Amoena (UK) Ltd v HM Revenue and Customs

This appeal concerns the appropriate classification, and hence level of customs duty, for a mastectomy bra (“MB”). This is designed principally to be worn with an artificial breast form by women who have undergone surgical removal of one or more breasts. The MB in this case is known as the “Carmen” MB. It is imported by the respondent (“Amoena”).

There are two issues: (1) Is the MB a part of or accessory to a breast form? (2) Is it an orthopaedic appliance? In each case the question is to be answered by the instruments relevant to the classification of MBs, as explained below. If the answer to either issue is yes, then, as the Upper Tribunal (“UT”) (Judge Greg Sinfield and Judge Nicolas Alexander) held, the Carmen MB carries no duty. If, as the appellants (“HMRC”) declared in the Binding Tariff Information (an official written notice of classification) (“BTI”) on 16 October 2009, and the First-tier Tribunal (“FTT”) held, it is an article of clothing, it carries duty of 6.5%.

In its application for a BTI, Amoena explained:

The FTT found that all the features of a Carmen MB were found in an ordinary bra with the exception of :

the positioning of the pockets to hold the breast form. The FTT considered that such pockets might be present in an ordinary bra but that the wearer would use such pockets to insert padding to create an appearance of a larger bust;

the high cup to cover the breast form;

the positioning of the straps to the centre of the cups.

The FTT found that the primary purpose of the MB was to hold the breast form(s) in place, not to compensate for a defect or disability. Another purpose was the lessening of the psychological impact of having a mastectomy. It would be worn by women who had had a mastectomy but also by women who had had a lumpectomy or who had had breast reconstruction. The appearance of the MB was so close to that of a normal brassiere that it could not be determined from its objective characteristics that it was worn to compensate for a breast which had been removed.

The FTT declined to find that the MB had a subsidiary function of preventing shoulder pain or problems arising from the absence of lymph nodes.

In the EU, customs classification is carried out under a system known as the Combined Nomenclature (“CN”). It is based on the customs classification scheme agreed and used internationally by a large number of countries, called the Convention on the Harmonised Commodity Description and Coding System (“HS”). The EU is a party to this Convention. The HS consists of some 5,000 groups of goods with 6-digit codes. The CN integrates the HS but in addition contains further subdivisions with 8-digit codes, specifically adapted for the EU. Both the HS and the CN have explanatory notes (HSEN and CNEN respectively), which are prepared by experts. Courts generally give weight to these notes even though they are not legally binding.

There are certain rules for interpreting the HS and the CN set out in those instruments which the court must apply. These are called the “GIR”.

The CN, including the GIR, is contained in Annex 1 to Regulation 2658/87. Annex 1 is updated annually. The update with which we are concerned is that contained in Regulation 927/2012.

The CN is divided into three parts of which only the first two are relevant. The first part, headed Preliminary Provisions, includes the GIR. The second part, headed Schedule of Customs Duties, is divided into sections which are then sub-divided into Chapters with notes, headings and subheadings. I have added underlining below for ease of explanation later.

The two contending Chapters in the CN in this case are:

Chapter 62 for “Articles of apparel and clothing accessories, not knitted or crocheted” in Section XI, headed Textiles and textile articles.

Chapter 90. This covers a large number of miscellaneous items: “optical, photographic, cinematographic, measuring, checking precision, medical or surgical instruments and apparatus; parts and accessories thereof”. Chapter 90 appears in Section XVIII, which covers all these items and in addition clocks, watches and musical instruments.

Chapter 62 starts with a number of explanatory notes which include the following:

There is then a table setting out the CN 8–digit code in the first column and then a description of the item under what are called headings and sub-headings. The final columns in the table show the rate of customs duty.

So far as relevant the table shows as follows: CN code 6212 has a heading “Brassieres, girdles, corsets, braces, suspenders, garters and similar articles and parts thereof, whether or not knitted or crocheted.” The subheadings then are for bras made up into a set with briefs and other bras (6212 10 90).

Chapter 90 is headed “Optical, photographic, cinematographic, measuring, checking precision, medical or surgical instruments and apparatus; parts and accessories thereof.”

Chapter 90 also starts with explanatory notes of which the following are relevant:

preventing or correcting bodily deformities; or

supporting or holding parts of the body following an illness, operation or injury.

