BARWICK C.J. The appellant, who is the manufacturer of an adhesive known as respondent in the Supreme Court of New South Wales for damage to the respondent's premises said to have been caused by the appellant's negligence in failing to give a purchaser of the adhesive adequate warning of its dangerous quality. The action was heard by a judge sitting without a jury pursuant to s. 3 of the Supreme Court Procedure Act, 1900-1965 (N.S.W.). A verdict was found for the respondent and judgment directed to be entered for $162,518, that being an agreed amount (Streets Ice Cream Pty. Ltd. v. Australian Asbestos Installations Pty. Ltd. (1967) 1 NSWR 50 ). This judgment was affirmed on appeal by the Court of Appeal Division of the Supreme Court (Streets Ice Cream Pty. Ltd. v. Norton Australia Pty. Ltd. (1967) 86 WN (Pt 1) (NSW) 345 ): see s. 5 (6), (7) and (9) of the Supreme Court Procedure Act. The appellant now seeks to set aside the verdict and judgment for the respondent and to obtain a direction that judgment be entered for it in the action. (at p637)

2. The respondent had let a contract to Australian Asbestos Insulations Pty. Ltd. (the contractor), a company engaged in the installation of insulating material, to insulate a cool-room on premises being constructed for the appellant at Turella. The cool-room was some one hundred and twenty feet long by eighty-four feet wide with a ceiling twenty feet high. Its walls were of rendered concrete and its ceiling of tongued and grooved timber boards. The method employed by the contractor to insulate this room, in performance of its contract with the respondent, was as follows: the walls were first treated with some rust preventative; then hessian was fixed to the walls with heated bitumen; thereafter three layers of cork in slabs were fixed to the hessian also by means of heated bitumen. This procedure was found inappropriate for the ceiling. Instead of hessian, aluminium foil was used and to affix it to the boards an adhesive was used. When the foil had been fixed to the ceiling, cork in slabs was applied to the foil partly by means of bitumen and partly by means of screws or strips of galvanized iron. (at p637)

3. At the time of the incident giving rise to this action, the ceiling was being treated. In the cool-room a scaffold had been erected which at the relevant time was some fourteen or fifteen feet from the floor of the room and extended across half its width and for forty feet of its length. The scaffold floor was then about six feet or so from the ceiling and enabled workmen to apply the foil and thereafter the cork to the ceiling. The surface of the scaffold was about forty feet by forty feet. On it the contractor had placed two trays containing heated bitumen. Under one of these trays at the time in question there was a lighted gas burner, fed with Portagas, with a naked flame some seven inches long playing on the under surface of the tray about its middle to keep the bitumen at the necessary temperature of about 180 degrees. Apertures were provided at each end of the metal frame holding the tray to enable the burner to be inserted beneath the tray and to allow air to reach the burner. The action of the burner was to draw air from the room to itself and, of course, to raise the temperature of the room. The cork slabs were dipped in the tray of hot bitumen before their application to the foil on the ceiling. (at p637)

4. At the same time foil was being prepared for application to the, as yet, uncovered parts of the ceiling. A quantity of the adhesive had been poured out of a five-gallon drum which was on the scaffold into a shallow tray which was also upon the scaffold. This five-gallon drum was one in which the adhesive had been delivered by the appellant to the contractor and bore the warning and directions for use to which I later refer. To prepare the foil which had been cut into suitable lengths, workmen dipped paint rollers into the shallow tray and spread the adhesive thinly on the surface of the foil, each piece of which represented about one hundred and forty square feet of surface. After application of the adhesive, a period of two or three minutes was allowed to elapse before placing the foil in position on the ceiling in order that the adhesive should become "gummy" or "tacky", a condition obviously brought about by the evaporation of the solvent in the adhesive. According to the finding of the learned trial judge a distance of some twenty feet separated this operation with the adhesive from the lighted burner under the tray of bitumen. On the day of the accident giving rise to this claim, these procedures had been in progress for some time. The cool room had only two relatively small apertures and "there was virtually no movement of air" passing through it. In consequence, the atmosphere within the room, again to quote the trial judge, had "become hot, oppressive and fume laden". The scaffold, which necessarily had been moved as the work progressed, was at the relevant time at one end of the room. At this stage the walls had been treated and only one sheet of foil remained to be placed in position on the ceiling. About ten a.m. as the sheet of foil to which the adhesive had been applied was about to be raised to the ceiling, it burst into flames, or, as it is said, "exploded in flames" whereupon fire spread throughout the whole area above the scaffold and great damage was caused to the respondent's premises. "There is no doubt that the fire was caused by vapour from the BM 226 coming into contact with the flame of the Portagas burner": so found the trial judge. It is also clear that it was the volatile quality of the adhesive which gave rise to the vapour or fumes, and its inflammable quality to their ignition by the naked flame. (at p638)

