New South Wales Dairy Corporation v Murray Goulburn Co-operative Co Ltd

1. (1968) 118 C.L.R. 128, at p. 133. 2. (1890) 15 App. Cas. 252, at pp. 263-264. 3. (1937) 57 C.L.R. 448, at pp. 458-459.

In an application for original registration, there is only one object to which effect must be given but, in an application after registration, effect must be given to both objects. The operation of s. 28(a) therefore depends on whether it is applied to an application for original registration or to an application for expunction of an entry which is attacked in s. 22 proceedings as wrongly remaining in the register. If it be right to say that, in s. 22 proceedings, s. 28(a) sometimes does and sometimes does not authorize the removal of a registered trade mark when its use has become likely to deceive or cause confusion, it is necessary to ascertain the condition upon which liability to removal depends. In my opinion, the Act does not contemplate that a registered proprietor should continue to enjoy the rights which s. 58 confers if he has produced the circumstances which make the further use of the mark likely to deceive or cause confusion: s. 58 cannot be understood as conferring rights which might be exercised by the holder to frustrate an object of the Act. But, where the registered proprietor has not produced those circumstances and where no provision other than s. 28(a) warrants expunction of the registered trade mark, there is no reason apparent on the face of the Act why the registered proprietor should not continue to enjoy the rights which registration of the trade mark confers. The Act does not make the avoidance of deception or confusion absolute, even in the case of an application for original registration: see, e.g., s. 34. It strikes a balance between the public interest in avoiding deception or confusion and the private interest of the person entitled to use the trade mark. The view that a registered proprietor may, by his conduct, become disentitled to claim the benefit of the rights conferred by s. 58 arises simply from the need to give to s. 28(a), when it is invoked in proceedings under s. 22 to expunge a registered trade mark on the ground that it is "wrongly remaining in the Register", an operation which will give effect to the provisions of s. 58 except to the extent necessary to deny to the registered proprietor the capacity to frustrate with impunity the statutory object of protecting the public against deception and confusion. A similar view would be reached if pars. (a) and (d) of s. 28 were so read as to make par. (d) govern par. (a) but I do not rest my view on par. (d) of s. 28. Paragraph (d) is not expressed to relate to the conduct of the applicant for original registration or of the registered proprietor after registration but to the character of the mark itself: a "mark which would otherwise be not entitled to protection in a court of justice". In this respect the Act differs from s. 11 of the Trade Marks Act 1938 UK which, though generally similar to the Act, speaks of a mark " the use of which would, by reason of its being likely to deceive or cause confusion or otherwise, be disentitled to protection in a court of justice".

In General Electric Co v General Electric Co Ltd [50] , Lord Diplock stated (with the concurrence of Lord Simon of Glaisdale and Lord Kilbrandon) four propositions derived from the 1938 United Kingdom Act, the third of which was:

If the likelihood of causing confusion did not exist at the time when the mark was first registered, but was the result of events occurring between that date and the date of application to expunge it, the mark may not be expunged from the register as an entry wrongly remaining on the register, unless the likelihood of causing deception resulted from some blameworthy act of the registered proprietor of the mark or of a predecessor in title of his as registered proprietor. Giving the term "blameworthy" the meaning to which I next refer, I would respectfully agree with his Lordship's proposition. I would take to be "blameworthy" any conduct (whether by act or omission) on the part of a registered proprietor or his predecessor in title which he knew or ought to have known would result in the likelihood that the use of the mark would deceive or cause confusion and which has in fact caused or contributed to that result. I would accept also his Lordship's next proposition [50] :

(4) Where a mark is liable to be expunged under (3) the court has a discretion whether or not to expunge it and as to any conditions or limitations to be imposed in the event of its being permitted to remain on the register.

It is unnecessary now to consider the scope of this discretion and the circumstances which may evoke its exercise, except to say that it may be relevant not only to consider the public interest but also to compare the respective contributions made by the conduct of the parties to expunction proceedings under s. 22 in creating a likelihood that use of a registered trade mark would deceive or cause confusion. 1. [1972] 1 W.L.R. 729, at p. 751; [1972] 2 All E.R. 507, at p. 526. 2. [1972] 1 W.L.R. 729, at p. 751; [1972] 2 All E.R. 507, at p. 526.

In the present case, the Full Court of the Federal Court acquitted the respondent of any blameworthy conduct. Their Honours were able to reach this conclusion by treating the non-use of the Moo mark while the Moove mark was in use in respect of flavoured milk (a period commencing in January 1979 and continuing until mid-August 1988) merely as a matter going to non-use of the mark - presumably, as a matter relevant only to an application for removal for non-use under s. 23. I am respectfully unable to accept that conclusion. If a proprietor of a registered trade mark abstains for a time from using it in respect of a class of products so that the mark loses its distinctiveness, and he permits another during that time to use a different mark so as to foster in the public mind a connexion between the latter mark and products of the relevant class to the exclusion of any connexion between the former mark and such products, with the result that any further use of his own mark would be likely to deceive or cause confusion, the conduct of the registered proprietor is, in my opinion, blameworthy in the relevant sense. Such conduct would be a cause of any deception or confusion which would be likely to follow the further use of his own mark. Section 58 could not support that further use for the conduct of the registered proprietor would have converted the statutory right to exclusive use of a registered trade mark into an instrument of likely deception or confusion. In evaluating the conduct of a registered proprietor in such a case, much depends on what he knew or ought to have known of the other person's use of a trade mark in respect of the relevant class of products, the length of time during which the registered proprietor remained supine, the opportunities for objection and the scale of the deception or confusion that is likely to result from the further use of his registered trade mark. It may be going too far to say that the registered proprietor must, in all circumstances, protect his statutory right by applying for an injunction; some other objection may suffice in particular circumstances.

In the present case, however, the registered proprietor of the trade mark Moo (the respondent's predecessor in title) did nothing while the appellant's predecessor and the appellant used the trademark Moove in respect of flavoured milk so that the subsequent use of the trade mark Moo in respect of flavoured milk would be likely to cause confusion. Gummow J. at first instance declined to find that the present respondent was guilty of blameworthy conduct but, as I understand his Honour's reasons, he used this term as equivalent to conduct of the respondent - excluding the conduct of the original registered proprietor - which equity would describe as unclean hands. His Honour said [51] :

The mark of the first respondent was one in respect of which it was not entitled to relief in equity. This was not because of any "blameworthy conduct" as understood when applying the maxim as to unclean hands. Rather, there was "otherwise" no entitlement to such protection, both for want of the necessary distinctiveness at 5 August 1988 and also as a consequence of the lengthy period of use of Moove before the commencement of use of Moo by the first respondent. Further, use of the Moo mark in relation to milk was likely to cause confusion by causing persons, in the sense I have explained, to wonder whether the Moo and Moove products might not come from the same source, and to entertain a reasonable doubt. In my opinion, such facts as his Honour found when taken in conjunction with facts which were not in issue (the period and extent of the use of the trademark Moove in respect of flavoured milk and the absence of any objection by the then registered proprietor of the trademark Moo) are sufficient to show that the registered proprietors of the trade mark Moo were guilty of blameworthy conduct in the relevant sense. An assignee of a registered trade mark stands in no better position than the assignor when an application for expunction is made and reliance is placed by the party seeking expunction on the blameworthy conduct of the assignor. 1. (1989) 86 A.L.R. 549, at p. 587.

I would allow the appeal. I agree with the order proposed by the Chief Justice.

The detailed facts are set out in the judgments of the Chief Justice and of Dawson and Toohey JJ. As those judgments explain, the appeal gives rise to two questions relating to the construction of s. 28 of the Trade Marks Act 1955 Cth ("the Act"). At the cost of some repetition, it is convenient to set out that section at the commencement of this judgment. It reads:

(a) the use of which would be likely to deceive or cause confusion;

(c) which comprises or contains scandalous matter; or

(d) which would otherwise be not entitled to protection in a court of justice,

The first of the two questions is whether the direct operation of s. 28 is confined to the original registration of a mark or whether the section extends to require or authorize the expungement or removal from the register of a registered mark which was properly entered in the register but which, by reason only of events occurring subsequent to registration, answers one or more of the designated descriptions. The second question is whether, by reason of contextual and policy considerations and the presence of the word "otherwise" in par. (d), par. (a) of s. 28 should be read as incorporating some implied additional requirement such as a requirement that the mark "be not entitled to protection in a court of justice" by reason of some blameworthy conduct on the part of the registered owner. Strictly speaking, the second question only arises in the present case if s. 28 is construed as prohibiting the continued registration of a mark which was originally duly registered but which has, by reason of subsequent events, become a mark of one of the relevant descriptions. The reason why that is so is that it is not contended that the original registration of the mark was contrary to s. 28 by reason of the mark being, at that time, of a kind referred to in that section. As will be seen, however, the two questions are closely related.

As a matter of mere language, it appears to me that the direct operation of s. 28 is confined to the act of registration. Prima facie, a provision that a "mark shall not be registered as a trade mark" if it is a mark of a certain description says nothing about the circumstances in which a mark which is a duly registered trade mark should be removed from the register. This prima facie effect of the words of s. 28 is supported by the section's context in a Part which is headed "Registrable Trade Marks" (a phrase traditionally used to refer to marks which are eligible to become, as distinct from remain, registered: see, e.g., Trade Marks Act 1905 Cth, Pt III) and which contains a number of sections, including the two immediately following sections, which are obviously concerned with applications for registration as distinct from prescribing conditions for the continued registration of a registered mark.

That prima facie construction of s. 28 is not precluded by authority, at least in so far as this Court is concerned. Indeed, it is strongly supported by what appears to me to be the only convincing analysis in judgments of persuasive authority, namely, that of Eve J. in W Woodward Ltd v Boulton Macro Ltd [52] . In speaking of a corresponding, though somewhat differently worded, United Kingdom section (s. 11 of the Trade Marks Act 1905 UK), Eve J. said [53] :

Am I entitled to say that a Trade Mark which, when tendered for registration, was on the evidence neither calculated to deceive nor otherwise disentitled to protection, must be judged years after the registration, and notwithstanding a user of it consistent throughout with that registration, by the existence of facts for which the owner of the mark may be in no way responsible and which indeed he may not have been able to control? I am not suggesting that that is either the position of this Trade Mark or that those are the facts relating to this Trade Mark, but in determining the moment of time to which the attention of the Court is to be directed I must assume, and am entitled to assume, the strongest case which might be put from the opposite point of view, and I put the case of a mark which it is not suggested was calculated to deceive when it was registered and not otherwise disentitled to protection, but which in course of time, through circumstances over which the owner of the mark had no control, has become calculated to deceive. I adhere to what I said in the case which was before me last Term of the Imperial Tobacco Co [54] , that I do not believe that that is the construction which the Court ought to put upon this Section. 1. (1915) 32 R.P.C. 173. 2. (1915) 32 R.P.C., at p. 198. 3. [1915] 2 Ch. 27.

