Appeal Outcome: Special leave dismissed with costs by the High Court, 12 February 2010 s201/2009 New South Wales Court of Appeal CITATION: Yarrabee Coal Company Pty Ltd & Anor v Lujans [2009] NSWCA 85 HEARING DATE(S): 23, 24,25 & 26 February 2009 JUDGMENT DATE: 27 July 2009 JUDGMENT OF: Allsop P at 1; Beazley JA at 7; Giles JA at 78 DECISION: (1) Appeal allowed; (2) Set aside the judgment for the respondent and in lieu thereof judgment for the appellants; (3) Respondent to pay the appellants' costs of the trial and of the rehearing of the appeal; (4) Respondent to have a certificate under the Suitors Fund Act if otherwise qualified in respect of the rehearing of the appeal; (5) Subject to order (6), respondent to pay the appellants' costs of the previous appeal to this Court; (6) Liberty to the respondent to file written submissions within 14 days in relation to the costs of the previous appeal and a certificate under the Suitors Fund Act in respect of the appeal to the High Court; appellants to file responsive written submissions within a further 7 days. CATCHWORDS: TORTS - negligence - motor vehicle accident - unsealed road - plaintiff's vehicle went onto shoulder - approaching guide post - loose material on shoulder - plaintiff corrected and lost control - vehicle went off road and plaintiff injured - road under defendants' control - used as haul road for coal trucks - regularly graded - issue of perception of boundary between shoulder and hard running surface - degree of uniformity of colour - whether road surface deceptive - if deceptive whether causation of accident because of deceptiveness proved - use of photographs as evidence - evidence given by reference to photographs - departure from trial judge's findings - need for inference of greater probability than other available inferences. Angel v Hawkesbury City Council [2008] NSWCA 130; Baulderstone Hornibrook Engineering Pty Ltd [2008] NSWCA 243; Beaton v McDivitt (1985) 13 NSWLR 134; Blacktown City Council v Hocking [2008] NSWCA 144; Blatch v Archer (1774) 1 Cowp 63; 98 ER 969; Bradshaw v McEwans Pty Ltd (High Court of Australia, 27 April 1951); Branir Pty Ltd v Owston Nominees (No 2) Pty Ltd [2001] FAC 1833; 117 FCR 424; C Van der Lely NV v Bamfords Ltd [1963] RPC 61; CSR Ltd v Della Maddalena [2007] HCA 1; (2006) 224 ALR ; Costa v The Public Trustee of New South Wales [2008] NSWCA 223; Flounders v Mitlar [2007] NSWCA 238; (2007) 49 MVR 53; Fox v Percy [2003] HCA 22; (2003) 214 CLR 118; Gett v Tabet [2009] NSWCA 76; 254 ALR 504; CASES CITED: Holloway v McFeeters [1956] HCA 25; (1956) 94 CLR 470; Jackson v Lithgow City Council [2008] NSWCA 312; Luxton v Vines [1952] HCA 19; (1952) 85 CLR 352; Lujans v Yarrabee Coal Co Pty Ltd [2008] HCA 51; (2008) 249 ALR 663; March v E & MH Stramare Pty Ltd (1991) 171 CLR 506; Reg v Travers (1958) 58 SR (NSW) 85; Roads and Traffic Authority v Royal [2008] HCA 19; (2008) 245 ALR 653; Schmidt v Schmidt [1969] QWN 3; Short v Barrett [1990] NSWCA 164;` Stillwell Trucks v McKay; Stillwell Trucks v TNT [2002] NSWCA 292; State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq) [1999] HCA 3; 73 ALJR 306; Warren v Coombes [1979] HCA 9; (1979) 142 CLR 531; Warren v Gittoes [2009] NSWCA 24; Yarrabee Coal Co Pty Ltd v Lujans [2007] NSWCA 342; (2007) 49 MVR 178. Yarrabee Coal Company Pty Ltd - First Appellant PARTIES: Jalgrid Pty Ltd - Second Appellant Matina Luise Lujans - Respondent FILE NUMBER(S): CA 40089/07 COUNSEL: C T Barry QC & J Davidson - Appellants B Toomey QC & S Longhurst - Respondent SOLICITORS: McCabe Terrill - Appellants Russell McLelland Brown - Respondent LOWER COURT JURISDICTION: Supreme Court - Common Law Division LOWER COURT FILE NUMBER(S): 20785/01 LOWER COURT JUDICIAL OFFICER: Cooper AJ LOWER COURT DATE OF DECISION: 6 October 2006 IN THE SUPREME COURT OF NEW SOUTH WALES COURT OF APPEAL CA 40089/07 SC 20785/01 ALLSOP P BEAZLEY JA GILES JA Monday 27 July 2009 YARRABEE COAL CO PTY LTD & ANOR v LUJANS Judgment 1 ALLSOP P: The reasons of Beazley JA and Giles JA relieve me of the need to express my reasons other than shortly and substantially by reference to what they have written. I agree with Giles JA. 2 I would only wish to add a number of comments which are to be read and understood by reference to the detailed evidential discussions in the reasons of my colleagues. 3 First, for the reasons expressed by both Beazley JA and Giles JA the primary judge had no relevant advantage over this Court in the resolution of the issues. I accept that the primary judge had an opportunity to absorb over the time of the trial the totality of the evidence and can be seen to have been in the position discussed by Kirby J in State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq) [1999] HCA 3; 73 ALJR 306 at 330 [90] and as discussed in the other cases set out in Branir Pty Ltd v Owston Nominees (No 2) Pty Ltd [2001] FAC 1833; 117 FCR 424 at 435-436 [24]. This Court, however, was exhaustively and meticulously taken through the relevant evidence over four days. The Court was assisted in this regard by, if I may say so, the considered, organised and thoughtful submissions of highly experienced senior counsel, aided by full and carefully drawn written submissions. The Court was able to evaluate the evidence assisted by not only the submissions, but also by the debate between Bench and Bar. This degree of assistance to, and perspective of, an appellate court fortunate enough to receive it and the opportunity then for reserving judgment and analysing the evidence with the advantage of discussion and debate among the appellate court are to be recognised as a species of appellate advantage not often available to a primary judge. This capacity for appellate synthesis and perspective is especially relevant in a case such as this where, as Giles JA's reasons show, the evidence of various witnesses was not always terminologically consistent or entirely lacking in confusion. 4 Secondly, the embedded requirement for the demonstration of error (see the discussion in Gett v Tabet [2009] NSWCA 76; 254 ALR 504 at 508-511 [10]-[23], Costa v The Public Trustee of New South Wales [2008] NSWCA 223 and Branir at 435-440 [21]-[39]) must be understood by reference to the nature and framework of the issues for debate. Where there is no ultimate advantage of the trial judge, that error can be embodied in the reaching of a conclusion different to that reached by the trial judge about a matter in respect of which there can be but one answer. For the reasons expressed by both Beazley JA and Giles JA that is the position here. 5 Thirdly, as to the question of the deceptiveness of the road as found by the primary judge, with respect, I have considerable difficulty in understanding it as a meaningful conclusion in terms of causal connection in this case. If one accepts, contrary to the conclusion of Giles JA, that the road was uniformly coloured so as to make it difficult to see while driving the border of the hard roadway and the shoulder, that is not deceptive. No deception was involved, such as by making the driver think that the hard roadway was somewhere where it was not. On the hypothesis of uniformity of appearance, the driver would know from what he or she could not see that he or she did not know precisely where the border lay. With a road of this width and a car of this size that would not present any inherent danger. One would merely drive safely well within the range of the likely position of the border. There was no suggestion in the evidence that in the vicinity of where the accident occurred the border cut in unexpectedly well within the usual width of the hard roadway. This was an unsealed road over which the respondent had driven some 300 times. Further, the clear arc of the road, the rill and the guideposts made the necessary direction to take to keep the car on the hard roadway tolerably clear. 6 Fourthly, as the reasons of both Beazley JA and Giles JA reveal, it is not possible to conclude that the accident happened because of one reason or another. The respondent, in my view, failed to demonstrate on the balance of probabilities that she swerved and crashed because of somehow being misled by some aspect of the surface of the road. For the reasons given by both Beazley JA and Giles JA that inference is simply not open to be drawn as more probable than not. 7 BEAZLEY JA: I have had the advantage of reading in draft the reasons of Giles JA in this matter. I have come to a different conclusion from his Honour in respect of certain factual findings of the trial judge. However, I have concluded, in agreement with the reasons of Giles JA, that the respondent has not established that any negligence on the part of the appellant caused her injury. 8 For my purposes, the starting point in this appeal is to recognise that it is an appeal from a trial judge in a circumstantial case where the findings as to how the accident occurred depended upon inferential fact-finding. The findings of the trial judge were based on the oral evidence of the witnesses called at trial, photographic evidence, engineering plans and a survey, the latter being of limited assistance. 9 The function of this Court on an appeal is governed by the Supreme Court Act 1970, s 75A. That section provides that an appeal is to be by way of re-hearing. The manner in which the Court is to undertake that task was examined in Fox v Percy [2003] HCA 22; (2003) 214 CLR 118. There, Gleeson CJ, Gummow and Kirby JJ stated at [25]: "Within the constraints marked out by the nature of the appellate process, the appellate court is obliged to conduct a real review of the trial and, in cases where the trial was conducted before a judge sitting alone, of that judge's reasons. Appellate courts are not excused from the task of 'weighing conflicting evidence and drawing [their] own inferences and conclusions, though [they] should always bear in mind that [they have] neither seen nor heard the witnesses, and should make due allowance in this respect' ( Dearman v Dearman [1908] HCA 84; (1908) 7 CLR 549 at 564 citing The Glannibanta (1876) 1 PD 283 at 287). In Warren v Coombes ([1979] HCA 9; (1979) 142 CLR 531 at 551), the majority of this Court reiterated the rule that: '[I]n general an appellate court is in as good a position as the trial judge to decide on the proper inference to be drawn from facts which are undisputed or which, having been disputed, are established by the findings of the trial judge. In deciding what is the proper inference to be drawn, the appellate court will give respect and weight to the conclusion of the trial judge but, once having reached its own conclusion, will not shrink from giving effect to it.' As this Court there said, that approach was 'not only sound in law, but beneficial in ... operation' ([1979] HCA 9; (1979) 142 CLR 531 at 551. See also Taylor v Johnson [1983] HCA 5; (1983) 151 CLR 422 at 426; Jovanovic v Rossi [1985] HCA 17; (1985) 58 ALR 519 at 522; cf Moran v McMahon (1985) 3 NSWLR 700 at 715-716 per Priestley JA)." 10 Their Honours further remarked, at [27]: "Such courts must conduct the appeal by way of rehearing. If, making proper allowance for the advantages of the trial judge, they conclude that an error has been shown, they are authorised, and obliged, to discharge their appellate duties in accordance with the statute." 11 The meaning of "error" in the preceding paragraph is illustrated by what their Honours next said, at [28]-[29]: Over more than a century, this Court, and courts like it, have given instruction on how to resolve the dichotomy between the foregoing appellate obligations and appellate restraint. From time to time, by reference to considerations particular to each case, different emphasis appears in such reasons. However, the mere fact that a trial judge necessarily reached a conclusion favouring the witnesses of one party over those of another does not, and cannot, prevent the performance by a court of appeal of the functions imposed on it by statute. In particular cases incontrovertible facts or uncontested testimony will demonstrate that the trial judge's conclusions are erroneous, even when they appear to be, or are stated to be, based on credibility findings. That this is so is demonstrated in several recent decisions of this Court. In some, quite rare, cases, although the facts fall short of being 'incontrovertible', an appellate conclusion may be reached that the decision at trial is 'glaringly improbable' or 'contrary to compelling inferences' in the case. In such circumstances, the appellate court is not relieved of its statutory functions by the fact that the trial judge has, expressly or implicitly, reached a conclusion influenced by an opinion concerning the credibility of witnesses. In such a case, making all due allowances for the advantages available to the trial judge, the appellate court must 'not shrink from giving effect to' its own conclusion. Finality in litigation is highly desirable. Litigation beyond a trial is costly and usually upsetting. But in every appeal by way of rehearing, a judgment of the appellate court is required both on the facts and the law. It is not forbidden (nor in the face of the statutory requirement could it be) by ritual incantation about witness credibility, nor by judicial reference to the desirability of finality in litigation or reminders of the general advantages of the trial over the appellate process." (Citations omitted) 12 This Court recently reviewed the principles governing an appeal by way of rehearing in Gett v Tabet [2009] NSWCA 76. Importantly for present purposes is the comment at [20], where, after having referred to Fox v Percy and other authorities, including Costa v The Public Trustee of New South Wales [2008] NSWCA 223, the Court said, "[e]mbedded within the approach required by all these authorities is the demonstration of error". 13 As I have said, this case was a circumstantial one and depended upon inferences being drawn from the evidence of witnesses, none of whom saw the accident, and who were being asked to recall events which had taken place nearly a decade previously. In Warren v Coombes [1979] HCA 9; (1979) 142 CLR 531, Gibbs ACJ, Jacobs and Murphy JJ said, at [18] 551: "… in general an appellate court is in as good a position as the trial judge to decide on the proper inference to be drawn from facts which are undisputed or which, having been disputed, are established by the findings of the trial judge. In deciding what is the proper inference to be drawn, the appellate court will give respect and weight to the conclusion of the trial judge but, once having reached its own conclusion, will not shrink from giving effect to it." Issues on the appeal 14 Giles JA has identified two essential questions requiring resolution on the appeal: first, whether the appearance of the road surface was deceptive and, secondly, whether the respondent's accident was contributed to by the deceptive appearance: see Giles JA at [51]-52]. I agree that these are the essential considerations for determination. The findings of the trial judge 15 The detailed findings of the trial judge are set out in the judgment of Giles JA. For my purposes, it is sufficient to refer to what I will characterise as the core findings that led to his Honour's conclusion that the appellant had breached its duty of care. Those findings are at [266]-[268], as follows: "Accordingly I am satisfied on the balance of probabilities that [the respondent] did allow the left wheels of her vehicle to enter onto the shoulder because she was paying attention but was misled by the deceptive nature of the road surface. I am further satisfied on the balance of probabilities that once her right hand wheels were on the hard compacted section of the road way and her left hand wheels were on the softer shoulders she experienced a different 'feel' in the passage of her vehicle as well as in the steering. It was this different feel which caused her to feel unsafe and to slacken off her speed. At the same time she had to get to her right to avoid the guidepost which was within about 1.5 seconds away. She turned her steering wheel to the right but, due to the difference between the adhesion of her right wheel to the road and those of her left wheels to the softer shoulder the vehicle slid out of control ending up on the wrong side of the road in the manner already described. In short, I am satisfied on the balance of probabilities that the appearance of the surface of the road was deceptive in the manner alleged and that it caused or materially contributed to [the respondent] driving with her left hand wheels on the shoulder and, further that this caused or materially contributed to [the respondent] losing control of her vehicle resulting in her injuries." 16 His Honour concluded that it was reasonably foreseeable that the surface of the road would develop such a deceptive appearance: see [269]-[272]; and further held that there was a reasonably practicable means available to the appellant to obviate the deceptive appearance in the road's surface: [273]-[275]. His Honour held, at [290], that the appellant was in breach of its duty because: "Nothing was done to check whether the known presence of coal dust on the road was creating the deceptive appearance of the road surface and, consequently, nothing was done by way of further grading to eliminate it." 17 Giles JA has examined the evidence of the various witnesses in relation to the appearance of the road surface and, in particular, whether the "colour" of the road on the morning of the accident was uniform from rill to rill. As his Honour explains, most of the evidence in respect of this was undertaken by reference to the photographic evidence in Exh 3 (and to some extent, Exh 4), being photographs taken on the day of the accident, and Exh K (and to a lesser extent, Exh L), taken a week after the accident. 18 Giles JA has concluded that the evidence of the various witnesses as to the uniformity of colour of the road surface on the day of the accident was mixed. This is correct, but is also to be expected. The witnesses were giving evidence nearly nine years after the accident and witnesses' recollection of such matters are notoriously variable. It was in this context that Exh 3, a photograph taken, three to six hours after the accident but at least within a few hours of the accident, was given such emphasis. Therefore, the proper evidential use of that photograph in particular had some criticality in the case. 19 Giles JA has also concluded that the evidence most supportive of the trial judge's finding that the road had a uniform appearance from rill to rill on the morning of the accident did not have the force and effect given to it by his Honour. As I understand the reasons of Giles JA, these two matters, that is, the proper use of the photographs and the force and effect of the evidence upon which his Honour relied, are, in part at least, interrelated. Accordingly, both need to be considered. Principles relating to use of photographic evidence 20 In Blacktown City Council v Hocking [2008] NSWCA 144 Tobias JA dealt with the use of photographic evidence at [167]-[170]. As his Honour noted, at [167], the admissibility of photographs as a visual reproduction is without doubt: Reg v Travers (1958) 58 SR (NSW) 85. However, as Lord Reid said in C Van der Lely NV v Bamfords Ltd [1963] RPC 61 at 71: "Lawyers are expected to be experts in the use of the English language, but we are not experts in the reading or interpretation of photographs. The question is what the eye of the man with appropriate engineering skill and experience would see in the photograph, and that appears to me to be a matter for evidence. Where the evidence is contradictory the judge must decide. But the judge ought not, in my opinion, to attempt to read or construe the photograph himself; he looks at the photograph in determining which of the explanations given by the witnesses appears to be most worthy of acceptance." 21 This passage was approved and applied in Short v Barrett [1990] NSWCA 164; and Stillwell Trucks v McKay; Stillwell Trucks v TNT [2002] NSWCA 292. 22 In Blacktown v Hocking, Tobias JA stated, at [169]: "It should be noted that the use by a trial judge of photographs is nothing new. They can, as the authorities to which I have referred confirm, be descriptive of what a witness says he or she saw, being a representation of the witness' knowledge and observations. But they should not be used by a judge to make findings of fact which are otherwise unsupported by the evidence and are therefore no more than conjectural: Schmidt v Schmidt [1969] QWN 3 at 6; Beaton v McDivitt (1985) 13 NSWLR 134 at 142." His Honour added, at [170]: "This is not to say that photographs can have no probative value of themselves even in the absence of an oral description of what is depicted; only that care must be taken to ensure that they are not used as the sole source from which a primary fact is to be inferred where that fact is not revealed on the face of the photographs such as the condition of a structure at a time significantly removed from that when the photographs were taken." 23 In the same case, Spigelman CJ, after noting that Tobias JA had set out the relevant authorities, commented, at [7]: "It is not entirely clear in what circumstances photographs can be used as direct evidence, as distinct from explicating the evidence otherwise given." His Honour then set out passages from Beaton v McDivitt (1985) 13 NSWLR 134 and Schmidt v Schmidt [1969] QWN 3, which are reproduced below. In Beaton , Young J (as his Honour then was) said, at 142-143: " … it seems to me that the law is a little unclear as to just what use may be made of photographs tendered at a trial. In a criminal trial, photographs of the victim may be tendered to enable the jury to understand the medical evidence: see, eg, R v Lobendahn (1980) 5 Petty Sessions Review 2484. Even if photographs are tendered in a motor car accident case, it may be that without concession the photographs can only be used by the judge as descriptive of what the witness who tendered them saw, and that the judge cannot himself make deductions from them: see, eg, Schmidt v Schmidt [1969] QWN 3. However, in R v Ames [1964-5] NSWR 1489 at 1491, the Court of Criminal Appeal seems to me to have made a greater use of the photograph than merely to use it to understand the evidence because their Honours said: '… the photographs … did have some probative value, as distinct from an oral description of what they depicted … They depict the direction of the flow of blood on to the clothing and establish the presence of blood on the mouth …' Very often in equity, photographs are tendered from the Bar table without a witness in the box to give complementary oral evidence, and in my view, when this happens, consistently with R v Ames , the judge treats the photographs as evidence and not merely as material to understand the evidence." 