who cares to look will regard that report as corroborative of statements the publisher has separately made itself on the website about the product and system. There is though no express adoption and the likelihood of awareness just mentioned is not to be equated with a necessary implication flowing from an objective assessment of the website content, reading it as a whole and in context. In the absence of Termicide's own language on its website either directly or indirectly suggesting an adoption, I am not persuaded that, by maintenance or reproduction alone, it has otherwise adopted as its own the representations alleged in para 8(c) and para 8(d) of the 2nd further amended statement of claim. Those representations are representations of the CSIRO, not Termicide. In short, in the face of the pleaded denial that Termicide has made these representations Granitgard has not made out this aspect its pleaded case. 19 As to the representation in para 8(e) about Termiglass and the Termiglass Termite Barrier System being tested by the CSIRO, the learned trial judge concluded that neither Termicide's website, nor the CSIRO Appraisal document contained any such statement. He said (at [61]): Termicide does not state on its website that the CSIRO has tested Termiglass as installed in its barrier system. Nor does the CSIRO in the reproduced or linked CSIRO Appraisal make such a statement. The language which the CSIRO uses as to the tasks it has undertaken is "assessed", "appraised" and "inspected". Further, it is explicitly stated in the CSIRO Appraisal which documents and inspections were used in carrying out the CSIRO Appraisal. The only "test of the system" to which the CSIRO refers is a test stated to be being undertaken by Ecospan Consulting Services. That reference does not constitute a representation that Ecospan Consulting Services is undertaking this test on behalf of the CSIRO. Further, and no less significantly, what is stated in the CSIRO Appraisal under the heading "Test of the System" with reference to Ecospan consulting Services (and a date 3 November 2003) is that: This is a trial of the Termiglass system for exposure to coptotermes acinaciformis and mastotermes darwiniensis. The trial has started and will be monitored during the term of the validity of this technical assessment. This is patently not a representation that the system has been tested but rather that a trial is underway.

The grounds of appeal on the representation issue 20 The grounds of appeal by which Granitgard seeks to challenge these findings are not particularly revealing. Four of them (1, 2, 11 and 12) simply allege error on the part of the learned trial judge without specifying the nature of that error. It is only the fifth one, ground 15, that contains any details of the alleged error. But even that ground fails to mention the CSIRO Appraisal document - which is said to contain the representations - and it is mainly directed to the performance, or falsity, issue. Ground 15 stated: The learned trial judge erred in that he misapprehended the true nature of the evidenced (sic) as to: (a) the expert testing of Termiglass carried out by the Applicant's experts; (b) the compliance by the Respondent with the requirements of the Australian Standard AS3660; (c) the representations appearing on the Respondent's website; and ought to have found that: (d) the representations as pleaded by the Applicant were in fact made by the Respondent; (e) the Termiglass as installed did not have the representative characteristics; (f) there had in fact been a number of breaches of the termiglass barrier system and the application chemical termiticide was a valid explanation for current lack of apparent breaches; (g) the Respondent engaged in misleading or deceptive conduct. 21 Despite these deficiencies in its notice of appeal, it did become apparent from Granitgard's written and oral submissions in this appeal what its complaints were. Before turning to consider those, it is appropriate to set out some of the factual background to explain how the CSIRO Appraisal document came to be on Termicide's website.

Factual background to the CSIRO Appraisal 22 The most convenient way to do that is to reproduce that part of the learned trial judge's decision where he sets out that factual background. None of this historical material is in dispute in this appeal. Excluding certain irrelevant facts and including some information [in square brackets] to further explain some aspects, his Honour said this (at [31]-[41]): 31. Termicide specialises in the provision of termite barrier systems in new residential buildings. It also provides a number of other pest control services, including the provision of chemical treatments. For a number of years, Termicide installed the Granitgard barrier system for builders in South East Queensland, purchasing the requisite components of that system from Granitgard. In 2003 Granitgard changed its business model such that it dealt directly with builders, consigning Termicide to the role of its installation subcontractor. This role was not one acceptable to Mr Paul Jeynes [Termicide's sole director]. To this end and drawing upon his experience with the Granitgard barrier system he set out to develop an alternative barrier system which might then be used by Termicide. He came to select crushed, recycled glass as the candidate material for use within an alternative barrier system. As a candidate that material had the advantages of being cheap and readily accessible. 32. In relation to whether Mr Paul Jeynes' idea might feasibly be translated into a viable termite barrier system Termicide came to take advice from Dr French [a former CSIRO scientist and an expert on termites in the built environment] and a Mr Schaffer. 33. By 2003 Mr Schaffer was in private practice having retired from the CSIRO after having spent a working lifetime employed by that organisation in the field of building materials and systems. For over a quarter of a century, until he retired in 2002, Mr Schaffer had been its technical manager responsible for the evaluation of innovative products in the construction industry. He is a tertiary qualified civil engineer. … 34. In 2003, after discussions with Mr Jeynes and inspection of the Vision Glass manufacturing plant, Mr Schaffer advised Mr Jeynes that he believed that his idea for the use of crushed glass as a termite barrier was viable. He also told him that, in order for the product appraisal process to proceed, he should commission some laboratory testing of the proposed crushed glass material so as to validate its performance as a termite barrier. Following this, Dr French's consultancy [Ecospan Consultancy Services] came to be retained by Termicide. 35. Thereafter, Termicide came to seek from the CSIRO an appraisal in respect of a particle barrier system which comprised graded glass screenings which Mr Jeynes came to call Termiglass, together with termite strip shielding which came to be called "Termiglass Quikstrip Shielding" used in accordance with an Australian Standard AS 3660.1-2000 Termite Management Part 1: New Building Work [referred to thereafter in the decision as the 2000 Standard]. 36. In December 2003 the CSIRO issued the CSIRO Appraisal. In that document the following opinion was expressed: In the opinion of CSIRO Appraisals, Termiglass Termite Barrier System will satisfy the requirements of Australian Standard 3660.1-2000 'Terminate management - New building work' as a suitable physical barrier against subterranean termite entry by use of a screenings system, in conjunction with Quikstrip strip shielding, under the following conditions: … [the conditions set out in the primary decision and do not require mention] … 37. A later version of the CSIRO Appraisal, issued in June 2006, updated Building Code of Australia (BCA) references. Subject to this, that document repeated the opinion, quoted above, which had been expressed in respect of the satisfaction of the 2000 Standard. Each of these CSIRO documents, as tendered, evidences the opinions of the CSIRO expressed therein. 38. … 39. Having received the original CSIRO Appraisal, Termicide began supplying and installing the Termiglass barrier system in South East Queensland in 2004. Since then Termicide has become the largest supplier of particle physical barrier systems in South East Queensland. 40. Prior to the commencement of these proceedings Termicide maintained and it continues to maintain a website, www.termicide.com.au. On the evidence to hand, the content and layout of that website has varied over time. In November 2007, under the heading "Termiglass" the following statement was made on the website: Developed with the environment in mind, Termiglass is crushed glass that makes use of a recyclable product (glass). The glass is crushed to a specific shape and density that is too hard to chew, too heavy to move and too small to crawl through. Used in conjunction with Termicide Strip Shielding, the Termiglass System forms a non-toxic physical barrier that eliminates the possibility of concealed entry by termites. Rated number one on environmental attributes the Termiglass System demonstrates significantly reduced environmental loads in comparison with other related termite barriers in Australia. A link with the CSIRO Appraisals logo opened up the CSIRO Appraisal. 41. By 2008, in both April and September and, I infer, in between, given the symmetry of content the following statement was being made on that website under the heading "Termiglass; The Natural Solution": The Termiglass System is the natural solution to termite protection. The crushed glass forms an impenetrable barrier that rates #1 environmentally. It's (sic) user friendly application and 60 yr durability is why Termiglass is the natural choice. Immediately underneath this statement appeared a link which allowed the downloading in PDF form of the CSIRO Appraisal. In this period it was also possible to go directly on the website to a page which reproduced the content of the CSIRO Appraisal. Further, the statement made in November 2007, quoted above, was still being made. 23 It should be noted that the role or effectiveness of the "Termiglass Quikstrip Shielding" is not in issue in this appeal.

The contentions on the representation issue 24 Dealing first with the representations pleaded in paras 8(c) and 8(d) of its statement of claim, Granitgard's counsel submitted that the learned trial judge erred in concluding (at [57] of the decision - see [18] above) that Termicide had not adopted the CSIRO Appraisal document as its own. He submitted that conclusion was in error for the following reasons: (a) the publication by [Termicide] of the CSIRO appraisal on its website amounted to advertising by [Termicide] that its product (including the glass itself and the system in which it was installed) satisfied the requirements of the relevant Australian Standard and had in fact been tested and appraised as complying with the Standard; (b) [Termicide] had a clear commercial interest as manufacturer and distributor of the Termiglass product, to promote the opinions of the CSIRO as relevant to a prospective user of [its] product, such that a reader of the website would, without considering it necessary to carry out its own tests of the crushed glass product and the system in which it is used, accept that the glass and the system complied with the relevant Australian Standard; (c) [Termicide] itself involved CSIRO in the appraisal of the development of the product, for the purpose of seeking an appraisal; (d) the very fact of the CSIRO's appraisals department having been asked by [Termicide] to appraise the product, and then report upon it, as appeared from the body of the report itself, meant that [Termicide] was at all times intending for the CSIRO's opinion to be promoted as an independent certification of [its] product; (e) [Termicide's] reliance on the CSIRO appraisal as providing reasonable grounds for making the representations; (f) the replication of the text of the CSIRO appraisal on the website itself was such that it: (i) was not immediately apparent that it was all the opinion of the CSIRO; (ii) presented on the website in a different format to the CSIRO's technical appraisal document; and (iii) was described on the website as "technical data"; (g) the opinion is distinguishable from a testimonial, in that the CSIRO's opinion was effectively put forward as a matter of fact; and (h) the trial judge found that the respondent was not a postman. 25 In relation to the representation pleaded in para 8(e) of its statement of claim, in addition to the submissions set out at [16] above, Granitgard's counsel submitted that the learned trial judge was in error in concluding that the words "assessed", "appraised" and "inspected" in the CSIRO Appraisal were not to be equated with the word "tested". He pointed to the various uses that were made of the words "test", "tested" or "testing" in the CSIRO Appraisal (see [16] above). He submitted that the learned trial judge had applied a narrow lawyer's interpretation of these words, as opposed to assessing how the various words would have been understood by the main target group, viz builders wishing to use the Termiglass Termite Barrier System. 26 In response, Termicide's counsel submitted that the learned trial judge had made no error in concluding that Termicide had not adopted the CSIRO Appraisal as its own. Responding to the points made by Granitgard's counsel (at [20] above), he submitted that: (a) The publication of the CSIRO Appraisal was not advertising that Termiglass "had in fact been tested and appraised as complying with the relevant Australian Standard". In any event, [Termicide] did no more than to rely on the "opinion" of the CSIRO; (b) and (d) [Termicide's] commercial interest and the fact that it had applied for the appraisal were factors taken into account by the Trial Judge …; (e) [Termicide's] reliance upon the CSIRO Appraisal (and its recent "revalidation") as providing reasonable grounds for making representations for the purposes of section 51A of the TPA is irrelevant to whether [Termicide], on its website, adopts the representations. Adoption or not must be determined with reference to the language appearing on the website and what that would convey to a reasonable reader. (f) The reproduction of the text of the CSIRO Appraisal on the website (as distinct from the hyperlink) does not detract from the fact that [Termicide] is simply reproducing the CSIRO opinion - see the heading "Technical Opinion" and the introductory words "In the opinion of CSIRO Appraisals …" …. (g) [Termicide] disputes that the CSIRO's "opinion was effectively put forward as a matter of fact". It is clear from an objective reading of the whole of the document that the CSIRO expresses an opinion and provides the basis for the opinion. The representation by the CSIRO (not [Termicide]) is that the opinion is honestly held and that it has proper grounds for the opinion. (h) That [Termicide] was "not a postman" (which is one extreme) does not mean that it makes the representations in the CSIRO Appraisal. 27 As to the question whether Termicide had represented that the CSIRO had tested the Termiglass Termite Barrier System, its counsel submitted that it did not make any such representation. He submitted that the only statement that related to testing in the CSIRO Appraisal document was a statement that Ecospan Consulting Services was conducting a trial. It stated: "a trial has started and will be monitored during the term of the validity of this Technical Assessment". As such, he submitted it was not open to find that either Termicide or the CSIRO had represented that Termiglass or the Termiglass Termite Barrier System had been "tested by the CSIRO". He submitted that, when one read the CSIRO Appraisal document as a whole, it was clear that the CSIRO had drawn a distinction between "testing" and "assessing" or "appraising". Finally, he submitted that an objective reading of the CSIRO Appraisal document did not suggest that the CSIRO had "tested" or purported to test the Termiglass Termite Barrier System, as distinct from "assessed" or "appraised" it.

