Societe des Usines Chimiques Rhone-Poulenc v Commissioner of Patents

Approaching the matter at the outset in a very general way, it would appear from the affidavit of Mr. Gallafent - and indeed from the specifications themselves - that in 1951 the applicant (which we may identify for practical purposes with the actual inventor) had made what was in effect a twofold discovery or invention. It had found that phenthiazine compounds of a certain general formula possessed useful therapeutic properties, and that one particular compound comprised within that general formula possessed very varied and valuable therapeutic properties. There were two isomers of this particular compound, the same utility attaching to each isomer and to a mixture of the two. This compound has been manufactured and widely sold and used under the trade mark "Largactil". What the applicant has been trying to do is to protect itself both in the wider field and in the narrower field, and it regards a grant of application no. 757 as necessary to give it complete protection in the narrower, which is apparently the more important, field.

Still looking at the matter in a very general way, the history of the relevant applications is briefly as follows. Applications nos. 6971 and 6972 were lodged on 17th December 1951. No. 6971 related to what is now known as Largactil, and no. 6972 to compounds of the general formula. Each was accompanied by a complete specification. It is common ground, and it appears to have been realised by the applicant at the time, that the specification of no. 6972 "overlapped" that of no. 6971. The notifications required by s. 38A of the Patents Act 1903-1950 were published in the Official Journal on 21st February 1952, and thereupon the specifications became open to public inspection and must be deemed, in accordance with sub-s. (2) of s. 38A, to have been "published". Each specification was amended in the Patent Office in respects which will have to be considered. Each application was ultimately accepted, and in due course a patent issued in respect of it. The date of acceptance of no. 6971 was 25th November 1954, and that of no. 6972 was 15th June 1955. The grant in each case was dated as of the date of the lodging of the application, viz. 17th December 1951. In the meantime, on 7th June 1954, applications nos. 756 and 757 had been lodged, each being accompanied by a complete specification. The priority date claimed in each case was the date of lodgment of nos. 6971 and 6972, viz. 17th December 1951. No. 756 also was ultimately accepted, and in due course a patent issued in respect of it with the priority date as claimed. The date of acceptance of no. 756 was 2nd May 1956. No. 757 was refused by the commissioner on 10th January 1957, and the reasons for the refusal were made available to the applicant shortly thereafter.

It is convenient, I think, before examining in further detail the history of the applications, to turn to the relevant legislation, and to consider certain questions of construction which arise out of the commissioner's reasons for refusing to accept no. 757. At the time of the lodgment of nos. 6971 and 6972 (17th December 1951) the Act in force was the Patents Act 1903-1950. On 27th September 1952 the Patents Act 1952 received the royal assent. Section 2 of that Act provided that it should come into force on a date to be fixed by proclamation. It was proclaimed to come into force on 1st May 1954. It repealed the Patents Act 1903-1950, for which it substituted a series of new provisions differing in many respects from the old. Section 5 provided that the new Act should apply to and in relation to all applications for patents lodged after its commencement, but that the repealed Acts should apply, notwithstanding their repeal, to and in relation to all applications for patents lodged before the commencement of the new Act. What may be called the old Act, therefore, applied to nos. 6971 and 6972, and the new Act to nos. 756 and 757.