24 In Schmidt, Douglas J, with whom Hart and Lucas JJ agreed, said, at 6: "[The primary judge] based a large part of his judgment on what he saw in [the photographs]. He was not entitled to use them in the way he did. He proceeded to make deductions of fact which on the evidence it was not competent for him to make. The greater part of his deductions made from the photographs can be described as conjectural, and to the extent that he used them as a basis for his findings they must be rejected." 25 Spigelman CJ noted that Douglas J had relied on the following passage from Wigmore on Evidence (3rd ed, 1940) vol 3 at 174-175: "… whenever such document is offered as proving a thing to be as therein represented, then it is offered testimonially, and it must be associated with a testifier. Two consequences plainly follow. On the one hand, the mere picture or map itself cannot be received except as a non-verbal expression of the testimony of some witness competent to speak to the facts represented. On the other hand, it is immaterial whose hand prepared the thing, provided it is presented to the tribunal by a competent witness as a representation of his knowledge." 26 However, the Chief Justice noted that in the subsequent edition of Wigmore, the following passage appears: "This theory which has been aptly dubbed the 'pictorial testimony theory of photographs' was advanced in prior editions of this work as the only theoretical basis which could justify the receipt of photographs in evidence. With later advancements in the art of photography, however, and with increasing awareness of the manifold evidentiary uses of the products of the art, it has become clear that an additional theory of admissibility of photographs is entitled to recognition. Thus, even though no human is capable of swearing that he personally perceived what a photograph purports to portray … there may nevertheless be good warrant for receiving the photograph in evidence. Given an adequate foundation assuring the accuracy of the process producing it, the photograph should then be received as a so-called silent witness or as a witness which 'speaks for itself'." 27 Spigelman CJ said, at [13], that given this commentary, the comments of Douglas J in Schmidt may need revision, as suggested in Beaton. 28 It is apparent from the comment of Tobias JA in Blacktown v Hocking at [170], to which I have referred, that his Honour accepted there was a wider basis for the admissibility of photographs than that stated in Schmidt. However, much will depend upon the photograph in question, the circumstances in which it was taken and what the evidence, if any, is in relation to the photograph. Thus, in Angel v Hawkesbury City Council [2008] NSWCA 130 the Court rejected that the photographs should have trumped the eyewitness evidence which, the Court held, was compelling: see at [69]-[70]. 29 Against that background, the evidence of the witnesses in this case, the photographs and the trial judge's reasons need to be assessed. That assessment itself needs to be considered in the context of this Court's function as governed by s 75A, discussed above. Trial judge's findings as to the colour of the road 30 At [259], the trial judge said: "Exhibit 3, the photograph looking south from the guidepost taken by Mr Rawlins between about three and six hours after the accident shows some, but not complete, uniformity of colour of the road surface from rill to rill. The strip of roadway extending about 1.5 metres to the east of the western rill is a slightly darker colour than the central part of that road. However, the contrast in colours is nowhere near as marked as appears in the photographs taken a week later, exhibits K and L." 31 His Honour also said, at [260]: "The tyre marks on the roadway depicted in the photographs exhibits 3 and 4 show that the surface of the shoulder on the western side was sufficiently dusty to allow the marks of tyres to remain but not sufficiently soft to permit a tyre to penetrate the road surface." 32 It is apparent, however, that his Honour did not rely solely on the photographs to reach his conclusion, as he said at [261], under the general heading, "Consideration": "In the light of all of this evidence I am satisfied on the balance of probabilities that a reasonably prudent driver travelling along the Haul road at the same time and place as [the respondent] was, could not readily distinguish where the hard compacted surface of the roadway ended and the relatively softer shoulder began. I am further satisfied that this state of affairs could materially contribute to [the respondent] so manoeuvring her vehicle that the left wheels entered up that shoulder." (Emphasis original) 33 His Honour then posed the question, namely, whether the lack of distinction between the hard compacted surface of the roadway and the relatively soft shoulder materially contributed to the respondent manoeuvring her vehicle so that the passenger side wheels entered onto the shoulder (at [262]). His Honour posed two possibilities as having caused that situation: (1) that the respondent was not paying attention to her driving; and (2) that she was paying attention, but was misled by the deceptive nature of the road surface. 34 The essential reason his Honour was satisfied that the second possibility was the probable reason was that up to that point, the respondent had been exercising care in relation to her driving. He instanced the manner in which she had overtaken a road train driven by Mr Routledge. His Honour also appears to have accepted the respondent's evidence that she was not changing stations on the car radio just prior to the accident. That finding negatived the suggestion that as the radio in the respondent's vehicle was blaring and between two stations when the first person arrived on the scene of the accident, that she had been occupied in changing radio stations immediately before the accident. 35 His Honour then made the following findings, at [266]-[267]: "Accordingly I am satisfied on the balance of probabilities that [the respondent] did allow the left wheels of her vehicle to enter onto the shoulder because she was paying attention but was misled by the deceptive nature of the road surface. I am further satisfied on the balance of probabilities that once her right hand wheels were on the hard compacted section of the road way and her left hand wheels were on the softer shoulders she experienced a different 'feel' in the passage of her vehicle as well as in the steering. It was this different feel which caused her to feel unsafe and to slacken off her speed. At the same time she had to get to her right to avoid the guidepost which was within about 1.5 seconds away. She turned her steering wheel to the right, but due to the difference between the adhesion of her right wheel to the road and those of her left wheels to the softer shoulder the vehicle slid out of control ending up on the wrong side of the road in the manner already described." 36 Before considering whether his Honour erred in finding the road had a deceptive appearance, it is convenient to deal with the appellant's contention that his Honour also erred in finding that the respondent experienced that "different feel" just before this collision, and that this had occurred just after the respondent had overtaken Mr Routledge's truck. In my opinion, his Honour was entitled to accept that the "different feel" the respondent felt in the manner in which her vehicle was travelling occurred at the point where the passenger side tyre marks were first seen about 33 m from the guidepost. However, to justify that conclusion, it is necessary to have regard to the respondent's evidence. 37 The respondent said, in evidence in chief: "Q. Do you recall what happened with your speed after you had finished overtaking? A. I remember lowering my speed as I was feeling unsafe for whatever reason I don't recall. I just recall feeling unsafe and dropping my speed back down to between 80 and 90. … Q. And after that what's the next thing that you can remember? A. Hitting my head on the window and then hanging upside down in the car." 38 The respondent recalled the road was wet, although she said that the road was not wet where she commenced overtaking the road train. She said she had a memory of being well clear of the truck. In cross-examination, there was some confusion in one answer about when she felt unsafe, but the respondent was given the opportunity to confirm her evidence in chief and she did so. 39 The respondent was no more specific in her evidence as to when or where she "felt unsafe" and reduced her speed than that it was after she had passed the truck. It is apparent that her recollection, whilst chronological, was episodic. The appellant sought to place reliance upon the respondent's answer above as identifying that she "felt unsafe" immediately after overtaking the truck. However, given the respondent's evidence, particularly as identified during the course of cross-examination, his Honour was entitled to make the finding that he did. 40 That leaves as the critical finding, his Honour's finding that the respondent was misled by the deceptive nature of the road. 41 Before deciding whether his Honour erred in finding that the road was "deceptive", it is necessary to understand what he meant by that term. It is apparent from his Honour's reasons that there was no clear delineation between the road pavement proper and the shoulder. There was evidence, independent of the photographic evidence, that supported that finding. It appears that his Honour then inferred that the surface was "deceptive" because there was thereby no means of knowing whether one's vehicle was close to the "softer shoulder". Accepting that the road was substantially the same colour all the way across, that inference was available to his Honour on the evidence that I discuss below. It appears that his Honour then deduced that as there was no clear distinction between the road pavement proper and the shoulder, the respondent was misled. 42 Giles JA has reviewed the evidence of the various witnesses which relate to these central questions. Nonetheless, it is necessary to give a brief overview of the evidence myself to explain the conclusions at which I have arrived. 43 Mr Rawlins was the managing director of the appellant from 1988 until June 2005. He attended at the scene of the accident at approximately 8.30 am. The respondent had been conveyed away from the accident site by that time. Mr Rawlins inspected the site, including the western side of the road, that is, the side of the road on which the respondent was originally travelling. 44 As Giles JA has explained, the witnesses used terminology differently when explaining various features of the road. For my purposes, it is convenient to use Mr Rawlins' evidence to understand the relevant features of the scene. Mr Rawlins referred to the "the crown", of the road as being the pavement of the road upon which vehicles normally travelled. He did not confine his use of that term to the slightly elevated portion in the middle of the road. The "rill" was the rounding of dirt and gravel which was pushed by the grader towards the outer edge of the shoulder. Rills ran along the full length of each side of the road. Mr Rawlins described "the shoulder" as being that part of the road that was on a slight incline. He described the shoulder as normally being outside the line of the guideposts. As he described it: "Part of the shoulder is … on the pavement side of the guidepost, in my definition of pavement and shoulder." 45 Mr Rawlins agreed that the graded portion of the road, that is, the distance of the road from rill to rill, exceeded 10 m. 46 Mr Rawlins said that on the morning of the accident, he carefully inspected the car tracks on the western side of the road and observed that: "… [t]he car tracks drifted to the left hand side of the road, the western side of the road, from the crown of the road onto the shoulder, and then suddenly changed direction and veered across the road." 47 He said that he tracked the passenger's side tyre marks for about 30 m before they commenced the swerve to the right just before the guidepost. He described the tyre marks as "parallel to the road but on the shoulder". 48 Mr Rawlins agreed that, generally speaking, he regarded the guideposts as delineating the margin of the road out to where it was safe to drive on. He qualified this answer in response to the next question, in which it was suggested that so long as a driver was within the line of the guideposts, it was safe. He said: "No, I can't make that statement. This was a dirt road. It depended on the conditions of the road and the speed that you were going. That road there is perfectly safe, in inverted commas, at 120 kilometres an hour in those conditions. In rain, it would be totally dangerous at that. So I can't say that you were perfectly safe to be within the guideposts. Depended on the conditions of the road." 49 Mr Rawlins said that on the morning of the accident, it would have been safe to drive at 120 km per hour on the road. As I understand it, his evidence was that on the day of the accident, the guideposts provided a safe line of travel, as, on the trial judge's finding, it had not been raining. That answer, however, was qualified as follows: "You don't drive on the edge of the road for the fun of it. You would drive off to that edge of the road if you were passing a big truck or there was a lot of traffic, but you wouldn't do it at full speed." 50 Mr Rawlins also refuted the suggestion that it was safe to drive at any line as long as the vehicle stayed within of the rills, because: "Once you go over the shoulder of the road you are at an incline on the side of the road." 51 Mr Rawlins described both the pavement and the shoulder as being very hard although he referred to gravel being "on the shoulder" of the road. He said that "[v]ery little" would happen if one drove on the gravel, "if you're going at a safe speed". He said there was nothing that morning he considered unsafe about the normal surface of the road. When he was asked whether in that regard he was referring to the pavement, he responded, "[a]nd most of the shoulder". 52 Mr Rawlins was asked about the uniformity of colour of the road surface as follows: "Q. Mr Rawlins, you'd agree would you not, that in terms of what we can see in exhibit 3, there is a uniformity of colour of the road surface from rill to rill with some minor variations where there are tracks in the road. By and large the road surface from rill to rill appears to be of the same colour? A. Yes, with your qualifications, I agree with that." 53 In my opinion, when Mr Rawlins agreed that, from what was apparent from Exh 3, there was a uniformity of the colour of the road surface, his agreement was not an answer to what the photograph depicted, but, rather, he was agreeing that Exh 3 depicted the substantial uniformity of the road surface. This is apparent from Mr Rawlins' evidence overall, namely, that when a photograph suggested something with which he was not in agreement, he made his disagreement apparent. One example is to be found in his evidence relating to Exh 33. The relevant questions and answers are as follows: "Q. And in Exhibit 33 where you have recreated the tyre marks, do you suggest that there is a fall away at that point? A. Yes, I do. Q. It's not apparent on the photograph, is it? A. Well I guess I'm familiar with the road and to me it is. From basically on the, as you look at the photograph, the right hand side of that orange part, the road is starting to fall away. The drainage of the road runs from right to left, from that orange, the right hand side of that orange strip. Basically the drainage runs that way to the left of the road and the drain on the other side would tend to run off to the right hand side and you can see that quite clearly because in some of the places you can see the wash-aways washing through the rill where the water is banked up there trying to run through and it has broken through the rill." 54 Mr O'Loughlin is a mining engineer and had been working at the Jellinbah Mine for just over two years at the time of the accident. He was already at the mine and logged a call that the accident had occurred. He then drove the mine ambulance to the scene of the accident, arriving there at about 6.45 am. He described the road surface as being hard and dry. He examined the road to see if he could find any indication as to what caused the accident. He said he observed tyre marks on the shoulder. He said he saw tyre marks for about 20 m along the shoulder, describing them as having "veered off to the left and then back to the right". 55 Mr O'Loughlin later prepared the incident report in respect of the accident which contained a description of the accident in the following terms: "Hi Lux was travelling to work in a northerly direction. For as yet an unexplained reason the vehicle found itself too far on the left hand side of the road between rill and shoulder (the left side wheels were on the soft ground – the right hand wheels were still on the haul road). The tyre tracks on the road indicate that the vehicle was attempted to be steered back onto the haul road causing the vehicle to slide across the road somewhat sideways." 56 He confirmed in his cross-examination that was what he observed. He said that he could tell the right hand wheels were still on the road because the left hand wheels were close enough to the road to mean that the right hand wheels must have been on the road. He also agreed that the shoulder at that point was about 1.5 m wide. He then gave the following evidence: "Q. … on that day, the appearance of the road from rill to rill, the appearance of the surface was uniform? A. The hard surface would be, yes. … Q. And what I am getting at is this: That the appearance of the road from rill to rill, that is, soft part, hard part, soft part, looked uniform? A. I believe it would, yes. Q. So you couldn't tell where the hard part ended and the soft part began could you? A. I believe you could, yes. Q. It looked uniform? A. Yes." 57 Mr O'Loughlin was asked further questions in cross-examination about whether, on the morning of the accident there was any differentiation between where the road ended and the soft shoulder began. He said there was, because: "… the main travel road is hard. It has been compacted and the softer area is the off-shoots from the grading material, so you can see small pebbles of different size." 58 He was then asked questions as to what he observed about the tyre marks once they ran off the road and his evidence was as follows: "Q. … The tyre tracks as they started to become visible in the soft shoulder would have been only at a very slight angle to the edge of the road? A. That's correct, yes. … Q. Let me suggest to you that the first part of the tyre track that you could see in the soft shoulder was only about one tyre width off the road? A. Yes. Q. And it ran along like that, parallel with the edge of the road and only about a tyre's width off the road for some of that 20 metres, didn't it? A. Yes. If I could explain it, with the road bending round to the right, I could see the tyre tracks veering slightly off to the left and then I saw the tyre tracks turn back in towards the road, if that answers your question. … Q. And for a short distance it seemed as if she was able to control the vehicle because the tyre tracks continued to run parallel with the edge of the road? A. Yes. Q. Then you saw the tyre tracks go further towards the rill, is that right? That is, further away from the road? A. Yes, yes. Q. And then you saw the left-hand tyre tracks turn back towards the road? A. That's correct." 59 He said that the maximum distance from the road that he observed the tyre tracks was 75 cm. He confirmed that he observed that the wheels of the respondent's vehicle had gone into soft ground. 