5. The respondent, in declaring in the action against the appellant, alleged that the appellant "manufactured and distributed for sale a certain dangerous and highly inflammable substance for use in connexion with the installation of insulating material" and laid the breach of the appellant's duty to the respondent in that the appellant "so negligently and unskilfully conducted itself in and about the failure to properly and adequately warn of the dangerous and highly inflammable nature of" the material that it "caused a fire to occur in the premises of the" respondent. (at p638)

6. It was proved and the trial judge found that the appellant had supplied the adhesive to the contractor in five-gallon drums carrying labels, which beside giving directions for the use of the adhesive, bore the warning: "CAUTION: HIGHLY INFLAMMABLE" and the direction "For further information ask for data sheet". This sheet the contractor had procured. It contained the following relevant information:

"Solvent: Petroleum - Benzene Free.

Inflammability: Highly inflammable, flash point below

73 degrees F.

Odour: Odourless when dry. When used in

refrigeration applications passes

butter test, providing all solvent

fumes have been exhausted.

Application: Stir adhesive thoroughly. Surfaces must

be clean and dry. Apply adhesive to

both surfaces using brush, spray or

roller-coater. Bond whilst adhesive

is tacky and ensure intimate contact

by firmly pressing surfaces together.

One surface only may be coated if high

wet strength is not essential, and

where at least one surface is porous." (at p639)

7. Though because of other circumstances nothing need turn upon it, the evidence disclosed that the contractor was itself a distributor of the appellant's adhesive. The contract superintendent employed by the contractor was experienced in insulating work of the kind contracted to be done at the respondent's premises and was in fact in charge of the work there on this occasion. There can be no question if the fact be relevant, that this employee knew that the adhesive was highly inflammable, that it contained a solvent which smelt like petrol and which evaporated and produced fumes. He knew that because of "the fumes of the material" smoking could not be allowed in the cool-room whilst the adhesive was in use and that a naked flame should not be brought near the adhesive. He said in substance that whilst he knew that a flame should not be brought "near or approximate to" a highly inflammable material, in this context "approximate" meant to him "round about three or four feet". It is not clear to me from the judge's reasons whether or not he believed this piece of evidence but I shall assume that this employee of the contractor did in fact so believe. Somewhat astonishingly, however, he saw no danger in carrying on the two operations of preparing and applying the foil and the cork at the same time over a substantial period of time, in what to all intents and purposes was an enclosed space. Clearly, the nature of the insulating operation on which the contractor was engaged did not require as a matter of necessity that these operations be carried on contemporaneously. Indeed, the application of the foil to the ceiling and of the cork to the surface of the foil were necessarily successive operations; in addition, there would seem to have been some advantage in allowing some interval between them in which the adhesive might harden before the cork sheets were applied. (at p640)