Section 33 of the old Act provided that an application for a patent should be for one invention only, and s. 65 provided that a patent should be granted for one invention only. The Act contained general provisions for the amendment of specifications before and after grant, but it contained no specific provision relating to specifications which claimed more than one invention. The matter was, however, dealt with by regulation. Regulation 11 of the Patent Regulations 1912 provided: - "(1) When a specification comprises several distinct matters, they shall not be deemed to constitute one invention by reason only that they are all applicable to or may form parts of an existing machine, apparatus, or process. (2) Where a person making application for a patent has included in his specification more than one invention, the Commissioner may require or allow him to amend such application and specification and drawings or any of them so as to apply to one invention only, and the applicant may make application for a separate patent for any invention excluded by such amendment. (3) Every such last-mentioned application may, if the Commissioner at any time so directs, bear the date of the original application, or such date between the date of the original application and the date of the application in question, as the Commissioner directs, and shall otherwise be proceeded with as a substantive application in the manner prescribed by the Act and by these Regulations." The new Act dealt itself specifically with the matter, incidentally taking away from the commissioner the discretion which reg. 11 had given him as to priority dates. By s. 51 it provided: - "(1) A person who has made an application for a patent may, at any time before publication of the complete specification, make one or more further applications in respect of an invention disclosed in the provisional specification or complete specification lodged in respect of the first-mentioned application." Section 45 (so far as relevant) provided: - "(1) Subject to this Act, the priority date of a claim of a complete specification is the date of lodgment of that complete specification. (4) The priority date of a claim of a complete specification lodged in respect of a further application made by virtue of section fifty-one of this Act, being a claim fairly based on matter disclosed in the provisional specification or complete specification lodged in respect of the original application, is the date which would have been the priority date of that claim if that claim had been included in the complete specification lodged in respect of the original application." The Patents Act 1955, by s. 5, amended sub-s. (4) of s. 45 so as to make the last words read "if that claim were a claim of the complete specification lodged in respect of the original application." Section 5 of the Act of 1955 also added to s. 45 a new sub-s. (5), which is of great importance in the present case, because it deals with cases (of which this is one) where an original application has been made under the old Act and what is commonly known as a "divisional application" has been made under the new Act. Sub-section (5) reads: - "Where, in respect of an application for a patent lodged under the repealed Acts, the Commissioner has required or allowed the applicant to amend the application and specification and drawings or any of them so as to apply to one invention only and the applicant has made an application under this Act for an invention excluded by the amendment, the priority date of a claim of the complete specification lodged under this Act, being a claim fairly based on matter disclosed in the provisional specification or complete specification lodged under the repealed Acts, is the date which would have been the priority date of that claim if that claim were a claim of the complete specification lodged in respect of the application under the repealed Acts." The applicant says that the present case falls within s. 45 (5), and that it is entitled to a grant of no. 757 with the priority date of 17th December 1951. There was nothing, so far as I can see, to prevent the applicant lodging no. 757 apart altogether from s. 45 (5), but it is only under s. 45 (5) that it can obtain the priority date which it desires. Section 49 (3) applies only where the original application has been made under the new Act.

Now, it is common ground, as the commissioner says, that the subject matter of no. 757 was described in the specifications of both nos. 6971 and 6972 as originally lodged. The commissioner, however, says that there was not in the claiming clauses of either of those specifications any specific claim to the subject matter of no. 757, and that, in the absence of any such specific claim, s. 45 (5) does not apply, even if (which he does not concede) the case otherwise falls within that sub-section. Up to this point the commissioner's reasoning would, I think, lead only to the conclusion that the priority date of a patent granted on no. 757 should be not 17th December 1951 but 7th June 1954. The commissioner, however, carries the matter further. The fact remains, he says, that the subject matter of no. 757 was disclosed in the original specification of no. 6971, and that specification was published when the notification under s. 38A of the old Act appeared in the Official Journal. It follows, the commissioner says, that the subject matter of no. 757 was not novel at the time of the lodgment of that application, and that that application must be refused. The applicant, as Mr. Aickin put it, "has anticipated itself". This conclusion may appear at first sight somewhat surprising, if not self-contradictory. It seems to me, however, to be a perfectly logical view. The conclusion appears inevitable if we accept both the legal premiss that s. 45 (5) does not apply unless the subject matter of the divisional application was specifically claimed by the original application, and the factual premiss that the subject matter of no. 757 was not specifically claimed under either no. 6971 or no. 6972.

The commissioner's view rests, however, on the truth of both of those premisses. I need not consider whether, as a matter of construction of the specifications, the factual premiss is true or not, because I am clearly of opinion that s. 45 (5) cannot be limited to cases where the subject matter of the later application has been made the subject of a specific claim in the earlier application. So narrow a view might lead, I think, to real hardship for inventors, and the result to which it has led in the present case is, as I have said, at first sight surprising. Such considerations disincline one so to construe the sub-section if the broader view is fairly open on its language. But in truth the language of the section is not merely fairly open to the wider construction. It contains, in my opinion, affirmative indications that the broader view is the correct view. It does not speak of amendment of the claims or of the claiming clause of the earlier specification. It speaks of amendment of "the application and specifications and drawings or any of them". It seems to me to contemplate (inter alia) that an amendment of a drawing may have the effect of excluding a separate invention. This indicates that disclosure without claim is enough. An invention may be disclosed by a drawing. Again, the claim in the later specification must be "a claim fairly based on matter disclosed" in the earlier specification. This requirement would be senseless if the sub-section really meant that the subject matter of the later claim must have been actually claimed in the earlier specification. Finally, reference may be made to s. 51, which is referred to in sub-s. (4) of s. 45, and which speaks not of claiming, but of "disclosure". There must, of course, be a real and reasonably clear disclosure.