On the other hand, support for the view that s. 28 should be construed as extending to prohibit the continuation of the registration of a duly registered mark is to be found in decisions and judgments of strong persuasive authority, including judgments in this Court in Berlei Hestia Industries Ltd v Bali Co Inc [55] and the decision of the House of Lords in General Electric Co v General Electric Co Ltd [56] .

1. (1973) 129 C.L.R. 353. 2. [1972] 1 W.L.R. 729; [1972] 1 All E.R. 507.

The leading speech in the General Electric Case was that of Lord Diplock, with whom Lord Simon of Glaisdale and Lord Kilbrandon agreed. His Lordship concluded that the United Kingdom equivalent of our s. 28 (Trade Marks Act 1938, s. 11: "It shall not be lawful to register") referred both to original registration of a mark and to retention in the register of a duly registered mark. The other member of the House, Lord Reid [57] , disagreed, preferring the view of Eve J. in Woodward Ltd v Boulton Macro Ltd Lord Diplock's view can be traced back to his judgment in the BALI Trade Mark Case in the Court of Appeal [58] . It was based, to no small extent, on the consideration that s. 13(1)(b) of the United Kingdom Act (like s. 61(1)(b) of the Act) contemplates that it might be "shown" that a registered trade mark "offends" against the provisions of s. 11 of the United Kingdom Act (i.e. the section corresponding to our s. 28). With due respect, however, that consideration is hardly persuasive since the use of the word "offends" can readily be explained by the fact that a registered mark "offends" against a provision such as s. 28 if the act of entering it in the register was prohibited by that provision.

1. [1972] 1 W.L.R., at p. 732; [1972] 1 All E.R., at p. 510. 2. [1968] R.P.C. 426.

In Berlei Hestia Industries Ltd v Bali Co Inc, it was common ground that, if the evidence established the existence of a likelihood of deception or confusion at the date of the proceedings, it also established the existence of that likelihood at the date of the original registration. That being so, the question whether s. 28 applied to prohibit the continued registration of a mark which had been initially duly registered but which had subsequently become a mark of the kind referred to in s. 28(a) did not necessarily arise. There are, however, statements in two of the judgments which indicate approval ( [59] , per Barwick C.J.) or acceptance ( [60] , per Mason J.) of the views expressed by Lord Diplock in the General Electric Case. I do not, however, read the judgments of the other two members of the Court (McTiernan and Stephen JJ.) as containing any definite indication of such approval or acceptance. In a context where both Barwick C.J. and Mason J. clearly indicated that the case did not turn upon that point (see [61] , per Barwick C.J., and per Mason J. [62] ), the views expressed in the judgments in the Berlei Hestia Case cannot properly be seen as constituting an authoritative determination by this Court of the proper construction of s. 28.

1. (1973) 129 C.L.R., at p. 356. 2. (1973) 129 C.L.R., at p. 360. 3. (1973) 129 C.L.R., at pp. 356-357. 4. (1973) 129 C.L.R., at p. 361.

The construction of s. 28 which confines the directive that a "mark shall not be registered as a trade mark" to the act of entering it in the register is, on balance, more in accordance with my own understanding of the basic notions and underlying policy of trade mark law than is a construction which reads that directive as extending to require expungement from the register of a trade mark which was duly registered in conformity with applicable statutory provisions at the time of registration. If the use of an unregistered mark is likely to deceive or cause confusion, one can see compelling reasons of policy why, subject to some limited exceptions (see, e.g., ss. 34, 36), registration of it should be denied. If, on the other hand, the use of an unregistered mark is not likely to deceive or cause confusion, registration as a trade mark should itself provide some protection for any goodwill attaching to the registered mark notwithstanding that subsequent events or circumstances may bring about a situation in which original registration of the mark would be precluded. It is true that there are considerations of policy which favour removal from the register of any registered mark whose use is likely to deceive or cause confusion. Those considerations are, however, modified by the availability of ordinary criminal or civil procedures to prevent dishonesty, fraud and passing-off and by the fact that registration of a trade mark does not ordinarily constitute a licence for what would otherwise be unlawful conduct: see, e.g., Lyle and Kinahan Ltd's Application [63] ; Van Zeller v Mason, Cattley & Co [64] ; and note the narrowness of the trade mark exception in s. 51(3)(c) of the Trade Practices Act 1974 Cth). On balance, it appears to me that the policy considerations favouring a construction of s. 28 which would make the section directly applicable to prohibit the continued registration of any duly registered mark which was brought within par. (a) by subsequent events are outweighed by the considerations militating against the lessening of the effective protection of a mark which due registration was, in my view, intended to provide.

1. (1907) 24 R.P.C. 249, at p. 262. 2. (1907) 25 R.P.C. 37, at p. 41.

The construction of s. 28 as not extending to require expungement from the register of a duly registered trade mark obtains, in my view, some support from a consideration of the section's history. The immediate predecessor of s. 28 in Commonwealth legislation was s. 114 of the Trade Marks Act 1905 Cth. That section, unlike the United Kingdom provision upon which it was based (Patents, Designs, and Trade Marks Act 1883, s. 73), was not confined to registration. It was contained in Pt X ("Miscellaneous") of the 1905 Act and provided:

No scandalous design, and no mark the use of which would by reason of its being likely to deceive or otherwise be deemed disentitled to protection in a court of justice, or the use of which would be contrary to law or morality, shall be used or registered as a trade mark or part of a trade mark (emphasis added). When the 1955 Act was enacted, s. 28 was introduced in terms which correspond more closely to the United Kingdom provision and which, like that provision, contain no reference to "use". The clear weight of authority at that time was to the effect that the corresponding United Kingdom provision did not extend to prohibit the continued registration of a trade mark which had been duly registered: see the cases referred to by Lord Denning M.R. and Salmon L.J. in the BALI Trade Mark Case [65] . In other words, the United Kingdom section had been construed as identifying circumstances which would preclude an unregistered mark from being a registrable mark. There is nothing to suggest that it was the legislative intent that our s. 28 should be differently construed. To the contrary, as has been seen, the section was included in the 1955 Act in a Part headed "Registrable Trade Marks". 1. [1968] R.P.C., at pp. 431, 438.

That construction of s. 28 is also supported by a consideration of other provisions of the Act. In particular, as the judgments of the Chief Justice and of Dawson and Toohey JJ. in the present case demonstrate, to treat s. 28 as requiring removal of a duly registered mark which, by reason of subsequent events, becomes a mark of the type referred to in par. (a) of that section gives rise, in the context of other provisions of the Act and presumed legislative policy, to a need to confine the effect of the plain words of s. 28(a) in what seems to me to be a most artificial way. I turn to explain why that is so.

Paragraphs (a), (b), (c) and (d) of s. 28 are clearly disjunctive. The word "otherwise" in par. (d) can be readily explained by the fact that each of the circumstances indicated in pars. (a), (b) and (c) could, depending upon the facts, deprive a mark of entitlement to protection in a court of justice. As a matter of mere construction of words, there is nothing in the section to suggest that the word "otherwise" should be used to confine the circumstances designated in par. (a) (or, for that matter, pars. (b) or (c)) to cases where the specified circumstances would lead to a loss of entitlement to protection in a court of justice by reason of some "blameworthy conduct" on the part of the registered owner. To the contrary, so to read down par. (a) would be effectively to render it otiose since the fourth paragraph would embrace the circumstances to which it refers. Yet, if s. 28 is construed as applying directly to prohibit continued registration of a mark if events occurring subsequent to its registration have brought about one or other of the circumstances referred to in the section, par. (a) must be confined by some such device since it would seem plain that it would not have been the legislative intent that, in a context where the Act contemplates deception or confusion by reason of continuing use of properly registered but "deceptively similar" marks in at least some circumstances (see, e.g., ss. 36, 58(3) and 6(3)), a registered mark should be expunged from the register merely because an infringing use of the mark by some other person had created a situation in which the continued use of the mark by the registered proprietor would be likely to deceive or cause confusion.

Obviously, due weight must be given to the views expressed in judgments in this Court in the Berlei Hestia Case and in the speech of Lord Diplock in the General Electric Case. Nonetheless, on balance, it seems to me that the words a "mark shall not be registered as a trade mark" in s. 28 should be read as directed to the act of registering and not to the state of remaining registered. So construed, the section prohibits the original entry in the register of a mark of the designated kind. The section, so construed, has an indirect continuing operation to the extent that a mark whose original registration was unlawful in the sense that it contravened the prohibition of s. 28 may be expunged from the register for that reason: see s. 22(1)(b) and Berlei Hestia Industries Ltd v Bali Co Inc The section does not, however, have a direct continuing operation in the sense that it requires what it does not say, namely, that a mark which was validly registered without contravention of the section must be removed from the register if the circumstances specified in s. 28 arise by reason of subsequent conduct or events. Once it is accepted that s. 28 of the Act should be so construed, there remains no reason why the words of s. 28(a) should be given other than their ordinary meaning and effect, that is to say, as precluding the actual registration of a mark "the use of which would be likely to deceive or cause confusion". That general prohibition must, of course, be read as subject to any special provision expressly or impliedly authorizing registration in circumstances where actual registration would otherwise be precluded by s. 28: see, e.g., ss. 34 and 36.

The effect of this construction of s. 28 of the Act is that the appeal must be dismissed. I am conscious of the fact that, no doubt because of the statements made in two judgments in the Berlei Hestia Case in this Court, it was not argued on behalf of the respondent that the direct operation of s. 28 of the Act is confined in the sense which I have explained. That fact does not, however, avoid the need to express what I see as the correct construction of an important general statutory provision. Nor, in circumstances where my views about the proper construction of s. 28 differ from those of the majority of the Court, does it give rise to a need to consider whether the matter should be relisted for further argument. In all the circumstances, however, it seems appropriate that I indicate that I would have been in favour of dismissing the appeal even if I had been of the view that s. 28 should be construed in the manner favoured by the Chief Justice and Dawson and Toohey JJ. If I had been of that view, the question would have arisen whether the respondent's registered Moo mark had, notwithstanding its registration and the failure of the appellant's case based on alleged non-use, become disentitled to protection in the courts by reason of some blameworthy conduct on the part of the respondent or its predecessor as owner of the registered Moo trade mark. In the light of the findings of the trial judge, including the specific finding to the effect that there was no belief (or intent) on the part of the respondent that its conduct would encourage deception or confusion or involve or lead to the appropriation to itself of any of the Moove goodwill, I would have agreed with Dawson and Toohey JJ. that that question should be answered in the negative. In view of that specific finding and of the failure of the appellant to make out its claim for expungement for non-use, the fact that laches or acquiescence on the part of the respondent may have given rise to personal equities in the appellant if the respondent had brought proceedings to prevent or restrain continued registration or use of the Moove mark by the appellant does not, in my view, mean that the respondent can, in the absence of such proceedings, be said to have been guilty of blameworthy conduct or that its registered Moo trade mark can be said to be disentitled to protection in a court of justice (cf. the Act, s. 58(3)).