8. Scientific evidence was given at the trial, which the judge fully accepted, that the flash point of the adhesive was "in the vicinity of minus 76 degrees or minus 78 degrees F., compared with the flash point of standard petrol of minus 60 degrees F". The petroleum solvent which was disclosed in the data sheets was in fact Shell X.60. It was also established to the judge's satisfaction that the adhesive evaporated at a fast rate, losing thirty per cent to sixty per cent of its volume by weight in a minute, such evaporation being made easier by spreading it thinly on a surface in the manner employed in the contractor's operations. A retired university lecturer and demonstrator in organic chemistry and in chemical engineering, having the degree of Bachelor of Science, who had had a "wide experience with inflammable liquids" said that the adhesive was one of the most dangerous materials he had handled. But whether or not this evidence was related to its use in a confined space or not is to my mind not very clear. This assertion, however, the judge accepted. The scientific evidence established that the solvent in the adhesive was a very heavy solvent, the vapour of which would tend to go downwards and to create in an enclosed space a sea of vapour. However, it was said in substance that its use as a dangerous material did not create a very great hazard in the presence of even a slow current of air but if used "in a very confined space with practically no air currents and possibly an ambient temperature that is fairly high, then the hazard can be very high". (at p640)

9. The evidence also established that the only suitable solvent for an adhesive of the kind in use in the instant case is one of the petroleum ethers. Such a solvent is hazardous to use in a warm and stagnant atmosphere, but of course only then if a source of ignition is present. The solvent in the adhesive was said to be more dangerous than petrol in an enclosed space, particularly if the adhesive is used in the particular way in which the contractor was using it in the cool-room. (at p640)

10. The meaning of highly inflammable in relation to a substance, such as the adhesive was said to be, is that the substance gives off sufficient vapour to form an explosive mixture at temperatures which will ordinarily be encountered; 73 degrees F. is accepted in industry as indicating an ordinary ambient temperature. It was stated in evidence that no classification is used in industry higher than "highly inflammable" in connexion with inflammable substances. (at p641)

11. Lastly, it was said that the two major precautions to be taken in handling solvents such as that in the adhesive were (1) adequate ventilation and (2) the absence of any source of ignition in the potential area of vapour. (at p641)

12. The crux of the trial judge's reasons for finding a verdict for the respondent is to be found in a short passage in his reasons for judgment. It is convenient to quote it:

"Having regard to the inherently dangerous qualities of this

adhesive, particularly to its extremely low flashpoint, to its

capacity for very rapid vaporization and dissipation and to the

enhanced dangers of using it where there is no real circulation of

air, I have come to the conclusion that the notification on the drum

label and the information contained on the data sheet did not amount

to a fulfilment of the third defendant's duty of care and were not

an adequate warning or notice. It is not for me to say what further

information or warning or advice should have been given and I

appreciate that problems might arise in determining what would be

the most suitable language to use. However, I think that some

information should have been conveyed to the purchaser as to the

rapid vaporizing features of the adhesive, as to the desirability of

maintaining an adequate draft of air current and possibly as to the

danger of using it in close proximity to flames, sparks or lamps. I

am of opinion that the failure to give adequate warning or notice of

the inherent dangers in the adhesive materially contributed to the

fire which broke out and to the loss which the plaintiff company

suffered. The statement that the adhesive was highly inflammable

was seen by Townsend and other employees mentioned in the evidence

but a more informative warning might well, as a matter of

probability, have been seen by them or other representatives of the

first-named defendant and have deterred them from using the adhesive

in the manner and under the conditions obtaining on the morning of

12th April 1962. As it is agreed between the parties that the

damage suffered by the plaintiff company amounts to the sum of

$162,518 there will be a verdict and judgment for it in that sum.

"I may say that my mind has fluctuated to some extent on the

question of what the proper outcome of this litigation should be. I

am not unmindful of the fact that the intimation that the substance

was highly inflammable is in itself a warning of some significance,

and I am not unmindful that the sale to the first-named defendant

was not to a housewife for the use of the substance in her home but

to a commercial undertaking which might be expected to have more

knowledge of such substances than many ordinary citizens. However,

having regard to the high degree of inherent danger in the product

and to the enhanced danger when the substance is used as it would be

contemplated that it would be used, by spreading thinly, and to the

circumstance that it should have been contemplated that it would be

used not in open air but in a room or building, I have come to the

conclusion that the lack of care alleged against the third-named

defendant has been established" (1967) 1 NSWR 50, at pp 52, 53

"The manufacturer and distributor of a substance which is

inherently dangerous owes a duty to warn people to whom it sells its

product of the dangers involved in its use. . . . The nature and

extent of the notice or warning sufficient to fulfil such a duty of

care must be commensurate with the degree of risk involved in the

use of the product" (1967) 1 NSWR, at p 52 (at p642)