Consideration of the representation issue - adoption by reproduction 28 Both counsel agreed that the question whether Termicide adopted the CSIRO Appraisal document as its own is a question of fact to be determined on all the relevant circumstances of the case. This generally accords with what the High Court said in Butcher v Lachlan Elder Realty Pty Ltd (2004) 218 CLR 592; [2004] HCA 60 ("Butcher") where the Court emphasised the need to look at the whole of the conduct of the person making the alleged representation to assess whether he or she was merely communicating information provided by another without adopting or endorsing it: see at [39]-[40] per Gleeson CJ, Hayne and Heydon JJ. In his decision in Butcher, McHugh J relevantly described the question as "whether the corporation assumed responsibility for or adopted (or endorsed or used its name in association with) the information so that it would be reasonable for a recipient to rely on the information": see at [115]. 29 Putting aside any advantages a trial judge may have in assessing the credibility of the witnesses or having the benefit of seeing the entirety of the evidence viewed as a whole, this Court is in as good a position as the learned trial judge to draw its own conclusions on a question of fact such as this: see Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 ("Fox v Percy") at [23]-[25] and [66]-[68]. Moreover, when this Court is conducting an appeal by way of rehearing such as this, it is required to conduct its own thorough examination of the record and to correct any errors of law or fact that it considers have occurred: see Seven Network Ltd v News Ltd (2009) 182 FCR 160; [2009] FCAFC 166 at [87]. These considerations apply with particular force in this instance because the learned trial judge's decision on this question of fact did not involve any assessment of the credibility of the witnesses before him and, instead, it was confined to an assessment of the use Termicide made of the CSIRO Appraisal document on its website. 30 Having myself examined all the relevant circumstances going to Termicide's use of the CSIRO Appraisal document on its website, I respectfully disagree with the learned trial judge's conclusion on this question. For the reasons that follow, I consider Termicide did adopt the CSIRO Appraisal document as its own. 31 Termicide actually reproduced the CSIRO Appraisal document on its website in two ways. It reproduced the body of the document there, and it separately provided a means whereby a person could click on an icon on the website and be hyperlinked to the original, or unaltered, Appraisal document. If it had limited the reproduction of the Appraisal document to the latter, I may have come to the same conclusion as the learned trial judge that it was simply making the CSIRO Appraisal document available to be read, without using any words or otherwise doing anything to adopt or endorse it. However, it is the former reproduction of the CSIRO Appraisal document on the website that I consider amounted to an adoption or endorsement of it by Termicide. It did that by, first, removing the CSIRO Appraisal's logo from the original Appraisal document at three places: at the head of the document; at the head of the section towards the end of the document where the history and purpose of the CSIRO Appraisal's project is described; and at the head of the Appraisal's form appearing on the last page of the document. It then took the remaining parts of the document and reproduced them on its website under its own logo "Termiglass 'the natural solution'". Thus, even though the content of the Appraisal document remained essentially the same, I consider it ceased to have the appearance of a CSIRO Appraisal's document and instead became a Termiglass "the natural solution" document. So, by altering and using the Appraisal document in this way, I consider Termicide took the additional steps necessary to adopt or endorse the Appraisal document as its own. 32 For these reasons, I consider that Termicide did make the representations pleaded in para 8(c) and 8(d) of the second further amended statement of claim.

Consideration of the representation issue - no statement about CSIRO testing 33 As to the representation pleaded in para 8(e), I consider the learned trial judge was quite correct (with respect) in concluding that Termicide did not make the representation that the CSIRO had tested Termicide or the Termiglass Termite Barrier System. Before expressing why that is so, it is appropriate to set out some of the relevant parts of the CSIRO Appraisal document. The body of the Appraisal document was divided into three main sections under the headings: "Technical Opinion", "Related Information" and "Appraisal". The "Technical Opinion" section came first. It set out the basic opinion of CSIRO Appraisals as follows: In the opinion of CSIRO Appraisals, Termiglass Termite Barrier System will satisfy the requirements of Australian Standard 3660.1-2000 'Terminate management - New building work' as a suitable physical barrier against subterranean termite entry by use of a screenings system, in conjunction with Quikstrip strip shielding, under the following conditions: (Emphasis added) 34 There followed a number of general and specific conditions. The general conditions were as follows: General Conditions: 1. The glass material is manufactured under license to the Termiglass specifications. The glass material ranges in size between 1.2mm and 3mm, is of consistently sound quality and is shaped and graded to the Termiglass specifications. Note: The Termiglass Termite Barrier System has not been appraised for use North of the Tropic of Capricorn. 2. The Termiglass Termite Barrier System is installed by licensed installers trained and approved by Termicide Pty Ltd. 3. It is installed in accordance with the Termiglass instructions contained in the Termiglass Installation Manual, November 2003. 4. Inspections are recommended, at least once a year, to check that no bridging or breaching of the barrier has taken place. 5. A durable notice in accordance with BCA requirements (Clause B1.4(i)(ii) Volume 1 and part 3.1.3.2 (b), volume 2 is attached to the building which states that the Termiglass barrier is installed. Notes: This notice should mention that the barrier should not be breached by tree roots, non-termite resistant material, or contaminated or disturbed by other means. 6. When used in conjunction with other termite barriers, all treatment is to comply with the requirements of AS3660.1-2000 'Termite management-New building work'. 7. There is no plant matter or building refuse where the Termiglass is to be placed. Notes: (i) As a guide the species mastotermes darwiniensis are believed to be located north of the Tropic of Capricorn. (ii) At the time of the Appraisal testing is being undertaken of the effectiveness of the barrier systems against mastotermes darwiniensis. The results of this testing will be reviewed with the possibility of removing this restriction during the term of validity of this appraisal. 35 The "Appraisal" section was in three parts: a description of Termiglass that had been supplied by Termicide; certain design information; and a part headed "Basis of Appraisal". The "Basis of Appraisal" part began with the following statement: CSIRO Appraisals has assessed the following aspects in undertaking the appraisal: (a) installation procedures (b) physical properties (c) relationship to Standards Australia, AS 3660.1-2000 'Termite management - New building work'. (Emphasis added) 36 It then set out the documents and inspections that were used in making the appraisal. They included the following. 37 Under the heading "Reports": TEST OF THE SYSTEM. Ecospan Consulting Services, 10 Miranda Street, Coulandra 4551. (8th November 2003) This is a trial of the Termiglass system for exposure to coptotermes acinaciformis and mastotermes darwiniensis.. The trial has started and will be monitored during the term of validity of this Technical Assessment. (Errors in original) 38 Under the heading "Inspections": CSIRO Appraisals representatives have inspected installations of the systems and found the level of performance satisfactory. As with the installation of all physical termite barriers, it was noted that a degree of care and skill, and attention to detail are necessary. Supervision would be needed to maintain the precision needed. 39 As the learned trial judge correctly observed (see [19] above), the only thing approaching a "test" of the Termiglass system that is referred to in the CSIRO Appraisal document is the trial that was stated to have been started by Ecospan Consulting Services (see [35] above). While there is also a reference to testing in Note (ii) on the first page of the Appraisal document under the heading "General Conditions" (see [34] above) that refers to "testing … the effectiveness of the barrier systems [sic] against mastotermes darwiniensis". It may also be noted that it is apparent from the reference to mastotermes darwiniensis in both Note (ii) and the "Test of the System" section (see above) that both these parts of the Appraisal document are referring to the same test or trial, viz that being conducted by Ecospan Consulting Services. 40 However, within the terms of the representation pleaded in para 8(e), there is nothing in the terminology used in either of these parts of the CSIRO Appraisal document to suggest that Ecospan Consulting Services was undertaking its trial on behalf of the CSIRO, or that the CSIRO had any involvement in that trial. Furthermore, even if it did, as the learned trial judge correctly pointed out, the CSIRO Appraisal document does not state that Ecospan Consulting Services had actually conducted and concluded any test or trial of the Termiglass system. Rather it states that a trial of the Termiglass system had started and was still in progress. 41 There is also no merit in Granitgard's other submissions on this issue. Note (ii) under the heading "General Conditions" plainly addresses the two earlier statements in those "General Conditions" that make it clear that the Appraisal does not apply to the use of the system north of the Tropic of Capricorn. The reason for this limitation is identified in Note (i) as follows: "As a guide the species mastotermes darwiniensis are believed to be located north of the Tropic of Capricorn." It appeared to be common ground at the hearing of this appeal that mastotermes darwiniensis is a more aggressive and robust species of subterranean termite. Thus, this testing was plainly directed to that express qualification on the use of the system in that region. It was not directed to the effectiveness of Termiglass or the Termiglass Termite Barrier System as a barrier against subterranean termites south of the Tropic of Capricorn, much less that the CSIRO itself was doing that testing. 42 I also do not agree with the submission that the words "assessed", "appraised" and "inspected" would have been understood by the ordinary or reasonable members of the class to whom they were directed - even if that class is restricted to builders, as Granitgard submits - as meaning "tested". In determining the meaning or effect those words had for the specified class, one must have regard to its effect on reasonable members of that class, which "may include the inexperienced as well as the experienced, and the gullible as well as the astute": Parkdale Custom Built Furniture Pty Ltd v Puxu Pty Ltd (1982) 149 CLR 191 ("Puxu") at 199 and Campomar Sociedad, Limitada v Nike International Ltd (1999) 202 CLR 45; [2000] HCA 12 at [103]. Furthermore, the words must be viewed in their context as a whole. In other words, "it would not be right to select some words only and to ignore others which provided the context which gave meaning to the particular words": Puxu at 199. 43 Here, the words in question were used in two quite specific and distinct ways. Insofar as the words "test", "tested" or "testing" are concerned, as I have demonstrated above, they were used to refer to a specific experiment or trial that was in progress at the time the Appraisal document was issued, which experiment or trial was aimed at measuring the effectiveness of the Termiglass system against two species of termites. Insofar as the words "appraised", "assessed" and "inspected" are concerned, as can be seen from the relevant parts of the Appraisal document emphasised above (at [31]-[36]), they were used to refer more generally to the process of evaluation or adjudication of Termiglass and the Termiglass Termite Barrier System to determine whether it complied with the relevant Australian Standard as installed. This latter aspect is important when one is considering the meaning or effect these parts of the Appraisal document have for builders, because there is a constant emphasis throughout that document on the importance of ensuring the system is properly installed. Thus, when one reads the CSIRO Appraisal document as a whole and considers the context in which the words "test", "tested" and "testing" were used on the one hand, viz a pending trial of Termiglass, and the words "assessed", "appraised" and "inspected" were used on the other, viz whether Termiglass will comply with the Australian Standard as installed, I do not consider the reasonable members of the class of builders identified by Granitgard would have understood the former words to have the same meaning or effect as the latter words. 44 For these reasons, I consider that the learned trial judge was correct in his conclusion that Termicide had not made any representation to the effect pleaded in para 8(e) that the CSIRO had tested Termiglass or the Termiglass Termite Barrier System and appraised it as satisfying the Australian Standard.