On 26 August 1966, Healthway Dairy Products Pty. Ltd. registered the trade mark Moo in respect of dairy products. The trade mark has been renewed from time to time. It was used in relation to a reduced fat cream product until 1974, but after that received little or no use for some thirteen years. In September 1987, Healthway Dairy Products Pty. Ltd. assigned the trade mark to the respondent, Murray Goulburn Co-operative Co. Ltd. Following the assignment of the trade mark, the respondent used it in relation to a packaged cheese product and, in August 1988, began to use it in relation to flavoured milk.

The appellant, the New South Wales Dairy Corporation, is the proprietor of the trade mark Moove, which was registered on 1 November 1978 in respect of, amongst other things, milk and milk products. It uses the trade mark in relation to a flavoured milk product which was launched in January 1979.

On 5 August 1988 the appellant brought proceedings in the Federal Court claiming relief of various kinds, including an application under s. 22 of the Trade Marks Act 1955 Cth ("the Act") for rectification of the register of trade marks by the expunction of the Moo mark. The trial judge ordered the register to be rectified by restricting the use of the Moo mark to cheese: see New South Wales Dairy Corporation v Murray Goulburn Co-operative Co Ltd [66] . The Full Court of the Federal Court allowed an appeal by the respondent and ordered that the Moo mark be left on the register with respect to cheese and flavoured milk, but not otherwise: Murray Goulburn Co-operative Co Ltd v New South Wales Dairy Corporation [67] . The appellant appeals by special leave from the decision of the Full Court.

1. (1989) 86 A.L.R. 549. 2. (1990) 24 F.C.R. 370.

The principal question which arises is the proper construction of s. 28 of the Act, which provides:

(a) the use of which would be likely to deceive or cause confusion;

(c) which comprises or contains scandalous matter; or

(d) which would otherwise be not entitled to protection in a court of justice,

shall not be registered as a trade mark. But before coming to the construction of s. 28, it is necessary to examine how it may be applied in this case where no attack is, or could be, made upon the original registration of the Moo mark. The attack which is made is based upon events occurring after the original registration, which, it is said, now make the use of the Moo mark likely to deceive or cause confusion. Rectification of the register as claimed by the appellant would have the effect of rendering the use of the Moo mark an infringement of the Moove mark.

Section 61(1) of the Act provides that in legal proceedings relating to a trade mark registered in Pt A of the register (as are the two marks in this case), including applications under s. 22 of the Act, the original registration of the trade mark shall, after the expiration of seven years from the date of the original registration, be taken to be valid in all respects, unless it is shown:

(a) that the original registration was obtained by fraud;

(b) that the trade mark offends against the provisions of section 28; or

(c) that the trade mark was not, at the commencement of the proceedings, distinctive of the goods or services of the registered proprietor.

It is apparent from the use of the present tense in par. (b) that, in considering whether the use of a mark would be likely to deceive or cause confusion within the context of s. 61, it is proper to look at its current use rather than its use at the time of the original registration. The use of the words "original registration" in par. (a), in contrast with the use of the words "trade mark" in par. (b), confirms this construction of par. (b). And in an application to rectify the register upon the ground specified in par. (b), it is use at the time when the application for rectification is made which is relevant: Berlei Hestia Industries Ltd v Bali Co Inc [68] . Thus, because it is an exception to the protection afforded by s. 61(1), a mark, the use of which would be likely to deceive or cause confusion within the meaning of s. 28 at the time of an application to rectify the register, is not deemed to be validly registered even after seven years. Paragraph (b) of s. 61(1) does make it necessary to give s. 28 a continuing operation. But s. 61(1) is concerned only with the validity of the original registration. When the sub-section applies, the original registration is to be taken to be valid in all respects. When it does not apply, there is no presumption of validity of the original registration; that is all. That is of little assistance in a case such as the present where the original registration of the Moo mark may be seen, without resort to any presumption, to have been clearly valid. The likelihood of deception or confusion is said to have arisen after the original registration and because of supervening events. 1. (1973) 129 C.L.R. 353, at p. 360.

It is s. 22 itself which establishes that in rectification proceedings attention need not be confined to the original registration. That section speaks of the expunging or amendment, not only of an entry wrongly made, but also of an entry wrongly remaining in the register, thus drawing a clear distinction between an entry wrongly on the register because it was wrongly made and an entry otherwise wrongly on the register. An entry can only be wrongly on the register, otherwise than because it was wrongly made, as a result of changed circumstances. There is then the question of what changed circumstances will make it wrong for a mark to remain on the register. Section 56, to which we shall refer later, provides one instance of a mark wrongly remaining on the register. In certain circumstances set out in that section, an entry is deemed to be wrongly remaining on the register. But that section is not intended to be exhaustive. In relation to the United Kingdom legislation, Lord Diplock in General Electric Co v General Electric Co Ltd [69] , expressed the opinion that "an entry which was lawful at the time it was originally made can only be expunged if, at the time of the application, its continued presence on the register is prohibited by virtue of some other provision of the Act". We see no reason to doubt that view in relation to the Australian legislation, though it may be closer to the language of the legislation to say that its continued presence is not authorized by any other provision of the Act. And, if it has a continuing operation, s. 28 is such a provision. It is at this point that reference to s. 61(1) is helpful because it demonstrates that s. 28 was intended to have a continuing operation, at least for the purpose of establishing the validity of the original registration of a mark, and we can see no reason, once the continuing operation of s. 28 has been recognized to that extent, why one should not recognize its operation for the purpose of establishing whether the registration of a mark continues to be valid. When the latter section says that a mark with the characteristics specified shall not be registered as a trade mark, it means not only that it shall not be placed on the register, but also that it shall not remain on the register. The language is capable of bearing that two-fold meaning and, if nothing else, s. 61(1) indicates that such is the meaning which it should bear.

1. [1972] 1 W.L.R. 729, at p. 739; [1972] 2 All E.R. 507, at p. 516.

The case turns upon which of two alternative constructions of s. 28 is the proper one. Should pars. (a), (b), (c) and (d) of that section be read entirely disjunctively, so that if the appellant can establish that, by reason of events which have occurred since registration, the use of the respondent's Moo mark in relation to flavoured milk would be likely to deceive or cause confusion, it is entitled to an order that the register be rectified by the expunction or amendment of the entry in respect of the Moo mark? Or should par. (a) be read together with or take colour from par. (d), so that the appellant must show, not only that the use of the Moo mark would be likely to deceive or cause confusion, but also that it would not be entitled to protection in a court of justice? The problem was recognized by Mason J. in Berlei Hestia Industries Ltd v Bali Co Inc, although it was unnecessary to offer any answer in that case, when he said [70] :

The meaning of the words "the use of which would be likely to deceive or cause confusion" in s. 28(a) is influenced by the language in which s. 28(d) is expressed when it refers to a mark "which would otherwise be not entitled to protection". Thus, a special problem may confront the Court when a mark which was not likely to deceive or cause confusion at the time of its original registration subsequently becomes so or when two marks substantially identical or deceptively similar have been registered following honest concurrent user (see s. 34). Berlei Hestia Industries Ltd v Bali Co Inc was, of course, a case in which it was agreed that, if a current likelihood of deception or confusion by the use of the two competing marks was established, it would establish the same likelihood at the time of the original registration of the mark alleged to offend. The problem in this case did not then arise. 1. (1973) 129 C.L.R., at p. 361.

That problem is whether the use of the expression "which would otherwise be not entitled to protection" in s. 28(d) indicates that the mere likelihood of deception or confusion arising from the use of a mark is insufficient to bring it within s. 28(a); that, in addition, the circumstances must be such as to disentitle the mark to protection in a court of justice. For it is not every mark, the use of which would be likely to deceive or cause confusion, which is disentitled to protection in a court of justice. That is the point of Mason J.'s reference to s. 34 which, in the case of honest concurrent user, allows the registration of trade marks which are substantially identical or deceptively similar, their use being governed by s. 58(3). And unless some such construction is given to s. 28(a), the word "otherwise" in s. 28(d) has no work to do; it is superfluous. On the other hand, s. 28 is, apart from this, clearly expressed in a disjunctive form.

The predecessor of s. 28 was s. 114 of the Trade Marks Act 1905 Cth which read:

No scandalous design, and no mark the use of which would by reason of its being likely to deceive or otherwise be deemed disentitled to protection in a court of justice, or the use of which would be contrary to law or morality, shall be used or registered as a trade mark or part of a trade mark. (Our italics.) Plainly the construction of s. 28 for which the respondent contends is supported by the wording of the old s. 114, if the breaking up of that section into separate paragraphs in the new s. 28 was not intended to effect a change in meaning. A committee known as the Dean Committee (the Committee Appointed by the Attorney-General of the Commonwealth to Consider What Alterations are Desirable in The Trade Marks Law of the Commonwealth) in 1954 recommended alterations in the law. It did so, having reviewed the recommendations of a previous committee, the Knowles Committee, which had proposed a Bill simply adopting the form of s. 114 of the 1905 Act with the addition of the words "or cause confusion". The Dean Committee proposed that s. 28 take its current form and commented at par. 63 that "Clause 28, while in different words, has in substance the same effect as section 114 of the present Act, but it relieves the Registrar and the Court from consideration of "morality"." It was the recommendation of the Dean Committee which was accepted. If the Dean Committee had sought to remove the requirement of disentitlement to the protection of a court where deception or confusion was relied upon, a requirement which was plainly present in the old s. 114, it is probable that they would have made their intention apparent. But they asserted the contrary. Thus, the circumstances in which the present legislation was enacted suggest that no such change was intended. Of course, it is the wording of the new s. 28 which is of primary significance, but the presence of the word "otherwise" in par. (d) creates an obvious ambiguity in the face of the otherwise disjunctive form of the section. It is in the resolution of this ambiguity that the history of the section is of assistance.

Section 114 of the 1905 Act was worded in a manner similar to that of s. 11 of the Trade Marks Act 1938 UK. The latter section provides:

It shall not be lawful to register as a trade mark or part of a trade mark any matter the use of which would, by reason of its being likely to deceive or cause confusion or otherwise, be disentitled to protection in a court of justice, or would be contrary to law or morality, or any scandalous design. That section was examined in an historical context by Lord Diplock, with whom Lord Simon and Lord Kilbrandon concurred, in General Electric Co v General Electric. One aspect of Lord Diplock's examination centred upon the nature of the legal protection afforded, initially by the Court of Chancery, to trade marks as a form of property. Sections which were the predecessors of s. 11 spoke of marks not being "entitled to protection in a court of equity". The reference to "a court of equity" was changed to "a court of justice" because of the passage of the Judicature Acts. Injunction was the principal means by which equity protected property in a trade mark. But an injunction has always been a discretionary remedy and, in some circumstances where there was the simultaneous use of deceptively similar trade marks, the owner of one would be refused relief against the owner of the other, provided the latter's use of the mark had been honest.

