England and Wales High Court (Chancery Division) Decisions

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
PATENTS COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		21/04/2008

Mr Justice Mann :

Introduction

1. This is a patent action in which the claimant ("Buhler") sues the defendant ("Spomax") for infringement of its patent, being a European patent for a milling process and associated apparatus, No 0336939 ("the 939 patent" or "the Patent"). If the patent is valid, then infringement is admitted. However, Spomax counterclaims for revocation on the grounds of obviousness, and seeks to attack the patent in an amended form on the basis of added matter. The claimant also makes an application for an amendment of the 939 patent; as will appear, I have heard this action on the footing of the proposed amended patent but have not fully considered the amendment application.
2. This action has changed its shape considerably since it was commenced. As originally brought it also involved another of Buhler's patents ("the 919 patent"), but once certain prior art was pleaded, it was accepted that that patent could not be sustained and, after a minor tussle between the parties as to the correct way of dealing with that, it is now accepted that I should make an order revoking the patent. I shall do so, and need say little or no more about it. Furthermore, after the close of evidence in the case, Buhler abandoned its maintenance of the validity of roughly half the claims in the patent in suit (apparatus claims) and accepted that they were invalid and fell to be revoked. That left just a method claim.

The background to the patent – milling technology

3. For present purposes a grain of wheat can be treated as having three parts – a coating of bran, surrounding a body of endosperm and a small germ. The trick of milling to obtain flour is to extract the endosperm by removing the bran, and then processing the endosperm so as to produce flour of an appropriate quality. The germ can be ignored for these purposes. It is the endosperm which ultimately makes the flour.
4. This process was originally done on millstones. Grain was fed in at the top and was crushed between the two stones (one of which turned on the other). The stones crushed the grain and the endosperm (flour) was released. This was a crude process. It did not permit much separation of the bran (and anything else that one would not want if one were producing white flour) from the endosperm, and the endosperm was not really processed. In about the mid 18^th century, probably in France, millers started to have the idea of successive grindings, setting the millstones farther apart than hitherto, and sifting in between, to produce a more refined flour with the possibility of removing bran. If bran is left in the flour, it is a wholemeal flour. The more bran that can be removed, the whiter and purer the flour. Since bran impedes the rising of bread, the rising qualities of the flour are also improved by bran removal, and a baker would want that too. This process of successive levels of grinding is now called "high grinding" as opposed to "low grinding" which describes a cruder process. The cruder processing in millstones is also sometimes called "sudden death" milling.
5. At some point in the first half of the 19^th century, millers started to use rollers rather than millstones. The process probably originated in Hungary, but it spread throughout the world by the end of the century. In this process wheat is fed into a pair of rotating rollers. The rollers are rotating against each other at different speeds. The grains are crushed by the rollers and as a result of the crushing and the operation of the rollers, the endosperm is scraped out. In a typical bread flour mill in 1987 (the year of the priority date of the patent in suit) there would be a whole series of rollers through which the increasingly processed product would pass, one after the other. The rollers were essentially of two kinds. The first ("break" rollers) performed breaking. This is a fairly crude process (in milling terms) in which the grain is passed between crushing rollers which have grooves or flutes in them. Having passed through several of these sets of rollers sequentially, the grain is gradually broken open. If white flour is being produced, bran has to be removed. This is done by sieving after each bit of rolling so as to remove the bran. There is also a process of purification involving air flow in order to remove bran, into which I need not go. This takes place after some of the rolling stages.
6. When the grain has been sufficiently processed at the break stage, the product passes through another series of rollers, this time generally smoother, which break down the product (by now mainly the endosperm) into smaller and smaller particles. This phase is called reduction. The rollers are set progressively closer together in order to achieve this. Again, for white bread flour there will be sieving or other separation after each set of rollers. Typically speaking, there will be between 3 and (at the most) 5 rollings in a break stage, and between 10 and 15 or more rollings in a reduction stage. The miller will judge how he sets his rollers, and the degree of processing, depending on, amongst other things, the nature of the product fed in and the quality of the product he wishes to extract. Part of the purpose of this stage in the process is to break down the starch in the milling process – the baker wants the starch broken down to aid water absorption.
7. That basically describes the process of extracting flour for bread-making, cake-making and other culinary purposes (which can be treated as one for the purposes of this action). However, as well as being the source of flour for bread-making and other culinary purposes, the endosperm of wheat grains can also be used for making flour from which starch is extracted. This is a relatively recent process. Until the 1960s maize had been the principal source of starch. It was extracted by a process involving vigorous breaking down of the kernels, followed by a wet extraction process involving some chemicals. From the 1960's Canadian and Australian millers looked to wheat for starch, and processed it for that purpose. The process involves milling the wheat to extract flour, and then subjecting the flour to a process involving the addition of water and the separation of the resulting paste into starch and gluten.
8. There is a significant difference between the processing of the flour required for bread (and other culinary purposes) and the flour required for starch. The endosperm contains the starch. The trick in milling for bread-making flour is to break down that starch, because to do so improves the extent to which water is absorbed. That is achieved by more and harder grinding of the product, particularly in the later stages when it is passing through the reduction process. A maker of starch, on the other hand, requires flour in which the starch has been damaged as little as possible. That means fewer processing steps, and in particular fewer reduction rollings. In addition, since the starch is extracted by a process involving water, and while the starch producer does not want a lot of bran in his flour, he is probably (and I find) more tolerant of bran in his flour than a customer of white bread flour would be. Nevertheless, in the 1980s in a typical process for producing starch flour, sieving was done at various stages in order to remove bran (and other large particles).
9. There is one other piece of terminology which it would be useful to introduce here. A rollermill is a machine comprising one or more pairs of rollers. A single rollermill contains just one pair, rotating against each other so as to grind in the manner referred to above. A double rollermill contains two pairs of rollers side by side, each with its own feed of stock going into it, processing in parallel. A 2 pair high rollermill contains two pairs, one above the other, with the output from the top pair passing into the pair below. A 3 pair high rollermill, as its name will now suggest, is three pairs stacked vertically, with the grain passing from the top to the middle to the bottom pairs in one flow.

The patent in suit – original and amendments

10. The patent in suit claims as its invention, in essence, that in the case of the production of flour for starch one can in fact do away with some of the screening (sieving) processes after some of the rolling stages. As appears above, the normal practice was, it is said, to screen after each rolling event. The object of the invention was to achieve a further reduction of the investment required for starch production whilst ensuring the quality of the product.

"The solution according to the invention is characterised in that the material is taken at least twice through double grinding stages without screening between the double grindings, and is in each case screened subsequently to the double grinding.

To the surprise of the technical world, it was possible to achieve the object completely and at the same time to provide the necessary pre-requisite for a further step forward in development for raw material preparation for obtaining starch. To do so, it was necessary to get away completely from the traditional attitude in milling practice using high-grinding operations….by the construction of specialised mills it is possible….to dispense admittedly not with the grinding stages per se but with some of the intermediate screening without thereby involving a reduction in the quality and yield of the fraction needed for starch production, whilst maintaining the same total throughput through the mill."