60 He was then asked about the colour of the road surface extending back some 50 or 100 m prior to the collision. Mr O'Loughlin described it as being "a dark colour, similar to our road surface but sort of just dirt". He agreed that the road could be uniform in colour from rill to rill and that he believed it would have been so on the morning of the accident: "Q. And the dark colour comes about because of the inevitable spillage of some amount of coal which then gets crushed by the trucks as they drive along? A. That's correct, yes. Q. Did you see the grader working from time to time on this roadway? A. Yes from time to time, yes. Q. And as the grading would take from rill to rill, then that dark colour would be spread right across the road? A. Yes. Q. And that's why you were able to say it was uniform in colour from rill to rill, is that right? A. Yes." 61 Mr O'Loughlin was then asked a series of questions about the colour of the road surface, based first on Exh K and Exh L, which were photographs taken a week after the accident, and then on Exh 3, taken by Mr Rawlins on the day of the accident. By reference to Exh L, Mr O'Loughlin said that there were variations in the colour of the road surface, in that there appeared to be areas of mud, some areas of black and a light area in the middle of the road. He was asked to compare that with Exh 3, which he had not previously seen, and was asked the following question: "Q. Now, you said earlier that the roadway at this point on that morning was uniform in colour from rill to rill. That is borne out by that photograph, isn't it? A. Yes, I believe so." 62 He was asked to confirm his earlier evidence that the colour of the road was "sort of brown colour, from the coal dust". Those answers did not depend upon anything derived from Exh 3. He was asked the following question: "Q. In fact, as you look at that photograph, if you were driving north along that way and the road looked like it looked in that photograph, it would be quite deceptive as to where the road ended and the shoulder began? A. Yes, it would be." 63 Mr Ross is a licensed surveyor and between 1992 and 1996, in the course of his employment, travelled along the road at least twice a week and between 1996 and the time of the accident, about five times a week. He described the road as being "a good dirt road". He arrived at the scene of the accident some time after the police had arrived. He drew a survey plan of the accident scene and measured the road from rill to rill at the point just before the guidepost as being 11.2 m wide. The road was slightly narrower further south, being about 10.8 m wide from rill to rill. 64 Mr Ross agreed the graded area stopped at the rill on each side, but he could not recall whether there was an area that could be called a shoulder. He said that the tyre tracks were able to be seen quite clearly. He considered the tracks touched the rill at one point. He said that the first section of tyre track, which was 33.3 m, was generally in a straight line. He was asked whether Exh 3 represented the colour of the road as he recalled it and responded, "[n]o, I am not going to say". 65 Mr Ross was generally resistant to describing a section of the road as being "an area of a shoulder", stating that he was "not an expert in the area". He agreed, however, that the road was graded, so that it looked the same from rill to rill. He was then asked: "Q. And if it was soft for the last 1.5 metres at the point where this accident happened, you certainly couldn't tell it visually, could you?" To which he responded: "A. I am no expert." 66 Mr Comiskey is a coal miner and earth moving contractor, having been in the mining and earth moving industry for 36 years. Between 1994 and 1998 he travelled along the road every day between Jellinbah Mine and Yarrabee Mine. At that time, he was a mine superintendent. He considered that the road was "very good". He was called on the two-way radio almost immediately after the accident occurred and went straight to the scene of the accident because he had a first aid certificate. When he arrived at the scene, there were one or two other people already there. The respondent was still trapped inside the vehicle. After the respondent was transported away by ambulance, Mr Comiskey walked back along the road, looking at the tracks, trying to work out what could have happened. He said he walked back to where "the vehicle had started to go into a drift and went across the road". 67 Mr Comiskey said that he only walked a little bit further south from the guidepost, but he could clearly see the tyre marks leading up to the post. He said the tyre marks were along the edge of the rill. He agreed that the respondent "didn't disturb the rill" "on the left hand side" and that it was only the passenger wheel marks that could be extending in the shoulder from the guidepost. He described the colour of the road surface as being "of a dark browny colour". 68 Mr Comiskey was shown Exh 3 and said he had not seen the photograph before. He was asked: "Q. Does it accord with your recollection of how the road looked in terms of colour at the time of the accident? A. Pretty well what it's like there in the photo." 69 Mr Routledge was also shown Exh 3 and agreed with a question put in cross-examination by the respondent's counsel that "the photograph [was] a true representation of the colour of the surface after coal dust [had] been ground into it". He agreed that on the morning of the accident, the road was graded, so that it had the appearance as shown in Exh 3. He then gave the following evidence: "Q. It's uniform in colour from rill to rill, is that right? A. Yes. Q. And that's how it looked on the morning of the accident? A. Yes, I would say so, yes." 70 Later in his cross-examination, he gave the following evidence: "Q. …. If it is graded the way it is graded in Exhibit 3, it would be hard to tell in the early hours of the morning, where the heavily compacted part ended and the softer part began, wouldn't it? … A. … Yes, it would be. … Q. Sometimes it would be more apparent, as it was in Exhibit K? A. Yes. Q. Other times it would be deceptive, as it is in Exhibit 3? A. Yes, that's correct." 71 These witnesses were frequent users of the road. In my opinion, their evidence, although given by reference to Exh 3, was evidence based on their own observations. The cross-examiner was not seeking to gain a concession that Exh 3 demonstrated that the road was relatively uniform in colour. Rather, it was used to ascertain whether the witnesses agreed that the uniform appearance of the road as it appeared in Exh 3 was the actual condition of the road on the day of the accident. The position was different in the case of Mr Stewart-Smith, who was an expert witness, and who was not at the scene on the day of the accident or otherwise familiar with the road. His evidence related to the appearance of the road as shown in a number of photographs, including Exh 3. Apart from his evidence, and that of Mr Ross, who was non-committal on this issue, overall the witnesses said that the colour of the road was relatively uniform on the day of the accident. As I have indicated, to the extent that reliance was placed on the photographic evidence, and in particular Exh 3, I am of the opinion that the photographic evidence was used as an aide to confirm or explain the recollections of the witnesses. Accordingly, to that extent I disagree with by Giles JA. 72 However, the fact that the colour of the road was of a mostly uniform colour does not of itself establish that the respondent was thereby misled so as to cause her to drive onto the shoulder. 73 Giles JA sets out the principles that apply in determining whether the respondent has established that the appellant's negligence, if any, caused the accident. I have also read the evidence that relates to this issue. Again, I am not always ad idem with his Honour in relation to his factual findings. For example, I consider that on the evidence, it was likely that the respondent felt unsafe sometime after overtaking Mr Routledge's vehicle and not immediately after doing so. I also would not draw the inference that she slowed down to 80 to 90 km per hour immediately after overtaking Mr Routledge, although she would have slowed down to some extent. 74 In the end result, these differences do not matter, as I consider that his Honour's conclusion in respect of causation is correct. In considering causation, Giles JA has made a number of assumptions, all favourable to the respondent. Those assumptions reflect the factual findings at which I would arrive, namely that the respondent was travelling at 80 to 90 km per hour as she was approaching the guidepost and immediately before swerving to the right and that the road surface was mostly of a uniform appearance. However, even assuming these factual matters in favour of the respondent, they do not lead to "a reasonable and definite inference": see Holloway v McFeeters [1956] HCA 25; (1956) 94 CLR 470 per Williams, Webb and Taylor JJ citing Bradshaw v McEwans Pty Ltd (High Court of Australia, 27 April 1951, unreported), that the onside wheels of the respondent's vehicle moved off the pavement proper and onto the shoulder because the appellant was misled by the appearance of the road. 75 Dixon, Fullagar and Kitto JJ, in Luxton v Vines [1952] HCA 19; (1952) 85 CLR 352, at [8] 358, cited the following passage from Bradshaw v McEwans: "In questions of this sort, where direct proof is not available, it is enough if the circumstances appearing in evidence give rise to a reasonable and definite inference: they must do more than give rise to conflicting inferences of equal degrees of probability so that the choice between them is mere matter of conjecture: see per Lord Robson, Richard Evans & Co. Ltd. v. Astley (1911) AC 674, at p 687. But if circumstances are proved in which it is reasonable to find a balance of probabilities in favour of the conclusion sought then, though the conclusion may fall short of certainty, it is not to be regarded as a mere conjecture or surmise: cf. per Lord Loreburn (1911) AC, at p 678." 76 In this regard, I agree with the reasoning and findings of Giles JA at [245]-[252]. 77 I also agree with his Honour's conclusion, at [257] in respect of breach of duty. The evidence established that the road was in good condition and was graded regularly by a competent driver. 78 GILES JA: The respondent was severely injured when the vehicle she was driving left the road and rolled over. The road was controlled by the first appellant and maintained under contract by the second appellant. Cooper AJ held that the appellants were liable in negligence to the respondent. His Honour assessed the respondent's damages at $8,759,510.55 after a reduction of 20 per cent for contributory negligence. 79 An appeal on liability, but not on damages or contributory negligence, was heard by this Court differently constituted, and was allowed (Yarrabee Coal Co Pty Ltd v Lujans [2007] NSWCA 342; (2007) 49 MVR 178). On further appeal the High Court held that while the conclusions in this Court's reasons were not necessarily wrong, "important steps in the Court's reasoning are not shown to be valid" and "[i]n that sense the Court of Appeal has not reheard the matter as it should" (Lujans v Yarrabee Coal Co Pty Ltd [2008] HCA 51; (2008) 249 ALR 663 at [31]). The appeal to the High Court was allowed. The matter was remitted to this Court for re-hearing of the appeal. 80 For the reasons which follow, in my opinion the appeal should be allowed. Introduction 81 The first appellant operated the Yarrabee and Jellinbah South collieries in central Queensland. The coal mines were some distance to the north of the Capricorn Highway and of the railway line which there ran adjacent to the highway. A haul road had been constructed for the carriage of coal from the mines to the highway, then to be taken to a railway siding at Boonal for loading onto railway trucks. 82 The respondent worked in a clerical capacity at the site of the Jellinbah South mine. She lived at Blackwater, a town on the highway about ten kilometres west of the Boonal siding. Her employer provided a vehicle, a Toyota Hilux twin cab four wheel drive utility, to drive between Blackwater and the mine. 83 At about 6.20 am on 18 September 1998, when driving on the road about 5.9 kilometres north of the junction with the highway, the respondent lost control of the vehicle, which left the road and rolled over. The respondent's injuries resulted in quadriplegia. 84 The road was held by the first appellant under lease. It was used by employees and others working at the mines and by local graziers, but the main users were large road trains carrying coal from the mines to the Boonal siding. The road trains were a prime mover and two or three trailers, each carrying up to 200 tonnes of coal. There were up to 500 road train movements each day. 85 At the time of the respondent's accident the road was unsealed. The maintenance by the second appellant included grading the road surface, of which more later. 86 The first appellant and the second appellant were treated as having the same liability, if any, to the respondent, and no distinction was drawn between them in the appeal. I will refer to them together as the appellants. The road 87 The road was constructed in or shortly after 1981. The engineering drawings provided for a width of 10.4 metres between the batters, with a slight crown and a slight super-elevation on curves. They provided that the central 7.4 metres should have a gravel sub-base topped with a gravel base, each of a stated depth and stated CBR, and that the outer 1.5 metres on each side should be shoulder gravel of a stated depth and CBR. The acronym CBR denotes an accepted measure of the strength of the material. 88 According to Mr Roger Stuart-Smith, a road engineer, the central 7.4 metres was correctly called the travelled way, being the lanes where movement was expected; the two outer 1.5 metres were the shoulders; the travelled way and the shoulders were the carriageway; and the entirety including the batters was the formation. This was not necessarily the way in which parts of the road surface was described by the witnesses at the trial. Mr Stuart-Smith said that the drawings showed that the road "was constructed to carry traffic over its full width of the two nominal lanes as well as the two shoulders". 89 Parts of the road were rebuilt and resurfaced from time to time. According to Mr Alistair McKellar, the managing director of the second appellant, in his time since about 1991 work of that kind had not been done in the location of the respondent's accident and it was "still the original surface that I believe was done by Thiess". 90 The road ran approximately north-south. The Jellinbah South mine was about 22 kilometres from the junction with the Capricorn Highway, and the Yarrabee mine was about 16 kilometres further north. The road was largely straight, but in the location of the respondent's accident it made a sweeping right-hand curve with a radius of about 2,000 metres passing through a deflection of about 22 degrees; the curve was about 680 metres long including transitions. This was known as the Yellow Corner, from the colour of the road material or of the land in that location (both explanations were given). 91 The maintenance by grading was carried out in association with a water truck. From 1992 onwards the grader operator was Mr Barry Jellick, and around the time of the respondent's accident the water truck driver was his uncle Mr John Jellick. 92 Mr Barry Jellick worked from Monday to Friday between about 8 pm and 5 or 6 am, each night grading a length of six or seven kilometres and the next night a similar length so that the whole road was graded over a period of a week. This did not mean that the same length of the road was graded on the same night of each week. The sequence could be interrupted by wet weather or occasion to carry out particular work on the road, and the sequence would be taken up after any interruption. 93 The water truck would heavily water the length of road on which the grader was to work, for dust suppression and to aid in the grading. The grader towed a multi-wheeled seven tonne roller. Mr Barry Jellick would pick up material from one side of the road with the grader blade and cut it across the road, and the roller would roll it into the road and pack it into any depressions. He would then do a second cut across the road, a procedure called a "double grade", to make sure that he was filling the pot holes and bringing the surface up to a good condition. The water truck would then put a fine spray over the road which had been graded. 94 The process of grading created windrows of heaped up material at or beyond the outside of the shoulders, formed from material spilling from the heel of the grader blade on the second cut. The grader operator could pick up material from a windrow to cut across the road. At the trial the windrows were generally called rills, although there was not uniformity in what was meant by a rill; sometimes a witness seems to have used the word more generally to refer to, or include, loose material on a shoulder, not necessarily heaped up. The rills in the vicinity of the Yellow Corner at the time of the respondent's accident were from about 150 to about 300 millimetres high. 95 As explained by Mr Stuart-Smith (and is common experience), on unsealed roads loose material "migrates" from worn wheel paths. The migration can lead to windrows, especially on the outside curves as a result of cornering forces, although the rills in the present case were the result of the grading rather than migration. However, loose material displaced from the travelled way could migrate to accumulate on the shoulders of the road over the period between one grading and rolling and the next. As well, pieces of coal (as carried, they were a little larger than golf ball size) would fall from the road trains. The coal would be crushed by vehicles into the worn wheel paths, particularly on the "loaded" as distinct from the "empty" side of the road, and would provide loose material on the shoulders otherwise than from degradation of the surface of the road. Many witnesses spoke of the loose material, in a variety of ways as well as by that description, referring for example to dust, light dirt, loose bits of coal, small pebbles, gravel, loose gravel, loose scree, marbles, and fine material. 96 I will return to the loose material on the shoulders when considering what the trial judge meant in his findings about a softer or relatively softer shoulder. For the present, his Honour used for it the description of a mixture of gravel and dust, plainly in distinction from a heaped-up rill. Early in his reasons his Honour said - "18 The central portion of the roadway, approximately 8 metres wide was heavily compacted by the road trains which ran along it in a northerly direction empty and in a southerly direction laden with approximately 200 tonnes of coal. 19 The shoulder of the road on the western side was approximately 1.5 metres wide and covered to a small degree by a mixture of gravel and dust. The tyre marks on the roadway depicted in the photographs show that the surface of the shoulder on the western side was sufficiently dusty to allow the marks of tyres to remain but not sufficiently soft to permit a tyre to penetrate the road surface." 97 After the accident Mr Matthew Ross measured the distance between the "edge[s] of road", meaning the inside toes of the rills, at 10.8 metres near where a tyre track from the near side wheels of the respondent's vehicle could first be seen on the western shoulder of the road, and at 11.2 metres where the tyre track deviated to the east and across the road. I will say more of the tyre track later in these reasons. Both distances were greater than the 10.4 metres in the engineering drawings. If the road was constructed according to the engineering drawings and remained as constructed, the rill on one side or the other was beyond the shoulder as correctly understood and on a further part of the formation, and if the grading was equidistant from the nominal centre line both rills were beyond the shoulders. 98 Guideposts were located at intervals along the road. They were made of white poly pipe over star pickets with attached reflectors and, from photographs, striped painted tops. To a driver approaching the Yellow Corner from the south, there was a guidepost on the western side of the road approximately 50 metres (as estimated by Mr Barry Jellick) or 100 metres (as estimated by Mr Robert Wilson) before where the accident occurred. There was another guidepost on the western side of the road at the location where the accident occurred. The deviation of the tyre track last mentioned was variously placed by witnesses between 10 inches and 10 to 12 metres to the south of the guidepost. The trial judge did not make a precise finding, but even at 10 to 12 metres the deviation was very close to the guidepost when (as later appears) from the tyre track the respondent's vehicle was heading straight for the guidepost and was travelling at at least 80 kilometres per hour. 99 The guidepost at the location of the accident was a little distance in from the toe of the rill, with material around the base. If correctly located, it should have been at the outside of the shoulder as correctly understood, and it was described by Mr Barry Jellick as "pretty close to the edge of the shoulder". The trial judge said (at [21]) that the photograph Exhibit 3 "shows a guidepost at the foot of which there is a small mound of material intruding into the shoulder on the western side". The material was consistent with the grader driver withdrawing the blade and then re-positioning it as he passed the guidepost. 100 The speed limit for the road was 100 kilometres per hour. There was no suggestion at the trial that this limit was too high. The respondent's accident 101 The respondent was familiar with the road, having driven over it in the order of 300 times. The weather was fine, but according to one witness overcast, and at the time of the accident (about 6.20 am) it was daylight. 