At common law the right to a trade mark was an adjunct of the goodwill in a business and incapable of an existence separate from it. If, however, the goodwill of a business was divided, the need to recognize more than one user of a trade mark arose, despite the possibility of confusion. For each proprietor to have been able to seek an injunction against the other would have been to undermine ownership. In addition, the increase in the ease of communications during the nineteenth century meant that goods displaying deceptively similar marks, which previously had been restricted to a particular locality, might be offered to a wider set of purchasers, thus producing a greater likelihood of confusion through simultaneous use. The doctrine of honest concurrent user, now embodied in the legislation, developed at common law as a response to this. It prevented one owner of a trade mark from bringing an action of infringement against the proprietor of a substantially identical or deceptively similar mark, but permitted either owner to obtain an injunction to restrain the use of the mark by a usurper: seeGeneral Electric Co v General Electric Co Ltd [71] ; Dent v Turpin [72] ; Alex Pirie & Sons Ltd's Application [73] , per Lord Tomlin; James F McKenzie & Co v Leslie [74] .

1. [1972] 1 W.L.R., at p. 743; [1972] 2 All E.R., at p. 519. 2. (1861) 2 J. & H. 139, at pp. 147-148 [70 E.R. 1003, at p. 1007]. 3. (1933) 50 R.P.C. 147, at pp. 159-160. 4. (1909) 9 C.L.R. 247, at p. 255.

The doctrine of honest concurrent user meant that a degree of confusion had to be tolerated within the system. The mere fact that the use of a mark would be likely to deceive or cause confusion was not necessarily sufficient to support an injunction restraining that use. That, Lord Diplock explained [1] , was how s. 11 and its predecessors came to be interpreted in such a way that something more than mere deception or confusion was required to disentitle a trade mark to protection in a court of justice.

1. [1972] 1 W.L.R., at pp. 745, 748; [1972] 2 All E.R., at pp. 521, 524.

The doctrine of honest concurrent user is also an explanation of the form of s. 114 of the Trade Marks Act 1905 Cth and if, as we think is the case, s. 28 was not intended to effect any relevant change in meaning, it is an explanation of the use of the word "otherwise" in s. 28(d). It is a reference back to s. 28(a). Section 28(a) must be read together with s. 28(d) so that the reference in s. 28(a) to a "use which would be likely to deceive or cause confusion" is a reference to a use which would not only be likely to deceive or cause confusion but would be not entitled to protection in a court of justice. The word "otherwise" in s. 28(d) is thus given some work to do. It means then that the inquiry must in every case be whether some remedy would or would not be available. As Dixon J. put it in relation to s. 114 of the 1905 Act in Radio Corporation Pty Ltd v Disney [2] :

If the circumstances are such that its adoption will give the applicants no right to protection by injunction or other remedy under the general law, then it should be kept off the register. It is consistent with that view that a mark should be removed from the register if circumstances arise after the date of original registration which leave its proprietor without protection. 1. (1937) 57 C.L.R. 448, at p. 459.

This view of s. 28 is, we think, supported by the scheme of the Act. Not only is it consonant with s. 34 which permits the registration of trade marks which are substantially identical or deceptively similar in the case of honest concurrent user, but any other construction would allow a person who "assiduously infringed" a trade mark to exploit the infringement by producing a situation in which the use of the registered mark would be likely to deceive or cause confusion: see Re Bali Brassiere Co Inc's Registered Trade Mark and Berlei Ltd's Application [3] , per Windeyer J. The greater the infringement, the greater the probability that the infringing user would have a ground for claiming rectification against the registered user. There may be a residual discretion to refuse relief, but it seems preferable to construe the provision in such a way as to give no encouragement to such conduct.

Furthermore, s. 82 of the Act allows the registered proprietor of a trade mark to assign the mark without the goodwill of the business concerned. This allows for the almost inevitable result that some confusion will occur concerning the origin of the goods or services in respect of which the trade mark is registered. The fact that such confusion must have been anticipated by the legislature suggests that it could not have intended that confusion alone should invalidate the registration of a mark. Of course, were the mark otherwise to become deceptive the validity of its registration might be impugned: see Pioneer Kabushiki Kaisha v Registrar of Trade Marks [4] , per Aickin J.; Heublein Inc v Continental Liqueurs Pty Ltd [5] , per Fullagar and Taylor JJ.

1. (1977) 137 C.L.R. 670, at p. 683. 2. (1960) 103 C.L.R. 435, at p. 451.

In Pioneer Hi-Bred Corn Co v Hy-Line Chicks Pty Ltd [6] , the New Zealand Court of Appeal took a different approach. The equivalent section of the Trade Marks Act 1953 NZ, s. 16, provides:

It shall not be lawful to register as a trade mark or part of a trade mark any scandalous matter or any matter the use of which would be likely to deceive or cause confusion or would be contrary to law or morality or would be otherwise disentitled to protection in a Court of justice. The Court pointed to the difference in wording of s. 16 in comparison with its predecessor, which bore obvious similarities to s. 11 of the United Kingdom Act of 1938. Richardson and Woodhouse JJ. concluded that the deliberate change meant that the new section had to be read disjunctively so that likelihood to deceive or confuse is not modified by the final phrase of the section. Richmond P. took the view that, if it was so modified, the likelihood of confusion or deception alone would ipso facto disentitle the mark to protection, so that the effect of the final phrase was not to impose any additional requirement. The Court concluded that the purpose of s. 16 is to protect the public rather than traders' interests in trade marks. 1. [1978] 2 N.Z.L.R. 50.

In previous legislation the emphasis may have been upon avoiding the deception of the public: Eno v Dunn [7] ; Radio Corporation v Disney [8] . But now it is clear that, whilst the protection of the public no doubt remains an important purpose, it is not the only purpose. Plainly, since the Act permits some deceptively similar marks to remain on the register, the protection of the proprietary interest in a registered trade mark is another. And the construction given to s. 16 by Richardson and Woodhouse JJ. in New Zealand renders otiose the word "otherwise" appearing in the final phrase. Finally, perhaps most significantly, there is not the same indication with the New Zealand section, as there is with its Australian equivalent, that no relevant change in meaning was intended by the change in wording.

1. (1890) 15 App. Cas. 252, at pp. 263-264. 2. (1937) 57 C.L.R., at pp. 458-459.

In Australia, authority favours the construction for which we have expressed a preference. It is consistent with the decision of this Court in Berlei Hestia Industries Ltd v Bali Co Inc In HTX International Pty Ltd v Semco Pty Ltd [9] , Fox J. adopted Lord Diplock's approach, saying [10] : "It seems inconsistent with the policy of the Act that the activity of a competing trader which would otherwise be regarded as infringement can bring about a situation where the proprietor's use of its mark, because of confusion, leads to its invalidity." And in Riv-Oland Marble Co v Settef SpA [11] , the Full Court of the Federal Court (Bowen C.J. and Lockhart J; Northrop J. dissenting) held that s. 28 should be construed in such a manner that the word "otherwise" in s. 28(d) serves to limit the circumstances in which s. 28(a) would invalidate the registration of a trade mark.

1. (1983) 78 F.L.R. 57; 49 A.L.R. 636. 2. (1983) 78 F.L.R., at p. 64; 49 A.L.R., at p. 643. 3. (1988) 19 F.C.R. 569.

In General Electric Co v General Electric Co Ltd, Lord Diplock, in summarizing the legal status under the United Kingdom Act of 1938 of a registered trade mark, the use of which is likely to cause confusion, gave as one proposition [12] :

If the likelihood of causing confusion did not exist at the time when the mark was first registered, but was the result of events occurring between that date and the date of application to expunge it, the mark may not be expunged from the register as an entry wrongly remaining on the register, unless the likelihood of causing deception resulted from some blameworthy act of the registered proprietor of the mark or of a predecessor in title of his as registered proprietor. The use of the term "blameworthy act" is, we think, no more than a shorthand method of referring to all those circumstances which might disentitle an applicant to relief in a court of equity. It was not intended to posit an all-embracing test. As Lord Diplock said [13] : It seems to me to be impossible to hold there was any such blameworthy conduct by the [appellants] as would amount to an equitable ground for disentitling them to protection for the use of the mark. It would, in our view, be quite wrong to use the term to erect an independent standard against which a registered proprietor's conduct must be judged. Not only would it be an impermissible extension of the language of the section, but it would shift the focus from the true question and introduce notions of a subjective kind. 1. [1972] 1 W.L.R., at p. 751; [1972] 2 All E.R., at p. 526. 2. [1972] 1 W.L.R., at p. 752; [1972] 2 All E.R., at p. 527.

It would clearly be a most significant circumstance, in defending a trade mark against expunction from the register, that the registered proprietor would have been refused relief in proceedings brought by him for infringement because he did not come with clean hands, being himself responsible in some way for the deception or confusion. That may be blameworthy conduct but it is not the only conduct which would disentitle an applicant to relief. And, in some circumstances, it may not disentitle an applicant to relief. For example, it may be more important that equity grant a remedy in the public interest than that it does not aid a wrongdoer: see Money v Money [No 2] [14] . But this is merely to emphasize that in equity the remedy was discretionary, notwithstanding that it was exercised along defined lines.

The only issue which is raised upon the appeal before us is that which arises under s. 28(a), namely, whether there should be an order for rectification of the register so as to preclude the respondent from using the Moo mark in relation to flavoured milk because such use would be likely to deceive or cause confusion. The trial judge found that there had been a non-use of the mark in relation to dairy products, other than cheese and flavoured milk, and would have made an order under s. 23, which provides for relief in respect of non-use, limiting the goods in respect of which the Moo mark was registered to those two products. However, the trial judge also found that the appellant had made out a case under s. 28 in respect of flavoured milk and so ordered that the register be rectified by limiting the goods in respect of which the mark was registered to cheese alone. The Full Court came to a contrary conclusion under s. 28 and, consequently, made an order for rectification which extended the limited class of goods in respect of which the mark was entitled to registration by adding flavoured milk to cheese.

The trial judge, applying the test formulated by Kitto J. and approved by the Full Court in Southern Cross Refrigerating Co v Toowoomba Foundry Pty Ltd [15] , found that there was a likelihood of deception or confusion in the use of the two marks. He found that "a substantial number of purchasers and prospective purchasers of the Moo U.H.T. flavoured milk product, being persons who know of Moove flavoured milk, including the U.H.T. Moove product, will be caused to wonder whether it might not be the case that the two products come from the same source; they would entertain a reasonable doubt" [16] . He found that the primary cause of confusion lay in the phonetic similarity of the two marks, particularly to people who knew only of Moove flavoured milk but had an imperfect recollection of it. The Full Court held that this finding was not only open but correct.

1. (1954) 91 C.L.R. 592, at pp. 595, 608. 2. (1989) 86 A.L.R., at p. 558.