13. It was rightly conceded on the hearing of this appeal that the appellant owed a duty of care to persons in the situation of the respondent as persons whom or whose property the purchaser of the adhesive might affect by his use of it. It could properly be said in this case that the use of the adhesive in the insulation of a cool-room was within the contemplation of the appellant and the contractor when the adhesive was sold to the contractor. But in so saying I do not mean that its use in a closed atmosphere in which there was a naked flame was or could possibly have been in such contemplation. To that situation I shall return later. No doubt because of the dangerous quality of the adhesive and the great risk of injury or damage if it were used in ignorance of its qualities, the duty of care resting on the appellant was high but it was not absolute; the appellant was in no sense an insurer. Its duty was yet a duty to take reasonable care in the circumstances which, of course, included the nature of the adhesive, the identity of the purchaser and the range of its use which might fairly be in contemplation when the sale was made. (at p642)

14. The respondent claimed and the appellant conceded, and again, in my opinion, rightly, that the relevant act in performance of the appellant's duty in the present circumstances was the giving to the purchaser of an adequate warning of the dangerous qualities of the adhesive. The respondent, however, wished to treat the contractor's workmen as being within the concept of the purchaser in this concession, but as to that I will say something later. (at p642)

15. The questions therefore which were before the trial judge were whether the warning in fact given by the appellant to the contractor was adequate to discharge the appellant's duty of reasonable care to the respondent and, if not, whether the inadequacy of the warning which was given caused or casually contributed to the damage to the respondent's premises by the fire which originated in the cool-room. (at p643)

16. What was the dangerous quality of the adhesive within these propositions? It was, in my opinion, that it contained a solvent which quickly evaporated in atmospheric temperatures whether in winter or in summer, that is to say, it was volatile and as it evaporated it created fumes which would ignite if a source of ignition were applied to them or to an area in which they were concentrated and was for that reason not merely highly volatile but highly inflammable. (at p643)

17. The trial judge seems to have taken the view that the tendency of the fumes given off by the adhesive as it became tacky to sink below the level of the adhesive or of the material to which it was applied was a separate dangerous quality. In my opinion, the creation of the fumes, which after all were of the nature of petrol fumes, the solvent being no more than a highly volatile form of petrol, was merely the concomitant of the inflammability which itself involved the volatility of the substance. Although there was evidence that the solvent had a high rate of evaporation, no precise comparisons were drawn between that rate and the rate at which, for example, any super grade of petrol or, for that matter, of standard petrol, evaporated. Nor was there any evidence that it was the rate of evaporation which caused the conflagration in this case. Nor was it said that, if the particular rate of evaporation had been known, other steps would have been taken which were not taken because the contractor was misled by the known nature of the substance. Indeed, the contractor committed itself to the concurrent performance of the operations with the bitumen and with the adhesive which I have described; and it was this decision which in the event brought about the fire. Lastly, it must be remarked that it would appear that the burner was above, not below, the level of the surface of the foil to which the adhesive was being applied so that the heaviness of the fumes given off by the solvent would not appear, in any case, to be particularly relevant to the conflagration in the present case. (at p643)

18. Where the dangerous quality of the substance sold can be adequately described so as thereby to communicate to the purchaser the knowledge of the dangers of its use, an adequate description of that dangerous quality is all that, in my opinion, is called for. In such a case, in my opinion, the manufacturer is not requires to instruct the purchaser in the use of the material or as to the safeguards which may be necessary in any of the great variety of circumstances in which the substance may in the ordinary course be used in order that the dangers involving such use may be contained. There may be substances whose dangerous qualities only manifest themselves in particular circumstances in which case it may well be that the warning must relate to those circumstances. It may also be that there are substances where adequate knowledge of the danger in using them can only be communicated by directions for use or even by directions as to particular safeguards to be taken during or in connexion with use either in general or in particular circumstances. Of course, in applying these statements, the identity of the purchaser is particularly significant because who and what he is will affect the question whether the description of the substance or, for that matter, any other statement by way of warning or direction, is adequate to communicate the relevant knowledge. The first question, therefore, in my opinion, in this case was whether or not the dangerous qualities of the adhesive were so capable of description when sold to the contractor and, if so, whether they were adequately described so as to alert it to the existence of those qualities. (at p644)