102 The respondent drove from Blackwater east along the Capricorn Highway to the junction with the road, and turned onto the road. A road train entered the road from the Boonal siding shortly ahead of her. It was driven by Mr Geoffrey Routledge, making his first run of the day, empty, to pick up coal. 103 As required by colliery instructions, the respondent used the CB radio in her vehicle to speak to Mr Routledge and ask if it was safe to overtake. According to Mr Routledge, this was about 100 metres along the road. He said that it was, and she overtook the road train. She reached a speed of 100 to 110 kilometres per hour. She continued driving north towards the Jellinbah South mine. 104 The respondent said that she "pulled back in" after passing the road train, and that she reduced her speed to between 80 and 90 kilometres per hour because she "was feeling unsafe for whatever reason I don't recall". The trial judge's finding that she reduced her speed at the Yellow Corner, when experiencing a different "feel" in the passage of her vehicle because its near side wheels were on the shoulder caused her to feel unsafe, was contested on appeal. The appellants submitted that any reduction in speed closely followed overtaking the road train, and that at the time of the accident the respondent was travelling at about 100 kilometres per hour. 105 Mr Baxter Smith turned onto the road shortly after the respondent, driving to service machinery at the Jellinbah South mine. He saw the respondent's vehicle ahead of him. In a statement recorded in a police notebook at the time he said that he was about 300 to 400 metres behind it. In his evidence he said that he was 500 metres behind it. He said nothing about a road train, and was not asked about it. 106 Mr Smith was travelling at 100 kilometres per hour and the distance from the respondent's vehicle did not change. In the statement he said that as he approached the Yellow Corner "I thought I saw the car … go off the road and wondered what happened". In his evidence he said "[t]hen I didn't see it as I approached a gradual curve in the road", but that he did not see the vehicle go off the road. He saw debris and the vehicle off the road, and stopped and gave assistance. 107 Mr Routledge, whose road train was travelling at about 80 kilometres per hour, did not see the accident. He said that the respondent's vehicle "just sort of disappeared up around the corner and out of sight". Within a very short time of the respondent's vehicle passing his road train he was passed by another vehicle, which on the evidence must have been that driven by Mr Smith. He said that the other vehicle "was probably 100 metres behind her so maybe 15 seconds, 20 seconds later", and that both vehicles went off into the distance. He came upon the accident a short time afterwards, and also gave assistance. 108 Assuming a speed of 90 kilometres per hour and five kilometres from overtaking the road train to the accident, it would have taken the respondent a little over three minutes to reach the Yellow Corner. The assumptions are open to variation, and the time may have been greater or possibly a little less. 109 The respondent had no recollection of the accident itself. Her first recollection was of hitting her head on the window of her vehicle and then hanging upside down in the vehicle. 110 No one saw the accident. What happened could, however, be reconstructed, principally from tyre marks on the road. 111 The respondent's vehicle came to rest on its side in a grassed area near some small trees about 22 metres to the east of the road. There were a roll mark and tyre marks for a total distance of about 41 metres after leaving the road. A curved tyre mark for 52.2 metres across the road, heading to where the vehicle crossed the rill on the eastern shoulder and left the road, was interpreted by the experts as a clockwise yaw, that is, side slipping with the rear of the vehicle moving to the left. Leading to the beginning of the yaw mark was the tyre track on the western shoulder of the road to which I earlier referred. It ran for a distance of 33.3 metres to a distinct deviation to the east at the beginning of the yaw, at a point (as earlier mentioned) variously placed up to 10 to 12 metres to the south of the guidepost. 112 The trial judge found - "167 Having regard to the totality of the evidence, I am satisfied on the probabilities that the plaintiff drove with her left hand wheels on the shoulder of the roadway and with her right hand wheels on the compacted surface for a distance of some 33 metres when she found herself approaching the guidepost. She then steered sharply to the right. Her passenger side rear wheel lost traction in the turn and slid sideways causing loss of control of the vehicle. The vehicle then, whilst moving sideways went through the rill on the eastern side and on to the rough surface of the adjoining paddock. It then rolled over coming to rest after having traversed in the rolling condition some 40 metres from where she crossed the edge of the roadway. 168 It is convenient to bear in mind that at 80 kph the plaintiff would cover 22.18 metres in a second and at 100 kph 27.77 metres in a second. At either of these speeds the time lapse while her passenger side tyres were on the shoulder of the roadway would have been just over a second. The evidence is that the distance between the near side and off side wheels is approximately 1.8 metres. Thus at all times that her near side wheels were on the shoulder, her off side wheels were on the compacted surface. 113 From the trial judge's [267] later set out, he considered that the respondent's vehicle lost traction in the turn "due to the difference between the adhesion of her right wheel [sic] to the road and those [sic] of her left wheels in the softer shoulder". His Honour referred at [261], also late set out, to the "relatively softer shoulder". 114 At the heart of the appeal was how the respondent came to be driving partly on the softer or relatively softer shoulder of the road and approaching (and very close to) the guidepost. After the accident 115 Messrs Smith and Routledge arrived at the scene of the accident within minutes. Mr Dale Bates and his passenger Mr Rodney Shannon were driving north on the road. Mr Gregory Majlinger was driving south. They all stopped to give assistance, also within minutes. An ambulance was called through 000 and from the Jellinbah South mine. 116 The 000 ambulance arrived at 6.44 am. At some time a fire brigade truck arrived. Mr Paul O'Loughlin, a mining engineer at the Jellinbah South mine, drove the mine's ambulance to the scene, arriving at about 6.45 am. Mr Joseph Comiskey went to the scene from the mine because he had a first aid certificate. The roof was cut from the respondent's vehicle, she was removed, and the 000 ambulance left with her at 7.23 am. 117 Senior Constable Anthony Turner arrived at about 7 am. While he was there Mr Ross came by on his way to work. Mr Ross was a surveyor, and at Senior Constable Turner's request took the measurements of the widths of the road and the lengths of the tyre marks earlier mentioned. The plan he prepared showing the measurements did not, however, purport to show the location of the 33.3 metre tyre track on the western shoulder or a precise relationship of the tyre marks to the rill or the guidepost. 118 Mr Gregory Martin, a mine employee, arrived at around 7am, and returned later in the day to retrieve the respondent's vehicle. Mr Barry Jellick and Mr John Jellick heard of the accident when at the Boonal siding and went to the scene, arriving while the ambulance was still there. Mr John Rawlins, the managing director of the first appellant, arrived at about 8.30 am. 119 Thus there was a number of persons potentially able to give evidence of the tyre marks and the condition of the road. As well, some photographs were taken. 120 Mr O'Loughlin took at least five photographs. Two of them became Exhibit 14, but the originals had been lost and were not available on appeal. All five were part of an incomplete report compiled by Mr O'Loughlin which became Exhibit F. The photographs of the road are small and indistinct. They are of no assistance in relation to the tyre marks and little if any assistance otherwise. 121 Mr Rawlins took photographs which became Exhibits 26, 27 and 28. Enlargements became Exhibits 3, 4 and 5 respectively. He was not certain of the time, but thought the photographs were taken at about 9.30 am; at another point he said before 9.00 am. The trial judge observed that one of the photographs (in fact Exhibit 5) showed Mr Martin with his crane at the scene, and that Mr Martin thought that he had arrived there to retrieve the respondent's vehicle at about midday; his Honour did not resolve this. Exhibit 5 showed the respondent's vehicle off the road. Exhibits 3 and 4 were photographs of the road, and were marked by Mr Rawlins to identify more clearly the curved near side wheel tracks as the respondent's vehicle went across the road. Neither clearly shows the tyre tracks on the western shoulder of the road prior to the deviation into the yaw. 122 Mr Rawlins returned to the scene of the accident approximately a week later. He took photographs of the road which became Exhibits 29, 30 and 31. Enlargements of the photographs became Exhibits K, E and L respectively. It appears that the road had been recently graded, and any tyre mark from the time of the accident would have been obliterated by grading or loose material on the shoulder. Mr Rawlins placed a hard hat, shown in Exhibit K, at the toe of the rill to indicate the outer edge of the shoulder and as a guide to the height of the rill. Its position did not purport to show the lateral location of the tyre mark he had seen. Mr Rawlins said that the rill was no different from the time of the accident. 123 In late October 1998 Mr Rawlins again returned to the scene of the accident, together with an investigator, Mr Lavin. He or Mr Lavin took photographs which became Exhibits 32 (4 photographs), 33 (3 photographs) and 34. By then the road had certainly been graded, and it was not necessarily in similar condition to its condition on 18 September 1998; even the rill may not have been the same as the rill at the time of the accident. Three of the photographs in Exhibit 32 showed tyre tracks in loose material on the western shoulder of the road, leading towards the guidepost and then deviating to the right across the road. Mr Rawlins was asked whether they were because of the tyre marks he saw on the day of the accident, and said - "I was asked to drive my vehicle approximately where I remembered the tyre marks so yes, those are marks in my vehicle approximately where I thought they were travelling that day." The trial judge's decision 124 The respondent's principal case at trial was that the accident occurred because the road surface in the travelled way was wet and slippery and covered with loose material from grading on the night of 17-18 September 1998, so that her vehicle lost traction and "drifted" onto the shoulder. 125 The trial judge rejected this case. It depended largely on the evidence of Mr Wilson, which the trial judge did not accept. His Honour was not satisfied that the relevant section of the road had been graded on the previous night, saying (at [214]) "[i]ndeed the evidence supports the contrary" although he did not say when he thought it had been graded. His Honour found - "232 In the light of all of this evidence I am satisfied on the balance of probabilities that the central compacted surface of the relevant part of the roadway was not in fact slippery and further that there was no loose gravel on the central compacted surface of the relevant part of the road of sufficient size to give rise to the effect of marbles on that hard surface. 233 It follows that the plaintiff has failed to satisfy me on the balance of probabilities that there was any slipperiness or loose gravel on the main central heavily compacted part of the road surface which caused or materially contributed to her accident." 126 The trial judge said (at [179]) that "[i]n addition the plaintiff argues that there was a lack of physical distinction between the edge of the compacted road surface and the shoulder". His Honour held in favour of the respondent on this basis. 127 The trial judge said - "238 On behalf of the plaintiff it was submitted that the colour and appearance of the road surface from side to side was uniform. Therefore she had difficulty in perceiving where the hard surface of the roadway ended and the softer western shoulder began. This, in turn led her to drive with her right hand wheels on the hard surface and her left hand wheels on the softer shoulder. The argument continues that once in this position she would have difficulty in maintaining control of her vehicle and it was this that led to it veering to the right and moving out of control to the opposite side of the roadway and rolling over. 239 The submission as to the deceptive nature of the road surface is sought to be supported by the contrasts appearing in photographs exhibits 3 and 4 taken on the day of the accident showing almost uniform colour across the road with those taken a week later (exhibits K and L) where the colour of the central part of the road differs markedly from that of the shoulders. This submission is further supported by photographs taken about a month later (exhibits 32 and 33) which also show marked contrast in colour between the central part of [sic: ? and] the shoulders. 240 These submissions raise the following questions: 1) Was the appearance of the surface of the road deceptive in the manner alleged? 2) If so, did that cause or materially contribute to the plaintiff driving with her left hand wheels on the shoulder? 3) Also if so, did that cause or materially contribute to the plaintiff losing control of her vehicle? 4) Also if so, was that deceptive appearance a breach of the defendants' duty of care to the plaintiff?" 128 The trial judge said at [241], "It is now appropriate to examine the evidence relating to these questions". After doing so over [242] to [260], to which I will return, the trial judge said - "261 In the light of all of this evidence I am satisfied on the balance of probabilities that a reasonably prudent driver travelling along the Haul Road at the same time and place as the plaintiff was, could not readily distinguish where the hard compacted surface of the roadway ended and the relatively softer shoulder began. I am further satisfied that this state of affairs could materially contribute to the plaintiff so manoeuvring her vehicle that the left wheels entered up[on] that shoulder. 262 This leaves the question as to whether on the probabilities it did materially contribute to her so manoeuvring her vehicle that the left wheels entered up that shoulder. In other words, did her left wheels enter that shoulder because she was not paying attention to her driving, or did it occur because she was paying attention but was misled by the deceptive nature of the road surface? 263 On behalf of the plaintiff it is submitted that the notion of momentary inattention is merely speculation, and flies in the face of her evidence that she felt unsafe and slackened off her speed. 264 The corroborated evidence of the plaintiff is that she appropriately contacted the driver of the road train ahead of her before overtaking it and overtook it in a proper manner. Furthermore, the evidence of Mr Baxter Ian Smith is that she was travelling at the same speed as he was – the speed limit for that road. There is no evidence to suggest that her manner of driving prior to the accident was in any way erratic or abnormal. 265 There is evidence from Mr Bates that when he went to her vehicle very shortly after the accident he noticed that the radio in the vehicle was blaring and was between two radio stations. The plaintiff was questioned on this and agreed that she had her radio on immediately prior to the accident but denied that she was occupied changing stations. The fact that the radio sounded as if it was between two radio stations could be because of the jolting the radio received during its rolling and/or it could have been due to the aerial becoming broken or dislodged at that time. 266 Accordingly I am satisfied on the balance of probabilities that the plaintiff did allow the left wheels of her vehicle to enter onto the shoulder because she was paying attention but was misled by the deceptive nature of the road surface. 267 I am further satisfied on the balance of probabilities that once her right hand wheels were on the hard compacted section of the road way and her left hand wheels were on the softer shoulders [sic] she experienced a different 'feel' in the passage of her vehicle as well as in the steering. It was this different feel which caused her to feel unsafe and to slacken off her speed. At the same time she had to get to her right to avoid the guidepost which was within about 1.5 seconds away. She turned her steering wheel to the right but, due to the difference between the adhesion of her right wheel to the road and those [sic] of her left wheels to the softer shoulder the vehicle slid out of control ending up on the wrong side of the road in the manner already described. 268 In short, I am satisfied on the balance of probabilities that the appearance of the surface of the road was deceptive in the manner alleged and that it caused or materially contributed to the plaintiff driving with her left hand wheels on the shoulder and, further that this caused or materially contributed to the plaintiff losing control of her vehicle resulting in her injuries." 129 His Honour went on to find (at [269]-[272]) that it was reasonably foreseeable "that the surface of the road would develop such a deceptive appearance", because the loaded road trains lost a fine dust and coal which was crushed onto the surface of the road so it went dark. His Honour asked (at [276]) whether the appellants were in breach of their duty of care owed to the respondent "by failing to have the hard compacted road surface clearly distinguishable from the softer adjacent shoulder", and after consideration of some cases concluded (at [290]) that they were because "[n]othing was done to check whether the known presence of coal dust on the road was creating the deceptive appearance of the road surface and, consequently, nothing was done by way of further grading to eliminate it". 130 The appellants complained on appeal that the trial judge had found against them on a case which they had not been required to meet at trial. They said that the case they had been required to meet was that the road had been graded on the night of 17-18 September 1998, the grading leaving the road surface in a wet and slippery condition, but that the trial judge had found against them because the road had not been graded on that night. The grounds of appeal did not extend to this complaint. The appellants applied for leave to amend the grounds of appeal, although not conformably with the complaint but rather to add grounds that the trial judge erred in finding that coal dust on the road gave it a deceptive appearance and that further grading would have made a difference to the appearance "in the absence of that case being put at trial". For reasons given at the time, leave to amend was refused. The softer or relatively softer shoulder 131 What did the trial judge mean by the shoulder, and what did he mean in describing it as softer or relatively softer? 132 A road engineer's distinctions between the shoulder of the road and the formation beyond it, on the one hand, and the shoulder of the road and travelled way, on the other hand, were understandably not maintained in witnesses' descriptions of the road. The rill seems often to have been taken to be part of, or at the least not distinguished from, the shoulder, and as I have said sometimes the word seems to have been used to refer to, or include, loose material on the shoulder. The shoulder was treated as something other than the central part of the road, which was given many descriptions - for example, the running track, the running surface, the hard road, the hard stuff, the hard part where the trucks ran all the time, the hard surface, or the "roadway itself". Mr Bates referred to "the main part of the road, the part you would normally drive on, the hardest, the clean part of the road", and distinguished it from "out to the rill". The shoulder was often referred to as such, although without precision, but also in other ways, for example as "out further" or "the edge part". 133 As has been seen, the trial judge referred at [18] to the central and heavily compacted portion of the road approximately 8 metres wide, and at [261] to "the hard compacted surface" of the road. Consistently with the witnesses' language, this was not identification of the travelled way according to the engineering drawings. It can be better understood from the trial judge's finding at [232] that there was "no loose gravel on the central compacted surface of the relevant part of the road of sufficient size to give rise to the effect of marbles on that hard surface". In short, the trial judge's reference to "where the shoulder began" (at [261]) was not according to the engineering drawings, although there may have been a rough correspondence, but from the usage of the road. 134 Many witnesses were asked about a soft shoulder. It was often not clear whether the questioner or the witness meant the softness of the heaped-up rill, softness from the presence of loose material on a firm shoulder, or softness because a shoulder less compacted than the central part of the road could give way under a vehicle's wheels. 135 In relation to the road in general, there was evidence of softness in all these senses. Illustrations are – Mr James McIver, that he had seen sections of the road "as one approaches the rill" that had been "damaged by truck wheels that had broken the surface"; he distinguished between "the running surface" and "the soft area along the edge"; Mr John Jellick, that the shoulders "would stand up to a truck running on them without sinking or anything but they are not part of the running track", but depending on "how far out on the rill" the road train went it could "break through"; Mr Comiskey, that the shoulders were not soft but there would probably be loose material along the edge; he distinguished between "the hard stuff" and where the surface was "softer than the central part of the road"; he also said that he had seen deep indentations in the shoulder left by a truck; Mr Bates, that the road still had a hard surface on the outside but had "fall-out", loose little bits of coal on top of the road, and was "the same compaction but a little bit slippery once you get outside the main running area of the road"; Mr O'Loughlin, that "the main travel road" was hard and had been compacted "and the softer area is the off-shoots from the grading material, so you can see small pebbles of different size"; Mr Majlinger, giving as one reason for the respondent's loss of control that her vehicle went too far out on the "marbles", a racetrack expression meaning "there is little bits of rolled-up rubber or little bits of debris or fine gravel"; Mr Rawlins, that the road was equally compacted from rill to rill "for all intents and purposes", as seen in Exhibit 3; and Mr Barry Jellick, that "the edges or shoulder which extended beyond the guideposts tapering towards the table drain were solid", which he knew not only from driving on the shoulders but also from walking on them when he had to replace guideposts dislodged through the grading process. 