For the reasons which we have given, the likelihood that use of the Moo mark would deceive or cause confusion was not, of itself, sufficient to disentitle it to registration by rectification of the register. The likelihood of deception or confusion had to be such that the mark would not be entitled to protection in a court of justice. Blameworthy conduct on the part of the respondent would have been sufficient for that purpose, but the trial judge found that the respondent's conduct was not blameworthy. There were other factors which, in the view of the trial judge, disentitled the mark to protection in a court of justice. He said [17] :

The mark of the first respondent was one in respect of which it was not entitled to relief in equity. This was not because of any "blameworthy conduct" as understood when applying the maxim as to unclean hands. Rather, there was "otherwise" no entitlement to such protection, both for want of the necessary distinctiveness at 5 August 1988 and also as a consequence of the lengthy period of use of Moove before the commencement of use of Moo by the first respondent. Further, use of the Moo mark in relation to milk was likely to cause confusion by causing persons, in the sense I have explained, to wonder whether the Moo and Moove products might not come from the same source, and to entertain a reasonable doubt. In that passage the trial judge appears to us to be saying that: (a) lack of distinctiveness of the Moo mark at the date of application for rectification was sufficient to disentitle it to protection in a court of justice, thus providing a ground under s. 28(d) for rectification; (b) non-use of the Moo mark or the failure to take steps to prevent infringement of Moo by the use of the Moove mark was sufficient to disentitle the Moo mark to protection in a court of justice, thus providing a ground under s. 28(d) for rectification; (c) the matters referred to in (a) and (b) rendered the likelihood of confusion or deception such that the mark would not have been entitled to protection in a court of justice, thus providing a ground under s. 28(a) for rectification. The order made by the trial judge contained a declaration that the use of the Moo mark was likely to deceive or cause confusion within the meaning of s. 28(a). It contained no declaration in relation to s. 28(d) [18] . 1. (1989) 86 A.L.R., at p. 587. 2. (1989) 14 I.P.R. 75, at p. 81.

The Full Court of the Federal Court took the view that neither lack of distinctiveness nor non-use, by itself or coupled with a failure to prevent infringement, was sufficient to disentitle the Moo mark to protection in a court of justice. Thus, neither in itself provided a ground for rectification under s. 28(d), nor did either do so under s. 28(a) when coupled with the likelihood of deception or confusion. We agree with the view taken by the Full Court.

The absence of distinctiveness is, of course, relevant to the likelihood of deception or confusion. As Kitto J. observed in Bayer Pharma Pty Ltd v Farbenfabriken Bayer Aktiengesellschaft [19] , they "tend to run into one another". But they are not the same thing and the Act deals with them quite separately. The distinctiveness required for the original registration of a trade mark is dealt with in ss. 24, 25, 26 and 27. Thus s. 28 does not impart any separate requirement of distinctiveness at the time of registration and it is hardly to be thought that it does so subsequently.

Lack of distinctiveness after registration is dealt with elsewhere in the Act. Under s. 26(2)(b), the distinctiveness of a mark rendering it registrable under Pt A may be by reason of its use. It may lose that distinctiveness by subsequent non-use. A mark registered under Pt B may fail to become distinctive through lack of use. The consequences of non-use are dealt with in s. 23. Distinctiveness may be lost otherwise than through non-use. Section 56 provides that, in certain circumstances, the registration of a trade mark becomes invalid because it contains a word or words which have come to be used in a generic sense to name or describe an article, substance or service. Under s. 64(1)(b), a person alleged to infringe a trade mark may have a defence if he uses the mark in good faith as a description of the character or quality of his goods or services.

The way in which the Act deals explicitly with lack of distinctiveness or the circumstances giving rise to lack of distinctiveness is a clear indication that s. 28 was not intended to embrace them by implication, at all events as grounds in themselves. Moreover, s. 61(1)(c), in denying the presumption of validity to a trade mark that is not distinctive, deals with lack of distinctiveness separately and does not include it as something which offends against the provisions of s. 28(a). It is, we think, clear that lack of distinctiveness does not, of itself, disentitle a registered trade mark to the protection of a court of justice, nor does it do so merely because it gives rise to the likelihood of deception or confusion. We have already referred to ss. 34 and 58 which contemplate the registration of trade marks which are substantially identical or deceptively similar in the case of honest concurrent user or other special circumstances.

As we have said, s. 23 deals with circumstances in which a trade mark may be removed from the register for non-use. The trial judge found that those circumstances did not exist in respect of the Moo mark in relation to cheese and flavoured milk and declined to make an order under s. 23 in relation to those products. That being so, it is not in our view possible to say that, because of non-use, that trade mark was not entitled to protection in a court of justice at the relevant time. In any event, the presence of s. 23 would, upon the ordinary canons of construction, exclude non-use as a ground of disqualification under s. 28. Of course, non-use may result in the likelihood of deception or confusion, but that is a different thing.

Equity may have refused protection to the Moo mark if the respondent had sought to benefit from the establishment of the Moove mark in the market by waiting for a substantial period before entering the market with Moo flavoured milk. But, in our view, the findings of the trial judge preclude such a conclusion. He rejected the appellant's claim that the respondent had deliberately set out to cause confusion in the market place with the intention or expectation of gaining a benefit from the deceptive similarity between Moo and Moove. He held that it was not the respondent's intention to gain part of the appellant's market share because of the resemblance between the two marks. Nor was the respondent, in the trial judge's words, "setting out to "cash in" on the reputation of Moove or to sail as close to the wind as it might without attracting legal liability" [20] . He said [21] :

The evidence satisfies me that when, over the period August-September 1987, the first respondent selected "Moo" for its cheese and milk products, it was aware that "Moo" was not in current use by the second respondent and had not been used for some time, but the legal significance of this was not appreciated by the first respondent. Further, the first respondent was well aware that there was significant aural similarity between "Moo" and "Moove" and that the latter was the name under which the applicant had produced flavoured milk products for a number of years. But I accept what was said by the witnesses for the first respondent to the effect that the first respondent was not acting fraudulently in the sense of choosing "Moo" in the hope or expectation of deriving benefit from its resemblance to "Moove". The first respondent's position was that the products would be directed at different "target" markets and be promoted accordingly. The acquisition of the registration of "Moo" confirmed the first respondent in the view that there would be no problem with the proposed course of conduct. 1. (1989) 86 A.L.R., at p. 594. 2. (1989) 86 A.L.R., at pp. 593-594.

These findings were open to the trial judge on the evidence. The words " Moo milk" were first suggested to the respondent by its advertising agents as suitable for attracting children as the consumers of dairy products. An alternative, "Farm Fun", was considered to be not suitable. The respondent quite properly instructed its patent and trade mark attorneys to search the register and, as a consequence, learnt that the Moo mark was registered. The respondent purchased the mark from the registered proprietor for a nominal consideration. The consideration was nominal because the mark had been dormant for some time and its registration was about to expire. The dormancy was, however, insufficient to amount to non-use which would justify the removal of the mark from the register. The respondent took no steps to protect the Moo mark against the Moove mark, although it was aware of the latter. But the evidence was that the respondent believed that the use of the Moo mark created no problem for either party "because it [Moove] was aimed "at an entirely different market" and was promoted so as to appeal significantly to that market" [22] . As we have said, the trial judge found that there was no blameworthy conduct on the part of the respondent and we are unable to see in the circumstances anything which would otherwise disentitle the Moo mark to protection. Moreover, in those circumstances, there is no justification for treating the absence of any activity on the part of the respondent's predecessor as a factor disentitling the Moo mark to protection.

The Act provides the means by which the registered proprietor of a trade mark may obtain relief for infringement. If he fails to act in circumstances where he ought to do so in order to protect his mark, there may be a basis upon which the mark should be denied protection, namely, that there is a likelihood of deception or confusion for which the registered proprietor is responsible in failing to pursue his remedies. But in this case, in the light of the respondent's explanation and the findings of the trial judge, the conduct of the respondent cannot, in our view, be regarded as disentitling the mark in question to protection.

On the basis that s. 28(a) of the Trade Marks Act 1955 Cth has an operation with respect to a trade mark which at the time of the registration was not likely to deceive or cause confusion but which has since become likely to do so, I agree with Dawson and Toohey JJ. that that operation must be confined by a requirement of blameworthy or disentitling conduct and that no such findings can be made in the present case.

The argument in the present case proceeded on the basis of an assumption that a properly registered trade mark could, by reason of some supervening event, offend against s. 28 which, in terms, does no more than specify those marks that "shall not be registered as a trade mark". That assumption is supported by the decision in General Electric Co v General Electric Co Ltd [23] and by statements by Barwick C.J. and Mason J. in Berlei Hestia Industries Ltd v Bali Co Inc [24] . However, as is pointed out in the judgment of Deane J., there is no authority requiring this Court to treat s. 28 as having that operation.

1. [1972] 1 W.L.R. 729; [1972] 2 All E.R. 507. 2. (1973) 129 C.L.R. 353, at pp. 356, 360.

The issues raised by the present case are such that I am content to decide it on the basis upon which it was argued. However, the assumption inherent in the argument does not appear to me to be well founded. It is contrary to the language of s. 28 and it seems to me that it creates a number of difficulties, including the difficulty in the present case, for which the Act provides no ready solution.

This appeal concerns the construction of s. 28 of the Trade Marks Act 1955 Cth ("the Act") which provides:

(a) the use of which would be likely to deceive or cause confusion;

(c) which comprises or contains scandalous matter; or

(d) which would otherwise be not entitled to protection in a court of justice,

1. Does s. 28 prohibit the continued registration of a mark which contravenes its provisions only by reason of events which have occurred after the mark has become registered?

2. Does par. (d) qualify the operation of par. (a)?

3. If par. (d) qualifies the operation of par. (a), is a mark which has become likely to deceive or confuse only liable to expunction if its registered proprietor is guilty of blameworthy conduct?

4. If the foregoing questions are answered in the affirmative, was the respondent guilty of blameworthy conduct?

In my opinion, the first question must be answered in the affirmative because of the way the case was conducted in this Court and the Federal Court. However, the second question should be answered in the negative. This conclusion has the consequence that the appeal must be allowed and it becomes unnecessary to answer the third and fourth questions.

The appellant was granted registration of the trade mark Moove in relation to Class 29 goods on 1 November 1978. In January 1979, it commenced to market flavoured milk under the trade mark Moove. Moove flavoured milk is available in both U.H.T. (long life) and fresh milk.

The respondent is the registered proprietor of the trade mark Moo. The respondent manufactures and supplies a large number of dairy products under the name "Devondale". Prior to August 1987, the respondent decided to expand its market. After consultation with market researchers and advertising agents, the respondent decided to market a children's cheese product and a school milk product. The advertising agents suggested the name Moo. A search of the trade mark Register revealed that this trade mark was held in relation to Class 29 goods by Healthway Dairy Products Pty. Ltd. ("Healthway"). At the instigation of the respondent, the registration was renewed on 26 August 1987. On 29 September 1987, Healthway assigned the mark to the respondent. The assignment was registered on 24 February 1988. Healthway had once used the trade mark in relation to a reduced fat cream product but had ceased production of that product in 1974. In November 1987, the respondent launched a packaged cheese product in Queensland under the name Moo. In February 1988, sales of Moo cheese were extended to New South Wales. In August 1988, the respondent began marketing flavoured U.H.T. milk in New South Wales under the trade mark Moo.

In the Federal Court, the appellant sought to remove Moo from the Register on the ground that, at the time of the institution of these proceedings, the registration contravened s. 28. The appellant contended that, although the Moo mark had been correctly registered, events occurring since registration have made that mark "likely to deceive or cause confusion" contrary to s. 28(a). It followed, so the appellant asserted, that the mark was an "entry wrongly remaining in the Register" within the meaning of s. 22(1)(b) and should be removed.