19. The learned trial judge in the passages from his judgment which I have secondly quoted, in my respectful opinion, in this connexion sufficiently expressed the relevant duty of the appellant. In the passage which I have firstly quoted he concluded that the warning in fact given did not fulfil the appellant's duty towards the respondent. (at p644)

20. But questions immediately arise in connexion with this finding, namely, first is whether the trial judge did in fact use the right criteria in deciding that the appellant was in breach of duty, and the second is whether, if he did, his finding is none the less shown to be wrong. (at p644)

21. Before turning to these questions, it is profitable, I think, to examine what the proved facts in the case amounted to. I have already set out the salient parts of the evidence, including the terms of the warning given to the purchaser. By them the knowledge was conveyed to the purchaser, that is to say, to the contractor, that the adhesive was highly inflammable, that its solvent was petroleum and that the solvent would vaporize in atmospheric temperatures. The purchaser was engaged in industry and in the use of adhesives for the purposes of its business. According to the evidence in the case, only petroleum solvents were suitable for adhesives of the class in question. Speed in developing a tacky or gummy condition of the adhesive was apparently of the essence of the success of the adhesive as an industrial product to be used in operations such as those in which the purchaser engaged. (at p645)

22. It was proved that the adhesive had an extremely low flash point with a capacity for very rapid vaporization and dissipation. The significance of the low flash point was first that it would only be at any extremely low temperature - lower than atmospheric temperature in Antarctica - that the adhesive would fail to vaporize, at least to the point where its vapour would ignite, or flash: second, that its rate of vaporization was faster than substances with higher flash points. This latter consequence is not categorically stated in the evidence but I think may fairly be inferred from it. But as I mentioned earlier, the evidence does not relate the rate of vaporization of this adhesive to the rate of vaporization of any similar or other adhesive. Further, as I have said, I find nothing in the evidence to suggest that it was the particular rate of vaporization of the adhesive which caused the conflagration. Not one of the witnesses even suggested that, if standard petrol had been allowed to vaporize in the enclosed space of the cool-room in the presence of the naked Portagas flame and the heated bitumen for the same length of time a different result would have ensued. Consequently, it seems to me that the flash point of the solvent, once it is known that the solvent will vaporize at atmospheric temperature in Australia, has, in my opinion, little, if any, bearing on the case. Its introduction with much elaboration in the evidence did no more, in my opinion, than intrude a seemingly scientific air into what was in truth a commonplace of human experience, namely, that the use of a highly inflammable substance in an enclosed, warm and airless space for a considerable period of time in the presence of a powerful naked flame is, to put the matter mildly, a highly dangerous operation. When it is known that there was no reason connected with the operation of insulating the cool-room which required the two operations of spreading the adhesive on the foil and heating the bitumen for the insulation of the cool-room to be carried on simultaneously in the enclosed space, doing so can only be regarded as foolhardy in the extreme. It was, I suppose, interesting to know that the adhesive would vaporize in Antarctica: it was important to know that it would vaporize in Sydney temperatures, a fact not made any more significant, it seems to me, by the other interesting information. (at p645)

23. Was the description of the adhesive "Highly Inflammable" in association with the other information of the data sheets apt without more to convey to the purchaser the requisite knowledge of the dangerous quality of the adhesive? (at p646)

24. The learned trial judge does not seem to have considered this question but rather to have taken the view that the use of the adhesive in the way it was used was or ought to have been in the contemplation of the appellant and that therefore the appellant was required to give to the purchaser some specific instruction as to safeguards to be employed to guard against the dangers of such particular use. Further, he seems to have thought that such directions ought to have been given to the appellant's employees engaged in such use or in such fashion that it would necessarily have reached them. In these views he was supported by the majority of the Court of Appeal Division. (at p646)