The same format of table follows. The relevant headings (shown in bold) and subheadings are as shown below:

Issues (1) and (2) arise from respectively subheading 9021 90 90 of, and EN 6 to, CN heading 9021. It is common ground that EN 1(b) to Chapter 62 does not apply.

Chapter 62 of the HS creates a category of bras but this does not contain an exception for orthopaedic appliances:

Brassières of all kinds…”

Chapter 90 of the HS has similar headings and subheadings to CN heading 9021. However it also provides a considerable amount of narrative about the items in heading 9021. For example, it contains the following narrative on orthopaedic appliances:

The relevant GIR in the CN provide:

As explained below, the Court of Justice of the European Union (“CJEU”) has given guidance on the interpretation of the CN. All the authorities cited below are from the CJEU.

The (amended) decision of the FTT released on 28 November 2011 did not deal with Issue (1).

On Issue (2), the FTT rejected the argument that the MB was an orthopaedic appliance within Chapter 90 because it was distinct from the breast form and did not perform any corrective functions on its own. The breast form’s function could not determine the correct classification for the MB. It was irrelevant that artificial eyes fell within 9021 even though the eyes could not take over the function of a normal eye since the question was whether the MB, and not the breast form, was an appliance.

The UT came to the opposite conclusion. In its judgment dated 12 August 2013, it held as follows:

The Carmen mastectomy bra is a part of or accessory to the breast form in that the bra is used in conjunction with the breast form to enable both to function as a prosthesis for the natural breast that has been wholly or partly removed or reconstructed.” (judgment, para 61).

Therefore (applying Note 2(b) to CN Chapter 90) the MB was an artificial body part to be classified under 9021 39 90 (judgment, para 62).

The FTT were correct to hold that the MB did not correct or prevent any deformity.

The second indent of Note 6 to Chapter 90, which defines orthopaedic appliances, was capable of including supports provided to an artificial body part (in this case the breast form).

The MB could be classified as an orthopaedic appliance under CN subheading 9021 10 10 “because it performs the function of the missing muscles and tissue in supporting or holding the breast form or forms, which are a part of the body, following the mastectomy” (para 69). The relevant part of the woman’s body being supported/held by the mastectomy bra was the breast form, meaning that the mastectomy bra fell within the second indent of the Note 6 definition (judgment, para 70).

The UT accordingly held that the MB could be classified either (per Issue (1)) under 9021 39 90 as an accessory, applying Note 2(b) to Chapter 90, or (per Issue (2)) under subheading 9021 10 10 as an “orthopaedic appliance” in its own right. Selecting between the two subheadings, the UT applied GIR 3(a) and held that the MB should be classified as an “orthopaedic appliance” within the meaning of 9021 10 10.

The UT dealt with a number of issues which do not arise on this appeal, including the question whether a MB can be classified under Chapter 90 when a bra is also used for non-medical purposes, and the question whether mass production disqualifies the MB.

I take the issues in the order in which they were argued and appear in the decision of the UT, though they could perhaps more easily have been taken in reverse order.

Mr Sarabjit Singh, for HMRC, submits that the UT was wrong to find that the MB was a part or accessory. Mr Singh submits that the correct approach to parts and accessories is to shown by Cases C276/00 and C250/05 Turbon International GMBH v Oberfinanzdirektion Koblenz, where the classification of a printer cartridge was considered. The CJEU held that the correct analysis was that the printer cartridge was an ink product since its essential characteristic was that it contained ink and it was therefore an ink product. It rejected the argument that the printer cartridge was a part or accessory for the purpose of the Chapter then under consideration because the printer could work even if a printer cartridge was not fitted (even though it would not be able to produce any readable copies). The printer cartridge was a significant product in its own right. On Mr Singh’s submission, Turbon is to be contrasted with a case which I shall consider in detail below, namely Case C-514/04 Uroplasty BV v Inspecteur van de Belastingdienst – Douana district Rotterdam [2006] ECR 1-67219, on which the UT relied, where the hydrogel was simply an ingredient, and not a finished product like an MB. Therefore, Mr Singh submits, the UT was wrong to place reliance on Uroplasty.