136 These are but illustrations: there was great variety, and it is not profitable to multiply them. It is clear, however, that road trains passing each other would move onto the shoulders, and that while there were occasions when on going close to the rill they could "break through" the shoulders generally bore their considerable weight. 137 The trial judge was referring to the shoulder at the location of the respondent's accident. I return to the tyre mark on the western shoulder: the trial judge's treatment of the tyre mark shows what he meant by a softer or relatively softer shoulder. 138 Early in the reasons was his Honour's statement at [19], set out above, that "[t]he tyre marks on the roadway depicted in the photographs show that the surface of the shoulder on the western side was sufficiently dusty to allow the marks of tyres to remain but not sufficiently soft to permit a tyre to penetrate the road surface". The trial judge had referred in the preceding paragraph to the photographs Exhibits 4 and 5, but only Exhibit 4 showed the western shoulder. As will be seen, he later referred specifically to the photographs Exhibit 3 and 4. 139 The trial judge later referred, under the heading "The Events of 19 September 1998", to evidence given by a number of witnesses. They included descriptions of the tyre marks on the road, not only those on the western shoulder, given by Mr Bates, Mr Shannon, Mr Majlinger, Mr Comiskey, Mr O'Loughlin, Senior Constable Turner, Mr Ross, Mr Barry Jellick, Mr John Jellick and Mr Rawlins. 140 Although the trial judge's account did not reflect it, Senior Constable Turner's evidence seemed to have both the near side and driver's side wheels of the respondent's vehicle off "the road surface", and he said that the vehicle "went over the rill". It was rather confused evidence, amongst other reasons because he considered that the rill and the shoulder were the one and the same. He was the odd man out, and I think he was implicitly not accepted by the trial judge. 141 I summarise the other descriptions of the tyre mark in the evidence which the trial judge recounted. Mr John Jellick said that the tyre mark may have touched the rill at one point. Mr Bates said that it went off "the road" about a metre, and did not go on the outer (or western) side of the rill; it went to about half a metre from the rill, and was then "trying to rejoin the road". Mr Majlinger said it appeared that the vehicle had been "on the marbles" at the extreme western edge of the road surface, and that the surface there was "hard but with loose stones on top, fine pebbles or whatever". Mr Comiskey said there were wheel tracks from the near side wheels on the shoulder. Mr O'Loughlin saw wheel tracks "veering from the road on the shoulder", and in a note on a photograph said that the wheels were "on the soft ground". Mr Ross said that the marks showed the vehicle veering off from the hard surface towards the edge. Mr Barry Jellick saw the left hand tyre marks probably a tyre width away from the top of the shoulder off the road surface, they appeared to come off the road gradually. Mr Rawlins could trace the near side wheel tyre mark along the shoulder of the road for about 30 metres. 142 At a later point in his reasons the trial judge referred, with apparent acceptance, to Mr O'Loughlin's evidence that the tyre mark went a maximum of about 75 centimetres off "the road". The trial judge did not refer to Mr Rawlins' recreation of the tyre track in late October 1998, as shown on the photographs part of Exhibit 32, although he did say that photographs were taken on that occasion. 143 Returning to the evidence which the trial judge recounted, Mr Bates said that the outside edge of the road was still hard but had loose little bits of coal on top of the outside edge of the road. The trial judge noted that Mr Bates agreed that the western shoulder was not actually slippery but it was loose. As in the earlier summary, his Honour referred to Mr Majlinger saying that it appeared that the vehicle had been "on the marbles". Mr Shannon said that the tyre marks "didn't stand out until you got close to the post", and the trial judge referred to his explanation that between the heavily compacted area and the rill was an area compacted only by the rubber tyred roller towed behind the grader. Mr Barry Jellick said that there was "dusty, loose material" on the western shoulder where he saw the tyre marks. Mr Rawlins described the shoulder of the road as firm and hard, although at this point Mr Rawlins appears to have meant by the shoulder the batter down to the table drain. 144 The trial judge's references to the evidence in these respects clearly underlay his findings of how the respondent's accident occurred, at [167]-[168] earlier set out. At a later point in his reasons he said, all but repeating what he had said at [19] - "260 The tyre marks on the roadway depicted in the photographs exhibits 3 and 4 show that the surface of the shoulder on the western side was sufficiently dusty to allow the marks of tyres to remain but not sufficiently soft to permit a tyre to penetrate the road surface." 145 While there were not explicit findings, I consider his Honour to have found that when the near side wheels of the respondent's vehicle went off the central and heavily compacted part of the road and onto what was, in distinction, called the shoulder, they did not touch the heaped-up rill. They ran along the shoulder inside the rill. From his Honour's references to gravel and dust and a dusty surface, I take his Honour to have meant by the soft shoulder or the relatively soft shoulder that there was loose material on a surface which, although not heavily compacted like the central part of the road, was firm and did not break through when the respondent's vehicle (which was of course very much lighter than a road train) went onto it. 146 As the trial judge did explicitly find, the shoulder was "not sufficiently soft to permit a tyre to penetrate the road surface". The relevant softness was not in lack of compaction of the shoulder, but because of the loose material, the gravel and dust, on the firm surface of the shoulder. 147 Findings so understood were in my view proper findings on the evidence, and to the extent to which the respondent contended for a different understanding of the trial judge's reasons – which was far from clear – I would not accept her contention. There was other evidence to like effect. It included Mr Barry Jellick describing tyre marks in loose material on top of a hard surface; and saying that on the day of the accident he walked over on the shoulder "and it was solid. The only loose material was on the shoulder where the vehicle had skidded"; Mr John Jellick saying that the area "beyond the edge of the road, bounded by the guidepost" was firm, with "some loose material on the shoulder for a distance of about 1 metre, beyond the white guide post"; Mr Comiskey saying that there were not troughs or indentations but just impressions of the wheel where the surface was softer than the central part of the road; and Mr Rawlins saying that where the tyre track was seen was similar in compaction to the "pavement" in a practical sense and the tyre marks were with in the "dusty material", the "gravel which exists on the shoulders". While Mr O'Loughlin referred to a soft shoulder and said that he "observed that [the respondent's] wheels had gone into soft ground", his evidence is consistent with a tyre mark in loose, and therefore soft, material on the shoulder. 148 Further, the understanding of what his Honour meant by the softer or relatively softer shoulder is consistent with, perhaps required by, the mechanism for loss of control found by the trial judge, namely loss of traction (at [167]) or the differential adhesion of the wheels (at [267]). There can be loss of traction or of adhesion due to loose material. It is not obvious that, and we were not referred to evidence supporting that, it could come from a tyre's penetration of an insufficiently compacted surface. 149 It should be said that there was little, if any, difference between the appellants and the respondent in relation to the position of the tyre track for the 33.3 metres prior to the deviation to the right. I have referred to Mr Rawlins' evidence that the tyre tracks shown in photographs in Exhibit 32 were approximately where he remembered the tracks were. Although there was some dispute over arcing or straightness of the tyre track, the appellants accepted that the tyre track on the western shoulder of which many witnesses spoke was approximately in that location and direction, and the respondent's counsel said that "we do not differ from the path – we don't differ greatly anyway – from the path as shown by Mr Rawlins". 150 The tyre track in the photographs is in what appears to be loose material approximately mid way between a clean and compacted surface over the central part of the road and the toe of the rill. A vehicle on the path shown by the tyre track would be headed directly for the guidepost. This last is consistent with the evidence of Mr Barry Jellick (to which the trial judge referred) that, if the vehicle had not changed its course, it would have "run the guidepost over". Mr John Jellick gave evidence to the same effect, also saying that "she would have hit to the right of centre of the bonnet, the right of centre of the vehicle". So did Mr Rawlins: he said that from the tyre marks the vehicle "would have taken out the guidepost" with the point of impact in the centre of the radiator. The central questions 151 As I have said, at the heart of the appeal was how the respondent came to be driving partly on the softer shoulder of the road and approaching the guidepost. The trial judge concluded that it was because the road surface was deceptive to a driver such as the respondent and because the respondent was in fact deceived, the deceptive appearance contributing to her near side wheels being on the shoulder of the road. 152 The respondent's counsel accepted that "unless she gets into the soft shoulder by reason of the deception we don't have a case". The respondent sought to uphold the trial judge's reasons; there was no notice of contention. The central questions were whether the appearance of the road surface was deceptive and whether the respondent's accident was contributed to by the deceptive appearance. 153 We were taken in considerable detail to the evidence at the trial. The deceptive appearance of the road surface 154 The trial judge's conclusion involved three related steps - (i) the road was uniform in colour from side to side; (ii) a reasonably prudent driver could not distinguish where the hard compacted surface ended and the relatively softer shoulder began; (iii) the appearance of the road surface was deceptive. 155 The trial judge came to the second step at [261], in the light of the evidence to which he referred over [242] to [260]. The first step must also be found in that evidence. The third step gave the characterisation "deceptive". 156 A lot of what the trial judge referred to over [242] to [260] was concerned with the tyre track on the soft shoulder. His Honour's references to evidence material to uniformity in colour and inability to distinguish between where the hard compacted surface of the roadway ended and the relatively softer shoulder began was fairly limited. It can be summarised as follows. At [244] the trial judge said that the evidence of all witnesses who regularly drove along the road was that they did not willingly drive along the shoulders but preferred to drive along the central hard part. This was not linked by his Honour to uniformity in colour or inability to distinguish the transition from hard to softer; on the evidence as a whole, it was because the shoulders did not provide a hard running track. At [245] the trial judge said that Mr Kelvin Pilcher, at the time a road train driver, said of the Yellow Corner that the edges were very soft and one could not see where hard part ended and the soft part began. The trial judge added that this was a general description not specifically related to the day of the accident. At [246]-[247] the trial judge referred to Mr Bates' description of the road as heavily compacted in the middle, and still hard on the outside but with loose bits of coal – Mr Bates also used the word "fallout" – on top of the road. At [248] the trial judge said that when shown Exhibit 3 Mr John Jellick said that "if you look you can notice that the shoulder is a bit more yellow than the centre". At [252]-[255] the trial judge set out evidence of Mr O'Loughlin, saying that the respondent placed great reliance on it. Two passages from the evidence are presently relevant. First, his Honour said at [252] that - " … [o]n the issue of the visibility of the distinction between the hard part and the shoulder of the road [Mr O'Loughlin's evidence] was (C288): - 'Q. You have agreed that the shoulder was 1.5 metres wide? A. It would be about that, yes. Q. And that it was soft? A. Yes. Q. Do you say that on the morning of the accident you could actually tell where the road ended and the soft shoulder began? Is that what [you] are saying? A. You can tell where the difference is because the main travel road is hard. It has been compacted and the softer area is the off-shoots from the grading material, so you can see small pebbles of different size." Secondly, part of a passage at [254] taken from the evidence of Mr O'Loughlin was - "Q. Now, what was the colour of the surface of the road say in the 50 or 100 metres leading up to where you saw the first tyre tracks? A. Between the accident and the start? Is that what you are referring to? Q. Imagine you are driving along the road from Boonal and as you are driving north, you get to the point where you first saw the tyre tracks, okay? A. Mm. Q. What was the colour of the road surface at that point and extending back south, say for 50 or 100 metres, on that morning? A. It would be like a dark, a dark colour similar to our road surface but sort of just dirt. Q. And you have said uniform in colour from rill to rill? A. Yeah it can be, yes. Q. Well it was, wasn't it? A. I believe it would have been, yes. Q. And the dark colour comes about because of the inevitable spillage of some amount of coal which then gets crushed by the tricks as they drive along? A. That's correct, yes. Q. Did you see the grader working from time to time on this roadway? A. Yes. Q. And that's why you were able to say it was uniform in colour from rill to rill, is that right? A. Yes." At [256]-[258] the trial judge referred to evidence of Mr Majlinger about driving on the "marbles". The evidence was not directly concerned with the appearance of the road, but it included Mr Majlinger saying that to him "anything in the marbles would be looking at that roadway, from the batter by the looks of it a metre or two towards the road from the batter would be sort of what I consider to have been the marbles". At [259] the trial judge said - "259 Exhibit 3, the photograph looking south from the guidepost taken by Mr Rawlins between about three and six hours after the accident shows some, but not complete, uniformity of colour of the road surface from rill to rill. The strip of roadway extending about 1.5 metres to the east of the western rill is a slightly darker colour than the central part of that road. However, the contrast in colours is nowhere near as marked as appears in the photographs taken a week later, exhibits K and L." 157 The trial judge's [259] just set out should be understood together with what his Honour had earlier said when considering whether the road at the Yellow Corner had been graded on the night before the respondent's accident. He had then said (at [209]) that the colouring of the road in Exhibits 3 and 4 "is much darker than in exhibits K and L, and is consistent with coal dust having been dropped on and crushed into the road surface", and (at [212]) that on looking closely at Exhibits 3 and 4 "it is seen that there is a certain uniformity in colour over the roadway consistent with the deposit of coal dust from road trains". His Honour said (also at [212]), and this must have been a substantial reason for his finding that there was no loose gravel on the central compacted surface of the road sufficient to give the effect of marbles, that Exhibits 3 and 4 showed corrugations which looked "like the tracks caused by the tyres of very heavy vehicles". While noting (at [213]) that Exhibits 3 and 4 did not show a track as shown in Exhibits 32 and 33, his Honour must have considered that the photographs were consistent with a central compacted surface free of the loose gravel. 158 A number of things may be said about the trial judge's conclusions from this evidence. 159 The only direct evidence of uniformity in colour from rill to rill to which the trial judge referred was that of Mr O'Loughlin. In the evidence set out by the trial judge Mr O'Loughlin first said of the uniformity in colour from rill to rill that it can be, then that he believed it would have been, then (by agreement with the suggested effect of grading) that it was. This rather qualified evidence called for some caution, particularly when the suggested effect of grading in making the road a uniform colour is difficult to reconcile with the trial judge's finding of negligence in failing to grade to eliminate the deceptive appearance of the road surface. 160 It also was not the whole of Mr O'Loughlin's evidence on the subject. 161 Mr O'Loughlin's first evidence on the point was when it was put to him that "[t]he road surface at that point, on that morning, was uniform in colour from rill to rill". He replied, "I don't recall what colour the road was", and there was then the evidence - "Q. But you would agree with me, wouldn't you, that at that point in the road, on that day, the appearance of the road from rill to rill, the appearance of the surface was uniform? A. The hard surface would be, yes. Q. But from rill to rill is not all hard, is it? A. No, no. Q. Because the shoulders are soft, aren't they? A. A. Yes. Q. And what I am getting at is this: That the appearance of the road from rill to rill, that is, soft part, hard part, soft part, looked uniform? A. I believe it would, yes. Q. So you couldn't tell where the hard part ended and the soft part began, could you? A. I believe you could, yes. Q. It looked uniform? A. Yes." 162 After the evidence in the passage set out by the trial judge Mr O'Loughlin was shown Exhibit L, with counsel's observation that the road surface seemed to have a yellowy colour or a light colour. His evidence was - "Q. Did you ever see it like that? A. Yes, I have. Q. Was it like that on the morning of the accident? A. I believe it was similar, from memory. Q. Well, let me just ask you something about that. The road surface depicted in Exhibit L, you wouldn't say that was uniform in colour from rill to rill, would you? There is colour variations in it? A. Yes, there is colour variations but it is not, you don't have large areas of black and this colour here. Q. No, I understand that but there are changes, aren't there, from rill to rill? There is some gravel on the shoulder, there is an area of what appears to be mud and it is quite light out there in the middle of the road, isn't it? A. Yes." 163 The cross-examiner was correct in that Exhibit L does not show uniformity in colour from rill to rill. He may not have expected the answer that the road was similar on the morning of the accident. Having received the answer, he showed Mr O'Loughlin Exhibit 3. The evidence was - "Q. Now, you said earlier that the roadway at this point on that morning was uniform in colour from rill to rill. A. Yes, I believe so. Q. That jogs your memory as to the uniformity of the colour, does it not? A. It does, yes. Q. And it is considerably darker than the road surface in Exhibit L, isn't it? A. Yes. … Q. And the grading of it from rill to rill has left the colour of the surface, from rill to rill, uniform? A. Yes." 164 Having regard to all the evidence, it is difficult to rely with confidence on Mr O'Loughlin's evidence set out by the trial judge for a finding that the road surface was uniform in colour from rill to rill. 165 Although most of the witnesses were not asked about uniformity in colour, there was evidence going to uniformity in colour from rill to rill to which the trial judge did not refer. It must also be taken into account. 166 Mr Pilcher said that the road was "mostly" the colour shown in Exhibit 4. This was not as at the day of the respondent's accident, and is imprecise. Mr Wilson drove along the road about an hour before the respondent's accident. He said that the road was "completely different" from Exhibit 3. He gave evidence of a muddy condition of the road, which the trial judge did not accept, and may be put aside. 167 Mr John Jellick was not asked for his recollection of the appearance of the road on that day. He was shown Exhibit 3, and said in evidence part of which was referred to by the trial judge (at [248]) - "Q. And you would agree with me, wouldn't you, that that photograph accurately depicts the fact that the graded surface, that is, from rill to rill, appear to be the same colour? A. Yes. Q. Shoulders and running track were the same colour? A. If you look you can notice the shoulder is a bit more yellow than the centre." 