Introduction 45 As I have already observed above (at [9]), the performance issue broadly goes to whether the representations made by Termicide were false. In turn that means whether Termicide's representations about the performance of Termiglass, or the Termiglass Termite Barrier System, as an effective barrier system against subterranean termites, were false. Given this, it is important to approach this issue by reference to the particular representations Granitgard has established that Termicide made about the performance of Termiglass, or the Termiglass Termite Barrier System. 46 For the purposes of this appeal, the representation pleaded in para 8(f) is not being pursued. Further, I have concluded (at [44] above) that Termicide did not make the representations pleaded in para 8(e). As to the balance, those pleaded in paras 8(a) to 8(d) of Granitgard's second further amended statement of claim were either admitted to have been made (see [12] above), or I have concluded were made (see at [32] above). It follows that these are the particular representations that fall to be considered in relation to this performance issue. While these representations have already been set out above (see at [11]), for the purpose of this exercise they bear repeating. They are: For the purpose of promoting "termiglass", Termicide has and continues to represent that the installation of termiglass: (a) forms "an impenetrable barrier" to termites; (b) provides a barrier that is "too hard to chew, too heavy to move and too small to crawl through"; (c) will satisfy the requirements of Australian Standard 3660.1-2000 'Termite Management - New building work' (the "Australian Standard"); and (d) provides a suitable physical barrier against subterranean termite entry. 47 To attempt to establish that these four representations were false, Granitgard has relied on three lines of attack, viz the second, third and fourth sub-issues, which it has raised under the broader performance issue above: see at [9]. To recap, they are the size and shape issue, the testing issue and the Lagoona Court installation issue. In turn, Granitgard has relied upon three bodies of evidence in support of these sub-issues. They are: Exhibit 24; Exhibit 47 and the Lagoona Court evidence, respectively. In summary: Exhibit 24 was said to show that Termiglass did not satisfy the size and shape requirements of the Australian Standard; Exhibit 47 was said to show that Termicide had not conducted any testing of Termiglass as is required, so it was claimed, by the Australian Standard; and the Lagoona Court evidence was said to show that the Lagoona Court installation had failed and this was claimed to demonstrate that Termiglass did not provide an effective barrier to subterranean termites.