Gummow J. said that, although s. 28 operates principally at the time of registration, it also operates to allow removal of marks which come to offend its provisions after registration and that such marks are marks "wrongly remaining" within the meaning of s. 22(1)(b). His Honour also held, following Berlei Hestia Industries Ltd v Bali Co Inc [25] , that the phrase "would otherwise be not entitled to protection in a court of justice" in par. (d) should be read as referring back to and qualifying par. (a) of s. 28. His Honour said [1] that "the words "or otherwise" [in s. 28] do cover cases where trade marks have lost distinctiveness by disuse (being marks not otherwise of strong inherent adaptation to distinguish) at least where the consequence is likelihood of deception or confusion" as well as cases of blameworthy conduct. His Honour was conscious of the fact that this conclusion was contrary to the holding in HTX International Pty Ltd v Semco Pty Ltd [2] . However, he drew support from Lord Diplock's comment in General Electric Co v General Electric Co Ltd ("the GE Case") [3] , that the words "or otherwise" may have been included to cover the cases of marks the title to which have been lost by disuse and from the comment of Kitto J. inBayer Pharma Pty Ltd v Farbenfabriken Bayer Aktiengesellschaft [4] , that the issues of deceptiveness and lack of distinctiveness could tend to run into one another.

1. (1973) 129 C.L.R. 353, at p. 361. 2. (1989) 86 A.L.R. 549, at p. 585. 3. (1983) 78 F.L.R. 57, at pp. 66-67; 49 A.L.R. 636, at p. 646. 4. [1972] 1 W.L.R. 729, at p. 746; [1972] 2 All E.R. 507, at p. 522. 5. (1965) 120 C.L.R. 285, at p. 303.

The above construction of s. 28 resulted in Gummow J. concluding that use of the Moo mark in relation to milk contravened s. 28; for his Honour found that the use of the Moo mark was likely to deceive or cause confusion and that it was not "otherwise entitled to protection in a court of justice" because of the lack of distinctiveness and the lengthy period of disuse. He ordered that the Register be rectified by limiting the goods in respect of which the Moo mark is registered to "cheese".

The judgment of the Full Court of the Federal Court

Lockhart, Pincus and von Doussa JJ. allowed an appeal from the judgment of Gummow J. [5] . They agreed with Gummow J. that s. 28 operates with respect to a mark that, although correctly registered, has since registration come to offend s. 28 and that par. (d) of that section qualifies the operation of par. (a). They also agreed that blameworthy conduct is not the only circumstance which renders a mark "not entitled to protection in a court of justice" under s. 28(d).

However, the Full Court held that lack of distinctiveness is not relevant under s. 28. First, the requirement of distinctiveness is dealt with at the time of registration by ss. 24, 25 and 26, not s. 28. Thus, it would be surprising, said the Full Court, if s. 28 were to deal with this issue after registration. Indeed, provisions in the Act other than s. 28 deal with lack of distinctiveness after registration. Thus, s. 23 applies to marks which lose their distinctiveness through non-use; s. 56 deals with marks that are no longer distinctive because they are generic. Secondly, the Full Court said that the interpretation of s. 28 proposed by Gummow J. would present difficulties in the case of marks in Pt B of the Register. A mark may obtain registration in Pt B even if it is not distinctive: s. 25. Such a mark may eventually be removed under s. 23 for non-use. But their Honours said that it is difficult to see how, consistently with the spirit of s. 25 of the Act, such a mark could be liable to expunction under s. 28 on the ground that it was not entitled to protection in a court of justice because of lack of distinctiveness.

In the courts below and even in this Court, the parties assumed that s. 28 applied to registered marks as well as to applications to register marks. But it is an assumption which should not be allowed to go unchallenged.

At first sight, s. 28 appears to operate only at the time of registration. The natural meaning of the words "shall not be registered" is that they refer to the act of registering rather than the state of being registered. This was the view that Lord Reid took of the corresponding U.K. provision in the GE Case. As Bradley, Loss of Distinctiveness, Deception and Expungement of Trade Marks, Sydney Law Review vol. 7 (1973) 62 points out, at p. 66 :

Standing alone, s. 28 appears to be designed only to prohibit registration of unregistered marks and thus to be addressed to the date of original registration. Section 28 speaks of "marks" simpliciter, not registered trade marks, and is found in Pt IV of the Act which is concerned to describe Registrable Trade Marks. This view of s. 28 is further supported by the subjunctive mood of "would be likely to deceive", "would be contrary to law" and "would otherwise be not entitled to protection". All these phrases indicate that the enquiry is to the prospective character of a mark as yet unregistered.

Nevertheless, it must be conceded that the word "registered" is capable of referring to a trade mark's continuing status. In fact, the use of the words "be registered", rather than "registrable" (see ss. 24 and 25) or "capable of registration" (see s. 33), gives support to the view that s. 28 is intended to have a continuing operation. Moreover, s. 11 of the Trade Marks Act 1938 UK ("the U.K. Act") (which corresponds to s. 28) has been interpreted so as to operate in respect of marks after registration. Section 11 provides:

It shall not be lawful to register as a trade mark or part of a trade mark any matter the use of which would, by reason of its being likely to deceive or cause confusion or otherwise, be disentitled to protection in a court of justice, or would be contrary to law or morality, or any scandalous design. The verb "to register" suggests that the section only applies as at the time of registration. Yet, in the GE Case, Lord Diplock thought [6] that as "a matter of ordinary usage of the English language" s. 11 could refer equally to "the act of making an entry on the register" or "the retention of an entry on the register". 1. [1972] 1 W.L.R., at p. 740; [1972] 2 All E.R., at p. 517.

Furthermore, other sections of the Act are said to support the argument that s. 28 was intended to have a continuing operation. Thus, s. 61(1) provides:

In legal proceedings relating to a trade mark registered in Part A of the Register (including applications under section 22), the original registration of the trade mark under this Act shall, after the expiration of 7 years from the date of the original registration, be taken to be valid in all respects, unless it is shown -

(a) that the original registration was obtained by fraud;

(b) that the trade mark offends against the provisions of section 28; or

(c) that the trade mark was not, at the commencement of the proceedings, distinctive of the goods or services of the registered proprietor.

(Section 13 of the U.K. Act does not have par. (c), otherwise it is substantially the same as s. 61.) In BALI Trade Mark [7] , Diplock L.J. drew attention to: The contrast between the past tense in paragraph (a) of section 13(1) "was obtained" and the present tense in paragraph (b) "offends", and the change from the words "that registration" in paragraph (a) to "the trade mark" in paragraph (b) and concluded: If the date of registration were the relevant date under paragraph (b), the natural way of expressing this would be to say, "unless that registration (a) was obtained by fraud, or (b) offended against the provisions of section 11 of this Act". See also the GE Case, per Lord Diplock [8] , and Berlei Hestia Industries Ltd v Bali Co Inc [9] , per Barwick C.J., and per Mason J. [10] . The use of the present tense in s. 61(1)(b), however, is explicable on the basis that, if contrary to s. 28 a mark has obtained registration, it "offends" against the provisions of that section. With great respect to those who have thought to the contrary, I do not think that the terms of s. 61 provide any sure foundation for cutting down the natural meaning of s. 28. 1. [1968] R.P.C. 426, at p. 434. 2. [1972] 1 W.L.R., at p. 749; [1972] 2 All E.R., at p. 525. 3. (1973) 129 C.L.R., at p. 356. 4. (1973) 129 C.L.R., at p. 360.

Support for the argument that s. 28 has a continuing operation is also said to arise from the terms of ss. 22(1)(b) and 56 of the Act. Section 22(1) provides:

Subject to this Act, a prescribed court may, on the application of a person aggrieved or of the Registrar, order the rectification of the Register -

(a) by the making of an entry wrongly omitted to be made in the Register;

(b) by the expunging or amendment of an entry wrongly made in or remaining in the Register;

(c) by the insertion in the Register of a condition or limitation affecting the registration of a trade mark which ought to be inserted; or

(d) by the correction of an error or defect in the Register.

(1) Subject to this section, the registration of a trade mark does not become invalid by reason only of the use, after the date of the registration, of a word or words which the trade mark contains, or of which it consists, as the name or description of an article, substance or service.

(2) The succeeding sub-sections have effect where -

(a) there is a well-known and established use of a word as the name or description of an article, substance or service by a person or persons carrying on a trade in that article, substance or service, not being used in relation to goods or services connected in the course of trade with the proprietor or a registered user of the trade mark or, in the case of a certification trade mark, in relation to goods or services certified by the proprietor; or

(b) the article or substance was formerly manufactured under a patent or the service formerly provided was a patented process, as the case requires, a period of 2 years or more after the patent has ceased has elapsed, and the word is the only practicable name or description of the article, substance or service.

(3) If the trade mark consists solely of that word, the registration of the trade mark, so far as regards registration in respect of the article or substance or of any goods of the same description, or of the services or of any services of the same description, as the case requires, shall be deemed for the purposes of section 22 to be an entry wrongly remaining in the Register.

The words "or remaining" in s. 22(1)(b) certainly demonstrate that a mark, originally validly registered, may nevertheless become one "wrongly remaining" in the register. However, in the present Act, it is possible to explain the operation of this paragraph by reference to s. 56(3) which deems certain marks to be "wrongly remaining in the Register". Those who rely on s. 22 as an indication that s. 28 applies to marks which become deceptive or confusing after registration point out, however, that there was no provision comparable to s. 56(3) in the Trade Marks Act 1905 Cth ("the 1905 Act") but that s. 22 had its equivalent in s. 71 of the 1905 Act. Nevertheless, the history of these provisions is not as illuminating as it appears to be. Under the 1905 Act, a registered trade mark was determinable with its goodwill: s. 58. Under that Act, therefore, a trade mark that was validly registered would wrongly remain in the register if the goodwill attached to the mark was lost. Consequently, s. 71 empowered the removal of such a mark because it wrongly remained in the register. Thus, the absence of any provision equivalent to s. 56(3) in the 1905 Act tells little about the construction of s. 22(1)(b). Its predecessor - s. 71 of the 1905 Act - operated in respect of marks which had lost the goodwill attached to them; and s. 22(1)(b) can operate in respect of marks which are deemed to wrongly remain in the register pursuant to s. 56(3). In any event, the terms of s. 56 are not necessarily inconsistent with the proposition that, under s. 28, likelihood of deception or confusion is a ground to be considered only at the date of registration. If that proposition is valid, s. 56 is simply an exception to the general rule that the likelihood of deception or confusion is a matter to be considered only as at the time of registration.

Enough has been said to demonstrate that neither the text nor the context of s. 28 provides any clear answer to the construction of s. 28.