The "double grinding stage" is illustrated in figure 1 of the patent, which appears at the end of this judgment. For present purposes, the viewer should concentrate on the left-hand half. That shows material to be milled coming in through a feed at the top and falling through two rotating rollers. Having passed through those rollers, it is then fed by gravity immediately into two further rollers before then passing out of the bottom of the machine. Thus there are two immediately successive grinding operations with no intermediate sieving (figure 1 in fact shows what is in effect two machines back to back – the same thing is going on on the right-hand side of the diagram). So what the patent was saying is that one can omit a sieving stage without impairing the final product. In fact, as will appear from the claims, that exercise can apparently be done twice.

11. The two relevant claims in the patent, at the start of this hearing, were claims 1 and 5. They are as follows:

"1. Process for the production of a raw starch material for the subsequent obtaining of pure starch from wheat rye maize or barley wherein:

(a) fractions of starch which is damaged as little as possible are produced from the material by repeated roller grinding and screening using the high grinding system, characterised in that

(b) the material is taken at least twice through double roller grinding stages, without effecting screening between the two roller grindings effected in each double roller grinding stage, and

(c) screening is effected only subsequently to the double roller grinding which occurs in each double roller grinding stage.

…..

5. Starch mill installation for the production of raw starch material from wheat, rye, maize or barley for the subsequent obtaining of pure starch, more particularly for carrying out the process according to one of the preceding claims, characterised by

(a) at least two double grinding passages with in each case two grinding roller pairs….arranged directly in succession to one another, and

(b) a screening device….consecutive to each double grinding passage."

12. Claim 1 is the process claim. Claims 2-4 are dependent. Claim 5 is the apparatus claim; claims 6-10 are dependent. It is claims 5 to 10 that were abandoned as unsustainable at the beginning of final speeches, and it is accepted that I should make an order revoking them.
13. The deleted words reflect deletions which are proposed in the amendment application; the underlined words reflect insertions similarly proposed. Buhler made an application to amend in the EPO to carry out those amendments. In a fax dated 31^st January 2008 an examiner in the EPO confirmed a prior telephone conversation and confirmed in writing that the examination division dealing with the amendment application intended to "limit" the patent in accordance with the amendments shown above, but "for internal procedural reasons it is at present not possible to issue the decision on the limitation of the Patent before April 2008". While that seems an extremely surprising state of affairs in a modern organisation, that is apparently the position. It is hard to conceive that the position will change pending the final issue, and in those circumstances I treated the patent as amended (all parties were prepared to argue the case on that basis) and did not hear any part of the English amendment application.

The inventive concept and the nature of the attacks on the patents – generally

14. The inventive concept is said to have its roots in a prejudice of millers of bread flour. Buhler claims that for decades, if not longer, the millers of bread flour had a rooted prejudice in favour of sieving after every roller grinding. They thought that it was necessary in order to remove the bran, so as to enhance the prospects of getting pure flour (uncontaminated by bran) at the end of the process. It was thought that that end would be less achievable if milled grain was able to pass from one grinding to the next without intermediate sieving (and in some cases additional purification steps). It is said that these prejudices would have carried into milling wheat for starch manufacture. The inventive step of the patent was to demonstrate that, so far as starch flour milling is concerned, one could have two successive roller grinding operations, without sieving in between, and do that twice in a milling operation, without affecting the quality of the end product. In final speeches Mr Richard Hacon, who appeared for Buhler, formulated the inventive concept thus:

"Grinding wheat for making starch using the high grinding system, where the wheat (or ground wheat stock) is taken at least through double roller grinding stages, in each case screening only after the double grinding stage."

15. It will be apparent from Claim 1 that it is in fact two such passages which is said to be the inventive step. Whereas high grinding was done in relation to starch flour milling, it had not (so it is said) been done without sieving between roller grinds. So this is said to be one of those cases in which a prejudice, thought to provide a "lion in the path" (to use the metaphor of the cases – see Pozzoli Spa v BDMO SA [2007] FSR 372) is said to be demonstrated to be wrong.
16. Spomax's riposte is to challenge the prejudice so far as it relates to starch flour milling. It says it did not exist, and raises queries about the extent of the prejudice in bread flour milling. It then points to certain prior art in the form of old machines or processes in which double grinding had taken place without sieving, and old textbooks, and says that in the light of modern starch milling technology which was part of the relevant common general knowledge, the idea was obvious. It seeks to rely on Australian starch flour mills which there were operation in 1987 (the priority date of the patent in suit) in which double milling without sieving was taking place, albeit that the second stage was a pin mill and not a roller mill. The first grinding was high grinding, so there were established machines and processes which carried out two roller grinds on flour without intermediate sieving. Since two immediately successive milling steps had taken place in Australia it would be obvious to try two roller-grinds in immediate succession; and once one had done that once, it would be obvious to do it a second time.

The issues arising

17. It was common ground that I should approach this case along the lines of the reformulation of the Windsurfer test propounded by Jacob LJ in Pozzoli at para 23:

"(1) (a) Identify the notional "person skilled in the art"

(b) Identify the relevant common general knowledge of that person.

(2) Identify the inventive concept of the claim in question or if that cannot readily be done, construe it;

(3) Identify what, if any, differences exist between the matter cited as forming part of the 'state of the art' and the inventive concept of the claim or the claim as construed.

(4) Viewed without any knowledge of the alleged invention as claimed, do those differences constitute steps which would have been obvious to the person skilled in the art or do they require any degree of invention?"

The witnesses

18. In order to address those points I heard evidence from two witnesses, one expert for each side:

(i) Mr Julian Garratt was called by Buhler. He retired in 2000, but had previously spent his entire working life in the milling industry, working for Spillers. He became the production director of Spillers in 1990, and during his time was responsible for the specification, purchase and installation of three complete flour milling plants. He was plainly extremely experienced in flour milling itself. From 1987 he ran a mill which milled for starch flour as well as other flour. His experience cannot be doubted, and he was a careful witness. His focus was, however, limited to the focus of essentially a bread (and other consumables) flour miller, and to the actual practical side of the milling.

(ii) Mr Bryan McGee was called by Spomax. He is now a self-employed consultant, but had had a long career in milling engineering, being involved in the design of milling machinery, including roller mills. Again, he was extremely experienced, but this time in the different field of engineering. He was generally a good witness, though sometimes lacking in a bit of focus in his answers. Mr Hacon accused him of departing from what is normally required and of actually becoming an advocate for his client's case, but I do not think that that is a fair criticism of his evidence.