25. There are, in my opinion, several errors involved in these views. In the first place, the purchaser was the contractor, his employee was not. If the knowledge of the dangerous qualities of the adhesive were communicated to the contractor, the responsibility of directing the contractor's employees in the use of the adhesive rested upon the contractor and not, in my opinion, upon the appellant. The case is quite unlike Donoghue v. Stevenson [1931] UKHL 3; (1932) AC 562 and Grant v. Australian Knitting Mills Ltd. (1936) AC 85 . In the former, the cordial was intended for consumption without examination by any person before the ultimate consumer, and in the latter the garment was intended to reach the ultimate user in the ordinary course of trade without intervening inspection or interference. But here, the reasonable contemplation, in my opinion, is that the purchaser, an industrial user of adhesives, apprised of the dangerous qualities of the adhesive, will know the simple safeguards to be observed and will accordingly arrange its use of the adhesive and appropriately direct his employees. Thus, in this case, the question I have propounded must be asked in relation to the contractor and not in relation to its employees. (at p646)

26. Also, was the use in fact made of the adhesive by the purchaser one which could fairly be said to be a use which ought to have been in the contemplation of the appellant, so that it ought to have given special directions as to safeguards to be taken when so used. The trial judge, as I have said, seemed to think it was. But, as I have already intimated, in my opinion, no one in the circumstances could have or ought to have contemplated that such a reckless use could be made of the adhesive as was involved in the contemporaneous operations carried on in this case by a purchaser who had been informed of the highly inflammable quality of the adhesive. It must be remembered that there was no danger in using this adhesive in a closed room, even in a warm and airless room. The danger arose by the introduction of a large naked flame. It is going too far, in my opinion, to require a manufacturer of such a substance as this adhesive, even if it really be the most dangerous substance a demonstrator and lecturer in a university has come upon, not only to contemplate that an industrial user aware of the liquid's inflammable quality would introduce a naked flame into a cool-room in the condition this one was but to give positive directions as to what steps should be taken when it is so introduced. But in any case I cannot think myself that the warning "Highly Inflammable" did not clearly and unambiguously inform the contractor that a naked flame should not be introduced into a closed and airless atmosphere in which the adhesive was being used in such quantities as would be involved in keeping the adhesive in the shallow tray and in spreading it on pieces of foil one hundred and forty square feet in area. To say that it does not is, in my opinion, to deny the significance of simple but dramatic words of the English language. It is nothing to the point that the employee Townsend did not comprehend the full import of such words, if indeed that is the right conclusion to draw from his evidence. I am unable to accept the view that it was incomprehension such as Townsend claimed to have and not that of a reasonably well equipped person in industry as its purchaser which the appellant ought to have had in mind when selling its adhesive to the contractor. (at p647)

27. In my opinion, the description of the adhesive as "Highly Inflammable" did adequately convey to the contractor the knowledge of the dangerous qualities of the adhesive: with that knowledge the appellant could fairly expect that the contractor would know of and be in a position to apply the simple safeguards which were spoken of in the evidence to which I have referred. The warning thus given by the appellant was, in my opinion, therefore sufficient to discharge its duty to the respondent. It further follows that, in my opinion, the cause of the respondent's damage was not a breach of duty by the appellant, but an act of the contractor in breach of its duty to the respondent. (at p647)

28. Lastly, I should say something about causation, though to refer to it at all in this case is, in my opinion, to indulge in a work of supererogation. But let it be supposed that the warning to the contractor was inadequate in that it did not convey to him the knowledge of the rate of evaporation of the solvent and of the density of the fumes. Did this deficiency cause or contribute to the cause of the fire? Of course, it is in general no answer to a person who has suffered injury or damage to say that the negligence of another person has contributed to cause the damage where the defendant or some person for whose acts he is responsible has negligently done or omitted something which is causally related to such injury or damage. But where the act or omission of another has intervened between the defendant's negligent act or omission so as to open up a new chain of causation, the defendant will not be responsible for the injury or damage caused by the intervening act or omission. The two principles seem often in competition. Once the defendant's negligence is proved, it will embrace within its attributable consequences all that was or ought to have been foreseen by the defendant or the actor for whom he is responsible at the time of the commission of the act or omission. If the intervening act or omission of another does not fall within the area of the foreseen, then, assuming it to be causally related to the resultant injury, the defendant will not, in my opinion, be responsible for that injury or damage. The person responsible for the new act or omission will bear that responsibility. (at p648)