The decision below on the performance of Termiglass 48 Before considering these three lines of attack, it is convenient to set out a summary of the learned trial judge's decision on the various aspects of this performance issue. Before doing so, I should point out that, while his Honour referred to the apposite Australian Standard as the 2000 Standard throughout his decision, for convenience, I will describe it (as I have already above) as the Australian Standard or, where appropriate, the 2000 version of the Australian Standard. 49 It is important, first, to record that Termiglass was manufactured at a factory called Vision Glass and the evidence showed that this factory was owned and operated by a subsidiary of Termicide. Further, while Vision Glass made a number of crushed glass products at its factory, Termiglass was unique because, as is detailed below (see at [50]-[51]), a specified portion (approximately 95%) of the crushed glass in Termicide had to fall into the 1.7mm to 2.4mm size range to meet the requirements of the Australian Standard. It was therefore important to be sure that the material that was said to be Termiglass was indeed the crushed glass product containing the sufficient proportion of this specified size range. This distinguishing characteristic of Termiglass led the learned trial judge to pose, at the outset of his examination of the effect of the representations pleaded, what he considered was a threshold question: whether each witness truly gave evidence with respect to Termiglass?: see at [67]. In turn, that caused his Honour to set out what he considered were the relevant requirements of the Australian Standard (see at [69]-[70]). His Honour described those requirements as follows: 69. To elaborate, Termiglass is supposed, based on advice received from Dr French, to consist of graded glass screenings comprised of particles in size generally between 1.7 mm and 2.4 mm (the so called "medium" size), which is within the 1.2 mm and 3 mm range mentioned in the appraisal. 70. Termiglass is also supposed to have other qualities the nature of which is dictated by what the CSIRO Appraisal incorporates by reference. Thus, the appraisal draws upon the 2000 Standard, para 2.3.3, which adds three performance criteria for particles: (a) termite resistant; (b) graded and shaped so that a sufficient proportion of the glass particles are of a size that cannot be transported by foraging termite species occurring in the region; (c) able to be placed in a manner so that the voids between the particles do not permit the penetration of foraging termite species occurring in the region. Termiglass is also supposed to meet these criteria. 50 Following this, his Honour elaborated on various aspects of the performance criteria in para 2.3.3 of the Australian Standard as they related to Termiglass, as follows: (a) The words "sufficient proportion" in the Australian Standard (see [70(b)] above) meant that Termiglass was supposed to contain in the order of a 95% proportion of "medium" particles, viz 1.7mm to 2.4mm size range: see at [73]. (b) The 95% requirement was based on research in relation to the graded stone particle barrier called granitgard. It was not a rigid figure in the sense that a barrier comprised of a lesser percentage of medium particles would certainly fail: see at [74] and also [58(f)] below. (c) Contrary to Granitgard's submissions, the Australian Standard did not require the glass particles to be of any specific shape, only that they be of a particular size and in a sufficient proportion with particular properties: see at [75]. (d) Apart from a requirement that there not be too many "flaky" particles - a requirement that particularly applied to crushed granite particles which contained mica - the shape of the particles was a distraction: see at [77]. 51 In summary, the learned trial judge found that, to comply with the relevant performance criteria set out in the Australian Standard, Termiglass should contain a 95% proportion, approximately, of crushed glass particles in the 1.7mm to 2.4mm size range. Further, he found that there was no requirement in those criteria that the crushed glass particles had to be of any specific shape. 52 The learned trial judge then turned to consider the evidence about the various samples of Termiglass, or what purported to be Termiglass, produced during the hearing. His Honour observed that much of this evidence was problematic because there was doubt as to whether most of those samples "were indeed Termiglass as supplied to the market and installed by Termicide in its particle barrier system": see at [81]. His Honour first considered what he referred to as the "unofficial samples" at [82]-[96]. He then turned to consider the samples obtained by Granitgard from the Vision Glass factory on two occasions in June 2008 at [98]-[105]. In relation to the "unofficial samples", his Honour was not satisfied that any of the test results provided a reliable indication of the properties of the product known as Termiglass: see at [83] re the Bain sample, [85]-[87] re the French samples, [91] re the Sapsford samples, [92] re the Salvado Drive sample, [93] re the Upper Coomera sample, [94] re the Gargal Court samples, [95] re the Skip Bin sample and [96] re the Springfield sample. 53 In relation to the Vision Glass samples that were obtained by Mr Hunt in June 2008, the learned trial judge was satisfied that two of the four sample bags contained the product known as Termiglass (see at [100]-[101]). Significantly, he was also satisfied that the test results for those two samples showed 94% and 95% of the crushed glass particles within the size range of 1.7 to 2.36mm: see at [102]. 54 The learned trial judge then turned to consider whether the product supplied by Termicide as Termiglass was what it was supposed to be. That required him to consider a number of issues in relation to the manufacturer of the product as supplied, viz the Vision Glass factory: see at [106]. His Honour undertook that examination at [107]-[119]. In the process, he noted a concession by Termicide that "because of the absence of documented controls, quality control at Vision Glass leaves something to be desired": see at [112]. In agreeing with that concession, his Honour made the following observation (at [112]) : I accept that Termicide does undertake batch sampling and that, resultantly, batches of crushed glass from the factory are occasionally rejected. I thought though that here, too, that, apart from retention of samples, evidencing of this process was problematic. I accept that the sampling undertaken did not suggest a systemic inability on the part of Vision Glass to produce medium range crushed glass particles of "sufficient proportion". 55 His Honour then reached the following significant conclusion (at [113]): When one takes [the results of the two samples obtained from the Vision Glass factory in June 2008] in conjunction with Dr French's evidence as to a "sufficient proportion" and, further, in conjunction with the results of testing and analysis conducted or caused to be conducted by Dr Holt in April/May 2008 tests and the known performance of the Termiglass barrier system in tens of thousands of installations over the years since 2004, it is improbable that this possibility has become as yet an actuality. That conclusion, in turn, is based on some conclusions which I have reached in relation to evidence concerning alleged breaches of the barrier system, a preference for Dr French's opinions where they differ from those of Dr Holt and Dr Ewart and also views expressed by Mr Schaffer and Mr Meadows. I detail these conclusions and the reasons for them at some length below. 56 In other words, on the whole of the evidence, his Honour concluded it was more probable than not that the Termiglass Termite Barrier System complied with the Australian Standard for size performance. 57 After rejecting Granitgard's submission about the colour and smell of Termiglass and finding that whatever contaminants present were immaterial, his Honour summarised his conclusions on this aspect (at [120]) as follows: In short then, I find that the product manufactured by Vision Glass and supplied to the market by Termicide as Termiglass for installation into its barrier system more likely than not has been and is what it is supposed to be. The product supplied as Termiglass does conform to the product as described in the CSIRO Appraisal. It is a crushed or "sintered" (which means much the same thing) glass product which is generally (in the order of 95%) between 1.7 mm and 2.4 mm in size. 58 His Honour then turned to consider the expert scientific evidence as to whether the representations Termicide had made about Termiglass as installed were false, misleading or deceptive. His findings in this regard are quite extensive and occupy approximately one-half of his reasons for decision. Without detailing all of them, the following are significant for the purposes of this performance issue: (a) Consistent with his conclusion (at [113]) (see [55] above), his Honour repeatedly accepted the evidence of Dr French, including where that differed with various other witnesses, including Dr Holt. He detailed Dr French's close involvement in the development of both Granitgard and Termiglass for use as a termite barrier system. He was clearly impressed with the qualifications, experience and evidence of Dr French. Indeed, he concluded that: "Dr French might be described as the father, in Australia, of particle barrier systems": see at [271]. (b) While the Australian Standard began in 1995 by referring to graded stone barriers, it was broadened in 2000 to refer to graded particle barriers: see at [158]. His Honour went on to detail how a graded particle barrier may be deemed to satisfy the requirements of the 2000 version of the Australian Standard as follows: Section 7 in the 2000 Standard set out the "deemed to satisfy" requirements for graded stone particulars. Section 7 did not include a specific provision dealing with particle size range. Instead, cl 7.3.1 referred to the performance requirements of cl 2.3.3. In turn and in summary, sub-clauses 2.3.3(a), (b) and (c) specified that (a) the particles be termite resistant, and (b) must be graded and shaped so that they cannot be transported by foraging termites species occurring in the region; and (c) able to be placed in a manner so that the voids between particles do not permit penetration of foraging termite species occurring in the region. Clause 7.4.1 of the 2000 Standard required a minimum depth of 75 mm while cl 7.4.2 (dealing with perimeter applications) required a minimum compacted depth of 75 mm when used in wall cavity applications. (c) In October 2003, Dr French was approached by representatives on behalf of Termicide to review and assess whether crushed glass particles would provide a satisfactory graded particle barrier under the 2000 version of the Australian Standard: see at [160]. Dr French produced a report in February 2004 to that effect. (d) The tests conducted by Dr Ahmed, which were provided to Dr French for the purposes of his February 2004 report, were genuine and accurate: see at [173]-[183]. Since this conclusion is particularly relevant to the Exhibit 47 issue, it is appropriate to set it out in full (at [182]): I see no reason to doubt that Dr Ahmed conducted the tests upon which he reported to Dr French for the purpose of that report. The alternative, in effect, was that Dr Ahmed either alone or in conjunction with Dr French was guilty of scientific fraud. So much was a theme in their cross examination. The evidence does not warrant the making of such a serious finding. It does not withstand objective scrutiny. Neither is it consistent with my observation of each of them in giving their oral evidence. I thought each of them was an honest and learned man. Quite why one or perhaps two of the very few scientists in Australia with relevant expertise would engage in that conduct either to assist or to deceive Termicide, especially knowing that their advice had been sought in the context of a proposal to market a glass particle physical barrier, continues to elude me. (e) His Honour considered the results of the concurrent field tests conducted by Dr French and Dr Holt in April 2008 (see at [191]-[196]) and concluded they showed no penetration of the replicate barriers, ie Termiglass (see at [194]). Further, while there were no significant differences in the results, to the extent that any of the results were considered different, his Honour preferred Dr French's evidence: see at [195]. (f) As to the sieve analysis tests of Termiglass conducted at about the same time, his Honour also accepted Dr French's evidence that about 90% of the sample was within the requisite range and this compared favourably with the 95% figure, accepting that this was not a rigid figure. Specifically, his Honour said (at [197]): The sieve calibration for the analysis conducted by Dr Ahmed and for Dr Holt is slightly different. Nevertheless, as Dr French observed and I find, the overall results are comparable in the sense of being consistent one with the other. Each test discloses that about 90% of the sample Termiglass was within the within the medium range of 1.7 mm - 2.4 mm. I accept as correct Dr French's opinion that the particle sizes of these Termiglass samples are appropriate for Termicide's barrier system. These results, when coupled with the absence of penetration, underscore the wisdom in Dr French's evidence, discussed above, as to 95% not being a rigid figure. They also further support the conclusions which I have reached, as set out above, concerning the consistency of production of Termiglass at the Vision Glass factory. (g) These various test results confirmed that Termiglass as installed met the requirements of the 2000 version of the Australian Standard. His Honour summarised his conclusions in that regard as follows (at [198]): The results of the April/May 2008 tests conducted by Dr Holt confirm the correctness of the opinion which Dr French gave to Termicide in his February 2004 report. Dr French also expressed the opinion, which I accept as the correct or preferable opinion, that these results confirm that Termiglass as installed provides an effective termite physical barrier and meets the performance requirements (cl 2.3.3) of the 2000 Standard. To summarise that opinion: by reference to those performance requirements and where necessary to construe the meaning of those requirements: (a) Termite resistant. Termites cannot eat glass; (b) Graded and shaped so that a sufficient portion of the particles are of a size that cannot be transported by foraging termite species occurring in the region. Termiglass mainly comprises particles within the medium range of 1.7 mm - 2.4 mm which is an effective particle size for coptotermes acinaciformis and similar termite sizes; and (c) Able to be placed in a manner so that the voids between the particles do not permit the penetration of foraging termite species occurring in the region. It is necessary, first, to construe the meaning of this requirement. I do not construe this requirement to mean that there can be no penetration at all by termites into the barrier, but rather that it means that the barrier material does not tallow (sic) penetration through to the surface. Termiglass meets this requirement. (h) These results were also consistent with the fact that more than 27,000 installations of Termiglass had been undertaken since 2004 without any evidence of termite penetration. In this regard, his Honour concluded (at [202]): The results of the April/May 2008 tests are consistent with and support Termicide's evidence that, after in excess of 27,000 installations since 2004, termites have never penetrated through a Termiglass barrier. Those results and that evidence are also consistent with the separately derived expert knowledge and experience of Messrs Schaffer and Meadows in the fields of building materials and pest management. In neither of these fields were Termicide's barrier system known to them as one generally reputed to fail. Yet if it had that general repute based on industry reporting each of these gentlemen was well placed to be aware of that. On this aspect, his Honour accepted Dr French's view and, in turn, rejected Granitgard's claim, that this outcome was explicable by the drought conditions that had subsisted in south-east Queensland in the 2004 to 2008 period: see at [225]. 59 Finally, his Honour assessed the evidence of the numerous witnesses about the alleged "breaching" of the Termiglass barrier installed at the Lagoona Court property (at [240]-[251]) and concluded (at [252]): My conclusion is that the reason why the barrier at Lagoona Court was breached was because of an installation error, particularly by leaving lagging around a pipe, not because the Termiglass in the barrier was penetrated. Termites also gained entry to the property by "bridging" at various locations. The latter though was an indication of a barrier which had to this extent worked by forcing termites into what is known as an "inspection zone" where their presence could be readily detected and treated. Evidence of post inspection surface chemical treatments is not an indication of a breach of a barrier but nothing more than standard practice in the pest management industry when the presence of termites is detected. Such treatments are not effective to neutralise such of those termites as are subterranean.