The identity of the skilled man

19. This is a key question in this case, because the case turns heavily on the existence and impact of the prejudice against removing sieves that I have referred to above. The extent of that impact will be significantly affected by the identity of the skilled man at whom the patent is taken to be directed. The parties were not wholly at one as to the attributes of this man.
20. Buhler's case on the skilled man was that he was a miller – something like the manager of a mill who is interested in making his mill run more profitably. However, his relevant knowledge (knowledge of techniques in grinding flour) was effectively going to be the same as that of a milling engineer, because millers and engineers would talk to each other and keep each other informed of developments. This depends on the factual supposition that, in the relevant area, they would share the same knowledge.
21. Spomax's original case was that the skilled person was a team – a milling design engineer and a process engineer involved in designing the process in which milling rollermills were to be used. The former (obviously) is a higher level generalist; the latter focuses on the implementation of designs in the workplace and has greater contact with day to day milling needs and practices. In many ways he is pretty close to the manager of a large mill (Buhler's skilled man). At this point the parties are fairly close in their respective formulations. Where they probably differ is in the influence of the bread miller (emphasised by Buhler) and the influence of the dedicated starch miller (emphasised by Spomax). Spomax's case was that the team would have known about dedicated starch milling as well as other milling.
22. The evidence of Mr McGee, which I accept, was that by 1987 there was a separate sector of the milling industry which was milling wheat to obtain starch. Mr Garratt had had experience of being asked to produce starch flour on machinery used primarily for bread flour (which is a perfectly possible process), but the starch flour industry was different. It involved plants which were more or less dedicated to starch flour, and which were not particularly suited to bread flour. The people who operated them were similarly dedicated, whatever their original background might be. There were plants in the UK (Corby and Ayrshire), Ireland, Belgium (though the nature of the machinery in this plant was not known in 1987), Finland, Canada, Australia and China. All this was known to the industry. It was also known that the milling process was different from that used in traditional bread flour (or multi-purpose) mills – it was shorter, amongst other things.
23. These things (apart from the techniques adopted in the Belgian mill) were not secret and while they were relatively new they were not so new that only a few people in the industry knew about them. In the light of that I consider that Spomax's formulation of the skilled person (team) is closer to the true position than Buhler's. This patent as granted has process claims and apparatus claims. The former would be more angled at the person performing the process – the process engineer, or miller. The latter would be more angled to the building or design engineer. But neither operates to the exclusion of the other. The patent is aimed at those who make flour and those who build. The combination of the two engineers is an appropriate formulation.
24. But, more significantly, it is aimed at those with a particular interest in starch. It is not aimed at the bread miller. The two types of engineer who make up the team would have some knowledge of starch manufacture and its requirements, though not to the exclusion of the knowledge of milling techniques generally (including bread milling). Mr Lykiardopoulos, who appeared for Spomax, did not say that a starch-only process engineer (that is to say, one without any knowledge of bread and other milling) would be a member of the team, and he is right not to say that; apart from anything else, it is unlikely that that class existed (or existed in any sensible numbers) in 1987. But the team would have, as one of its members, someone who really knew something about the then current state of starch milling. At the end of the day I am not sure that Buhler would gainsay that, but the dispute probably came down to one of emphasis. I certainly find (if Buhler was otherwise averring) that a bread miller, as such, was not part of the team.
25. If it matters (and I doubt if it does), I do not accept Mr Hacon's submission that the miller and the engineer can be treated as being the same for these purposes because they would have the same general awareness of milling generally. I do not think that that is necessarily correct as a matter of fact. The engineer is not tied to a particular plant, which a miller would be. He will have a more industry-wide view, since he is dealing with a number of customers and projects. He will be more likely to keep abreast of developments as they occur; a miller himself who has just installed a plant will be less likely to be interested. Mr Garratt said that any individual miller may or may not have the same degree of knowledge of historical matters that an engineer would have, and I accept that evidence; it accords with the probabilities, but it does not mean that the two classes should be treated as one.

The common general knowledge

26. There was no dispute about a large section of common general knowledge. It was accepted by both sides that the general background to milling, set out above, and the differences in the requirements of bread flour milling and starch flour milling, would be part of the general common knowledge of the skilled man. The principal dispute under this head was the extent to which the skilled man would approach the patent and the prior art with the bread miller's prejudice in favour of intermediate sieving after every roller passage.
27. Buhler's case depended very heavily on this prejudice. For most of the trial Spomax accepted that it existed in the mind of a bread flour miller. At one point in his final speech Mr Lykiardopoulos seemed to qualify that in relation to the United States (which has a very large bread milling industry), but I do not think that the evidence allows him to draw such a distinction. All the evidence points to the existence of the firmly held view of bread millers that high grinding of wheat flour for bread making (or other culinary use) involved sieving after every grind unless wholemeal flour was being made. Mr McGee accepted that the bread miller would have the prejudice, and he accepted that the engineer (by which he must have meant the mechanical engineer rather than the process engineer) would have it though less strongly. This view of his about the engineer makes sense. Engineers will to some extent take their cue from experienced millers (or their process engineers) as to the actual processes, but since they had a wider remit than just bread flour milling the prejudice will not have been so strong or entrenched. They will have had a broader perspective.
28. The remaining relevant disputes about common general knowledge turned on what the skilled man would know about double roller grinding, the force of the prejudice and the extent to which it would be operating when the skilled man was considering milling for starch flour. In this context there was evidence of what appeared from textbooks and of what happened in dedicated starch flour mills.
29. Care must be taken in treating textbooks as common general knowledge. As Laddie J said in Raychem Corp's Patents [1998] RPC 31 at page 40:

"The common general knowledge is the technical background of the notional man in the art against which the prior art must be considered. This is not limited to material he has memorised and has at the front of his mind. It includes all that material in the field he is working in which he knows exists, which he would refer to as a matter of course if he cannot remember it and which he understands is generally regarded as sufficiently reliable to use as a foundation for further work or to help understand the pleaded prior art. This does not mean that everything on the shelf which is capable of being referred to without difficulty is common general knowledge nor does it mean that every word in a common text book is either. In the case of standard textbooks, it is likely that all or most of the main text will be common general knowledge. In many cases common general knowledge will include or be reflected in readily available trade literature which a man in the art would be expected to have at his elbow and regard as basic reliable information."

Mr McGee produced a book published by Mr J H Scott in 1972 entitled "Development of Grain Milling Machines". Mr Scott was a well-known and respected senior figure in the milling industry. His stated object was to fill a gap left by the absence of a "comprehensive, unified and up-to-date account of the development of all the principal [milling] machines". He believes that such an account would be welcomed by the industry. He suggests that as well as providing a useful historical record:

"… re-study of certain discarded earlier designs and methods may suggest novel apparatus or ways in which modification of these designs, using the newer materials and techniques now available, could rewardingly be made."

All that suggests that the book was not reflecting common knowledge, but supplying an omission in the materials. If that were the case it would tend to indicate that it was not the sort of textbook which would amount to common general knowledge. However, in a Foreword, Mr E Alan Williams, another senior figure in the milling industry, points out the value of placing the history of milling on record, and records the fact that:

"the author has provided students and all those associated with flour mills and the grinding of grain, with much material of value. This will add to the interest and knowledge of millers everywhere.

A proper appreciation of the ideas put forward by the early inventors of milling machinery is required to enable students to trace how modern machines have been improved and developed over the years and to understand why some ideas had to be rejected and replaced by better designs. The study of the achievements of the past is a necessary part of the modern production manager, and all those whose occupation is associated with milling will greatly benefit from reading this book."

These words suggest that the book ought to become common general knowledge, but do not demonstrate that it necessarily did. Mr Garratt said he was aware of its existence, and that his company had one in its library, but he had not seen it himself. On the other hand, Mr McGee said that it was widely known and referred to. Mr McGee confirmed that its general account of the history was what had been known anyway.