29. In the present case, the carrying on simultaneously by the contractor of the two operations, in my opinion, as I have already said, clearly did not fall within the area of the foreseeable for which the appellant accepted responsibility when selling the adhesive to the contractor. That the contractor was negligent in carrying on those operations is, to my mind, beyond all question. It undoubtedly had the knowledge of the dangerous quality of the adhesive. Let it be supposed that that knowledge did not embrace the fact that the fumes of the solvent were heavy, and that the solvent was particularly volatile, none the less the unnecessary performance of the two operations at the same time in the motionless atmosphere of the cool-room was clearly negligent. If the knowledge as to the particular volatility and the weight of the fumes was significant, to carry on those operations simultaneously in the circumstances in ignorance of those facts, given the knowledge of the highly inflammable quality of the adhesive, was, in my opinion, in breach of the contractor's duty to the respondent. To bring a naked flame, particularly of the dimension of that used by the contractor, into the presence of a known highly inflammable substance clearly could not be excused, in my opinion, by an assertion of ignorance of such relative refinements as the rate of evaporation and the weight of the fumes given off. (at p649)

30. It would follow, in my opinion, that even if the warning given to the contractor by the appellant could be said to be deficient, an act or omission of the contractor, which the appellant could not possibly have or be required to have foreseen, intervened so as to be the cause of the damage resulting from the fire. (at p649)

31. In my opinion, therefore, however the matter be viewed, the judgment of the trial judge for the respondent cannot be sustained and ought to be set aside. Judgment should be directed to be entered in the action for the appellant. (at p649)

McTIERNAN J. I concur with the Chief Justice in his opinion that the appeal should be allowed for the reasons he has stated. (at p649)

KITTO, MENZIES AND OWEN JJ. The appellant was the defendant in an action brought against it by the plaintiff, the present respondent, in the Supreme Court of New South Wales which was tried by Maguire J. sitting without a jury and in which a verdict was found and judgment entered for the plaintiff for $162,518 (1967) 1 NSWR 50 . The Court of Appeal (1967) 86 WN (Pt 1) (NSW) 345 , by a majority (Wallace P. and Holmes J.A., Jacobs J.A. dissenting), dismissed an appeal from that decision and the defendant now appeals to this Court. (at p649)

2. The facts are that the plaintiff, the owner of a factory, had entered into a contract with Australian Asbestos Insulations Pty. Ltd. (which we will call "the contractor") under which the contractor was to carry out the work of insulating a cool-room which was being built at the plaintiff's factory. The room was about one hundred and twenty feet long, eighty-four feet wide and twenty feet high and at the time with which this case is concerned the walls had been erected, the timber ceiling had been completed and the contractor had carried out the work of insulating the walls to a height of about two feet below the ceiling. The ceiling was to be insulated by covering it with sheets of thin aluminium foil and to these sheets layers of cork board were to be affixed. To enable this work to be done a timber platform had been built on which the men who were putting the foil and the cork boards on the ceiling could work. In order to bond the foil to the ceiling an adhesive called BM 226 General Purpose Adhesive was being used. It was a product manufactured by the defendant and supplied by it to the contractor. It was put up in five-gallon drums, each of them bearing a label which referred to the bonding qualities of the product, gave directions for its use and contained the statement, in prominent lettering, "CAUTION: HIGHLY INFLAMMABLE". It was indeed, as the label said, highly inflammable. The petroleum solvent which it contains vaporizes rapidly and the more lightly the adhesive is spread, the more quickly it evaporates. The vapour is heavier than air and very inflammable, particularly in an enclosed space where there is little air circulation. (at p650)