The Lagoona Court installation and the 27,000 unbreached installations of Termiglass 60 Turning then to Granitgard's three lines of attack on this performance issue (see [47] above), it is convenient to deal with the last first, viz the Lagoona Court installation. This is so because, as distinct from the other two, it does not directly involve the Australian Standard. Moreover, it serves to underscore a significant difficulty that overshadows the whole of Granitgard's case in this appeal. It stems from this finding made by the learned trial judge: since 2004, Termicide had installed Termiglass as a part of its Termiglass Termite Barrier System on more than 27,000 occasions in south-east Queensland without any evidence of subterranean termite penetration: see at [202] set out at [58(h)] above. 61 To attempt to address this significant difficulty, Granitgard's counsel submitted that there was "incontrovertible" evidence of breaching by termites at the Lagoona Court installation and this had not been addressed by the learned trial judge. He also submitted that Termicide's claim to have made 27,000 installations of Termiglass since 2004 without any instances of breaching was not valid because it was not supported by any documents or records, but instead was merely based on estimates made by Termicide's general manager. Granitgard's counsel also claimed that, contrary to the Australian Standard, Termicide's staff had incorrectly characterised what was clearly "breaching" as being "bridging" and, as a result, Termicide was able to make the claim that there had not been any breaching among the 27,000 installations it had undertaken. Finally, Granitgard's counsel submitted that the learned trial judge had not addressed these various aspects of its challenge to Termicide's claims. 62 Termicide's counsel submitted that the "breaching" at the Lagoona Court installation had occurred, as the learned trial judge found, because of an installation error - lagging around a pipe - not as a result of penetration of the Termiglass by termites. He submitted that the learned trial judge had properly and comprehensively dealt with this issue and rejected all of Granitgard's claims in relation to it. As to the challenge to its claim to have undertaken 27,000 installations of Termiglass since 2004 without any termite penetration, Termicide's counsel submitted that there was a clear distinction between "bridging", which essentially meant the termites were visible, and "breaching", which essentially meant that the termites were concealed. In this regard, Termicide relied on the definitions of "bridging" and "breaching" in the Australian Standard. It submitted that its staff was well aware of the difference. Further, Termicide's counsel submitted that its calculations about the 27,000 installations were detailed and comprehensive and were all based on evidence given by its administration manager, Ms Shaw, who produced a spreadsheet detailing the 27,342 installations of Termiglass undertaken up to 18 February 2008, which spreadsheet was based on, and referred to, the books and records of Termicide. 63 I reject all of Granitgard's submissions on this issue. First, it is clear that his Honour carefully and comprehensively considered the evidence given by Messrs Roff, Walsh, Walters, Gabriel and McFarland going to Granitgard's claims about the Lagoona Court installation (at [240]-[251]) and concluded that the termite infestation at that site had occurred as a result of an installation error; specifically, "by leaving lagging around a pipe": see at [252] set out at [59] above. Applying the principles set out in Fox v Percy (see [29] above), I can detect no error in his Honour's assessment of this evidence, nor the reasoning process leading to his ultimate conclusion. Accordingly, Granitgard's claims on this aspect must be rejected. 64 As to Granitgard's challenges to Termicide's 27,000 unbreached installations claim, at the outset, it is worth making the point that, even if Granitgard were correct about the Lagoona Court installation, that would only establish one failure of Termiglass in more than 27,000 installations. Indeed, in oral submissions, Granitgard's counsel could put it no higher than there had "quite possibly" been one failure of Termiglass at the Lagoona Court installation. 65 Nonetheless, contrary to Granitgard's submissions, it is clear from his decision that the learned trial judge also carefully and comprehensively considered Granitgard's challenges to the validity of Termicide's 27,000 unbreached installations claim and rejected all of them. He began by expressly referring to this issue at the outset of his consideration of the Lagoona Court installation (at [240]) where, discussing Granitgard's claims about that installation, he said: Another aspect of its case was that this infestation (and those it sought to evidence with respect to other properties) rendered unreliable the evidence otherwise apparent from Termicide's records of many thousands of installations without subsequent breaching by termites. 66 After reaching his conclusion about the Lagoona Court installation (above), he turned to consider the evidence about breaching at the other properties (at [253]-[255]). In the result, he accepted the evidence of Mr Wallace, corroborated by Messrs Jeynes, Schaffer and Meadows for Termicide, in preference to that of Mr Gabriel for Granitgard, and rejected Granitgard's claims: at [254]. Again, I can detect no error in his Honour's assessment of the evidence nor the reasoning for costs he employed. Furthermore, I cannot detect any misuse of his advantage as the trial judge in assessing the credibility of these witnesses: see Fox v Percy at [21]-[31]. 67 Next, the learned trial judge assessed Granitgard's claims - based on the evidence of Mr Gabriel - that Termicide had a policy of characterising "breaching" as "bridging" thereby "covering up" breaches of the Termicide barrier system. First, his Honour noted the denials of this policy by four members of Termicide's staff: at [255]. He then went to the evidence of Mr Jeynes about Termicide's "no quibbles" policy when dealing with any complaints of breaching and concluded that Mr Gabriel had been confused about the effect of this policy, observing in the process that he was "not a reliable historian". Accordingly, he concluded that Termicide did not employ a policy of covering up any breaches of Termiglass: at [258]. Furthermore, he referred to the history of the dispute that had arisen between Mr Gabriel and Mr Jeynes after Mr Gabriel left his employment with Termicide and concluded his "sense of grievance had affected the reliability of [his] memory": at [265]. Yet again, I can detect no error, or misuse of any advantage, of the kind I have described above. 68 For these reasons, I consider Granitgard's submissions in relation to the termite infestation at the Lagoona Court installation and its challenges to Termicide's 27,000 unbreached installations claim, must all be rejected. It follows that Granitgard has failed in this line of attack in relation to the performance issue. 69 It might be thought that this conclusion that Termicide had undertaken 27,000 installations of Termiglass since 2004 without any evidence of termite penetration is, by itself, fatal to any claim Granitgard may make that Termiglass does not perform as an adequate barrier against subterranean termites. More specifically, in terms of the representations I have accepted that Termicide made, it might be thought that this 27,000 unbreached installations conclusion provides powerful proof of the truth of, at least, Termicide's representations that Termiglass: (a) forms "an impenetrable barrier" to termites; (b) provides a barrier that is "too hard to chew, too heavy to move and too small to crawl through"; and (c) … (d) provides a suitable physical barrier against subterranean termite entry. 70 Granitgard's response to this, as I understand it, is that, if it can show that Termiglass does not satisfy the Australian Standard, it follows that these representations must also be false. This response seems to me to be based on the assumption that, if Termiglass does not satisfy the Australian Standard, then it necessarily follows that it will not provide an effective barrier against termites in the terms of these representations. In my view, that does not follow. There is an obvious marketing advantage in being able to say a product complies with an applicable Australian Standard, particularly where a person is endeavouring to introduce a new product to a market, or needs to comply with some tender condition that requires such compliance. There is also a number of Australian Standards or Codes that are adopted in the Building Code of Australia 1990 with which compliance is required under that regime. However, for present purposes, I do not understand that there is any requirement under that Code, or anywhere else, that made it compulsory for Termicide to comply with Australian Standard 3660.1 or 3660.3. It follows, in my view, that Granitgard's assumption in this regard is false and, therefore, establishing non-compliance with these Australian Standards does not lead to the conclusion that Termiglass does not provide an effective barrier against termites in the terms of those representations. 71 This is further supported by the fact that the representations in paras 8(a) and 8(b) above do not mention Australian Standard 3660.1, or Australian Standard 3660.3, or the CSIRO Appraisal document and they are not sourced from any of those documents. Instead, they appeared elsewhere on Termicide's website. Thus, they are quite independent of both the Australian Standards and the CSIRO Appraisal document for the purposes of the representations they make about the effectiveness of Termiglass as a termite barrier. The representation in para 8(d) is somewhat different. The representation pleaded does not mention the Australian Standards or the CSIRO Appraisal document but, as I have observed above (at [13]), it is sourced from the CSIRO Appraisal document, which does mention Australian Standard 3660.1, but not Australian Standard 3660.3. Nonetheless, I consider the representation as pleaded stands apart from the Australian Standards insofar as it contains statements about the effectiveness of Termiglass as a termite barrier. 72 If all this is so, it is difficult, if not impossible, to see how Granitgard can succeed on this performance issue. Indeed, if the three representations above are true, but Granitgard could show that Termiglass did not satisfy the Australian Standard, and therefore the representation in para 8(c) is false, the best Granitgard could obtain would be a technical victory to that effect, ie a declaration that Termiglass does not satisfy the Australian Standard. But that victory would be pyrrhic because it would be off set by the overwhelming fact of the demonstrated performance of Termiglass (its main competitor's product) in the terms of these three representations. It is these considerations that led me to observe at the outset of this issue (see at [60] above) that there was a significant difficulty overshadowing the whole of Granitgard's case in this appeal. Nonetheless, since they could possibly lead to this technical victory, I will proceed to consider Granitgard's other two lines of attack on this performance issue: Exhibits 24 and 47.

Exhibit 24 - Introduction 73 First, it is worth reiterating how it is that Exhibit 24 relates to this performance issue. As I have already observed above (at [47]), Granitgard seeks to use it to show that Termiglass does not satisfy the size and shape requirements of the Australian Standard. In turn, it claims that this establishes the falsity of the representation pleaded in para 8(c) of the second further amended statement of claim (see [46] above) that Termiglass will satisfy the requirements of the Australian Standard. As I have observed above (at [70]), Granitgard also claims that this will also establish the falsity of the representations pleaded in paras 8(a), 8(b) and 8(d). However, for the reasons I have given there (at [71]-[72]), I consider this claim is based on a false assumption. If that is so, then Exhibit 24 will only go to prove the falsity of this somewhat technical representation and not the other, more telling, representations about Termiglass' performance as an effective barrier against termites. Nonetheless I will proceed to consider it for the reasons indicated above.

Exhibit 24 - what it was and how it came to be tendered 74 Exhibit 24 was a bundle of 22 pages comprising weekly test results undertaken at the Vision Glass factory over a 2½ year period from 9 January 2006 to 19 June 2008. Each page of Exhibit 24 is headed "Sieve analysis test for Termiglass based on 200g sample". As this heading suggests, the exhibit recorded the results of a series of sieve analysis tests of samples of the crushed glass product that purported to be Termiglass showing, among other things, the percentage component of those samples that fell within six sieve size ranges, ranging from .300mm to 1.0mm at the smallest, to 2.8mm to 3.0mm at the largest. 75 Exhibit 24 was tendered by Granitgard's counsel during the cross-examination of Mr Jeynes, Termicide's CEO. The particular part of the cross-examination where the tender occurred proceeded in this way. First, Mr Jeynes was taken to the Quality Plan Summary for the Vision Glass factory. This document was annexed to an affidavit of Mr Murray Lockheed, the manager of the Vision Glass factory. That affidavit had earlier been filed by Termiglass, but had not been tendered in its case. I will return to this aspect below. Mr Jeynes' attention was directed to a number of pages of that Quality Plan, including pp 18 and 19, which detailed the procedure for taking samples of the Termiglass product for sieve analysis. He was then asked about the sieve analysis process and the six sieve sizes involved. He confirmed the six sieve sizes used were: 3.0mm to 2.8mm; 2.8mm to 2.36mm; 2.36mm to 1.7mm; 1.7mm to 1.4mm; 1.4mm to 1.0mm; and 1.0mm to 0.3mm. Following this questioning, Mr Jeynes was taken to a document that he was told had been produced on subpoena from Vision Glass. Immediately thereafter, this document was tendered as Exhibit 24. Mr Jeynes was then questioned about the contents of Exhibit 24. In the course of that questioning, he confirmed that it related to samples taken in accordance with Vision Glass' Quality Plan Summary that he had been taken to earlier. He was then taken, at random, to three of the test results shown in Exhibit 24, and he agreed that they showed that a much larger percentage of the test results fell outside of the 1.7mm to 2.4mm size range - indeed, 14%, 21% and 43% of those three results respectively. As is detailed in the decision of the learned trial judge (see at [50]-[51] above), the requisite prescribed proportions for the purposes of the Australian Standard were that approximately 95% should be within this size range, or 5% outside it. Mr Jeynes was also asked about the "load rejected" note beside the test results for 25 April 2008 and he confirmed that that test result showed that 74% of that sample was outside of the 1.7mm to 2.4mm size range. 76 At the end of this part of his cross-examination, Mr Jeynes offered this explanation for the test results in Exhibit 24, in response to a question from the learned trial judge: HIS HONOUR: The question for you is that, the bundle of test records, the sieve analysis which is now exhibit 24, you have no reason to believe these test results not to be accurate?---I believe what may have happened is, I believe there may be a mix up with the figures and what's actually - I believe they may be using - they may have changed the sieve or something on these documentation. If I get the vast majority within the first, the centre three columns, I know when I go and sieve them, I get the majority in the 1.7 to 2.4 and when I say the majority, I'm talking about 95 per cent, to 98 per cent in that range. So, there may be an error with the sieve sizes that are placed on the top of this form and they may have continued just to use it. (Emphasis added) 77 On the next day, when Granitgard's counsel returned to Exhibit 24, Mr Jeynes gave a somewhat different explanation, as follows: All right. And I suggest to you that the product that in fact you have been installing has the specifications as recorded in the sieve analysis test, that was the subject of the cross-examination yesterday, and which is exhibit 24?---No. Now, I've slept on that sieve analysis. That sieve analysis didn't indicate the tonnage that we were using, four tonne a week. I'd use that - one truck, we'd use that in a day. So I was trying to work out, and the variants were all wrong. That's not an accurate indication of our sieve analysis. When I was trying to go through it, I think what we've been sent, and it may have been an error, is other products that are produced for the company. I slept on that. They are not - and there is an error in this, the volumes just aren't there, this 4.5 tonne a week, one truck, we'd use that in one day. So my volumes weren't there and the grades weren't correct. So I know I was questioned over this, but it's not right. I've lost a lot of sleep over it, and what I'm thinking of is he sent me other forms for other glasses. I know we submitted it as evidence, but they don't accurately reflect Termiglass. There has been a major error with that. (Emphasis added)