30. Taking all the evidence, I think that this book is on the cusp of common general knowledge and specialised information, or perhaps more accurately it spans the two. It does not contain or reflect common general knowledge in the sense that its detail would be generally independently known. It contains a lot of detail of particular machines which an engineer or miller would not actually remember, but its main thrusts would be known and it would be treated as a known work of reference which would be referred to by the skilled man in the sense identified by Laddie J in Raychem (as Mr McGee said, there were very few books in the area) and I think it likely that the aspirations of the foreword writer were probably fulfilled.
31. I therefore turn to the relevant parts of the book. It contains a detailed history of milling, and while it does not confine itself to bread or flour milling, much of it (on the basis of what I have seen) is given over to that. Chapter 12 deals with roller milling of flour and contains a description of "Unconventional Roller Mills". He observes that "three and four pair high machines had a continuing market for wheat and other cereals in small country and overseas mills", but he considers these to be "irregular designs" when compared with the "principal normal form of roll assembly; the 'single passage' two or four roller mills." The book then describes various two, three and four pair high mills in some detail, and comments on their drawbacks:

"But one can well imagine that outweighing these various factors would be increasing appreciation that optimum grinding performance was not possible when stock was successively passed between the rolls of three or four-high assemblies. … In such case, the fine material created in the first transit tends to cushion and confuse the grinding operation in the subsequent transit(s). In this regard it has long been standard in milling to size-separate or grade stock immediately after each passage so that particles of markedly different size may be more efficiently re-ground when segregated.

"However, in small mills, it was apparently not uncommon to use machines having three or four superimposed single rolls as just described, or alternatively several superimposed pairs of rolls, the top pair acting as a first break, the second pair as a second break and so on. In this way a near complete breaking operation could be made on one machine. In this regard it may be noted that [in a publication by a Herr Kick in 1883 he] states that several constructions of roller mills were intended for a low grinding, or for at least abridged systems of high (ie gradual reduction) grinding."

He then goes on to describe various machines, at least one of which had successive rollings without intermediate sieving, and others of which had an intermediate sieve. Then he observes:

"Special machines containing mostly two, three or four-pair high rolls continued to be made and improved to some extent, but their less efficient adjustment mechanisms and the inherent disadvantages of successive rolling prevented them from coming into general use for large scale high quality flour milling. However, such machines continued to give good service for the production of wheat and other cereal meals … "

This book, and particularly these passages, reflects the common general knowledge of at least engineers, and more certainly process engineers, that stacked systems had been in use in the bread industry, that some of them had no intermediate sieves and that they had generally been displaced in the manufacture of high quality bread flour. I also find that it was plainly suggested that stacked roller machines, with no intermediate sieving, had been used in an abridged form of high grinding ie a shortened form, with fewer passages. At least one machine (by Mechwart) was identified as having been used for this. This would probably have been interpreted as producing less pure flour than an extended form, but it was still regarded as high grinding. Mr Hacon sought to show that these passages were consistent with the use of these machines only in a low grinding process, and while that is true of some of the references, overall I do not think that that is how they would be taken, especially in the light of the reference to abridged high grinding. The thrust of these passages (including the parts that I have not set out at length in this judgment) is not to make that distinction, but to point out that modern processes, with the pursuit of a higher quality flour, had tended to move away from that sort of machine so far as bread flour is concerned.

32. Other machines were relied on by Spomax as demonstrating a use of 2 pair high roller mills for grinding wheat without sieving. By final speeches they had been distilled to a smaller number, some of them among the prior art. It was not suggested that all of these machines were actually known to engineers (or millers) in 1987, so to that extent they were not themselves part of the common general knowledge, but Mr McGee said that some of them would have been known specifically, and that generally speaking engineers would have been aware of the use of 2 pair high stacked rollers in bread flour milling. I find that he is right about that, and I also accept his evidence that such mills were used to mill other cereals (which was not seriously disputed).
33. Mr Hacon sought to make the case that such stacked mills were not used for high grinding of wheat, and that that was part of the common general knowledge. I think that he is partly right and partly wrong about that. It was plainly the case that mainstream millers had moved away from such equipment because they preferred to sieve after every grinding stage. Mr McGee said as much in his initial report. To that extent Mr Hacon is right. However, he is not right to say that they were never used for high grinding. I think that they probably were (both for bread flour and for other purposes) and that engineers would have known that too. The machines were not all very historic (though many were). One of them dates from the 1950s (I deal with it in relation to prior art), and is from a well-known American supplier. The collective memory of their use will not have totally faded by 1987.
34. This is really by way of background to the main issue which divides the parties in relation to common general knowledge, which is the extent of prejudice which the skilled man would have against grinding twice without intermediate sieving. Mr Hacon said that the real question was whether the skilled person harboured a technical prejudice against grinding flour using a high grinding process, for the purposes of starch production, without sieving after every grind. This prejudice lies at the heart of his case. It should be noted that he relies on the prejudice in relation to milling for starch flour, not for bread flour, though the former is said to flow from the latter. He said that if the prejudice was established, then it was then necessary to turn to the historical material to which I have referred above in order to see whether it dispelled the prejudice. He said it was also necessary to turn to the actual experience of starch flour millers and consider whether that affected the prejudice too.
35. I do not consider that this is the correct approach to ascertaining prejudice. One does not look at the evidence which, by itself, is said to establish it, and then look for evidence dispelling it. The enquiry is an overall one – does the prejudice exist, weighing all the evidence together and not starting with one piece and then testing the rebutting quality of others. It is therefore necessary to consider the remainder of the evidence before coming to a conclusion as to whether the prejudice relied on by Buhler was part of the common general knowledge of the skilled person.
36. Thus far the evidence establishes, and I find, the following, in relation to the position in 1987:

(a) There was a very widely held view amongst bread flour millers that, in a high grinding process, it was best if not necessary to sieve the product after every grind to produce high quality bread flour. They had what counsel described as a mindset to that effect.

(b) That prejudice was known to milling engineers, and so far as bread flour is concerned they would generally be concerned to cater for it, and to a lesser extent it was part of their mindset too.

(c) That does not mean that in every high grinding process for bread flour there was sieving after every stage. In the 1950s Allis-Chalmers offered two, three and four pair high stacked roller systems. I have seen the brochure for these machines, and a preceding brochure from the 1940s (they were both relied on as prior art). They were offered for sale for, inter alia, bread flour milling, and although the brochures do not in terms reflect whether they were anticipated to be used for high or low grinding, I think it unlikely that none of them were used for bread flour milling. This does not, however, mean that the view in (a) was significantly weakened. It merely means that some millers were prepared to take a more modified view and were satisfied with the product accordingly.

37. However, it does not follow that that prejudice is to be automatically extended to milling wheat for starch flour. The evidence did not establish that it necessarily did.

(a) Milling wheat for starch was a relatively recent thing in 1987, but it was not simply carried out by bread millers who all adopted the same approach as they did for their bread flour because they knew no better. If that had been the case, then the likelihood that the prejudice would have carried over would have been greater; but it was not. I have indicated above that starch flour milling had been carried out by dedicated starch millers for some years. That will not have escaped the attention of engineers (design/manufacturing or process).

(b) Mr McGee's evidence, which I accept, was that starch flour mills carried on a form of double grinding without intermediate sifting in several plants in Australia, again known to European engineers. He produced a flow diagram of a pilot plant that his own company installed there in 1986. These plants carried out high grinding. Typically, for present purposes, these plants had at least one phase on the reduction side at which there was double grinding without intermediate sieving. The first grind was by rollers. The second grind was by pin mills. The flow chart produced by Mr McGee showed two such phases. Thus there were plants that carried out something which was, subject to Mr Garratt's attempts to distinguish the situation, apparently contrary to the bread miller's prejudice. Mr McGee said he was aware of other plants which had a similar arrangement for both starch flour and bread flour production.