168 It was put to Mr John Jellick that the surface appeared to be a different colour in Exhibit K from Exhibit 3. He answered, "Obviously taken at a different time of day". 169 Mr Comiskey agreed that "by the time of the accident" the road surface, which he called a dark browny colour, was uniform from rill to rill. When shown Exhibit 3 and asked for his recollection of "how the road looked in terms of colour at the time of the accident", he said that it was "[p]retty well what its like there in the photo". 170 Mr Ross agreed that the road was "graded so that it looked the same from rill to rill", although this was not specifically put in terms of colour. To a surveyor, the question may have been taken geometrically. 171 Mr Majlinger said that where the "marbles" area began could easily be seen in Exhibit K, and that that was "how clearly [he] could see that area … on the morning of the accident". Shown Exhibit 3, he agreed that it showed the road surface "on the way it was on the morning of the accident" and said that "in terms of the road surface" there was no difference from Exhibit K. He said - "Q. Do you say that Exhibit K and Exhibit 3 appear to be pictures of an identical road surface. A. It could be just completely different colouring." 172 Exhibit K does not show uniformity in colour from rill to rill. Mr Majlinger appears later to have accepted that there were different colours in Exhibits K and 3, and gave the evidence - "Q. The way it shows in Exhibit 3 is the way it generally looked after there had been coal dust ground into the surface, does it not? A. Yeah, just a normal haul road. Q. The darker colour in Exhibit 3 is because there is a layer of coal dust on top of the road, isn't there? A. I don't know. Q. You don't agree with that? A. I don't disagree or agree, I just don't know. It just looks like a normal road and both pictures are taken in different light, same as that one there is completely, those two colours there is completely different and this one is completely different. Q. Do you say Exhibit K – you have that in front of you? A. Yes, sir. Q. Do you say Exhibit K and Exhibit 3 are the same in terms of the rill? A. They look the same on the inside but the outside you can't tell because it is different, kind of, two different photos. Q. Do you say that, looking at the guideposts in Exhibit K, do you say that that's exactly how the road looked on the morning of the accident? A. All I can say is that it was just the basic old road. I can't tell you like, what it looked like. I wasn't attention to it and even when I was looking at the road like, when we were looking at skid marks, I wasn't looking at the road. I am just saying, 'Oh, I wonder what happened?' You know what I mean? 'Oh, this looked like what happened'. I am sorry -- ". 173 Mr Routledge agreed that the road might have "a coating of ground-up coal on it", although not "if they had graded it", and that sometimes it would look as in Exhibit L and sometimes as in Exhibit 4. He agreed that Exhibit 3 was "a true representation of the colour of the surface after coal dust has been ground into it", and - "Q. And at least on the morning of the accident [the road was] graded so that it had the appearance that it does in Exhibit 3? A. Yes, I would say that's it. Q. It's uniform in colour from rill to rill, is that right? A. Yes. Q. And that's how it looked on the morning of the accident? A. Yes, I would say so, yes." 174 Mr Robert Davis, who had worked at the Yarrabee mine for many years and became the registered mine manager in 1987, agreed in answers to the respondent's counsel that there was "very little visible coal on the road" in Exhibit 3, indicating that it had been graded "recently". It was put to Mr Bates and Mr Rawlins by the respondent's counsel that Exhibit 3 showed that that section of the road had been graded the night before, and they agreed. It should be recalled that the respondent's principal case was that the road had been graded on the night of 17-18 September 1998, and was left wet and slippery and covered with loose material. This evidence was not specifically directed to colour, but in a case that the road should have been graded to do away with uniformity in colour it suggests that, on the witnesses' readings of the photographs, there was not uniformity in colour. 175 Mr Rawlins said in chief that there was "no significant difference" between the road as shown in the photographs in Exhibit 32 and as it was on the morning of the accident, save that - "Q. Can you see any difference at all? A. There is a little more loose material on the shoulder or the edge of the road just before it goes over the shoulder perhaps. That is just a function of precisely when the last time the grader passed there which, whether this was three days versus the other one, two days, I don't know. There is a slight amount more loose material on the edge of the road." 176 This evidence also was not specifically directed to colour. In cross-examination Mr Rawlins said - "Q. Mr Rawlins, you'd agree would you not, that in terms of what we can see in Exhibit 3, there is a uniformity of colour of the road surface from rill to rill with some minor variations where there are tracks in the road. By and large the road surface from rill to rill appears to be of the same colour? A. Yes, with your qualifications, I agree with that." And later - "Q. The road is quite a different colour isn't it in exhibit K compared with Exhibit 3? A. Yes. Q. It's a much lighter colour? A. That's correct. Q. It's clearly been graded between the taking of photo Exhibit 3 and the taking of Exhibit K? A. That's correct. Q. Is that right? A. Yes, that's correct." 177 Mr Stuart-Smith agreed that the road surface depicted in Exhibit 3 was "pretty much" uniform in colour from rill to rill, but "a bit darker close to the rill". 178 The evidence was mixed. Uniformity in colour is a matter of degree, and the witnesses generally gave their evidence of uniformity in colour, or gave content to their recollection of uniformity in colour, by regard to what Exhibits 3 or 4 and K or L showed. The trial judge could see for himself what they showed, and expressed what he saw at [259]; for convenience, I repeat it - "259 Exhibit 3, the photograph looking south from the guidepost taken by Mr Rawlins between about three and six hours after the accident shows some, but not complete, uniformity of colour of the road surface from rill to rill. The strip of roadway extending about 1.5 metres to the east of the western rill is a slightly darker colour than the central part of that road. However, the contrast in colours is nowhere near as marked as appears in the photographs taken a week later, exhibits K and L." 179 Photographs can be used to understand other evidence. Their use as primary evidence from which the judge draws conclusions is more contentious, see the recent discussions in Blacktown City Council v Hocking [2008] NSWCA 144 at [7]-[14] per Spigelman CJ and [167]-[170] per Tobias JA, Angel v Hawkesbury City Council [2008] NSWCA 130 per Beazley and Tobias JJA at [71]-[72], and Warren v Gittoes [2009] NSWCA 24 at [2]-[3] per Hodgson JA at [54]-[56] per Tobias JA. 180 Taken with the oral evidence, the trial judge was entitled to act upon what the photographs showed. But it was necessary to bear in mind that colour, and uniformity in colour, could be affected by the quality of a photograph and the conditions in which it was taken. I have referred to Mr John Jellick's comment about photographs taken at a different time of day. As was said by Mr Stuart-Smith of the different colours of the road surface in Exhibits 3 and K, "Exhibit 3 the sky is very hazy which could be overcast and it could be facing into the sun in which case you get different colours". Mr Rawlins referred to different light conditions and possible differences from the development process. The evidence of the witnesses who were asked about what photographs showed, or who gave content to their recollection by reference to what photographs showed, was ultimately to be assessed with regard to what the trial judge thought the photographs showed. 181 It is also necessary for this Court to assess the evidence with regard to what it thinks the photographs show. Considering them with the oral evidence, I do not entirely agree with what the trial judge saw in the photographs. The trial judge referred only to Exhibit 3 and its contrast with Exhibits K and L. In Exhibit 4 the strip of roadway extending to the west of the eastern rill appears of a distinctly different colour. What amounts to uniformity in colour can not be closely described; as I have said, it is a matter of degree. For my part, I would elevate the trial judge's description of a "slightly" darker colour and "incomplete" uniformity in colour to something more than slight and incomplete. While not as clear a contrast as in Exhibits K and L, I consider that there is a well discernable difference in colour between the central part of the road and the shoulders. 182 Going to ability to distinguish between the hard compacted surface and the relatively softer shoulder, it is to be noted that in the evidence set out by the trial judge Mr O'Loughlin said, despite his acceptance of uniformity in colour, that it could be told where the road ended and the soft shoulder began. He explained that the "main travel road" was hard and "the softer area is the off-shoots from the grading material, so you can see small pebbles of different size". Other than the general statement of Mr Pilcher, nowhere else in the evidence of the witnesses to which the trial judge referred was there evidence of inability to distinguish the transition from hard to softer, and the evidence of Messrs Bates and Majlinger to which the trial judge referred supported Mr O'Loughlin's explanation. 183 A finding, from this evidence, that a reasonably prudent driver could not distinguish the transition from hard to soft can not readily be made. If the trial judge's conclusion came from the photograph Exhibit 3, that is not made clear, and what the trial judge saw in the photograph had to be taken together with the other evidence to which he referred. 184 There was other evidence going to the matter. 185 Mr O'Loughlin was asked further about telling the difference between the where the road ended and the soft shoulder began. It was put that "you couldn't tell where the hard part ended and the soft part began", and he replied, "I believe you could, yes". When shown Exhibit K, he gave the evidence - "Q. (Witness shown Exhibit K) When you said that you could tell where the shoulder, that is, the soft area began because there would be gravel there, is that the sort of thing you had in mind? A. Yes. Q. In that photograph, you can see the roadway proper, can't you? A. Yes. Q. You can see what appears to be some type of mud at the edge of the road? A. Yes. Q. Then you can see a gravelly area which you describe as the shoulder? A. Yes. Q. And there is gravel clearly visible in that photograph, isn't there? A. Yes. Q. On the soft shoulder but not on the road itself? A. That's correct. Q. So when you said earlier that you could distinguish where the hard part ended and the soft part began, you referred to gravel, that is a reasonable depiction of what you were attempting to describe in words? A. Yes." 186 The cross-examination later brought the evidence - "Q. What I suggest to you is this: That whilst in Exhibits K and L you can see the loose gravel, which you say helps you pick up where the shoulder is quite clearly? A. Yes. Q. You can't see that in Exhibit 3, can you? A. It is less well defined in Exhibit 3. Q. Well, it is not defined at all actually, is it? A. It is very difficult to. Q. You would agree with this, wouldn't you: That driving north on that road that morning, if that's how it looked, it would be very difficult to pick up visually, where the road ended and the soft shoulder began. You would agree with that, wouldn't you? A. I would say it would be less difficult, yes. Q. More difficult? A. Sorry yes, more difficult. … Q. In fact, as you look at that photograph, if you were driving north along that way and the road looked like it looked in that photograph, it would be quite deceptive as to where the road ended and the shoulder began? A. Yes, it would be." 187 This was not evidence of observation or experience from his driving on the road; his experience was rather to be found in the evidence that it could be told, from the "gravelly area", where the road ended and the soft shoulder began. It depended on what was properly to be seen in the photograph. 188 Mr McIver said, speaking generally, that he saw coal dust along the length of the road the effect of which was that "[i]t was difficult to determine where the edge of the hard running surface was". 189 Mr Routledge agreed that loose gravel "on the last metre or so of the edge of the road, was a constant feature of this road", and that it could be seen quite clearly in Exhibit K but not in Exhibit 3; that it was "[e]asy to see in K, hard to see in 3"; that "if it was graded in the way it is graded in Exhibit 3, then if you were driving towards the cameraman, it would have been difficult to see where the heavily compacted part of the road ended and where that loose, gravelly part commenced"; and that in Exhibit 3 the road was in that respect deceptive. This also was not evidence from observation or experience over the period he was driving a road train. 190 It was put to Mr Stuart-Smith that, in Exhibit 3, "if you were travelling towards the photographer, there is nothing about the colour of the surface that allows you to determine where the pavement ends and the shoulder begins". He answered, "Not really, no". 191 To the opposite effect, it was put to Mr Shannon by counsel for the respondent - "Q. Normally could you see where the very hard part of the road ended and the less hard part began? A. Most definitely, yes. Q. Was that because it was normally graded in such a way to make it obviously where the very hard part ended and the less hard part began? A. No, it wasn't indicated by the grader but any country roads you can see where the actual line is or the hard area on a country road is." 192 Mr Shannon agreed that this was not a normal country road because of the road trains, but this evidence remained. 193 The respondent led evidence at the trial from Mr Scott Beatty, who had worked at the Yarrabee mine for about six months in 1997 and had gone off the road when passing a road train. It was a quite different accident; he went too wide and drove into soft material a foot or so deep that had been graded into the table drain as an "extension of the flat part of the road". He was asked whether he was able to "see any difference between the hard-packed material and the soft material", and answered in the negative and said that that applied "along the whole length of the road". The soft material was not loose material on the shoulder, and in the photograph Exhibit B tendered to show the skid marks in his accident there is a clear distinction between the graded road surface and the soft material. The respondent relied on Mr Beatty's evidence, but I think it tends against her case. 194 Mr Gregory Dalliston, a union official who had visited the collieries a dozen times, observed in March 1998 that "there were few 'delineators'" to notify where "the edges of the running track" were, and said that there was "a lot of bull dust, like soft dust", which without the delineators made it hard to define where the softer shoulders were. He could not say if all the road was like that. Ordinarily delineators would refer to the guideposts, and he said that the guideposts he saw were "on the outside edge of the total road width and not delineating where the running surface was" and that he told colliery management that the road "does not have the same amount of delineators on the side of the road" as the haul roads on the mine sites. I refer to this for completeness. I do not think his evidence was of inability to distinguish, in relation to uniformity of colour, the transition from the running track to the shoulders. 195 Again the evidence was mixed, and a deal of it was evidence of what a witness saw or explained from photographs rather than what the witness described from observation and experience. The respondent relied on appeal particularly on the evidence of Messrs O'Loughlin and Routledge. Mr O'Loughlin's agreement that, if the road looked as in Exhibit 3 on the morning of the respondent's accident, it would be "more difficult" to "pick up visually where the road ended and the soft shoulder began" was not expressly linked with uniformity in colour, but in the light of earlier questions uniformity of colour was probably understood as the basis for what was put to him. Similarly as to Mr Routledge's evidence. But neither had explicitly accepted that the road looked as in Exhibit 3, and the force of this evidence turns on what this Court makes of that Exhibit. 196 Although not articulated, his Honour appears to have decided by regard to Exhibit 3 that, because of the incomplete uniformity in colour, a reasonably prudent driver could not distinguish between the hard compacted surface and the relatively softer shoulder. But Exhibit 3 could not be taken alone; it should have been taken together with all the other evidence. Giving due regard to the trial judge's conclusion, this Court should decide for itself. 197 I respectfully differ from the trial judge. Attention should not be confined to the colour of the road surface. The appearance of the road surface, and ability to distinguish between the hard compacted surface and the relatively softer shoulder, involves also the different texture (for want of a better word) provided by the loose material on the shoulder. As Mr Shannon said, "any country roads you can see where the actual line is or the hard area on a country road is", and Mr O'Loughlin gave a like explanation for his belief that you could tell where the hard part ended and the soft part began. Mr Stuart-Smith also spoke of the importance of the different appearance in the surfaces - "An unsealed road, all the driver can tell is the apparent edge between the travelled way and the shoulder, as determined by where other vehicles travel. And that is, in fact, the most important distinction because whether or not the theoretical shoulder is, is not as important as this distance between where vehicles travel and where they don't travel, because you get loose material where they don't travel and harder material where they do travel. That's the important distinction. In this case, it tends to coincide with where the theoretical area is between the travelled way and the shoulder." 198 What the trial judge meant by the softer or relatively softer shoulder is of some significance. The difference between the hard compacted surface, free from loose gravel, and the loose material on the shoulder can mark the transition from hard compacted surface to relatively softer shoulder. It will do so, because the shoulder is softer by reason of the loose material, unless there is such uniformity in colour that the difference in texture is largely lost. 199 The distinction between the running track and the shoulders, to revert to some of the language used in the evidence, is plain from loose material on the shoulders in the photographs Exhibit K and L and in photographs in Exhibits 32, 33 and 34. The loose material is less evident in Exhibits 3 and 4, but on my examination of the photographs is also apparent. A number of the witnesses saw it, see earlier in these reasons in relation to the tyre track on the western shoulder, giving evidence of it sometimes indirectly in descriptions of the tyre mark on the soft ground and in descriptions of the marbles (Mr Majlinger) and the loose little bits of coal (Mr Bates). In his evidence of his observation of the western shoulder at the time Mr O'Loughlin referred to the "off-shoots from the grading materials … small pebbles of different sizes". I bear in mind that in context the evidence was of what the witnesses saw when on foot, not when driving along the road, but this evidence assists in appreciating what the photographs show. 200 As I have said, in my view there is in Exhibits 3 and 4 a well discernable difference in colour between the central part of the road and the shoulders. Taking account of all the evidence, I consider that a reasonably prudent driver could distinguish where the hard compacted surface of the roadway ended and the relatively softer shoulder began; put in the trial judge's language, I decline to conclude that the driver could not readily do so. 201 I finally go to the characterisation "deceptive". It is important, not least because it adds content, for the purposes of negligence and causation, to the perception of a driver on the road at the time of the respondent's accident. 