The contentions on Exhibit 24 78 Granitgard submitted that Exhibit 24 was the "most critical evidence" on the size and shape of Termiglass and whether or not that product complied with the relevant requirements of the Australian Standard. Granitgard claimed that Exhibit 24 established that on many occasions over the 2½ year period to which it related the sieve analysis results for the Termiglass produced at the Vision Glass factory showed that in excess of 20% of the product did not fall within the requisite 1.7mm to 2.4mm size range. According to Granitgard's counsel, this evidence was critical because: the entire efficacy of particulate barriers as termite barrier systems depended upon the particles being within a particular size, namely 1.7mm to 2.4mm in diameter (apart from shape, which is dealt with further below). The reason for precision in this regard was that particles of that precise size are too large for termites to grasp with their mandibles and move aside to enable a path through them to be created, and when a barrier of a sufficient depth is created the voids are too small for termites to crawl through. 79 Notwithstanding its critical importance, Granitgard submitted that the learned trial judge had failed to mention Exhibit 24 in his decision. It submitted this was despite the fact that Mr Jeynes, Termicide's CEO, was cross-examined on it at some length and the exhibit was expressly relied upon in Granitgard's final submissions. It claimed that, had the learned trial judge taken Exhibit 24 into account, its case that Termiglass failed to comply with the Australian Standard would have been made out. 80 Granitgard also submitted that the learned trial judge failed to take into account the fact that "there is and never has been any testing by [Termicide] for the 'shape' of" [Termiglass]" and it claimed that the learned trial judge had failed to deal with the evidence it had adduced on that issue. Instead, Granitgard claimed that the learned trial judge had wrongly focused upon whether the particles had to be of a particular shape, when Granitgard's case was that there was no testing for shape, let alone any particular shape. 81 For its part, Termicide submitted that Exhibit 24 was not critical evidence in the case and therefore it did not need to be considered by the learned trial judge. As I note below, Termicide's counsel did not maintain this position when pressed during oral submissions. Nonetheless, Termicide submitted that the most critical evidence as to whether Termiglass complied with the Australian Standard for size was the results of the samples of Termiglass that were randomly selected by Granitgard personnel at the Vision Glass factory in February and June 2008. Those results showed that approximately 95% of the Termiglass samples fell into the 1.7 to 2.4mm size range. Termicide submitted all this evidence was considered by the learned trial judge. It submitted Exhibit 24 was not a document prepared by Mr Jeynes and, while it was put to him in cross-examination, he did not unreservedly adopt it. It submitted that Vision Glass also supplied Termicide with other products apart from Termiglass and it pointed to what, it submitted, were "serious anomalies" in the quantities in Exhibit 24, claiming that the weekly volumes set out in it represented about one day's use of Termiglass. Thus, it submitted that Exhibit 24 was not reliable and it was not clear on the evidence that it even related to Termiglass. Finally, even if Exhibit 24 related to Termiglass, Termicide submitted that the relevant size range pleaded by Granitgard in its second further amended statement of claim was 1.2 to 3.0mm and it submitted that the results shown in Exhibit 24 generally fell into that size range.

Consideration of Exhibit 24 and the size and shape issue 82 Two of these submissions can be disposed of briefly. First, I consider the learned trial judge was quite correct, for the reasons explained at [49], [50(c) and 50(d)] and [51] above, in rejecting Granitgard's submissions that the Australian Standard prescribed any performance criteria for shape. It follows that size was the only relevant prescribed performance criterion. It also necessarily follows from this that any evidence Granitgard adduced about testing for shape was irrelevant and that the learned trial judge was quite correct in ignoring that evidence. 83 Secondly, I reject Termicide's submission that the relevant size range for the purposes of complying with the Australian Standard was 1.2mm to 3.0mm. While Termicide is correct in its claim that Granitgard pleaded this as the requisite size range, I consider it is clear, for the reasons given by the learned trial judge (see above at [49] and [50(a) and 50(b)]) above, that the relevant size performance criterion was whether approximately 95% of the crushed glass particles fell into the 1.7mm to 2.4mm range. Thus, the 1.2mm to 3.0mm range pleaded by Granitgard, and seized upon by Termicide, was irrelevant. 84 Turning then to Exhibit 24 itself, Granitgard is quite correct in its claim that the learned trial judge did not expressly mention that exhibit anywhere in his reasons for decision, much less conduct any analysis of it. 85 Plainly enough, Exhibit 24 constituted prima facie evidence that much more than 5% of the Termiglass samples tested fell outside the 1.7mm to 2.4mm size range. Indeed, most of the test results did not come remotely close to showing that 95% of the Termiglass samples fell inside the 1.7mm to 2.4mm size range. The test results ranged between 82% at best on 27 May 2008 and, putting aside the load that was rejected on 25 April 2008, 46% at worst on 5 June 2008. On its face, Exhibit 24 therefore contained important evidence in Granitgard's case on the size performance issue. 86 That being so, I consider it was incumbent upon the learned trial judge to expressly deal with Exhibit 24 in his reasons for decision. Indeed, when he was pressed on this point in oral submissions, Termicide's counsel did not submit otherwise. I therefore consider the learned trial judge fell into error in failing to deal with that exhibit. 87 The combined effect of this conclusion and the fact this is an appeal by way of rehearing, is that now it falls to this Court to undertake that task: see the Seven Network decision referred to at [29] above. I do so hereunder. 88 The first thing to be observed is that, despite Exhibit 24 being squarely raised in cross-examination with Mr Jeynes, Termicide did not call anyone from Vision Glass to explain the anomalies raised by it. As I have already noted above, Termicide had filed an affidavit by Mr Lockheed, the manager of the Vision Glass factory, but it eventually elected not to tender that affidavit in its case. As a consequence, Granitgard submitted that a Jones v Dunkel inference should be drawn against Termicide to the effect that Mr Lockheed's evidence would not have assisted Termicide on this issue. In response to this submission, Termicide submitted that Granitgard did not ask the learned trial judge to draw such an inference. Whether or not that is so, since this Court is now required to assess Exhibit 24 itself, there is no reason why it cannot draw such an inference for itself. 89 In cross-examination, Mr Jeynes offered two explanations for the anomalies shown in Exhibit 24: an error in the sieve sizes (see [76] above), or that some other product was tested instead of Termiglass (see at [77] above). In the absence of the readily available evidence from Mr Lockheed, I consider that a Jones v Dunkel inference should be drawn to the effect that Mr Lockheed's evidence would not have assisted Termicide's case on this particular aspect. Furthermore, in the somewhat unique circumstances I have mentioned above, I consider it should also be inferred that Mr Lockheed's evidence would not have provided support for either, or both, of Mr Jeynes' explanations for the results shown in Exhibit 24: see Jones v Dunkel (1959) 101 CLR 298 at 308, 312 and 320-321. In turn, this means that Termicide's challenges to Exhibit 24 that rely upon Mr Jeynes' explanations must be rejected (see [81] above). However, that does not mean that Granitgard has thereby established the affirmative position, viz that Exhibit 24 shows that Termiglass does not satisfy the size requirements of the Australian Standard. It may assist in drawing that conclusion, if the evidence is evenly balanced on the point, but it is first necessary to consider all the evidence to assess whether that is so: see Sagacious Legal Pty Ltd v Wesfarmers General Insurance Ltd [2011] FCAFC 53 at [79]. 90 Essentially, that means it is necessary to consider Termicide's claims that the objective evidence showed that the figures in Exhibit 24 were wrong. According to Termicide's counsel, this objective evidence came from the results of the tests conducted on the randomly selected samples of Termiglass taken at the Vision Glass factory in February and June 2008. As to the former date, the evidence shows that Granitgard's solicitor, Ms O'Neill, took random samples of Termiglass at the Vision Glass factory in February 2008. She then sent three bags of those samples to Dr Holt for analysis. Dr Holt's analysis of these three bags showed that 95%, 93% and 94% of the contents fell within the 1.7mm to 2.4mm size range. Since those samples were taken at random by Granitgard's solicitors and analysed by its expert witness, Dr Holt, they do provide what may be fairly described as objective evidence of the fact that the Termiglass produced at the Vision Glass factory fell within the requisite size range for compliance with the Australian Standard. 91 However, by itself, that does not necessarily show the figures in Exhibit 24 were wrong. That may follow if the random samples were taken at, or about, the same time as the results shown in Exhibit 24. As it turns out, they were. The closest set of weekly tests shown on Exhibit 24 were those undertaken on 7 and 21 February 2008. They showed the following results: 7 February 2008 3.0mm 2.8mm 2.36mm 1.7mm 1.4mm 1.0mm to to to to to to 2.8 mm 2.36 mm 1.7mm 1.4mm 1.0mm 0.30mm Sample 1 0% 42% 112% 40% 6% 0% Sample 2 0% 40% 110% 42% 8% 0% Sample 3 0% 44% 110% 38% 8% 0%

21 February 2008 3.0mm to 2.8mm to 2.36 mm 2.36mm to 1.7mm 1.7mm to 1.4mm 1.4mm to 1.0mm 1.0mm to 0.30mm 2.8 mm Sample 1 0% 38% 116% 36% 10% 0% Sample 2 0% 42% 114% 38% 6% 0% Sample 3 0% 40% 114% 38% 8% 0%

92 The percentage figures shown above obviously require some explanation, because in this context a 112% result is nonsensical! However, Mr Jeynes accepted in cross-examination that since each is a 200g sample, the correct percentage figure is obtained by dividing each result by 2. Thus, for example, 110 = 55%, 112 = 56%, and so on. Applying these calculations, Exhibit 24 shows that between 55% and 58% of the sieve analysis samples on these two dates fell within the requisite 1.7mm to 2.36mm size range. It may also be noted that the requisite size range under the Australian Standard is actually 2.4mm, not 2.36mm, as shown above. However, this 0.04mm difference is most unlikely to explain the enormous deviation in the results from the required proportion of approximately 95%. Thus, when the two sets of test results - Dr Holt's and Exhibit 24 - are compared, they are so wildly different that one of them must be wrong. 93 But even this comparison does not necessarily establish that it is Exhibit 24 that is wrong. It is possible that the random samples selected by Ms O'Neill in February 2008 and/or Dr Holt's test results were anomalous and Exhibit 24 accurately reflects the true position in relation to Termiglass as at that date. However, there are three other matters that lead me to conclude that it is more likely than not that it is Exhibit 24 that is wrong. 94 The first matter is the tests of the other random samples of Termiglass taken from the Vision Glass factory in 2008. Those samples were taken by Mr Hunt in June 2008. The learned trial judge concluded that two of those samples were, indeed, Termiglass and, when tested, they showed 94% and 95% of each sample, respectively, was within the 1.7mm to 2.4mm size range: see at [102]. I do not understand this finding to be challenged by Granitgard in this appeal. These results also tend to confirm Dr Holt's results of the February 2008 random samples mentioned above. Moreover, like Dr Holt's results, they stand in stark contrast to the test results shown in Exhibit 24 at about the same time. The test results shown in Exhibit 24 for June 2008 are as follows: Date Delivery Docket Number Quantity In Sample Number 3mm 2.8mm 2.36mm 1.7mm 1.4mm 1.0mm to to to to to to 2.8mm 2.36 mm 1.7mm 1.4mm 1.0mm 300mm 2.6.08 1580 36.000 Ton 1 0 % 8 % 160 % 30 % 2 % 0 % 2 % % % % % % 3 % % % % % % 4.6.08 1583 36.000 Ton 1 0 % 6 % 154 % 36 % 4 % 0 % 2 % % % % % % 3 % % % % % % 5.6.08 34022 36.620 Ton 1 14 % 50 % 92 % 38 % 6 % 0 % 2 % % % % % % 3 % % % % % % 14.6.08 34031 35.180 Ton 1 0 % 4 % 142 % 46 % 8 % 0 % 2 0 % 1 % 140 % 48 % 8 % 0 % 3 0 % 0 % 114 % 68 % 18 % % 19.6.08 35590 36.180 Ton 1 0 % 2 % 158 % 34 % 6 % 0 % 2 % % % % % % 3 % % % % % %