(c) Mr Garratt said that those mills were doing something different which did not impact on the prejudice. He pointed to the fact that this form of double grinding, using pin mills on the second grind, was occurring only where there is endosperm-rich stock; or in other words the feed was of material which was basically endosperm with the bran removed. It was not done at the break stage; rollers were used at the break stage, with sieving after each break. In addition, pin mills operated differently from rollers. He did not accept that this use demonstrated that rollers could or would have been used instead of the pin mills.

38. What this evidence shows so far, in my view, is that what Mr Garratt would wish to present as the bread flour miller's mantra was apparently not operating per se in relation to starch milling. By 1987 the starch industry was at least to some extent moving away from it. These plants did not demonstrate that it had completely lost its way. It is probably not without significance that the double grinding took place only on the reduction side, but take place it did.
39. The strength of the prejudice probably has to be tested partly by reference to the reasoning underlying it. Two principal reasons were given.
40. The first was contamination with bran. This contamination can take the form of bran dust or larger particles. A baker (unless making wholemeal bread) does not want bran in his flour because it alters the colour and affects the rising. This is so in relation to both bran dust and the larger particles. Bran dust arises from grinding down larger particles of bran. Most of it will arise on the break stage. One of the objectives of milling (for both starch and bread flour) is to get the bran out at the break stage so far as possible, though total exclusion is probably not possible. The break stage is the stage at which bran is predominantly removed. The more there are large particles left over from any grind, the more likely bran dust is to be created on the following one; so sieving at each stage was thought necessary. Bran dust could not be removed from the flour. That was Mr Garratt's evidence in relation to bread flour, and Mr McGee did not gainsay it. I accept it.
41. The position is not quite the same with starch. The evidence on this, which was mainly Mr McGee's, was sometimes confused, but in the end it seemed to me to amount to this. Starch is extracted by a wet process. Water is added to the flour in order to separate the starch from the gluten. The starch is washed away in a milky flow; the gluten remains (and is separately exploited). Any bran dust is washed away with the starch. Contaminants, including bran, can be removed at this stage. Bigger particles of bran can be removed too. Mr McGee called fine particles, about the same size as the flour particles, bran dust. Bigger particles he called bran contamination.
42. Obviously the starch maker is not concerned about the effect of bran dust on baking. He does not want bran in his starch, but Mr McGee's evidence, which I accept, is that he was more tolerant of it in the flour that he used because he could remove it in his manufacturing process (unlike the baker). He did not welcome it, and he preferred to have less rather than more of it, but he could cope with it. He did not think so much in terms of bran dust as the contaminant, but bran in larger particles. A different, and less rigorous, approach to bran removal is already evident in the fact that it was accepted that when milling wheat for starch flour, some of the intermediate purification steps which would have been carried out on bread flour, were omitted. Since this reason for sieving after every grind was less compelling in the case of starch, that would go to the extent to which the prejudice existed in the mind of a starch flour miller, and therefore those engineers who had an eye to his processes.
43. The second reason for the prejudice in the case of bread flour is the cushioning effect of flour. Mr Garratt's evidence was that millers across the board perceived that if you did not sieve them out, fine particles from one grind would cushion the effect of the rollers; hence the need to sieve them out. He was unable to say whether the belief would be shared by an engineer, but since Mr McGee himself accepted that bread millers had that perception, and therefore implicitly accepted he knew about it, that perception would be likely to affect the thinking of such engineers. He did not, however, accept that those involved in milling for starch flour would have thought in terms of such a problem. He and his company (Henry Simon) had not turned their minds specifically to the cushioning effect in starch manufacture in 1987, but he said that if they had they would not have assumed an undesirable cushioning effect in double grinding for starch flour because there were starch flour mills doing it at the time. In addition, starch flour milling was less aggressive so the particles were coarser than in flour milling. Accordingly, since the cushioning arises because of fine particles cushioning larger particles, there would be fewer fine particles to cause the problem. The thrust of his evidence was that the skilled person who thought about it would not be so concerned about cushioning in the context of starch milling as a bread flour miller would be.
44. I think that a proper view of the cushioning point lies closer to Mr McGee's point of view than Mr Garratt's. Mr Garratt was coming to this point from the perspective of the bread miller, though he claimed that the cushioning view was held across the board. However, Mr McGee's wider perspective demonstrates what is in my view the truer view which is that the view was not held so firmly, or perhaps at all, outside pure bread flour millers. In particular, those thinking about starch would not necessarily hold it, for the reasons that he gave, although the existing double-grinding mills used pin mills on the second grind and not roller mills. It is also the case that two pair high and three pair high rollers had been on the market for a considerable period of time, without intermediate sieving being incorporated in those machines. They must have been used for something, and they demonstrate that cushioning cannot have been a universally regarded undesirable effect across milling.
45. Bearing all that in mind, I therefore turn to make findings on the prejudice relied on by the claimant. I find that it existed in bread flour millers – they, or at least the big commercial ones, would tend to think in terms of sieving after every grind, as the textbooks suggested. If all wheat starch flour millers had been the same people, and if they had narrow mindsets, then that prejudice might have carried over into starch flour milling. However, they were not all the same people, and the evidence of what was happening in starch flour milling was that they apparently did not share the prejudice. Looking at the reasoning said to underlie it, there is no technical reason why they should share it, or at least to the same extent. So far as the relevant skilled man is concerned, informed as he will be about starch milling, the prejudice did not really exist. While it is true that the Australian starch mills did not use double roller grinding without intermediate sieving, they did disclose two lots of grinding without intermediate sieving. That clearly points away from the prejudice existing. Mr Hacon suggested that knowledge of those mills would not "sweep away" the prejudice. I do not consider that that is the correct approach. In assessing the extent, if any, of the prejudice in relation to starch flour milling in 1987 it is not an appropriate approach to start with the prejudice and consider whether things have happened to remove it ("sweep away" is a little over-dramatic anyway). That might have been appropriate had the question been arising in relation to bread flour milling, because it was demonstrated that it had existed for decades. However, the starting point is not bread flour milling. I do not consider that there is any particular starting point. Whatever might have been the position if wheat starch flour milling were just starting out (as to which I say nothing), it was an established and known industry by 1987, and the proper approach to ascertaining the general common knowledge of the skilled man is to take an overall view of the evidence with no particular starting point or presumption in terms of the prejudice. The prejudice of the bread miller merely provides one element in the overall picture. There is no lion in the particular path down which the skilled man travels in this case (or at least not the particular lion alleged). There is at most an interesting small terrier which yaps a little to attract attention. It does not provide much distraction, let alone deterrence.
46. There is an additional point that it will be useful to dispose of at this stage. Mr Lykiardopoulos took the point that even if there was a prejudice in fact, the removal of it could not count as an inventive step because the patent did not (he said) mention it, or mention it sufficiently. He relied on Jacob LJ in Pozzoli at para 28:

"Where, however, the patentee merely patents an old idea thought not to work or to be practical and does not explain how or why, contrary to the prejudice, that it does work or is practical, things are different. Then his patent contributes nothing to human knowledge. The lion remains at least apparent (it may even be real) and the patent cannot be justified."