As to the question of accessory, Mr Singh, on discovering C152/10 Unomedical v Skatteministreet (Unomedical) at a late stage in his submissions, submits that the word “accessory” is to be given a restricted meaning as a matter of law in accordance with that case. The breast form was capable of functioning without the MB and so the MB was not an accessory to it.

For Amoena, Mr Tim Eicke QC emphasises that this is an appeal on the point of law and that therefore that HMRC have to show an error of law. If there is no definition in the GIRs or in the notes, then the ordinary language must be applied. The UT concluded that the MB was an accessory.

Mr Eicke agrees that the judgments in Uroplasty and Unomedical provided the most assistance. He submits that Turbon is distinguishable because the breast form could not perform its function without the MB. The test was correctly set out at the end of paragraph 57 of the judgment of the UT as follows:

On Mr Eicke’s submission, paragraph 32 of the decision of the FTT can only be correct through the combined action of the breast form and MB. The UT was looking too narrowly at the individual parts of the MB.

Mr Eicke also relies on the definition of “accessory” in the Medical Devices Directive (93/42/EEC) (“the MDD”), but the UT held that this was not relevant. The term “accessory” is defined in the MDD as follows in terms which are sufficiently wide to include the MB:

I start by setting out the reasoning of the UT on this issue in full to show how, understandably, they reached their conclusion:

It is apparent from this extract that the parties did not cite the relevant authorities on “part” or “accessory” for the purposes of Note 2(b) to Chapter 90 to the UT. The CJEU interpreted these terms most recently and relevantly in Case C-152/10 Unomedical. The issue was the correct classification of urine drainage bags designed to be used in connection with dialysers and catheters. The function of the bag was to collect urine. It did not contribute to the process of dialysis or the work of the catheter. The bags were plastic and therefore provisionally fell within CN Chapter 39 (plastics and articles thereof), but Chapter 39 contained an exclusion for articles within Chapter 90. One possibility, therefore, was that the bags were parts or accessories of the dialyser or catheter, which were classified under Chapter 90.

The CJEU applied Turbon, noting that in that case the CJEU had held that:

The CJEU further held that this was the meaning of parts or accessories in Chapters 84 and 85, and that the same meaning applied in Note 2(b) to Chapter 90 (para 34).

The bags could not be a part of the dialyser or catheter since they were not essential to the functioning of either instrument. Those appliances could function without a bag.

The bags were also not “accessories” since they did not improve the performance or functionality of the dialyser or catheter.

As to this case, the MB is not a part of the breast form in the sense given in Unomedical since it is not essential to the operation of the breast form. (The breast form may be, and in some instances is, simply strapped to the body.) The function of the breast form is to simulate the appearance of the natural body. It is not a question of whether the wearer perceives the breast form to be, or have become, a part of her body as the question of classification must always be determined by reference to the objective characteristics of the article in question.

In my judgment, contrary to the submission of Mr Eicke, the MB is also not an accessory of the breast form for the purposes of Chapter 90. It does not improve the performance of the breast form or give it any additional functionality which it is not capable of having without the MB.

As to the definition of “accessory” in the MDD, like the UT, I see no basis for applying this definition to the CN.

Note 6 to Chapter 90 contains a definition of “orthopaedic appliance”: see paragraph 17 above. This issue turns on the meaning of that Note.

Mr Singh submits that the first indent cannot apply because it is the breast form, and not the MB, which prevents or corrects any deformity.

As to the second indent, Mr Singh’s short point here is that an orthopaedic appliance must support a natural part of the body, and the UT were wrong to conclude otherwise. Supports for artificial parts of the body are not within the heading 9021 or its subheadings. The only subheading under which it could fall is 9021 90 90, but the heading requires that the appliance should “compensate for a defect or disability”. This does not, on Mr Singh’s submission, apply here because (1) it is the breast form, and not the MB, which compensates for the defect; and (2) the absence of a breast is not a “disability”. Even if it is, the MB does not compensate for it.

Mr Singh submits that HMRC’s interpretation of 9021 90 90 is consistent with the CNENs for heading 9021 which state that:

Moreover, Mr Singh informs the court that HMRC is not aware of any country in which the MB is treated as an orthopaedic appliance. He notes that Amoena has not produced any evidence that it is so treated even though it exports MBs to some 60 countries. Mr Singh further informs the court that there is evidence that it is subjected to duty under Chapter 62 in two states of the USA and in Ireland.