202 Mr O'Loughlin agreed with that characterisation "as to where the road ended and the shoulder began", and Mr Routledge also agreed with the adjective "deceptive". The deceptiveness was not because the transition between the compacted part of the road and the relatively softer shoulder was not where it appeared to be. It was because the uniformity in colour was said to bring that the transition could not readily be seen. 203 There was what I have called the texture provided by the loose material on the shoulder. And deceptiveness should not be judged only by the appearance of the surface of the road. A driver would be assisted also by the visible rills: Mr Stuart-Smith said they "provided a definite 'edge' to the travelled way". As well, there were the guideposts (it will be recalled that where the respondent's accident occurred there was a guidepost towards which her vehicle was directly heading). Even a considerable degree of uniformity in colour does not take away these guides to the driver's course on the road. 204 The characterisation "deceptive" can be questioned by assuming that there was no dust or detritus from coal carried by the road trains. The road would have the uniformity in colour from rill to rill provided by the material from which it was constructed, in all probability a considerable degree of uniformity in colour – reddish, yellowish, or similar, as is commonly encountered in unsealed roads. Apart from the rills and the guideposts, a driver would (in the words of Mr Shannon) see "where the actual line is or the hard area on a country road is". One would not ordinarily characterise an unsealed road as deceptive because the uniformity in colour made it difficult to distinguish the precise boundary between the running track and the shoulder. Dark colour from coal dust and detritus changes the colour, but in the fine and daylight conditions in which the respondent was driving I do not think it brings deceptiveness. Even if it be the case that one could not determine from the appearance of the road where the boundary was that is not deception. One is told that one cannot tell where the boundary is. There is nothing deceptive about that. 205 Thus I respectfully differ from the trial judge's conclusion that the nature of the road surface was deceptive. Contribution to the respondent's accident 206 At this point I assume that there was such uniformity of colour across the road that it was difficult to see where the hard compacted surface of the roadway ended and the relatively softer shoulder began, such that the nature of the road surface could be regarded in that sense as deceptive. What I have said in differing from that conclusion should not be put aside, however, because in addressing causation – how the respondent came to be driving partly on the relatively softer shoulder of the road and approaching (and very close to) the guidepost – what I have said about discernable difference in colour and matters other than uniformity in colour must be borne in mind. 207 I have earlier set out the trial judge's reasons at [262]-[268] for his finding on causation. It should be noted that at an earlier point the trial judge had said, when posing the question of what caused the respondent's near-side wheels "to encroach upon the shoulder" - "170 There is no evidence of the presence of any other vehicle on the relevant part of the roadway leading up to the movement to the wrong side of the road. The nearest vehicle was that driven by Mr Ian Baxter Smith [sic] some 500 metres behind her. 171 There is no evidence of any moving object such [as] an animal on the roadway which may have caused her to change her course. 172 There is no evidence of any mechanical defect with the vehicle she was driving which contributed to the course which the vehicle took. 173 There is evidence that no object was found on the roadway after the accident which could have caused the plaintiff to proceed along the course that she did." 208 The basis for the trial judge's satisfaction expressed in [266] is, with respect, not entirely clear. I take his Honour to have accepted the respondent's denial that she was occupied changing radio stations; that finding should remain for the purposes of the appeal. The reasoning appears to have been that the respondent was driving appropriately when contacting Mr Smith and overtaking his road train and in maintaining the speed limit; that there was "no evidence to suggest that her manner of driving prior to the accident was in any way erratic or abnormal"; and that those matters led on the balance of probabilities to the conclusion that, in the choice between inattention and being misled, the respondent was misled by the deceptive nature of the road surface. 209 In the conduct of the trial the parties appear to have presented the trial judge with the choice between inattention and being misled. The question was more correctly whether the respondent had established, on the balance of probabilities, that she was misled by the deceptive nature of the road surface. It was relevant that there was no evidence of another vehicle, an animal, a mechanical defect or an object on the road to cause what happened, but it remained for the respondent to establish the cause of the accident. That she was driving appropriately prior to the accident proved little about what caused the accident. 210 Some time was spent on appeal on two matters said to go to inattention or being misled: arcing or straightness of the tyre track on the western shoulder of the road, and speed at the time of the accident. (a) Arcing or straightness 211 Senior Constable Turner said that "prior to where the incident initiated the road was straight", and that the vehicle "continued along the shoulder in a straight line". Mr Ross said that the 33.3 metres was "generally in a straight line". I have referred to the evidence of a collision course with the guidepost, which the appellants treated as involving driving in a straight line. Other witnesses said to the effect that the tyre track went for a distance parallel to the edge of the running track and then diverged away, before the sharp deviation to the right. 212 There was no necessary inconsistency. "Straight" in the witnesses' usage was relative, and could accommodate some curvature. But the parties took competing positions. 213 In the evidence of Mr O'Loughlin set out by the trial judge at [263]-[255] he agreed that the tyre track first ran "parallel to the edge of the road and only about a tyre's width off the road", and explained that "with the road bending round to the right, I could see the tyre tracks veering slightly off to the left and then I saw the tyre tracks turn in towards the road … ". He said that the tyre tracks went towards the rill to a maximum distance of 75 centimetres off "the road", and agreed that they "describe[d] an arc away from the road" to a maximum of 75 centimetres. The respondent relied on the description of an arc away from the road; the appellants contended for straightness. 214 The point of the difference, in the parties' eyes, was this. In support of the position that the respondent came to have the near side wheels of her vehicle on the shoulder through inattention, the appellants said that she travelled straight having through inattention failed to take the curve in the road, and came at the last second to turn sharply to the right in order to miss the guidepost. In support of the position that she had been misled by the deceptive nature of the road surface, the respondent said that inattention should not be accepted: first because the accident occurred part way around the curve and she would have been off the road prior to the accident had she gone straight, and secondly because the arcing of the tyre track was explained by the softer shoulder causing the respondent's vehicle to veer to the left. In the respondent's submission, the arcing supported an explanation other than inattention for the path taken by the vehicle. 215 The dispute as to arcing and straightness does not appear in the trial judge's reasons, nor does any use of arcing towards the conclusion that the respondent was misled by the nature of the road surface. The respondent may be taking too much from Mr O'Loughlin's agreement with the description of arcing, since his explanation may have attributed the veering to the road's curvature ("with the road bending round to the right, I could see the tyre tracks veering slightly off to the left"). Further, that encountering the loose material on the shoulder would have caused the respondent's vehicle to arc to the left had doubtful support in the evidence; Mr Keramidas' report referred to a dragging effect, but from deep soft material. 216 I do not think this dispute is of particular significance. It must be remembered that the distance in question is 33.3 metres, and any arcing over (say) 20 metres at 90 kilometres per hour would occur in about three-quarters of a second. Further, by the time of the arcing the respondent's vehicle was already partly off the running track, and the arcing did not shed light on why that came about. In the state in which the evidence was left, arcing is as consistent with the near side wheels moving onto the shoulder through inattention as it is with being misled into allowing them to move onto the shoulder; more correctly, it does not materially support the latter explanation. 217 The question of the accident occurring part way around the curve took the submissions beyond arcing. As I have said, the appellants suggested failing to take the curve through inattention. The respondent submitted that taking the curve at the Yellow Corner at least until the point she went on to the western shoulder suggested that she was attentive, and so the probability was that she was misled by the nature of the road surface. 218 The trial judge did not consider whether failing to take the curve would have caused the respondent to go off the road at an earlier time, or whether the fact that she did not go off the road at an earlier time suggested attentiveness material to why the near side wheels of her vehicle came to be on the shoulder. Nor did the evidence directly address the first of these matters. We were provided by way of submission with a diagram of the curve and the path of the respondent's vehicle, from which it would not be safe to draw any conclusion, but not with any calculations which perhaps could have been made from the evidence. 219 On the skimpy indications in the evidence, I think it probable that the respondent's accident occurred sufficiently far around the curve of the Yellow Corner that, if she had not taken the curve at all, she would have gone off the road at an earlier time. Mr Stuart-Smith said that the deviation of the tyre mark to the right was about 180 metres after the commencement of the curve, and with full appreciation of the danger of relying on photographs for judgments of distance it seems to me that Exhibits 3 and 33 show that it was sufficiently far around the curve. To this may be added the evidence of Mr Routledge that the respondent's vehicle disappeared around the corner, and perhaps Mr Smith's evidence although in the statement to the police he said he thought he saw the car go off the road. 220 However, even if that be so I do not think sufficient attentiveness to begin a curve sweeping over more than half a kilometre translates to attentiveness while going around the curve. Something happened part way around the curve, and I do not see particular significance in successful entry into the curve. The dispute over arcing or straightness and the attendant question of the accident occurring part way around the curve do not in my view materially contribute to determining why the near side wheel of the respondent's vehicle came to be on the shoulder. (b) Speed at the time of the accident 221 It will be recalled that the trial judge found at [267] that the respondent experienced a different "feel" in the passage of her vehicle and its steering when the near-side wheels were on the softer shoulder and "[i]t was this different feel which caused her to feel unsafe and to slacken off her speed". The trial judge thus attributed the respondent's evidence that she felt unsafe and reduced her speed to 80 to 90 kilometres per hour to the moments immediately before the accident when her vehicle travelled the 33.3 metres, prior to the sharp deviation to the right which caused loss of control. 222 Although not so stated at that point in the reasons, it seems that the trial judge regarded this as a matter favouring being misled over inattention. I so infer from what his Honour said when considering the respondent's primary case - "222 She could have had that feeling of being unsafe and a resulting need to slow down because she realised that she was on a collision course with the guidepost. 223 Her Senior Counsel argued that her own words, slowing down because she didn't feel safe, militate very strongly against a lack of attention. If she had said 'I was listening to the radio, I was thinking about the movie I had seen last night, I probably wasn't paying much attention', but in a very poignant way the plaintiff has said 'I can't tell you now why, I know the road was wet, I can't tell you now why, but I actually slowed down'. Her evidence was that she normally drove, like everybody else, at a hundred or so, 'for some reason I didn't feel safe, I slowed down to 80 or 90'. That single piece of evidence militates very strongly against the momentary lack of attention hypothesis, particularly since it is strongly to be inferred that it was in the very process of starting to drift out of control that the plaintiff felt unsafe." 224. This argument would have some force if there were acceptable evidence that there was in fact an adverse condition of the central compacted part of the road surface which was capable of causing or contributing to her vehicle's drift onto the shoulder and then out of control.' 223 The trial judge did not accept an adverse condition of the central part of the road surface. But he found an adverse condition of the road surface, its deceptive nature, and so is likely still to have seen force in the argument that feeling unsafe and reducing speed militated against inattention. 224 The respondent submitted on appeal that feeling unsafe and reducing speed supported that she was driving attentively at the time, because she reacted to the feeling. The appellants submitted that the trial judge erred in attributing feeling unsafe and reducing speed to the last moments before the accident. They submitted that the correct finding on the evidence was that feeling unsafe and reducing speed came at the time the respondent overtook the road train driven by Mr Routledge, and (as I have earlier noted) that it should be found that the respondent thereafter increased her speed to 100 kilometres per hour . 225 I do not think it greatly matters which of these views is correct. A reaction to feeling unsafe when the near side wheels went onto the shoulder would not really signify attentiveness. It could be that feeling unsafe was what brought attention back to where the vehicle was being driven; put another way, by the time the different "feel" from the softer shoulder came to the respondent, inattentiveness might have led to the drift onto the shoulder. Equally, travelling at 100 kilometres per hour at the time of the accident, after an earlier short reduction in speed when feeling unsafe, could stand with being misled by the nature of the road surface. 226 I should nonetheless deal with the issue. I respectfully differ from the trial judge in his attribution of feeling unsafe and reducing speed to when the near side wheels of the respondent's vehicle were on the softer shoulder. 227 I go first to the respondent's evidence. In her evidence in chief she said - "Q. Do you recall what happened with your speed after you had finished overtaking. A. I remember lowering my speed as I was feeling unsafe for whatever reason I don't recall. I just recall feeling unsafe and dropping my speed back down to between 80 and 90. BARTLEY: Q. And after that what's the next thing that you can remember? A. Hitting my head on the window and then hanging upside down in the car." And in similar terms - "Q. Just one last matter. I take your mind back to the day of the accident. You have described to his Honour what you can recollect about your driving on that day, that you can remember passing this truck, tucking back into the road, your speed dropping to about 80 or 90. The next thing you can remember about the journey, I think, is hitting your head on the windscreen? A. Yes." 228 In cross-examination the respondent said that she had a memory of looking in her rear vision mirror after she overtook the road train, which led to the evidence - "BARRY: Q. Let us go back. You have a recollection of overtaking the truck? A. Yes. Q. You have a recollection of coming back to the left-hand side of the road? A. Yes. Q. You have a recollection of decreasing your speed from 100 to 110 back to, what did you say? Was it 80 to 90? A. Yes." 229 Further in cross-examination, the respondent was asked to clarify "the true position as to what you remember when you overtook the truck". She gave the evidence - "A. That I pulled in safely. I thought the road was wet, in patches. I hit my head on the window. I felt unsafe but I don't know if it was immediately after the truck or some time before it. Q. At line 26 page 23 you were asked this question, 'Do you recall what happened with your speed after you had finished overtaking?' And your answer was, 'I remember lowering my speed as I was feeling unsafe, for whatever reason I don't recall, I just recall feeling unsafe and dropping my speed back down to between 80 and 90'? A. Yes. Q. Was that answer true? A. Yes. Q. Did you look at your speedometer or was that an estimate. A. No, I felt, I took my foot off the accelerator and felt myself slow down. Q. But my question was, did you look at the speedometer or was the 80 to 90 an estimate as to the speed that you had reduced the car's travel to? A. I would have briefly had a glance at it. Q. When someone says 'would have', that usually means they are speculating. I want to know whether you actually have a recollection, one way or another, of actually looking at the speed and observing the speed you were actually doing? A. Yes. Q. And you looked at the speedometer? A. Yes. Q. And you observed the speed? A. Yes. Q. What speed were you doing after you had overtaken the truck and come back on to the left-hand side of the road? A. Immediately after overtaking the truck? I don't recall that. Q. I am trying to work out when it was that you made the observation that you were doing between 80 and 90 kilometres per hour. That is when you made the observation of the speedometer that you were doing between 80 and 90 kilometres an hour? A. It was after I had overtaken the truck and I don't recall exactly when that was. Q. And do you recall whether, after you had looked at the speedometer, do you have any recollection of whether you had increased your speed thereafter, reduced it thereafter? Do you have any recollection from that time on until when you recall hitting your head? A. I had reduced my speed. Q. I don't want to be at cross-purposes. I am asking you about the period of time when you said you observed you were doing 80 to 90 kilometres per hour after you had overtaken the truck. My question was from that time until you have the recollection of hitting your head, do you remember anything about the speed you were then doing? A. I was doing under the speed limit. Q. But is that an answer based on what your usual practice was or based upon what you can remember now of this morning? A. Of that morning and yeah, and usual practice. I don't speed. Q. How are you able to recall the speed that you were doing after you reduced to 80 or 90 kilometres per hour if you didn't remember anything from that point on until you remember hitting your head? A. I can't explain it, I am sorry. That is just one definite memory. Q. And a memory of reducing your speed to between 80 to 90 kilometres per hour is a vivid memory? A. Yes. Q. And the memory of looking at y our speedo and observing that you had reduced your speed to 80 to 90 kilometres per hour is a vivid memory. A. Yes." 230 The respondent was later taken though an affidavit she had earlier sworn, and amongst other matters said that the paragraph was true which said - "Once I overtook the truck, I decreased my speed to about 80 to 90 kilometres. I can recall that I did this because I didn't feel safe on the road for some reason. I can't recall why. However, I just remember feeling that way." 231 The respondent later said that just before her vehicle left the road she was "doing between 80 to 90", but when challenged that she did not remember she said that she knew she "wasn't doing more than a hundred". 232 It seems to me that the better understanding of the respondent's evidence of her recollection of feeling unsafe and reducing speed is that it closely followed overtaking the road train. While the respondent said that she did not know "if it was immediately after the truck" and did not recall "exactly when that was", the flow of her recollection was that the reduction in speed was immediately after pulling in after overtaking the road train. She said that thereafter she was doing under the speed limit, suggesting (but no more) that there was a period of time before the accident occurred. 233 I accept that the respondent's evidence of her recollection of feeling unsafe and reducing speed could be otherwise understood, although I will later refer to the significance of her further recollection of feeling herself slow down and looking at the speedometer. I go to the evidence of Messrs Smith and Routledge. 234 Mr Smith was doing 100 kilometres per hour and was not gaining on the respondent. Mr Routledge was doing 80 kilometres per hour and the respondent went out of his sight. Their evidence in these respects, or at least Mr Smith's evidence, is not consistent with the respondent reducing her speed to 80 to 90 kilometres per hour and maintaining that speed for the three minutes or so until the accident. However, it is consistent with the respondent briefly reducing her speed after overtaking the road train and then returning to driving at about 100 kilometres per hour. The respondent's evidence could not extend to whether the reduced speed was maintained from reduction to the accident. 235 I turn to the expert evidence of Mr Stuart-Smith and Mr William Keramidas. The appellants and the respondent each relied on it, the appellants for a speed of about 100 kilometres per hour at the time of the accident and the respondent for a speed of about 80 kilometres per hour, with consequences for when the respondent felt unsafe and reduced speed. 