Mr Lykiardopoulos says that the patent makes no mention of the prejudice, and merely patents an old idea thought by millers to be a bad one, but does not explain how or why things are different.

47. This point is not a good one. The immediately preceding words of Jacob LJ contain an explanation of why it is that an old idea, once thought not to work but which it transpires does work, is patentable:

"A patentee who contributes something new by showing that, contrary to the mistaken prejudice, the idea will work or is practical has shown something new. He has shown that an apparent 'lion in the path' is merely a paper tiger. Then his contribution is novel and non-obvious and he deserves his patent."

Before that, he had set out what he had said in Union Carbide Corp v BP Chemicals Ltd [1998] RPC 1:

"Invention can lie in finding out that that which those in the art thought ought not to be done, ought to be done. From the point of view of the purpose of patent law it would be odd if there were no patent incentive for those who investigate the prejudices of the art."

I do not think that those dicta require, in all cases, that the patent should explain in almost scientific terms why the prejudice is wrong and how the invention works in some sort of scientific language. It is sufficient if it demonstrates with sufficient clarity that it does work. What Jacob LJ is saying is that merely patenting the idea, without more, is not sufficient. It is necessary to disclose the inventive step. Where an old prejudice is concerned, the inventive step is the discovery that, contrary to expectation, the invention has the (unexpected) effect that it in fact has. That is what has to be disclosed.

48. The patent in this case does disclose the prejudice, albeit in a form which is more oblique than one might have expected. The section of the specification dealing with "State of the art" does not, somewhat surprisingly, refer to it. It refers to two "essential perceptions" – the first being that starch grains which are damaged during grinding yield less starch, and the second being the wish to keep the number of grinding passages as low as possible to keep the investment in plant low. The next section is headed "Representation of the invention", and it contains the wording already set out in paragraph 10 above, which I will repeat in a fuller context. The relevant part reads as follows (it has probably lost a little in translation from the German):

"The invention, then, had as its object to achieve a further reduction in the investment required for a starch mill but whilst ensuring the quality of the product, more particularly the fractions required for the starch factory, and more especially the heavy fraction which contains the A starch, but likewise the complete starch containing A and B starches.

The solution according to the invention is characterised in that the material is taken at least twice through double grinding stages without screening between the double grindings, and is in each case screened subsequently to the double grinding.

To the surprise of the technical world it was possible to achieve the object completely and at the same time to provide the necessary pre-requisite for a further step forward in development for raw material preparation for obtaining starch. To do so, it was necessary to get away completely from the traditional attitude in milling practice using high grinding operations. It is still true that, for achieving maximum results as regards all the usual grinding products, the grinding has to be carried out extremely gently and therefore in several stages: opening the grain, detaching the endosperm particles, gentle grinding of the coarse semolina particles, producing the necessary fine fraction without starch damage, etc. By the construction of specialised mills it is possible – this has been found by most recent perceptions – to dispense admittedly not with the grinding stages per se but with some of the intermediate screening without thereby involving a reduction in the quality and yield of the fraction needed for starch production, whilst maintaining the same total throughout the mill. This means that the grinding continues to be carried out in many stages and very gently.

Dispensing with a substantial proportion of the intermediate sifting (almost half) achieves a saving in all the conveying units, filters etc hitherto required for that purpose, so that if the novel underlying idea of the invention is applied in thorough-going manner a mill is obtained which is substantially less expensive in cost. More particularly the operating of a mill according to the invention is also more economical because of the fewer apparatus to be maintained and the reduced energy needed for conveying and driving. In addition there are detachable savings due to the reduced building structure volume."

That section does not in terms identify a prejudice, but it does so implicitly, and almost in terms, in referring to "surprise", the "traditional attitude", achieving the objective, and "dispens[ing] … with some of the intermediate screening". When read fairly by the skilled man, it would be taken as referring to a differing state of affairs from the traditional, and if there had been the prejudice which I have ruled against it would have been taken as referring to that. So it identifies the prejudice (assuming there is one) and indicates how the invented practice gets away from it, successfully. What it does not do is say why it works in scientific or causation terms. But I do not consider that that is necessary in order to comply with Jacob LJ's dictum. I do not see why a person who identifies a prejudice and demonstrates its falseness by means of practical steps should be barred from claiming a patent if he cannot identify why, in terms of actual causation, the invention works, but another man can claim the patent if he can carry out some scientific analysis and show how it does. The former adds to human knowledge by identifying the prejudice and indicating with sufficient clarity how it is bypassed, or falsified, by the invention. Mankind is that much the wiser.

49. Accordingly, had I found the prejudice, this patent would not have fallen foul of the requirements in this respect in Pozzoli.

The inventive concept

50. That finding goes a very long way to determining this action against the claimant, because the prejudice, and its removal, lies at the heart of the inventive concept. However, I shall continue down the Via Pozzoli, the next stop on which is the identification of the inventive concept.
51. The inventive concept as defined in Mr Hacon's written final submissions has been set out above. In the light of some of the prior art, it becomes necessary to construe certain phrases in the patent, namely "high grinding", "grinding", and "opening up" the wheat. It is common ground that the test is what the words would have meant to the skilled man.
52. There was no material dispute as to what "high grinding" meant in the patent. It is the process described above, whereby grain (in this case wheat) is milled using a series of gradual steps (breaks and reductions) as opposed to one or a small number of more vigorous steps (sudden death or low grinding). It is not necessarily distinguished by an absolute minimum number of steps but on the evidence sudden death would be an absolute maximum of 5 steps, and usually three or less.
53. An issue arises as to what is meant by "grinding" because of one of the items of prior art in relation to which it is said that it does not disclose grinding, and "roller grinding" is one of the integers of Claim 1. The question is whether grinding has to involve the rendition of particles to a greater number of smaller particles, or whether it can cover the passage of wheat through rollers in which the wheat is opened up without being reduced to a greater number of particles, or in which the bran is not even actually broken but merely somehow squeezed to remove dirt (see below).
54. In his second report Mr McGee said that he considered that grinding generally meant the production of reduced size particles. However, he also agreed that a first break (which he would bring within the definition of grinding) might be "little more than opening up the wheat", to take a phrase from a publication. Mr Garratt accepted this latter point. The context of Mr McGee's evidence suggests that this "opening" did not always connote a breaking of the kernel into two or more parts, as opposed to cracking it without breaking into parts. His evidence suggested that he would exclude from grinding the passage of kernels through rollers with no material effect on the kernels (not surprisingly), but otherwise if there were an effect then grinding would have occurred. To that extent his general definition has to be qualified.
55. "Grinding" is not defined in Claim 1, but guidance is given in the part of the specification quoted above. It refers to grinding "extremely gently" and in stages, the first of which is "opening the grain". This does not necessarily suggest fracturing the particles, and is more in line with Mr McGee's qualified views of what grinding might be. I find that in the context of this patent "grinding" does not necessarily and in every case involve the fracturing of particles, though of course in practice this is what will usually happen. The relevance of this debate will become apparent in considering the first piece of prior art referred to below.

The prior art

56. The next two stages in the Pozzoli procedure involve considering the disclosure of the prior art and then considering whether the differences constitute steps which would be obvious to the skilled man, or whether they would require a degree of invention. It will be convenient to take these steps together in relation to each item of the prior art.