Mr Eicke relies on the approach of the CJEU and Advocate General in Uroplasty to the interpretation of heading 9021. He submits that the expression “appliance” has to be interpreted to include a wide range of supports, even those which are not technical. Also, in examining the article to be classified, regard is to be had to its objective characteristics, including its intended purpose, and, on Mr Eicke’s submission, this was the case even if the purpose was not objectively ascertainable at first inspection. The broad approach, approved by the CJEU, means that the MB should be treated as for the purpose of compensating for a defect or disability. This follows, on Mr Eicke’s submission, from paragraph 61 of the Opinion of the Advocate General in Uroplasty, which states:

The CJEU specifically agreed with the Advocate General’s point here, and held that an appliance did not have to have any particular element of technical structure:

Mr Eicke submits that the UT were right to hold that the MB is an orthopaedic appliance because it substitutes for the muscles that hold up the breast form “which are part of the body, following the mastectomy.” (UT judgment, para 69). This gives effect to the broad approach to interpretation mandated by the CJEU. The wide interpretation of Note 6 is, on his submission, reinforced by the references in Chapter heading 9021 to the expressions “the body” and “any artificial part of the body” in line 2 of the heading. He submits there is no suggestion that the appliance to hold an artificial leg in place is outside Note 6.

Lastly Mr Eicke relies on Case C-260/00 to 263/00 Lohmann GmbH & Co KG and Others v Oberfinanzdirektion Koblenz. In this case, the national court faced a dilemma because various supports were said to be orthopaedic appliances when the same items could be used for ordinary purposes, e.g. knee supports. The essence of the guidance given by the CJEU on the question how to determine whether these items were orthopaedic appliances was in paragraph 45 of its judgment:

Mr Eicke submits that the MB has features which distinguish it from an ordinary bra. Further, the MB is worn for a different purpose from that for which an ordinary bra was worn. In addition the MB has to be fitted by a nurse. Mr Eicke submits that HMRC look too closely at the components of the appliance instead of it as a whole, including its purpose, and that the objective criteria show differences which exist beyond mere appearance. It is, he submits, wrong to focus on the fact that it can be used to perform some other function. This must be disregarded, as where an actor wears an appliance (e.g. a crutch) for some theatrical role. The market reality is that the MB is virtually only worn by women who have had parts of a breast removed. It is plainly more relevant to them. As stated above, they are fitted for an MB by a nurse. It is clear from the evidence in the market that there is specificity of purpose.

I start with the general approach to interpretation of the CN.

It is clear from the Opinion of Advocate General Kokott in Uroplasty that the court must apply a structured approach. At the first stage it must determine the intended use and material composition of the article. Next the court must make a provisional classification by reference to section and chapter headings. Then the court must make a combined examination of the headings and Notes, applying GIRs 2 to 5 in case of conflict. The interpretation of the headings and EN should be consistent with the HS. Finally the article must be placed under the appropriate subheading. The relevant paragraphs in the Opinion are as follows:

The CJEU emphasised that the determination of the characteristics and properties of the article must be an objective one, and that the wording of the CN must prevail over the EN, which cannot alter the scope of the headings:

The Advocate General also states that the term “appliance” should be interpreted broadly: see paragraph 61, set out above. Some insight into the point she was making is to be derived from her footnote 16, which states:

The article must still, therefore, have a medical purpose.

The UT correctly drew attention to the fact that Uroplasty contains further important guidance on the interpretation of the term “orthopaedic appliance”. In the opinion of both the Advocate General and the CJEU, the term “orthopaedic appliance” had to be given a wide meaning. At the time there was no Note 6, which was introduced subsequently, but there is nothing to indicate that Note 6 was intended to affect the reasoning in Uroplasty. The case concerned silicone flakes which were injected into the body, with a substance called hydrogel, to alleviate incontinence. They could not perform the same function as a muscle but they substituted for its failure. Moreover, an orthopaedic appliance did not have to be a technical apparatus.