Moreover the purpose of the legislation as a whole provides no real guidance as to whether s. 28 should be construed as operating after the act of registration. It has often been said that the purpose of s. 28 itself is the protection of the public: see, e.g., Eno v Dunn [11] ; Radio Corporation Pty Ltd v Disney [12] . If no more appeared, that purpose would be a powerful reason for giving the section a continuing operation, for the public is as easily deceived or confused by a mark which has become deceptive or confusing after registration as it is by a mark which would be deceptive or confusing if registered. But the Act has other competing purposes. Since the decision in Eno v Dunn , the scope of the legislation has changed. Thus, while the Act has the purpose of protecting the public by indicating the origin or nature of goods or services, it also has the purpose of protecting the valuable rights which a proprietor acquires in relation to a trade mark. Since the decision in Eno v Dunn , the enactment of provisions with respect to registered users (ss. 73-81) and assignment of trade marks without goodwill (s. 82) has changed the focus of the Act. The registered user provisions, the assignment provisions and the limited indefeasibility given to registered proprietors tend to suggest that, after registration, the interests of traders are preferred to the competing interests of consumers in being protected from the use of marks which are likely to deceive or confuse. Thus, the presence in the Act of ss. 73-82 tends to neutralise the inference to be drawn from s. 28 standing alone.

1. (1890) 15 App. Cas. 252. 2. (1937) 57 C.L.R. 448.

The Dean Committee (1954), on whose recommendations the present Act was adopted, stated that s. 28 "while in different words, has in substance the same effect as section 114 of the [1905] Act, but it relieves the Registrar and the Court from consideration of "morality"." Unfortunately, this statement does not assist in resolving the questions about the true construction of s. 28. First, the exact meaning of s. 114 of the 1905 Act is a matter of debate. Secondly, there is a significant difference in the wording of the two sections. Section 114 provided:

No scandalous design, and no mark the use of which would by reason of its being likely to deceive or otherwise be deemed disentitled to protection in a court of justice, or the use of which would be contrary to law or morality, shall be used or registered as a trade mark or part of a trade mark. (My emphasis.) The words "used or registered" in s. 114 suggest that that section had a continuing operation. If that is correct, the language of the report of the Dean Committee would require s. 28 to have a continuing operation. But this would mean that the Committee believed that "used or registered as a trade mark" in s. 114 was equivalent to "registered as a trade mark" in s. 28. This seems so unlikely that the omission of the words "or used" in s. 28 suggests that the Committee did not regard use after registration as any concern of s. 28.

An examination of the case law shows that the courts have also had trouble construing s. 28 of the Act and s. 11 of the U.K. Act. In R Thorne & Sons Ltd v Pimms Ltd [13] , Neville J. held that a trade mark should be expunged on the basis that it was "wrongly remaining" under s. 35 of the Trade Marks Act 1905 UK ("the 1905 U.K. Act") (the equivalent to s. 22) because it was "calculated to deceive" under s. 11 of the 1905 U.K. Act. His Lordship thought that the mark only became calculated to deceive by reason of facts occurring after registration. That is, he did not think the mark was "calculated to deceive" at the time of registration. Section 6 of the Trade Marks (Amendment) Act 1937 UK amended s. 11 so that, like s. 28, it now reads "likely to deceive or cause confusion" instead of "calculated to deceive". However, this amendment did not change the meaning of s. 11: in In re McDowell's Application [14] , it was said that "calculated to deceive" did not mean "intended to deceive" but meant "likely (or reasonably likely) to deceive or mislead".

1. (1909) 26 R.P.C. 221. 2. (1927) 44 R.P.C. 335, at p. 341.

Section 11 of the 1905 U.K. Act was also given a continuing operation by the Court of Appeal in Bowden Wire Ltd v Bowden Brake Co Ltd [15] . The marks of the plaintiff had become deceptive after registration because of a licensing arrangement. The Court of Appeal held that the plaintiff's trade marks should be removed from the Register because they were not distinctive and were calculated to deceive. The House of Lords upheld this decision [16] . In In re Pan Press Publications Ltd's Applications [17] , Wynn-Parry J. said that the real basis of this decision was that the licensing agreement was contrary to s. 70 of the 1883 U.K. Act or s. 22 of the 1905 U.K. Act. These sections prevented the assignment or transmission of trade marks without the goodwill of the business. Whether the comment of Wynn-Parry J. is or is not correct, the decision in Bowden Wire Ltd illustrates the need to treat many of the earlier cases with caution. The current Act sanctions the use of a trade mark by someone other than the registered proprietor (see ss. 73-81). In that context, the current Act arguably prefers the rights of registered trade mark proprietors to the interests of consumers.

1. (1913) 30 R.P.C. 580. 2. (1914) 31 R.P.C. 385. 3. (1948) 65 R.P.C. 193, at p. 200.

In In re Imperial Tobacco Co's Trade Marks [18] , Eve J. found that the trade marks under consideration were not deceptive, but added [19] :

In my opinion s. 11 has no application to, and it would be an abuse of language so to construe it as to make it have any application to, a mark which was in no sense contrary to law when registered and the use of which to-day is identical with that which has obtained ever since it was registered and for years before that. It does not, of course, follow that such a mark might not be so used as to become contrary to law or a deceptive mark within the section, but that is a wholly different case from the present. This statement is equivocal. However, in W Woodward Ltd v Boulton Macro Ltd [20] , Eve J. stated that s. 11 applies only at the time of registration. 1. [1915] 2 Ch. 27. 2. [1915] 2 Ch., at p. 40. 3. (1915) 32 R.P.C. 173.

In Pan Press, Wynn-Parry J. had to consider the validity of the respondents' two registered trade marks, a device consisting of a representation of the Greek god Pan and the words "Pan-Books". These marks were registered on 17 November 1943. The appellant was incorporated on 12 August 1943 under the name "Pan Press Publications Ltd." and commenced publishing in 1945. The appellant complained that the respondents' marks wrongly remained in the Register because they were, contrary to s. 11, likely to deceive or cause confusion. Wynn-Parry J. said [17] :

I confess that, had the matter been res integra, I should have had considerable doubt as to whether s. 11 could apply at all, having regard to the circumstance that, as was pointed out in the later case of Woodward v Boulton Macro, s. 11 only refers to the moment of registration. However, he felt bound by the decision in R Thorne & Sons Ltd v Pimms Ltd He concluded [21] , however, that: the Court will not direct a mark to be expunged of which it cannot be suggested that it was calculated to deceive when it was registered or was not otherwise disentitled to protection, but which in course of time, through circumstances over which the owner of the mark has had no control, has become calculated to deceive, unless, possibly, as was suggested by the Assistant-Comptroller, in a case where the public interest is strongly concerned and the damage which would be inflicted upon the registered proprietor by the expunction of his mark would be comparatively slight. I do not desire to be taken as expressing any view as to this possibility or doing more than leaving the matter open for consideration if such a case should arise. He held that the respondents were not responsible for any actual or potential confusion and upheld the registrations. This compromise is similar to that reached by Lord Diplock in the GE Case. 1. (1948) 65 R.P.C. 193, at p. 200. 2. (1948) 65 R.P.C., at p. 201.

The issue as to whether s. 28 has a continuing operation was again considered in the cases concerning the Bali trade mark. The applicant was the registered proprietor of the trade mark Berlei which was first registered in England in 1924 and in Australia in 1917. The respondent had registered the trade mark Bali in England in 1938 and in Australia in 1947. Both companies applied the mark to brassieres. The applicant ultimately succeeded in having Bali removed from the register in both countries on the basis that it was likely to deceive or cause confusion. In BALI Trade Mark, Lord Denning M.R. said that s. 11 only applied at the time of registration and Salmon L.J. said [22] that he was "far from convinced that the view expressed by Eve J. in Woodward Ltd v Boulton Macro Ltd [23] and by Wynn-Parry J. in Pan Press Publications Ltd's Application [24] , was wrong". Diplock L.J., on the other hand, considered that the section had a continuing operation and said that one had to assess the likelihood of confusion at the date of the application for expunction. Diplock L.J. largely based this conclusion on inferences drawn from the wording of s. 13(1) (the U.K. equivalent to s. 61(1)). In the House of Lords, the parties agreed that the case should be approached on the basis that the relevant inquiry was whether the Bali mark was correctly registered in 1938: Berlei (UK) Ltd v Bali Bra Inc [25] .

1. [1968] R.P.C., at p. 438. 2. (1915) 32 R.P.C. 173. 3. (1948) 65 R.P.C. 193. 4. [1969] 1 W.L.R. 1306, at pp. 1311, 1321; [1969] 2 All E.R. 812, at pp. 815, 825.

This Court did not decide the comparable Australian case until after the House of Lords decision in the GE Case. The judgments of Barwick C.J. and Mason J. were influenced by Lord Diplock's reasoning. Both their Honours said that s. 28 has a continuing operation. Mason J. said [26] :

for persuasive reasons the view that s. 28 (s. 11 of the 1938 [U.K.] Act) was merely a prohibition against original registration was rejected. Although there are differences in the expression of the United Kingdom and Australian provisions, they are not of such significance as to justify a departure from the conclusion, conformably with what has been said, that the words "shall not be registered" in s. 28 refer not only to the act of making an entry in the register but also to permitting the continuance of the entry of a trade mark in the register. The observations in Berlei Hestia Industries Ltd were obiter. The likelihood of confusion between Bali and Berlei was said to arise from their phonetic similarity, which was obviously the same at the date of registration as at the date of application for expunction. That is, the likelihood of confusion existed both at registration and at the time of the application for expunction. 1. (1973) 129 C.L.R., at p. 360.

The major case on s. 11 of the U.K. Act is the House of Lords decision in the GE Case. Lord Reid expressed agreement with the view of Eve J. in W Woodward Ltd v Boulton Macro Ltd Lord Simon of Glaisdale and Lord Kilbrandon, however, agreed with Lord Diplock who reiterated the view he had expressed in BALI Trade Mark. Lord Diplock held that s. 11 could in certain circumstances operate with respect to marks that had come to offend its provisions only since registration.

Since the GE Case, Australian courts have treated the debate about the scope of s. 28 and the meaning of the words "shall not be registered" as resolved. They have considered themselves bound by the obiter dicta of Barwick C.J. and Mason J. in Berlei Hestia Industries Ltd v Bali Co Inc and have held that s. 28 has a continuing operation: see HTX [27] ; Riv-Oland Marble Co v Settef SpA [28] ; and The Ritz Hotel Ltd v Charles of the Ritz Ltd [29] . Likewise, the text writers treat the issue as settled: see Shanahan, Australian Law of Trade Marks and Passing Off, 2nd ed. (1990), p. 138, and Lahore,Intellectual Property in Australia: Patent, Designs and Trade Mark Law, par. 4.3.023. However, the question is still open in this Court. Unfortunately, because of the course which the argument took in the Federal Court and in this Court, the question as to whether s. 28 operates by reference to events which occur after registration will not be settled by this case. My present inclination is that the section should be seen as operating as at the time of registration and not after. But the conduct of this case requires that the Court should proceed on the basis that the section applies to events occurring after registration.

1. (1983) 78 F.L.R., at pp. 63-64; 49 A.L.R., at p. 643. 2. (1988) 19 F.C.R. 569, at p. 594. 3. (1988) 12 I.P.R. 417, at p. 469.