The Ellesmere Port Mill 1910

57. In 1910 the trade journal "Milling" published an article about a newly installed milling plant built by Messrs Frost & Sons at Ellesmere Port. It describes the plant in some detail, and in glowing if not gushing terms ("A truly fine sight … the green and white of the walls add a most pleasing effect to the whole scene"). The roller floor is described as follows:

"There are three long lines of roller-mills and these are arranged on a system of four breaks, thirteen reductions and a dual scratch. The first break is effected on two of Simon's eight-roll mills with 'roll' scalping sieves underneath the two bottom pairs. The two pairs of rolls in the top portion of the frame are for cracking the wheat which, after passing through the exhaust, which draws away nothing but the dark stive, is broken heavily on the lower rolls. The second and third breaks are each effected by three double mills and the fourth by two doubles."

It is necessary to understand that "stive" is dirt which is trapped in the crease of the wheat. It has to be removed at some early stage in the processing. What is described here is a process of dislodging the dirt and then blowing it away in the exhaust.

58. Spomax says that this discloses a process for the production of flour from wheat grain by repeated roller grinding with screening only taking place subsequent to the first double grinding. It is said that it would be obvious in 1987 to apply this to the grinding of starch flour, and that it would be obvious to apply it to starch flour milling, as Mr Garratt said it would. Once one had got one of these stages in the mill, it would be equally obvious to add another (to do it again).
59. Buhler's case is that this mill does not disclose grinding in the first set of rollers because all that happens is that stive is released without any flour being released. The skilled man would infer that no flour is released at that stage because only stive is referred to and the miller would not want to lose any flour. Accordingly, he would understand that the kernel is not even cracked, or even if it is this does not amount to grinding. Furthermore, the skilled man with the prejudice in favour of sieving after every grind would not think there was grinding. He would see there was no flour produced, so the problems of bran dust and cushioning would not arise. His prejudice would not be dislodged, and would operate to prevent its being obvious that a proper grind could take place on the first rollers for the purpose of starch manufacture.
60. One of the difficulties in considering this piece of prior art is working out exactly what it does teach. The experts disagreed as to what was going on in the top rollers. Mr Garratt considered that the kernels were merely being "knocked" in order to remove the stive, and were not being cracked, despite the use of the word "cracked" in the description. If that is the case then he says there was nothing which could be regarded as grinding – see the debate above on the use of that word. Mr McGee did not think that that was the factual scenario. He found it hard to imagine how you could remove the stive without cracking the wheat to some extent. "For the crease to open the bran, the coat, has to be broken." He had never been able to operate a roller mill with the degree of sensitivity which would be required to release just stive and not crack the kernel, though at the end of the day he is (as he accepted) an engineer and not a miller.
61. This piece of prior art probably leaves the skilled man a little uncertain as to whether flour was released or not in this process. On the other hand he knows that the mill has deployed a serious milling machine, capable of doing vigorous milling, and that something has happened that the journalist (in his specialist field) has called "cracked". In the industry this would signify something going beyond "knocking" (which was Mr Garratt's word). I think that the skilled person would hesitate before he accepted the word "grinding" in relation to this activity, though he would not reject it. (He would probably wonder why the characterisation in terms of words matters.) However, I do not consider that that necessarily deals with the point. Mr Hacon said that this would not dispel the bread miller's prejudice. He may be right about that, but in the light of the correct identification of the skilled man (who is not completely dominated by that prejudice, because he knows about starch flour milling) there is no prejudice to dislodge. The skilled man approaching this piece of prior art would see some form of process being carried out on this milling machine in which the upper rollers were dong something, and that it was described as cracking the wheat. It is plainly doing something. The top roller is also fully adjustable so as to grind hard, or not at all (the particular machine was identified as the Simon machine that is the next piece of prior art). With starch flour milling in mind, where the prejudice merely exists as part of the background to bread flour milling, and where double grinding (with pin mills as the second mill) has already taken place for some years, I think that it would be obvious to try what both experts would consider to be a full grind with the top rollers when he has seen this piece of prior art. If it can be done once, then it is obvious to do it a second time – the contrary was not really contended for by either expert.
62. In the circumstances I consider that the invention is obvious over this piece of prior art.

Henry Simon "Reform" 8 roller mill 1914

63. This is a piece of machinery which appears in a brochure of this well-known milling firm dating from 1914. It is common ground that this is the mill referred to in the article about the Ellesmere Port mill. It has two two-pair high sets of rollers in the one machine, side by side, rather like the diagram in the patent. There is a photograph, and underneath it is said:

"The 'Reform' Eight-Roller Mill has been designed so as to procure double treatment on the first break. The wheat is opened out on the upper pair of rolls, and so prepared for heavier treatment on the lower pair. By this means intermediate handling of the stock is reduced to a minimum, more semolina and broader bran being produced."

It is apparent that the rollers are fully and independently adjustable. They can work very lightly, or heavily. The phrase "opened out" can cover anything from just opening the kernel to releasing over 25% of the endosperm. As Mr Garratt agreed, it can be used as part of a high grinding system. These machines do not have sieving between the upper and lower rollers. He also conceded that it would have been obvious in 1987 that this machine could be used for starch flour milling. He sought to make the point that the top rollers would not be undertaking a break in the sense that they were not going to produce a lot of flour – that is what he took to be meant by "opened out" - so his concession probably has to be seen in that context.

64. Mr Hacon submitted that the disclosure in this brochure was the same as in the Ellesmere Port article. If that is so then again he fails on obviousness for the same reason that he fails in relation to that article, but I do not consider that he is right. Whatever the uncertainties flowing from the article might be, I consider that this disclosure is a little clearer. It refers to "double treatment", which suggests something active on the first break. It does not confine itself to something particularly gentle which removes the stive but otherwise leaves the kernel unaffected; the expression "double treatment" suggests more than that.
65. Mr McGee thought that using this for the break stage, and to try it for the reduction stage, of starch milling, would have been obvious in 1987. I agree with him on the evidence. The bread miller's prejudice would not operate to prevent it. It does not matter whether this machine would dispel the prejudice; that is not the relevant test on the facts as I have found them. I agree with Mr Lykiardopoulos that the patent is obvious in the light of this disclosure. It discloses double grinding as part of a high grinding process; sieving after the lower rolling is implicit; it is obvious to try it for starch; and obvious to try it twice.

Allis-Chalmers 1955 brochure

66. This brochure illustrates and advertises the products (or some of them) of Allis-Chalmers, whom it describes as "the nation's [ie the US] largest producer of flour mill equipment. It advertises "compact, straight-line roller mills" in one pair high, two pair high and three pair high versions, which "as well as [in] flour mill installations" are used in other applications – distilleries, malt mills, soybean plants and chemical plants. For present purposes the reference to flour mills is the important one. All three of the versions just described are illustrated, and against the two pair high model it says:

"The Two-Pair High roller mill is a two stage grinding mill for making reductions where a separation between reductions is not necessary … the material passes from the first pair of grinding rolls by gravity to the second pair, where another, finer grinding takes place."

An illustration makes it clear that there is no internal sieving between those pairs of rollers.