In my judgment, the facts of Uroplasty do not provide much assistance on the facts of this case. Nothing is said in Uroplasty about supporting artificial parts of the body. The UT took the view that in Uroplasty the flakes substituted for muscle and that the hydrogel performed the function of the MB (see para 49 of the UT’s judgment). However, as the UT pointed out, the classification of the hydrogel was not in issue and so the analogy is of limited use. Mr Singh makes the further point, with which I need not deal, that the analogy between hydrogel and the MB is also not correct because, unlike hydrogel, the MB is not a single use item. It could be used without the breast form, or by women who have had a lumpectomy rather than a mastectomy.

I turn next to the structured approach in Uroplasty explained above, starting with the characteristics and properties. The FTT has found these, as set out above. In particular, the FTT found that the primary purpose of the MB was to hold the breast form(s) in place, not to compensate for a defect or disability. Thus the first indent in Note 6 cannot apply.

I can deal swiftly with the sections and the Chapters as no-one has suggested that the narrative in them is important in this case. The candidate sections are Sections XI and XVIII, but under GIR 1 the wording here does not play any part in interpretation. The candidate chapters are Chapters 62 and 90.

The next task is to compare each of the headings and notes. The heading of, and notes to, Chapter 62 disclose that this Chapter applies to clothing and not to orthopaedic appliances. The fact that the HS does not perceive any overlap between bras and orthopaedic appliances suggests that no significant overlap is contemplated. The heading of, and notes to, Chapter 90 disclose a wide range of objects. Note 1(b) shows that textile support articles are excluded but only where they depend solely on elasticity. The reference to Section XI in brackets at the end of this Note is no doubt so that the user is specifically referred to Chapter 62. No-one has suggested that Note 1(b) constitutes the limit of the overlap.

Note 6 specifically defines “orthopaedic appliances” for the purposes of heading 9021 and does so by reference to their function or purpose, which I take to be their primary purpose, and not their physical characteristics. Mr Eicke repeats his submission that the term “orthopaedic appliance” is to be interpreted broadly. He submits that the MB qualifies because it has a medical purpose.

While the HSENs are not binding on the CJEU, they are important as an aid to interpretation as a means of ensuring the uniform application of the EU Customs Code: see, for example, Lohmann where the CJEU held:

As Mr Singh submits, the HSENs make it clear that the heading 9021 does not apply to products which merely alleviate a condition. They must compensate for a defect or disability even though this is not expressly stated in the heading to be a required characteristic of each of the items mentioned in the heading. It would follow that an article within the particular category on which Amoena relies – 9021 90 90 - would also have to compensate for a defect or disability in the sense given in the HSENs.

In my judgment, given that (1) heading 9021 does not indicate that there can be articles within heading 9021 which do not compensate for a defect or disability, and (2) when the Advocate General opined that the expression “orthopaedic appliance” should have a broad meaning, she was particularly addressing the ambiguity in the word “appliance” as used in the English language version of heading 9021, I consider that the correct interpretation is that, to be an orthopaedic appliance within heading 9021, the product must compensate for a defect or disability. That is the function of the breast form. The MB clearly does not do this.

Mr Eicke repeatedly submits that the FTT found that a woman wore the MB in order to compensate for a defect or disability. But the FTT, as I read their decision, did not say this. In the judgment of the FTT the MB was worn to hold the breast form in place and the breast form compensated for the defect or disability.

I move to the second point: does Note 6 apply to supports for artificial body parts, as the UT held? In my judgment, the answer to this question is no. I start by making some observations about Note 6. It does not prescribe the material composition of orthopaedic appliances. Rather orthopaedic appliances are defined by their function. Their functions are limited to the two indents in Note 6. Note 6 exhaustively defines the functions and therefore the range of orthopaedic appliances for the purposes of Chapter 90.

The UT held that Note 6 was satisfied on the basis that “parts of the body” in the second indent included artificial parts of the body. On this appeal, reliance is only placed on the second indent.

The scheme of the two indents is important. The first covers a deformity which can be prevented or corrected by the appliance. This seems to be directed at remedying the situation caused by a deformity, e.g. loss of teeth. The second indent appears to include measures of a supportive nature, which may be only temporary such as splints and slings. Heading 9021 is consistent with this although it reverses the order of the indents. Taking Mr Eicke’s example of a strap holding up an artificial leg, there is no reason to mention the express purpose of supporting a prosthesis because in an appropriate case this would be a part of the orthopaedic appliance in question to which Note 2(b) would apply. In fact, the range of items that can be within the second indent because they support a body part is reduced to some extent by the fact that purely elasticated items have already been excluded by Note 1(b) to Chapter 90.