3. On 12th April 1962 the contractor's employees were on the platform using the adhesive to fix the foil to the ceiling. The temperature in the room was high and there was little or no movement of air. The method of doing the work was as follows. A quantity of the adhesive was poured from the drum into a shallow tray about eighteen inches square. Paint rollers were dipped into the tray and were then used to spread the sheets of foil with a thin coating of the adhesive. The foil was then left for a few minutes until, as the solvent vaporized, the adhesive became tacky and the sheets would then be pressed onto the ceiling. While this work was proceeding, other workmen were also working on the platform fixing cork boards to the ceiling and for this purpose using heated bitumen. The bitumen was heated outside the coolroom and then brought to the platform where it was placed in a tray about three feet six inches long and thirteen or fourteen inches wide. To keep the bitumen in a sufficiently liquid state to be used a Portagas burner with an open flame was placed under the tray and this was being done only about twenty feet from where the workmen were applying the adhesive to the foil. (at p650)

4. A man named Townsend who was in the employ of the contractor was in charge of the work of insulating the cool-room. He knew of the warning notice on the drums and knew also that the adhesive smelt like petrol and petrol was, he agreed, also highly inflammable. He said that he understood the warning "Highly Inflammable" to mean "not to have a flame - not to have a match or a flame anywhere near, approximately near it". Nevertheless he apparently failed to realize the danger involved in the use of a naked flame within twenty feet or so from the place where the adhesive was being used. While the work was proceeding a sheet of foil, which had been coated with adhesive and was being lifted to the ceiling, burst into flames and the fire immediately spread and caused damage to the building. There is no doubt, as the learned trial judge found, that the fire was caused by vapour from the solvent in the adhesive coming into contact with the flame of the Portagas burner. (at p650)

5. The plaintiff's claim was based upon the allegation that the defendant had negligently failed to give adequate warning of the dangerous nature of its product. (at p651)

6. It was not disputed that the plaintiff owed a duty of care to the defendant nor was it disputed that, in the case of such potentially dangerous material as this was, a high degree of care in warning users of it would reasonably be required. The learned trial judge concluded, although with some doubt, that the warning "Highly Inflammable" was not reasonably adequate. He expressed the opinion that some warning should have been given that the material vaporized quickly ; that when it was being used it was desirable to maintain an adequate current of air ; and, possibly, that it should not be used in close proximity to "flames, sparks or lamps". Accordingly he found for the plaintiff. In the Court of Appeal, Wallace P. was of opinion that the notice "Highly Inflammable" was an inadequate warning - at least it had not proved to be a sufficient warning to Townsend - "because of the propensity of BM 226 to quickly emanate heavier-than-air explosive vapour which rapidly travels along a floor such a distance as twenty-two feet and cause an explosion in a room one hundred and twenty feet long". Other than this, his Honour did not indicate what further warning was, in his view, reasonably required. Holmes J.A., as we read his reasons, thought that warning should have been given of the explosive nature of the vapour which would be produced as the solvent evaporated. Jacobs J.A. said (1967) 86 WN (Pt 1) (NSW), at p 356:

"It is my conclusion that if one has a product marked highly

inflammable and which smells of petrol no stronger warning can be given

that an open flame should be kept well away from any area within the fumes of the product may be."

"It is quite impossible to state in any case what that area will be

because it will be infinitely varying depending upon the particular

conditions, particularly the condition of enclosure of the area to which the fumes are allowed to escape."

With this we agree and it seems to us to supply the answer to what is, reality, the plaintiff's contention, namely that the defendant had committed a breach of its duty of care because it had failed to provide a definition of the area, surrounding the place where the adhesive might happen to be used, into which it would be dangerous to introduce a naked light. For the reasons stated by Jacobs J.A., we are of opinion that to say that it was the duty of the defendant to do this would have required it to do the impossible and, with all respect to those who have reached a contrary conclusion, we are of opinion that the warning given, particularly when given in connexion with a product which smells like petrol, cannot be said not to have been reasonably adequate. (at p652)

7. For these reasons we would allow the appeal. (at p652)

Appeal allowed with costs. Order of the Supreme Court of New South Wales (Court of Appeal Division) set aside and in lieu thereof order that the appeal to that Court be allowed with costs. Verdict and judgment for the plaintiff set aside and order that judgment be entered in the action for the defendant with costs.