67. Mr Garratt's objection to considering this for starch flour grinding was that the skilled person would not read this brochure as indicating that this machine was suitable for high grinding. Because he would think in terms of intermediate sieving, the skilled person would think this was good for wholemeal only, or low grinding. He was, of course, looking at it through bread millers' eyes with the bread millers' prejudice. He was unable to speak as to how an engineer would use it. Since those are the wrong eyes, and the prejudice was not operating as he said it would, his view does not assist.
68. Mr McGee said that the two pair high roller mill was eminently suitable for starch milling, and a starch milling engineer designing the process would consider using it. I accept that evidence. I note in particular that this machine is said to be suitable for reductions (and does not refer to breaks). Bran contamination is less at this stage because most of the bran is taken out at the break stage, so even if it would otherwise have been perceived as a problem for double grinding it would have been much less of a problem at this stage. Furthermore, double grinding was taking place at the reduction stage in Australia, making it more obvious to try this machine. Again, it would be obvious to try more than one of these machines (and in particular two of them) in the process line. The alleged invention in the patent is obvious in the light of this machine.

Allis Chalmers 1947

69. This is an earlier version of the 1955 brochure, with a different presentation. The mills are again said to be single, double, two pair and three pair high. The two pair high mill is again clearly described as one where the material passes from the first pair of rollers to the second pair, and it is apparent from a drawing that there is no intermediate sieving, though this time there is no reference to the avoidance of handling. They (and the other versions) are described as standard equipment for flour mills. There is a photograph of a two-pair high roller mill being used in a soybean plant.
70. This brochure discloses a double-grinding machine for flour. Mr Hacon suggested to Mr McGee that it did not disclose this being used for flour, but I find that on a proper reading of the brochure it was. Mr Garratt thought that this could not be used for making white flour, but in forming that view he was relying on the bread miller's prejudice again. He accepted that if an engineer would think it could be used for milling bread flour, he would think it could be used for milling starch flour; and he accepted that if the engineer would put one in, then it would be obvious to put in more than one and sieve in between.
71. Taking the disclosure in this piece of prior art, and viewing it through the eyes of the properly identified skilled person, I again consider and find that the patent is obvious in the light of the disclosure. The relevant skilled person does not come with the bread miller's prejudice colouring his entire view. He knows of the Australian processes which have dispensed with intermediate sieving (albeit with pin mills). There is nothing to suggest that these mills could not be used as part of a high grinding process, and while there is nothing explicit on the point, and bearing in mind that high grinding had been a well-known and standard practice for many decades, it would be very surprising if these machines were actually understood to be for low grinding only. It must be remembered that high grinding describes an entire process, with many milling stages. It does not, in itself, require sieving if a process is to qualify as being high grinding. Once pin mills had been successfully tried, it would be obvious to try a roller mill in their place. It would be obvious to try these machines in a high grinding starch milling process.

Dedrick

72. In 1924 an American Professor called Dedrick published a book called "Practical Milling". Mr McGee says it was a widely known textbook in the US and was used as a reference source. I accept that, though Mr Garratt had not encountered it prior to his involvement in these proceedings. Dedrick gives examples of working mills in industry and one of them is a "Small Corn Mill" (his fig 205). Corn translates as maize in this jurisdiction. It shows a "complete corn mill outfit, having two reductions, each being a 2-high roller mill. A double aspirator [which is a purifying mechanism] with a grading sieve is shown. The purified grits can be ground on the second pair of rolls to fine meal." Thus the mill shows two double grinds, with a sieving or separation between each of them, but not in the middle of each double grind. After the milling the various fractions pass on for appropriate consumption later. Advice is then given about keeping products for table use free from cob, husks and silks.
73. So this diagram certainly shows cereal passing from one roller to another without sieving in between the two, then a sieving, and then another passage between two rollers with no sieving in between them. The grits produced by this process is endosperm by another name. Mr McGee's evidence was that it would be self-evident to an engineer looking at this that he could use this mill to produce raw starch material for a subsequent wet process. Corn milling involved separating the husk (bran) from the endosperm, just as wheat milling did. Dedrick's disclosure demonstrated that. Mr Garratt said that this procedure had nothing to do with flour milling – it was concerned with milling maize and would not have led his skilled person in 1987 to think it could be applied to flour milling, though he could not say what an engineer would have thought on looking at it.
74. I do not agree that the patent is obvious over this piece of prior art. It seems to me to involve an imaginative step or two too far. It is a relatively crude process, explicitly involving only 4 milling phases, involving maize not wheat, and to produce something that is not flour but grits (as Mr McGee agreed). While one can argue logically from each of those steps by a couple of hops and arrive at the invention, the process is not, in my view, something that would have occurred to the skilled man just looking at this in 1987 in the context of milling wheat for starch flour. The differences between this and the invention are too great to be covered by obviousness.

If obvious, and if no prejudice, why has it not been done before?

75. Mr Hacon ran the argument that if there were no prejudice against taking out intermediate sieves, and if the invention is obvious, why had no-one done it before? It is a plain money-saver in terms of saving on the cost of equipment and the movement of stock between machines, and if were such an obvious idea it is odd that it had not been implemented before.
76. I do not think that that point holds good in this case. The "starch from wheat" industry was relatively new, so the failure to exploit the idea was not a long-standing one. Furthermore, the main obviousness point arises out of the prior art which is of some age. It may be that engineers had not yet sat down and considered it but when they do that they see its obvious application. In the circumstances I do not think that this point has any weight.

Added matter

77. In the light of my conclusions on obviousness I do not strictly need to deal with this point, but it was argued, albeit briefly, and I will therefore deal with it, commensurately briefly. I was not shown the application, but both sides agreed that I could assume that the application corresponded with the patent as originally granted for all material purposes. The patent as deemed to be amended limits the source product to wheat, the other cereals having been taken out (apparently as a result of the citation of Dedrick as prior art). The question is whether the resulting patent thereby adds matter not disclosed in the application. This is pleaded as a ground of invalidity (on the assumption that the patent is actually amended), but it would also be an objection to the amendment itself. I do not need to distinguish between those aspects for present purposes.
78. Spomax says that the absence of the other cereals from the patent adds matter because the original application did not disclose that there was any difference between all those cereals, or that the invention was applicable only to wheat. It was not disclosed that there was anything special about wheat which differentiated it from the other cereals. Mr Lykiardopoulos' written final submissions put his point thus:

"If by the amendment, Buhler seeks to say that the alleged invention is only applicable to wheat and would not be equally applicable to rye, maize, or barley, then it is submitted matter is being added by the amendment".

The short answer to the point is that Buhler is not making that point. The patent as amended does not seek to say that wheat is somehow different; it seeks to apply the invention to wheat without saying anything about the other cereals (other than, perhaps, that Buhler does not wish to extend the claim to those cereals, which is not relevant to this debate). Nothing seems to me to be added; no subject matter relevant to the invention is added (see Aldous J in Bonzel v Intervention Ltd [1991] RPC 553 at 574). There is no added feature; nothing special is said about wheat. So far as the process of amendment is concerned, the purpose of the amendment is not to add subject matter; it is to remove references which, it is conceded, would make the patent revocable in the light of Dedrick.

79. Accordingly this point fails.

Conclusion

80. I therefore conclude that this patent should be revoked.
Buhler Patent fig 1