Most importantly, it is clear that other references to a body in Note 6 and heading 9021 cannot include artificial body parts: see the reference to “bodily deformity” in the first indent and “body” in heading 9021, which must refer to the natural body. These terms are used in contrast to the words “artificial parts of the body” in heading 9021. Where the drafters of the CN wish to refer to artificial body parts, they say so.

The MB in fact falls outside Note 6 on the findings of the FTT because it supports the breast form and it is the breast form which corrects the deformity. There is no difference for this purpose between a disability and a deformity.

That takes me to Mr Eicke’s point on Lohmann. Effectively he argues that this case lays down a separate test for an “orthopaedic appliance” where various types of the same article are used for ordinary purposes. The test on the basis of this case, he submits, is whether there are characteristics which distinguish them from articles in ordinary use. He refers to the evidence given before the FTT.

I read Lohmann as providing guidance on the question how heading 9021 applies where there are in ordinary use articles of the same types as that for which relief from duty under heading 9021 applies. I do not read it as dispensing with the other requirements for an orthopaedic appliance. But even if I am wrong on this point, the CJEU’s holding in paragraph 45 is focused on what the article reveals about itself, i.e. its appearance.

As Mr Singh submits, the material used was the same as for an ordinary bra, and the fastening operated in the same way as a normal bra. The MB had pockets but then according to the finding of the FTT so did some normal bras so this was not a distinguishing characteristic. The three distinguishing features found by the FTT are set out in paragraph 4 above. Despite these features, the FTT found:

In this passage, the FTT asks itself whether the objective characteristics of the MB disclosed that the article was worn to compensate for a defect or disability. It would appear that the CJEU was not as precise as this and envisaged that the fact-finder would ask whether there were features to distinguish the bra from one in ordinary use. But I do not consider that there is any significant difference in this context between those two questions. If a person could not tell that it was a MB, he could not tell that it was not for ordinary use.

In the light of the appearance of the MB, Lohmann does not assist Amoena.

Finally, Mr Eicke relies on the fact that the MB would be within the meaning of orthopaedic appliance in the MDD. As HMRC accepts, the MB is zero-rated for VAT purposes under the MDD because the MB falls within the definition of “medical device” in the MDD. This provides in material part that “medical device” means “any instrument, apparatus, appliance, material or other article, whether used alone or in combination… to be used for human beings for the purpose of… diagnosis, monitoring, treatment, alleviation of or compensation for an injury or handicap…”.

Mr Singh submits that the fact that medical devices are zero-rated for VAT as aids to the handicapped does not help for customs classification which concerns countries outside the EU. I agree. I do not consider that this answer can be different because the MB is treated as a medical device for VAT purposes.

In addition, in agreement with the UT (para 55), I do not consider, in the absence of authority, that it is correct to say that legal certainty requires that the term “orthopaedic appliance” must be defined in the same way in the CN as “medical device” for the purposes of the MDD. This Directive deals a with very different subject-matter.

I would therefore allow the appeal on this issue.

For the reasons given above, I would allow this appeal. The UT were in error in holding that the MB was an “orthopaedic appliance” within heading 9021 of Chapter 90 in the CN. It therefore falls within Chapter 62 of the CN.

I am most grateful to Arden LJ for her comprehensive citation of the various regulations, notes and guidance which fall for consideration and for her statement of the facts and procedural history of the case. In this judgment, I will adopt the definitions of terms that Arden LJ uses.

I admit to having found the case a difficult one to decide, not least because of the extraordinarily labyrinthine quality of the instruments that we have to construe and what appears to me to be some challenging logic in parts of the decided cases that were cited to us.

On issue (2) (orthopaedic appliance), I find myself in broad and respectful agreement with Arden LJ that the MB is not an orthopaedic appliance. As a matter of strict construction of the CN, the MB could only fall within the description:

Turning to issue (1), however, I find this more difficult still. It is common ground and agreed on all hands that the breast form itself is an artificial part of the body and would qualify for inclusion in 9021 under 9021 39 90. In general, accessories to parts included in Chapter 90 are to be included in the Chapter. In any normal parlance the MB would surely be an accessory to the breast form. It holds it up. Certainly, there was evidence that there were other means, apart from MBs, for securing breast forms, but in one way or another something has to be done to prevent the form simply sliding down the body and out of place. The MB is, therefore, in my view, as a matter of simple common sense, an accessory designed to perform that imperative function in relation to the breast form.