236 In analyses of the yaw mark across the road in reports prior to the hearing each expert concluded that the respondent's speed was about 100 kilometres per hour. The analyses turned on the radius of the circular path taken by the respondent's vehicle, a friction co-efficient for the surface of the road and a level road. Mr Stuart-Smith took a friction co-efficient for "a typical hard gravel road surface" of 0.6 to 0.65, and calculated a "probable range" of 95 to 105 kilometres per hour. Mr Keramidas took a friction co-efficient, which he also called a skid resistance factor, of about 0.6, and calculated a "likely speed" of about 101 kilometres per hour. 237 In his oral evidence Mr Stuart-Smith said that his typical hard gravel road surface was one which "doesn't have an excessive amount of loose material on it", explaining that a fine layer of loose material would be pushed out of the way and the grip was the same as on a hard surface; he agreed that it was a matter of degree. His assumptions included that the road was in good condition and without "excess loose material" on the running surface, and that the surface was firm with "no excess moisture". He was taken to Exhibit 3, but was not asked by counsel for the respondent whether what it showed affected his assumptions. In answer to counsel for the appellants he said that (with a presently immaterial qualification) what he saw was consistent with his assumptions, and that the yaw marks were consistent with his friction co-efficient of 0.6 to 0.65 "because [the road surface] appears to be very hard". 238 In his oral evidence Mr Keramidas was taken to Exhibits 3, 4 and 5 by counsel for the respondent, and was asked whether they modified his opinion on the friction co-efficient. He said they caused him to reduce it to about 0.4, and if the road was damp or moist even further. He saw lighter material in the photographs which appeared to him to be sub-base, of a tan or yellowy colour like clay. (In passing, he therefore did not see complete uniformity in colour.) He also saw corrugations. He saw the same clay in Exhibit L and in photographs in Exhibits 32 and 33. He concluded that the gravel base had been graded off over the years, and that the sub-base material did not have the same friction co-efficient. In cross-examination he explained that his original opinion had assumed that the road surface was "full and dry", by full meaning in thickness, and that "under the conditions as I see them in the photographs now [the vehicle's speed] would be marginally over 80 kilometres per hour". He said that, depending on the photographs, an expert opinion could confidently be given on what is shown in photographs, and that he was "happy that I have enough photograph material from many different angles and over different time periods to give the answers that I have given today." 239 The trial judge did not make findings, or address this evidence at all. Mr John Jellick, who would have known, did not agree that there was clay at the Yellow Corner: he said it was ironstone. Mr Davis described construction and maintenance of the road with gravel, and said that there should not have been clay at the Yellow Corner although there was "a little bit of talc type material" with the ironstone, which compacted well and did not get slippery when wet. Mr Comiskey did agree that he saw a yellow clay substratum in Exhibit K, but it was put to him by counsel for the respondent that he could not see it in Exhibit 3 and he again agreed; he said Exhibit 3 showed "the natural road surface" and the road must have been graded between the photographs. Contrary to Mr Comiskey so far as he saw clay in Exhibit K, Mr Barry Jellick said that the "white material" in Exhibits K and L was "the road base, that's what the gravel is underneath the road". I have great difficulty in regarding Mr Keramidas' revised opinion as soundly based on his interpretation of the photographs. 240 But there is also uncertainty in the assumptions for Mr Stuart-Smith's opinion with which he thought what he saw in Exhibit 3 was consistent. How good is good condition? How much loose material is excess loose material? When is moisture excessive? In that connection, the evidence did not explore whether Mr Keramidas' "dry" allowed for non-excessive moisture, and there were a number of references in the evidence to moistness, dampness or wetness of the road and similar rather indeterminate descriptions difficult to relate to either expert's assumptions. 241 Both experts had to work with assumptions plus what they thought they saw in photographs, but the room for legitimate divergence was considerable. It seems to me that the evidence as a whole tends towards a friction co-efficient in the order of that taken by Mr Stuart-Smith, and it should not be forgotten that the trial judge found a central compacted surface of the road which was not slippery and did not have loose gravel of sufficient size to have a marbles effect. But the speeds calculated by the experts can not have their apparent mathematical certainty, and the range (from 80 kilometres per hour to 105 kilometres per hour) is so affected by assumptions which can not be precisely established that, while I do not think Mr Keramidas' revised speed should be found, I do not think a speed of about 100 kilometres per hour can reliably be found. In the result, the expert evidence should be regarded as neutral in relation to speed at the time of the accident. 242 There remains a matter which in my view is of significance. The respondent recollected feeling herself slow down and looking at the speedometer. The reduction in speed was by taking her foot off the accelerator. She observed the speed to which she had slowed down. Her vehicle travelled 33.3 metres with the near-side wheels on the shoulder, before the sharp deviation to the right. Assuming for calculation purposes a speed of 90 kilometres per hour, there was about 1.2 seconds for the feeling to register with the respondent, for her to take her foot off the accelerator and slow down, and for her to glance at the speedometer and observe the speed to which she had slowed down; and then she must have reacted to the guidepost 10 to 12 metres or less in front of her at the time she steered sharply to the right. Even if, as the respondent suggested in argument, there was a transition time and some forewarning as the near side wheels began to go onto the shoulder, it is highly unlikely that all this could have occurred in the available time. There had to be an element of reaction time to the feeling, then time for slowing by 10 to 20 kilometres per hour without application of the brakes, then further reaction time. All this must have taken more than a few seconds. 243 In my view, this favours the initial understanding of the respondent's recollection of feeling unsafe and reducing speed, and is against feeling unsafe and reducing speed when the near side wheels went onto the softer shoulder. Having regard to all the matters to which I have referred, in my opinion on the probabilities the feeling unsafe and reducing speed was immediately after overtaking the road train; and, from the evidence Messrs Routledge and Smith, on the probabilities the respondent thereafter increased her speed and was driving at about 100 kilometres per hour at the time of the accident. 244 I repeat that I do not think it greatly matters. But so far as the trial judge rested his finding of causation on a reaction to feeling unsafe when the near side wheels went onto the shoulder, I consider that he was in error. 245 Although the respondent did not refer to it, in determining whether the respondent established that she was misled by the deceptive nature of the road surface it should be remembered that "all evidence is to be weighed according to the proof which it was in the power of one side to have produced, and in the power of the other to have contradicted": Blatch v Archer (1774) 1 Cowp 63 at 65; 98 ER 969 at 970. The respondent was not able to give direct evidence of how the accident came about. While recognising that the appellants were in a like position, the Court should be prepared to draw inferences in the respondent's favour if the circumstantial evidence warrants them. 246 But it remains that the Court must be "satisfied that the case has been proved on the balance of probabilities" (Evidence Act 1955, s 140(1)). The relevant principles were summarised by Allsop P, Basten JA and Grove J agreeing, in Jackson v Lithgow City Council [2008] NSWCA 312 - " 9 Before dealing with the facts it is helpful to remind oneself of the terms of the task at hand. In Holloway v McFeeters [1956] HCA 25; 94 CLR 470 at 480-481, Williams, Webb and Taylor JJ cited and applied passages from the joint judgment of Dixon J (as he then was), Williams, Webb, Fullagar and Kitto JJ in Bradshaw v McEwans Pty Limited (unreported, High Court of Australia, 27 April 1951), which were as follows: ' … you need only circumstances raising a more probable inference in favour of what is alleged … where direct proof is not available it is enough if the circumstances appearing in evidence give rise to a reasonable and definite inference; they must do more than give rise to conflicting inferences of equal degree of probability so that the choice between them is [a] mere matter of conjecture … . All that is necessary is that according to the course of common experience the more probable inference from the circumstances that sufficiently appear by evidence or admission, left unexplained, should be that the injury arose from the defendant's negligence. By more probable is meant no more than on the balance of probabilities such an inference might reasonably be considered to have some greater degree of likelihood.' 10 In Jones v Dunkel [1959] HCA 8; 101 CLR 298, the High Court discussed Holloway v McFeeters in an appeal concerning a truck accident on a hill. The proof of the accident was circumstantial. The cause had been heard before a jury. The High Court divided on the question whether the evidence permitted the jury to conclude that the defendant driver had been negligent and caused the accident. (Dixon CJ and Taylor J thought not; Kitto, Menzies and Windeyer JJ thought it could.) Dixon CJ, at 101 CLR 305, having referred to Holloway v McFeeters and Bradshaw v McEwans said the following, referring to the passage from Bradshaw v McEwans : 'But the law which this passage attempts to explain does not authorise a court to choose between guesses, where the possibilities are not unlimited, on the ground that one guess seems more likely than another or the others. The facts proved must form a reasonable basis for a definite conclusion affirmatively drawn of the truth of which the tribunal of fact may reasonably be satisfied.' Though Dixon CJ was in dissent, this passage can be taken to reflect the essential content of Holloway v McFeeters , and of the views of the justices in Jones v Dunkel . See also Flounders v Millar [2007] NSWCA 238 at [4]-[35] (Ipp JA, with whom Handley AJA agreed) and in particular [32]-[35]. 11 On the basis of these authorities, Mr Morrison SC, leading Mr Elliott on behalf of the appellant, submitted that the appellant did not have to prove on the balance of probabilities what occurred, rather he said, ' all the appellant has to do is to show that the appellant's scenario was in the smallest degree more likely than either of the other eventualities.' 12 With respect, that is to misapprehend these authorities. It is to say no more than the "appellant's scenario" is the most likely guess. The inference must be available and be considered to be more probable than other possibilities. To put the matter as Mr Morrison did is to commit an error of the kind to be found in the reasoning of the trial judge in Rhesa Shipping Co SA v Edmunds (The 'Popi M') [1983] 2 Lloyd's Rep 235. … " 247 The respondent did not submit to the effect that breach of duty in permitting a deceptive road surface, followed by an accident within the area of foreseeable risk thereby created, was sufficient for causation. On the analyses of High Court authority undertaken in this Court in Flounders v Mitlar [2007] NSWCA 238; (2007) 49 MVR 53 (and see subsequently Roads and Traffic Authority v Royal [2008] HCA 19; (2008) 245 ALR 653 at [31], Baulderstone Hornibrook Engineering Pty Ltd [2008] NSWCA 243 at [240]; Gett v Tabet [2009] NSWCA 76 at [250]-[256]; and Warren v Gittoes at [45]-[52]), it is still necessary for the respondent to establish causation on the facts. The respondent relied on March v E & M H Stramare Pty Ltd (1991) 171 CLR 506, referring in particular to a passage in the judgment of Deane J at 521, for a submission to the effect that any negligence on her part did not preclude recovery if the appellants' negligence had created the risk of the accident occurring. However, as was pointed out in Roads and Traffic Authority v Royal at [32] in relation to a similar argument from March v E & M H Stramare Pty Ltd, it must first be established that the appellants' negligence (at this point assumed) has been causative. 248 To return to the facts, that there was no evidence to suggest that the respondent's manner of driving prior to the accident was erratic or abnormal did not enable an inference that she was misled. As I have said, that the respondent was driving appropriately prior to the accident proved little about what caused the accident. Other matters must be considered. Much more than uniformity in colour of the road went to how the respondent's vehicle came to have the near-side wheels on the shoulder. 249 The respondent had driven over the road in the order of 300 times. In an affidavit she said that she was quite familiar with the road, having travelled the route two or three times per day. She described it as made out of hard dirt which had been well compacted, although it had "soft patches", and as a typical dirt road. While she referred to difficulty in seeing "where the compacted road ended and where the soft verge started", the verges seem to have been the rills or at least the very edges of the road, and she was aware of their softness - "With respect to the dirt verges on the side of the road, these would often become raised when the grader pushed all the soft, loose dirt to the sides. It would build up and in places the verge would become raised like a ramp. It was also difficult to see where the compacted road ended and where the soft verge started." 250 The respondent must have experienced the road in all conditions, from recently graded to a week or more since grading, and the Yellow Corner was no more or less affected by coal and coal dust from the road trains than other parts of the road. She must have experienced the hard compacted surface of the running track, free from loose material, and must have come to know that, depending on the recency of grading, there would be loose material on the shoulders and a difference in that respect between the running track and the shoulders. The rill provided a prominent border to the carriageway and a guide to the lateral location of the respondent's vehicle. There were guideposts, including the guidepost towards which the respondent's vehicle was heading at the time the near-side wheels went onto the shoulder. Their principal purpose was not to tell the course of the road in daylight, but they did so. 251 The lastmentioned guidepost is of some importance. At the time the near side wheels of the respondent's vehicle went onto the shoulder she was about 45 metres or less from the guidepost, travelling at at least 80 kilometres per hour: a little over 1.5 seconds until hitting the guidepost. The guidepost would have told the respondent that her vehicle should not be heading where it was. That she was so close to it when the near side wheels went onto the shoulder, and had to steer sharply to the right, is a reasonable indication that coming to have the near side wheels on the shoulder was not because of the appearance of the road surface. 252 When there were these guides to the respondent's driving, but the nearside wheels of her vehicle went onto the shoulder and to within about 75 centremetres of the rill so close to the guidepost and heading for it, it is difficult to accept that she was misled in her driving by uniformity in colour across the road. The uniformity in colour, also, was not complete, and there remained what I have called the different texture from the loose material on the shoulder, which must have been within the respondent's experience from driving on the road. In my opinion, that the respondent was misled by the nature of the road surface is not a conclusion which can affirmatively be reached, as an inference more probable than other possibilities. Again respectfully differing from the trial judge, I am not satisfied that her case in this respect has been proved on the balance of probabilities. Departing from the trial judge's findings 253 I do not overlook the respondent's submission that the trial judge had the advantage of hearing and seeing the many witnesses, and arrived at a "synthesis" of competing evidence which must have involved regard amongst other things to their demeanour and his assessment of their credibility. In the respondent's submission, the appellate restraint considered in cases such as Fox v Percy [2003] HCA 22; (2003) 214 CLR 336 and CSR Ltd v Della Maddalena [2006] HCA 1; (2006) 224 ALR 1 applied, such that this Court should depart from the trial judge's conclusions as to soft shoulder, uniformity in colour "and so on" (it was left rather open-ended) only if there was no evidence to support them, they were contrary to incontrovertible evidence or the result was glaringly improbable. 254 The appeal to this Court is by way of rehearing (Supreme Court Act 1970, s 75A(5)), and subject to the aforementioned appellate restraint should be "a real review of the trial … and of [the trial judge's] reasons" (Fox v Percy at [25] per Gleeson CJ and Gummow and Kirby J). Save in relation to Mr Wilson, the trial judge's reasons did not express views on the reliability or credibility of witnesses. Nor were his Honour's conclusions evidently founded on preference for one witness over another or others; there was a deal of evidence with shades of consistency and inconsistency, and in the "synthesis" the photographs, equally able to be understood by this Court, were important. Further, causation in particular was scarcely a matter of preference for one witness over another or others. While (as I have accepted) respect should be afforded to the trial judge's conclusions, I do not think that the respondent's submission should be accepted. 255 There is further reason to decline to accept the submission. The trial judge came to his findings and conclusions on a limited consideration of the evidence in the course of his judgment. I have sought to explain, in relation to uniformity of colour, that the evidence of Mr O'Loughlin to which the trial judge referred was qualified, and was not the whole of his evidence on the subject, and that there was much other evidence to which his Honour did not refer. In relation to ability to distinguish between the hard compacted surface and the relatively softer shoulder, I have sought to explain that the evidence to which the trial judge referred did not readily support the trial judge's finding, and again there was other evidence going to the matter. No judgment can take up every detail of the evidence, but there was much relevant evidence left without mention. The judge's conclusions were in these respects flawed, and for that reason did not command acceptance unless contrary to incontrovertible evidence or glaringly improbable. 256 We have had the benefit of extensive attention to the evidence through counsel's written and oral submissions, and if in our real review conclusions are reached differing from those of the trial judge they should be given effect. If error on the part of the trial judge needs to be found (see the discussion in Costa v Public Trustee [2008] NSWCA 223), it may sufficiently be found in what I have said in the preceding paragraph. In my opinion, it is open to this Court to depart from the trial judge's findings in the manner appearing in these reasons. The result 257 Since the appearance of the road surface was not relevantly deceptive, there was no relevant breach of duty by the appellants in failing by further grading to eliminate uniformity of colour across the road. Nor was any such breach of duty shown to be causative of the accident. The appeal should be upheld. 258 No reason was proffered against the ordinary order that the respondent pay the appellants' costs of the re-hearing of the appeal. The High Court ordered that the costs of the previous appeal to this Court be in the discretion of the Court on the rehearing. Subject to any further submissions, the respondent should also pay those costs. 259 The success of the appeal can entitle the respondent to a certificate under the Suitors Fund Act 1951 in respect of the rehearing of the appeal. Her entitlement in respect of the previous appeal and the appeal to the High Court is less clear. 260 The respondent should have the opportunity to file written submissions within 14 days in relation to the costs of the previous appeal and a certificate under the Suitors Fund Act in respect of the appeal to the High Court. If she takes up the opportunity, the appellants should have a further 7 days to file responsive written submissions, and the matter will be decided on the written submissions. 261 I propose the orders - 1. Appeal allowed. 2. Set aside the judgment for the respondent and in lieu thereof judgment for the appellants. 3. Respondent to pay the appellants' costs of the trial and of the rehearing of the appeal. 4. Respondent to have a certificate under the Suitors Fund Act if otherwise qualified in respect of the rehearing of the appeal. 5. Subject to order 6, respondent to pay the appellants' costs of the previous appeal to this Court. 6. Liberty to the respondent to file written submissions within 14 days in relation to the costs of the previous appeal and a certificate under the Suitors Fund Act in respect of the appeal to the High Court; appellants to file responsive written submissions within a further 7 days. ****** DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.