2. Are the paragraphs of s. 28 to be read disjunctively?

In the GE Case, Lord Diplock thought that a limitation on the prohibition of the registration of a deceptive or confusing mark was imposed by the requirement that its use "would be disentitled to protection in a court of justice". He concluded [30] that:

If the likelihood of causing confusion did not exist at the time when the mark was first registered, but was the result of events occurring between that date and the date of application to expunge it, the mark may not be expunged from the register as an entry wrongly remaining on the register, unless the likelihood of causing deception resulted from some blameworthy act of the registered proprietor of the mark or of a predecessor in title of his as registered proprietor. The requirement of blameworthy conduct was derived from the equitable doctrine of unclean hands which Lord Diplock perceived as subsumed under the words "disentitled to protection in a court of justice". 1. [1972] 1 W.L.R., at p. 751; [1972] 2 All E.R., at p. 526.

Although the requirement of blameworthy conduct was at the heart of the dicta of Eve J. in Imperial Tobacco [31] and Wynn-Parry J. in Pan Press [32] , it cannot be said that, prior to Lord Diplock's judgment in the GE Case, it was established doctrine that marks which became deceptive after registration would only be expunged if the registered proprietor was guilty of blameworthy conduct. For example, in the GE Case in the Court of Appeal [33] , Salmon L.J. said:

I do not consider that the criterion as to whether it wrongly remains on the register because it has become likely to deceive or cause confusion can depend upon whether or not this deception or confusion has been caused by a wrongful act of the proprietor. The public mischief flowing from deception or confusion is the same whether or not it has been caused by a wrongful act. No doubt, however, the proprietor's conduct may be taken into account, amongst other things, when it comes to exercising the court's discretion as to whether the entry should be expunged from the register. 1. [1915] 2 Ch., at p. 40. 2. (1948) 65 R.P.C., at p. 201. 3. [1970] R.P.C. 339, at p. 368.

Further, the requirement of blameworthy conduct has not been universally accepted since the decision of the House of Lords. In Pioneer Hi-Bred Corn Co v Hy-Line Chicks Pty Ltd [34] , the New Zealand Court of Appeal said that the phrase "likely to deceive or confuse" was not modified by the words "disentitled to protection in a Court of justice". The Court drew attention to the difference in wording between the U.K. and New Zealand sections. Section 16 of the Trade Marks Act 1953 NZ provides:

It shall not be lawful to register as a trade mark or part of a trade mark any scandalous matter or any matter the use of which would be likely to deceive or cause confusion or would be contrary to law or morality or would otherwise be disentitled to protection in a Court of justice. This section is very similar to s. 28 of the Australian Act. Pioneer Hi-Bred Corn was an opposition case. In the GE Case, Lord Diplock appears to say that blameworthiness is only relevant where a mark had become deceptive or confusing since registration. Nevertheless, the comments of the New Zealand Court of Appeal were expressed in general terms. 1. [1978] 2 N.Z.L.R. 50.

Australian courts, however, have adopted the approach of Lord Diplock in the GE Case. In Berlei Hestia Industries Ltd v Bali Co Inc [35] , Mason J. said:

The meaning of the words "the use of which would be likely to deceive or cause confusion" in s. 28(a) is influenced by the language in which s. 28(d) is expressed when it refers to a mark "which would otherwise be not entitled to protection". In HTX [36] , Fox J. expressed agreement with Lord Diplock's judgment in the GE Case. In Riv-Oland, the majority of the Full Federal Court held that the word "otherwise" in s. 28(d) served to limit the circumstances in which the operation of s. 28(a) would disentitle a trade mark to the protection given by registration. In the present case, all members of the Federal Court accepted that the phrase "would otherwise be not entitled to protection in a court of justice" should be read as referring back to and limiting par. (a) of s. 28. They went on to say, however, that blameworthy conduct is not the only circumstance which would render a mark "not entitled to protection in a court of justice". 1. (1973) 129 C.L.R., at p. 361. 2. (1983) 78 F.L.R., at p. 64; 49 A.L.R., at p. 644.

In my opinion, an interpretation of s. 28 which holds that a mark does not offend its provisions even though the use of the mark is "likely to deceive or cause confusion" ignores the wording and structure of s. 28. As Northrop J. said in Riv-Oland [37] :

According to normal canons of construction the use of the word "or" between par. (c) and (d) makes each paragraph a true alternative to each of the other paragraphs. On this construction par. 28(a) stands alone. Similar comments were made in Pioneer Hi-Bred Corn in relation to s. 16 of the New Zealand Act which is almost identical to s. 28 [38] . The structure of s. 28 is so different from its predecessor, s. 114 of the 1905 Act, and its U.K. equivalent that the English cases, including the GE Case, cannot be regarded as persuasive. 1. (1988) 19 F.C.R., at p. 582. 2. [1978] 2 N.Z.L.R., at pp. 52 and 72.

Nor can I accept the conclusion that the word "otherwise" in par. (d) requires that par. (d) be read as limiting the operation of pars. (a) to (c). In Riv-Oland [39] , Lockhart J. said:

The use of the word "otherwise" in s. 28(d) appears to me to be otiose unless it refers back to the preceding pars. (a), (b) and (c). In my view, the syntactic function of that word requires that par. (d) be taken to involve such reference back, which in consequence will limit the circumstances in which the factors specified in pars. (a), (b) and (c) will disentitle a trade mark to protection. However, the word "otherwise" merely indicates that a mark which comes within par. (a), (b) or (c) is ipso facto considered to be not entitled to protection in a court of justice. As Northrop J. said in Riv-Oland [40] : The word "otherwise" when used as an adverb usually denotes "in another way" or "other ways" or "in other respects": see the Shorter Oxford English Dictionary. In the context of the syntax of s. 28 of the Act, the word "otherwise" is used correctly and does give a direct force to s. 28(d) to operate in its own right and not as a qualification on s. 28(a), (b) or (c). On this view s. 28(a), (b) and (c) each operate independently of par. (d) which is a catch-all provision proscribing the registration of a mark where that mark would not be entitled to protection in a court of justice even though the mark did not come within s. 28(a), (b) or (c). 1. (1988) 19 F.C.R., at p. 598. 2. (1988) 19 F.C.R., at p. 583.

Nor do I think that the fact that "some degree of deception and confusion seems to be inherent in the system when you have provisions such as s. 34 and s. 82" (Riv-Oland [41] , per Bowen C.J.) is a ground for reading down s. 28(a). Those sections together with ss. 73-81, which provide for a system of registered users, are the foundation of the argument that to remove a registered mark simply because it is confusing or deceptive is to contradict a basic assumption of the Act, viz., that registration is not inconsistent with deception and confusion. Accordingly, so it is argued, something more than confusion or deception must be required before a mark can be removed. However, as Lahore points out (at par. 4.3.030): "the present assignment and registered user provisions do not and were never intended to authorize the use of a mark in a way which was likely to cause deception and confusion." And the existence of s. 34 provides no ground at all for giving s. 28 an operation in respect of events which have occurred since registration.

Many of the recent Australian decisions which have held that a registered trade mark is not liable to expunction merely because it is likely to deceive or cause confusion have in truth been motivated by a desire not to reward "the assiduous efforts of an infringer". In Re Bali Brassiere Co Inc's Registered Trade Mark and Berlei Ltd's Application [42] , Windeyer J. asked rhetorically:

Is a mark which was valid when registered to be removed because later events, perhaps the assiduous efforts of an infringer, have created a likelihood of confusion? In HTX [43] , Fox J. said: It seems to me to be contrary to the scheme and policy of the Act to allow what Windeyer J. in the Bali Brassiere Case called "the assiduous efforts of an infringer" [42] to invalidate (or render void) a registered trade mark unless, at least, there has been "blameworthy" conduct on the part of the registered proprietor. Doubtless a registered proprietor is expected to take due steps for the protection of his mark, but the infringer must also be expected to take such steps as the Act allows in order to establish an equal or superior right to the use of the mark, or to have the register appropriately rectified. While registration by no means creates indefeasibility, it gives notice, and encouragement should not be given to the trader who fails to search it, or disregards the results of a search, and creates a substantial business using the mark, to the point that the registration is to be regarded as invalid, and liable to be expunged. In Riv-Oland [44] , Bowen C.J. said: I am, I think, conscious of leaning against a construction of s. 28 in regard to blameworthy conduct on the part of a proprietor which would lead to a situation described by Windeyer J where the proprietor of a mark might be denied or lose registration by reason of "the assiduous efforts of an infringer" , or to adapt the words to cover the period before registration "the assiduous efforts of a misappropriating user". In the same case [45] , Lockhart J. said: If s. 28(a) has an operation which is not restricted by reference to s. 28(d), then a trader who uses the registered trade mark of another trader may exploit that infringing use of the mark in the market-place to the point where it can be said that the use of the registered mark by the proprietor of the mark would be likely to deceive or cause confusion. It would be an extraordinary result if this conduct, which constituted infringing conduct, could produce either the invalidity of the mark or its liability to expungement. It is no answer to say that a court has a residual discretion to decline to order expungement even where a mark is shown to be invalid. The fact that the infringing conduct of a rival trader, in the circumstances to which I have referred, could produce the consequence that a registered mark becomes invalid is itself so extraordinary a result that its correctness could only be accepted as a last resort. 1. (1968) 118 C.L.R. 128, at p. 133. 2. (1983) 78 F.L.R., at p. 64; 49 A.L.R., at p. 644. 3. (1968) 118 C.L.R. 128, at p. 133. 4. (1988) 19 F.C.R., at p. 574. 5. (1988) 19 F.C.R., at pp. 598-599.

But as the author of an article in An Annual Survey of Australian Law 1988 points out (p. 295):

It is difficult to see the need for this massive swing from one extreme to the other: from denying the [infringer] her or his ill-gotten fruit to granting the [registered proprietor's] mark this presumptive immunity unless he or she was at fault. If the court had said that the [registered proprietor] would not lose a mark which had become deceptive as a result of an infringer's wrongful conduct, this would have taken care of the assiduous efforts of an infringer; but the proprietor would still lose the mark if it was deceptive for any other reason - that is, irrespective of who was to blame.

In any event, the registered proprietor of a mark which is constantly infringed by the assiduous efforts of an infringer is, to some extent, also guilty of blameworthy conduct. Registered proprietors have a duty to be vigilant in protecting their rights. If they are not, it is possible that s. 56(3) or s. 64(1)(b) will eventually diminish their rights. This caused Gummow J. in the present case to ask [46] :

Why should not s. 28 produce a comparable result where a registered mark becomes likely to deceive or cause confusion? 1. (1989) 86 A.L.R., at p. 577.

Moreover, if s. 28 is considered to have a continuing operation, it must be because its purpose is to protect the public in respect of marks which have been registered as well as those which seek registration. On that hypothesis, any requirement of blameworthy conduct on the part of the registered proprietor before that section can be utilized appears to be contrary to its essential purpose.

Accordingly, s. 28(a) should be given its natural meaning. That is, a mark offends against it if, at the time of registration or later, it is "likely to deceive or cause confusion".

Gummow J. found that the Moo mark was likely to deceive or cause confusion. The appeal should be allowed.