However, is that the correct analysis under the CN, as interpreted in the cases decided in the CJEU?

Again, as a matter of language/construction of the instrument, I would conceive the answer to this question to be “Yes”. That answer would depend upon Note 2 at the beginning of Chapter 90. I would consider that the MB would properly fall within the description of an accessory

We derive from the opinion of the Advocate-General in Uroplasty (supra) (paragraphs 59 and 60) that the word “appliance” in this part of the instrument is of wide meaning, including “apparatus”, “instrument” or “device” or the equivalent of the French term “article et appareils”. “Appareil” can be rendered in German as “Apparat” or “Gerät” and “article” corresponds to the German “Ware” (product) or “Gegenstand” (object). But, of course in any event, whether an appliance or not, the breast form is a qualifying artificial body part and any proper accessory to it falls within the same classification.

Among the meanings for “accessory” in the Oxford English Dictionary is:

As a matter of language alone, therefore, I would be inclined to hold that the MB is an accessory for an artificial body part, namely the breast form.

Is this, however, compatible with cases concerning the concept of “accessory” in the CN, as decided by the CJEU?

While a cursory internet search reveals many entries for printer ink cartridges under “printer accessories”, the CJEU has held that such a cartridge is not an accessory to a printer.

In Turbon (supra), we find in paragraph 29 of the judgment that in view of the fact that a cartridge such as this was specifically designed for, and could only be used by, a particular type of printer, and had a technically sophisticated ink flow mechanism accordingly, the national court wondered (perhaps not surprisingly) whether the cartridge should not be regarded as “part” or “accessory” of the printer.

The CJEU (in the passage which I found logically challenging) held (in paragraphs 30 and 31) that the cartridge was not a “part” of the printer and, on the question of whether it was an accessory, (in paragraph 32) said as follows:

The court analysed the question as it did by reference to an HS explanatory note relating to heading 8473 which it quoted in paragraph 12 of the judgment as follows:

For my part, I would have thought that a printer cartridge was classically something which “performs a particular service relative to the main function of” the printer, namely to enable it to print, but apparently not.

I was inclined to think that the problem caused in my mind by this decision on heading 8473 would not matter in the present case because it depended to a significant degree upon an explanatory note to heading 8473. We are concerned with headings 9021 and following. However, we then get to the decision of the CJEU in Unomedical (supra) which concerned plastic dialysis drainage bags (intended for exclusive use with dialysers) and plastic urine drainage bags (intended exclusively for use with catheters). The competition here was between headings 9018 (within Chapter 90 with which we are concerned) and 3926 (“plastics and articles thereof” with which we are not).

The court held that these items were not accessories for the dialysis machines or the catheters. They fell within 3926. In so deciding, the court said this in paragraphs 29 and 30 of its judgment:

So, is the MB an accessory to the breast form in the sense that it “perform[s] a particular service relative to the main function” of the breast form. As might be understood from the above, I think that, on any natural meaning of these words, it does precisely that; it holds it in place and prevents it sliding off the body, even if there may be other devices that are capable of achieving that result.

Is that enough for us to hold that the MB is an “accessory” to the breast form, if the law that we have to apply is that an ink cartridge for a printer does not perform such a service for the printer? Regrettably, I feel constrained to hold that the answer is “No”.

On this basis, the MB would not be an accessory to the breast form within the meaning of Chapter 90.

With reluctance, therefore, I agree that the appeal must be allowed.

I am also most grateful to Arden LJ for her comprehensive review of the relevant regulations, notes and guidance together with her statement of facts and procedural history. I, like McCombe LJ, find the logic in parts of the decided cases testing and indeed I too would have thought that a printer cartridge was typical of something which 'performs a particular service relative to the main function of' a printer, namely to enable it to print.

The analysis of Arden LJ and McCombe LJ however lead me to the reluctant conclusion that, notwithstanding any view I may hold as to the natural meaning of the relevant words, the law is such that I also can only conclude that the MB would not be an accessory to a breast form within the meaning of Chapter 90.

Accordingly I too would allow the appeal.