95 Applying the same calculations as are referred to above (at [92]), these results show that between 80%, on 2 June 2008, and as little as 46%, on 5 June 2008, fell within the requisite 1.7mm to 2.36mm size range. 96 The second matter flows from the fact that a sieve analysis test is but one of the means of testing Termiglass for compliance with the Australian Standard. Another is to conduct a field test of Termiglass for termite penetration. Granitgard said as much in its written submissions (see [103] below) and Termicide's counsel claimed in oral submissions, without demur from Granitgard's counsel, that this was a more rigorous test than a sieve analysis test. As it happens, such field tests were conducted by Dr French and Dr Holt, at about the same time, in 2008. Significantly, the results of both these field tests separately established that the Termiglass tested was not penetrated by termites. The learned trial judge accepted the two sets of results as consistent and accurate, but insofar as there was any significant difference between them, he accepted Dr French's results in preference to those of Dr Holt: see at [194]-[195] summarised at [58(a)] above]. Again, I do not understand this conclusion to be challenged by Granitgard in this appeal. It is also worth recording, as the learned trial judge pointed out, that these results confirm the opinion about Termiglass that Dr French gave in his February 2004 report: see at [198] set out at [58(g)] above. It might also be noted that there was another set of sieve analysis tests undertaken concurrently, in about April 2008, by Dr Ahmed (on behalf of Dr French) and Dr Holt: see at [197] set out at [58(f)] above. While his Honour's conclusions about the results of these tests was criticised by Granitgard's counsel in oral submissions, I consider it is of some significance that the results are broadly consistent with all the other results mentioned above. 97 Finally, there is the "significant difficulty" overshadowing the whole of Granitgard's case in this appeal that I mentioned at the outset of my consideration of this performance issue (see at [60] above), viz the learned trial judge's conclusion in favour of Termicide's claim that more than 27,000 installations of Termiglass had been undertaken in south-east Queensland since 2004 without any evidence of termite penetration. In a practical sense, this provided 27,000 tests of the effectiveness of Termiglass as a termite barrier in actual use over a four to five year period in what is obviously a highly competitive commercial marketplace.

Conclusion on Exhibit 24 98 Taking all these matters into account, but ignoring Mr Jeynes' explanations for it, I consider the only reasonable conclusion, on all the evidence, is that Exhibit 24 is, indeed, wrong. For these reasons, I reject Granitgard's contentions that Exhibit 24 provided evidence that Termiglass failed to comply with the requisite size performance prescriptions for the purposes of the Australian Standard. It follows that Granitgard has failed to establish, by reference to Exhibit 24, that the representation pleaded in para 8(c) is false. That leaves Exhibit 47.

Exhibit 47 - Introduction 99 As with Exhibit 24, Granitgard has relied on Exhibit 47 in this appeal to attempt to show that Termiglass does not satisfy the Australian Standard. The same observations as I have made above (at [73]) about Granitgard's use of Exhibit 24 apply equally to its use of Exhibit 47: even if Granitgard is successful, that success will not significantly advance its fundamental attack on the effectiveness of Termiglass as a termite barrier. However, Granitgard's approach to Exhibit 47 is quite different to its approach to Exhibit 24. Whereas it sought to use the test results shown in Exhibit 24 to directly attack the effectiveness or performance of Termiglass, with Exhibit 47 it seeks to attack the tests result shown in that exhibit to show they are invalid so that it can then claim that Termiglass has not been tested as satisfying the Australian Standard 3660.1 and it was, therefore, false for Termicide to represent that it did satisfy that Standard. As a critical part of this approach, Granitgard claims that the requirement to test Termiglass arises from a combination of AS 3660.1 and AS 3660.3. Granitgard has referred to this sub-issue (of the broader performance issue) as the "testing" issue. I will turn to consider these claims, beginning with a fundamental question that was raised in oral submissions on this appeal: was this testing issue properly pleaded and pursued in this form at the trial of these proceedings? Before I do that, I will briefly describe what Exhibit 47 is.

Exhibit 47 - what it is 100 Exhibit 47 was a two-page document entitled "Laboratory Bioassay Termite Test Inspection Sheet". These termite tests were undertaken in December 2003 and January 2004 by Dr Ahmed, for Ecospan Consulting Services. In fact, they are the trial that is mentioned in the CSIRO Appraisal document: see at [37] above. They were undertaken in conjunction with Dr French, for the purposes of Dr French's February 2004 report. They involved field tests for termite penetration of the same kind as those I have mentioned in [96] above. The details of these tests are set out at the head of each page of Exhibit 47. They show that the material tested was the Termiglass physical barrier. They also show that: Experiment set: 18.12.03. Inspection period 15.1.04. Termite species: see C. acinaciformis (10g/unit) Testing unit: 1 litre glass jar. 101 In simple terms, these tests were directed to establishing the extent to which, if at all, termites of the species shown, viz coptotermes acinaciformis, would tunnel through, or penetrate, a Termiglass barrier, placed in glass jars, to gain access to a food source placed on the other side of the Termiglass barrier. The results were shown for various size ranges of the Termiglass particles as follows: fine > 1.0mm; medium 1.7mm-2.4mm; coarse > 2.4mm; and, finally, 5% mixed medium. The learned trial judge found that these test results showed that there was no tunnelling through, or penetration of, the Termiglass particles in the medium, 1.7mm to 2.4mm, size range: see at [170].

Was this testing issue a part of Granitgard's case at trial? 102 In my conclusion on the representation issue (at [44] above), I found that the representation about testing pleaded in para 8(e) of the second further amended statement of claim (set out at [11] above) was not made by Termicide. In oral submissions, Granitgard's counsel conceded that, if the representation in para 8(e) was not established, Granitgard had not expressly pleaded any other representation about testing. A cursory examination of the balance of the representations pleaded in paras 8(a) to 8(d) of the second further amended statement of claim (see at [44] above) will reveal that this concession was well made. Thus, the testing issue Granitgard wishes to pursue on this aspect of its appeal was not pleaded in its second further amended statement of claim. 103 Ordinarily, this would prevent Granitgard now being able to raise this issue on appeal: see, for example, Coulton v Holcombe (1986) 162 CLR 1 at 7 and SZKCQ v Minister for Immigration and Citizenship [2009] FCA 578 at [7]. However, Granitgard claimed that, notwithstanding this testing issue was not expressly pleaded, it was implicitly pleaded in the representations it had pleaded and, moreover, this testing issue was run as a part of its case at trial. In relation to its pleading, Granitgard submitted that this representation about testing is implicit in the representation it pleaded in para 8(c) of the second further amended statement of claim. This is so because, so it submitted, to satisfy the requirements of Australian Standard 3660.1-2000, clause 1.3 of that standard required the Termiglass barrier system to be tested, as a new system, under Australian Standard 3660.3. In its written submissions, Granitgard identified four different ways in which Australian Standard 3660.3 called for testing, as follows: (a) laboratory assessment using jars with the barrier material in them; (b) field assessments; (c) testing of the complete system; and (d) use of particular species of termites for the testing. 104 For its part, Termicide's counsel responded that, in one respect, testing was always an issue at the trial because it had been expressly raised as a representation under para 8(e) of Granitgard's second further amended statement of claim. However, it submitted that, absent that representation, testing in any other sense was not raised as an issue and it strongly rejected any suggestion that compliance with any testing requirement of Australian Standard 3660.3 was ever raised as an issue in the way now suggested by Granitgard. 105 In my view, an appeal court should be extremely reluctant to entertain a claim by one party to an appeal that an issue that it accepts was not pleaded is still alive on appeal because that is how that party conducted its case at the trial of the proceedings. An appeal court should only do so in the clearest of cases or, perhaps, where there is a consensus between the parties that this is what happened at trial. To do otherwise would significantly undermine the main purpose of pleadings: to define the issues, to give notice to the other party of the case that it has to meet and to limit the trial to evidence relevant to defined issues: see Dare v Pulham (1982) 148 CLR 658 at 664, Multigroup Distribution Services Pty Ltd v TNT Australia Pty Ltd (1996) ATPR 41-522 at 42,679-42,680 and Council for the City of the Gold Coast v Pioneer Concrete (Qld) Pty Ltd (1998) 157 ALR 135 at 146-149. In the process, it might have the potential to do a grave injustice to the other party by allowing an issue to be agitated, of which it had no notice, or a proper opportunity to address. It would also allow, perhaps encourage, a self-serving idiosyncratic redefinition of the issues after defeat at the trial in the hope of finding another path to victory at the appeal level. That could not serve to ensure finality in litigation, or the quick, inexpensive and efficient disposal of it, as required by Pt VB of the Federal Court of Australia Act 1976 (Cth). And it would mean that, instead of a narrower and more defined task of correcting error in the trial judge's decision, the appeal court would have to conduct a free-ranging inquiry into the conduct of the trial, looking at the possible purposes of submissions made, questions asked and tactics employed, to try to determine what issue was being pursued. This is a task that is likely to be steeped in ambiguity and subjectivity. 106 The difficulty and complexity of this task is underscored, in this case, where the trial lasted 19 sitting days; 2,023 pages of transcript were involved; and some 97 exhibits were tendered. Given that a testing issue was always a live issue under para 8(e) of the second further amended statement of claim and no other testing issue was pleaded, this could not be assessed as one of those clear cases where a party has unmistakably raised an unpleaded issue at trial and conducted its case accordingly. Further, given the length of the trial and the volume of the evidence, there is no warrant for an inquiry being conducted by this Court to ascertain whether the issue was in fact raised at trial at all. All the more so where the respondent, Termicide, is quite emphatic that a testing issue in the form now proffered by Granitgard was not run by it at trial. It might also be noted that Granitgard was given the opportunity to amend its pleadings shortly before the original trial dates allocated for this matter in June 2008 and that resulted in those trial dates being vacated: see the reasons at [17]. Despite this, Granitgard was apparently still unable to properly plead its case. In his decision, the learned trial judge was quite critical of Granitgard's failure to follow proper pleading practice, eg pleading conclusions based on material facts it had not pleaded: see at [14]-[16]. Furthermore, Granitgard was allowed to reopen its case some months after the trial concluded, to call further evidence going to the Lagoona Court installation: see the reasons at [18]. This history suggests that Granitgard has had a somewhat ambivalent approach to this litigation. For all these reasons, I do not consider that this Court should determine the testing issue that Granitgard agitated before it. Nor should it now embark upon the sort of inquiry I have referred to above, to establish whether Granitgard ran this unpleaded case at trial. That necessarily means its final line of attack based on Exhibit 47 should be rejected. 107 It follows that Granitgard has failed in its attempts to rely upon both Exhibit 24 and Exhibit 47 to establish the falsity of the representations pleaded in para 8 of its second further amended statement of claim.

Introduction 108 Finally, I turn to consider the costs issue: did the learned trial judge err in making the indemnity costs order against Granitgard? 109 Termicide actually made two offers of compromise before the trial of these proceedings: one on 14 January 2008; and the other on 25 September 2008. Each offer set out, in some detail, Termicide's arguments as to why it should be accepted. Granitgard did not accept either offer. The learned trial judge rejected Termicide's claim for an indemnity costs order based on the first offer (see [2010] FCA 381 at [28]), but made such an order based on the second offer. It is only this order that is in issue in this appeal.

The relevant principles 110 The principles applicable to the review, by an appeal court, of a discretionary costs order made by a trial judge are well-established. They were succinctly summarised in the Full Court decision of Gore v Justice Corporation Pty Ltd (2002) 119 FCR 429; [2002] FCA 83 as follows (at [50]): Questions of costs are normally determined on discretionary principles by the trial judge. The discretion is unfettered, save that it has to be exercised judicially and in accordance with general legal principles covering awards of costs. If an appellate court is to interfere, it must be upon the clear understanding that it is not merely replacing its view for that of the trial judge. The well known and accepted principles for interfering with a trial judge in the exercise of his or her discretionary powers are to be found in House v The King (1936) 55 CLR 499 at 504-5: "The manner in which an appeal against an exercise of discretion should be determined is governed by established principles. It is not enough that the judges composing the appellate court consider that, if they had been in the position of the primary judge, they would have taken a different course. It must appear that some error has been made in exercising the discretion. If the judge acts upon a wrong principle, if he allows extraneous or irrelevant matters to guide or affect him, if he mistakes the facts, if he does not take into account some material consideration, then his determination should be reviewed and the appellate court may exercise its own discretion in substitution for his if it has the materials for doing so. It may not appear how the primary judge has reached the result embodied in his order, but, if upon the facts it is unreasonable or plainly unjust, the appellate court may infer that in some way there has been a failure properly to exercise the discretion which the law reposes in the court of first instance. In such a case, although the nature of the error may not be discoverable, the exercise of the discretion is reviewed on the ground that a substantial wrong has in fact occurred." 111 Two other Full Court decisions to similar effect are: Queensland Wire Industries Pty Ltd v Broken Hill Proprietary Co Ltd (1987) 17 FCR 211 at 222 and, more recently, Australian Competition and Consumer Commission v Prouds Jewellers Pty Ltd ACN 073 053 273 [2008] FCAFC 199 at [69] per Black CJ, Ryan and Gordon JJ.

The contentions 112 Granitgard submitted that the learned trial judge erred in finding that, the offer of compromise that Termicide made on 25 September 2008, was a genuine offer. It submitted that that offer was simply an invitation to capitulate and it was, therefore, not a genuine offer of compromise. In response, Termicide submitted that the learned trial judge had made no such error. It submitted his Honour was quite correct in concluding that it was no longer sufficient for the recipient of an offer of compromise to show that it was reasonable to reject an offer of compromise, instead that party had to show that there were special circumstances that justified an order for indemnity costs not being made against it.

Intervening Rules amendment 113 It is of some significance on this issue that, during the period between Termicide's two offers, the applicable Federal Court Rule: O 23 r 11, was amended with effect from 2 August 2008, to insert subrule (6) as follows: If: (a) an offer is made by a respondent and not accepted by the applicant; and (b) the respondent obtains an order or judgment on the claim to which the offer relates as favourable to the respondent, or more favourable to the respondent, than the terms of the offer; then, unless the Court otherwise orders: (c) the respondent is entitled to an order that the applicant pay the respondent's costs in respect of the claim incurred up to 11 am on the day after the day the offer was made, taxed on a party and party basis; and (d) the respondent is entitled to an order that the applicant pay the respondent's costs in respect of the claim incurred after that time, taxed on an indemnity basis.

The new Rule created a "prima facie" position 114 The learned trial judge variously described this new rule as giving rise to a "presumptive entitlement", "prima facie" position, or a "rebuttable presumption" that an indemnity costs order should be made: see [2010] FCA 381 at [32]. He concluded that Granitgard had not displaced this prima facie position, summarising his conclusions as follows ([2010] FCA 381 at [51]): Having regard to each of the factors which I have mentioned, especially the then recent re-issuing of the CSIRO appraisal, Granitgard has not, in my opinion, in the face of these factors, displaced the prima facie position for which the rules provide. In my opinion, a judicial exercise of the discretion as to costs requires, in these circumstances, that effect be given to that prima facie position so far as the basis upon which taxation is conducted is concerned. That is so whether, as I apprehend the weight of authority, it is for Granitgard to demonstrate compelling and exceptional circumstances or whether, alternatively, it must show that it was not imprudent or unreasonable not to have accepted the offer. 115 In coming to this conclusion, the learned trial judge, first, noted that he had a wide discretion under s 43 of the Federal Court of Australia Act 1976 (Cth) to order costs, but that discretion had to be exercised judicially: see [2010] FCA 381 at [9]. He reviewed a number of relevant authorities and rejected the "genuine offer" approach saying that it carried with it an "unnecessarily pejorative quality": see [2010] FCA 381 at [34]. Instead, he considered that a party: "may displace the 'presumptive entitlement' or 'prima facie' position or 'rebuttable presumption' created by O 23 r 11(6) … by showing that the offer truly was so derisory that, even having regard to the risks of litigation, it effectively gave nothing away such that the Court should 'otherwise order'": see [2010] FCA 381 at [34]. 116 In passing, it may be noted the learned trial judge's approach is consistent with a Full Court decision that does not appear to have been mentioned by him in his review of the relevant authorities: see Futuretronics.com.au Pty Ltd v Graphix Labels Pty Ltd [2009] FCAFC 40 at [10], where an earlier decision of Heerey J was referred to with approval by the Full Court as follows: In dealing with rule 11(4), which also uses the expression appearing in rule 11(6) - "unless the Court otherwise orders" - Hely J in Port Kembla Coal Terminal Ltd v Braverus Maritime Inc (No 2) (2004) 212 ALR 281 at [17], after referring to Heerey J's observation in an earlier case that "compelling and exceptional circumstances" must exist before the Court would "otherwise order", said: Once an offer is made, and a judgment no less favourable obtained, a rebuttable presumption in favour of indemnity costs is created. It then becomes incumbent on the defendant to show reason why the presumption should not crystallise. Correctly understood, Heerey J was explaining the operation of the Rule, rather than impermissibly attempting to place a fetter on the exercise of the court's discretion. … [H]is Honour was not seeking to do more than to convey that the prima facie position should only be departed from for proper reasons which, in general, only arise in an exceptional case… To similar effect, see also Vawdrey Australia Pty Ltd v Krueger Transport Equipment Pty Ltd (2010) 83 IPR 1; [2009] FCAFC 156 at [187] per Lindgren J and Nutrientwater Pty Ltd v Baco Pty Ltd (No 2) [2010] FCA 304 at [23]-[24] per Kenny J.

Granitgard fails to displace the prima facie position 117 In deciding that Granitgard had neither demonstrated exceptional circumstances, nor shown that it acted imprudently or unreasonably in not accepting the offer, the factors the learned trial judge took into account, included the following (all references are to [2010] FCA 381): (a) Prior to receipt of the second offer, Granitgard had obtained samples of Termiglass from Termicide and had them analysed: see at [35]. The results from these samples were consistent with the statements made in the original CSIRO Appraisal and the recent reissue of that Appraisal: see at [48]. (b) In addition, Granitgard had the benefit of reading and analysing the affidavit evidence-in-chief in Termicide's case, including two affidavits of Dr French: see at [36]. With the benefit of those two affidavits, Granitgard must have appreciated that its case "necessarily entailed persuading the Court that the opinions of the man whose scientific expertise in no small measure underpinned the barrier system which it marketed were not to be preferred in relation to Termicide's barrier system": see at [40]. Further, Granitgard must have had that appreciation in the face of the CSIRO reissuing its Appraisal on the strength of Dr French's opinions: see at [40]. (c) Furthermore, Granitgard had the results of the jar tests of Termicide conducted earlier in 2008, which did not support Granitgard's case and were consistent with the original CSIRO Appraisal and its recent reappraisal to the effect that Termiglass as installed in Termicide's barrier system met the Australian Standard: see at [38]. (d) Whilst putting forward different figures as to the value of Termicide's second offer, both parties agreed that the offer yielded more than a trifling sum (see at [43]). However, the offer did otherwise invite capitulation based on the evidence then available and the recent reissuing of the CSIRO Appraisal: see at [49]. (e) Finally, both parties were "bitter commercial rivals" (see at [46]) and, in not accepting the second offer and proceeding to trial, Granitgard "took a commercial risk in its litigious dealings with a competitor", and chose to press ahead with a case that "went to the very heart of the efficacy of a competitor's product and system". In such circumstances, it was to be inferred that "viewed against its wider commercial contexts, there was much at stake in the litigation in terms of market repute and share and thus profit": see at [50].

Consideration and conclusion on the costs issue 118 It is apparent that the learned trial judge undertook a careful analysis of the relevant authorities and undertook a comprehensive review of all the relevant factors. There is, therefore, no basis for concluding that he exercised his discretion other than judicially. Furthermore, I am unable to discern any mistake of fact, or error of principle, that his Honour committed in this process. As I noted at the outset on this costs issue, since this decision was a discretionary decision, unless such a mistake or error is shown, there is no warrant for this Court to interfere with his Honour's decision. For these reasons, I do not consider there is any merit in Granitgard's appeal on this cost issue.

Conclusion 119 To sum up, while Granitgard has been partly successful in relation to two aspects of the representation issue, it has failed to establish that any of the representations that Termicide made, were in breach of the relevant provisions of the Trade Practices Act 1974 (Cth), in the sense that they were false, misleading, or deceptive. Its appeal against the substantive decision of the learned trial judge must therefore be dismissed. Furthermore, Granitgard has failed to establish any vitiating error on the part of the trial judge in making the indemnity costs order against it, so its appeal in relation to that decision must also be dismissed. 120 As to the costs of this appeal, Termicide's counsel was given leave to refer the Court to authorities that established that the indemnity costs order made below should flow through to this appeal. In the event, he failed to do so. In those circumstances, the appropriate order is that Granitgard pay Termicide's costs of this appeal calculated on a party and party basis. I certify that the preceding one hundred and eighteen (118) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Reeves.