The question whether s. 45 (5) should be construed as limited to cases where the application under the new Act is made before publication of the original specification seems to me to be a separate and distinct question. Perhaps the strongest way in which the argument for so limiting sub-s. (5) can be put is to say that the second application to which it refers is an "application under this Act", that the only authority to be found in the Act for making a divisional application is that given by s. 51, and that s. 51 authorises such an application only "before the publication of the complete specification". The specification here referred to is, of course, the specification relating to the original application. The argument can be (and was) supported by general considerations of convenience. I do not regard the question as free from difficulty, but I am of opinion that the text of the relevant provisions of the Act is against the argument, and that no sufficient warrant can be found for the suggested limitation of the application of sub-s. (5).

The opening words of s. 51 (1) are: "A person who has made an application for a patent may ". The application here referred to is, in my opinion, an application under the new Act. It is true that the perfect tense is used - "has made". But I cannot read the opening words as referring to something done before the Act came into operation. The applications to which it applies are applications made after its commencement (s. 5), and generally, wherever it enacts something with respect to "applications", it refers, in my opinion, to original applications made after its commencement and under it. So, for example, when s. 54 says: "Where an application and specification have not been accepted ", it must, in my opinion, be read as referring only to an original application and specification lodged after the commencement of the new Act and subject, by virtue of s. 5, to the provisions of the new Act.

Then sub-s. (4) of s. 45 specifically refers to applications "made by virtue of s. 51". It speaks of "further applications", which is the expression used in s. 51, and it seems plainly intended to apply, and to apply only, to cases where the original application has been made after the commencement of the new Act and under that Act. So far no provision has been made for cases, which will necessarily be limited in number, where an original application has been made under the old Act, and a divisional application has been lodged after the commencement of the new Act and therefore, by reason of s. 5, under and subject to the new Act. What is now called the "priority date" of divisional applications had been dealt with before the commencement of the new Act by reg. 11, which did not require the divisional application to be made before publication of the original specification but did leave to the commissioner, within limits, a discretion as to priority date. No doubt the fact that an original specification had been published before the divisional application was lodged would be a matter affecting the discretion of the commissioner. The regulations, of course, disappeared with the repeal of the old Act, and one cannot doubt that the purpose of sub-s. (5), which was enacted in 1955, was to provide for this limited number of intermediate cases, as they may be called. That new sub-section clearly left no discretion to the commissioner, as one might perhaps have expected it to do in the case of a divisional application lodged after publication of the relevant original, but, in the absence of express words or any clear basis for an implication, it would, in my opinion, be wrong to say that the benefit of s. 45 (5) is lost to an applicant if his original specification is published before his divisional application is lodged.

There is one other question of construction which has seemed to me to arise on s. 45 (5). The amendment to which it refers must be made "so as to" make what is amended "apply to" one invention only. The conjunction may be thought to be ambiguous. Does it refer to purpose or to consequence? I think the latter is the correct view. I think it is enough if the effect of the amendment - whether that effect is or is not intended or realised at the time by the applicant or by the commissioner - is to make a specification henceforth comprise one invention only.

I should perhaps add, in conclusion on this aspect of the case, that I have considered whether the true view is not that the case is governed altogether by the old Act and reg. 11. I think, however, that there is clearly no escape from the view, assumed throughout the argument before me, that a divisional application is a new substantive application, and, by virtue of s. 5, is, if made after the commencement of the new Act, governed by and subject to the new Act.

I have thought it best to deal first with these general questions of construction. My answers to them, however, though in favour of the applicant, do not dispose of this appeal, because the commissioner does not concede that the case falls at all within s. 45 (5). It is, of course, common ground that both nos. 6971 and 6972 were amended. It does not seem to matter whether those amendments were "required" or merely "allowed" by the commissioner: they were either required or allowed. But the commissioner denies that either of them was amended "so as to apply to one invention only", and he denies that no. 757 is an application "for an invention excluded" by any amendment of no. 6971 or no. 6972. The two questions thus raised are the questions which I now have to consider. They are questions of considerable difficulty, especially to one who is unfamiliar with the subject of organic chemistry, but I have had much assistance from counsel, and also from a clear and concise statement (Exhibit E) of the elementary principles involved, which has been attributed to the commissioner, and which I have found most helpful. I have tried to heed the warning implicit in the words of Fletcher Moulton on Patents (1913), p. 146: "There is no class of cases where the thorough education of the Court by means of expert testimony is so necessary as in chemical cases."

In considering the history of the relevant applications there are two things to be borne in mind, which it will be well to mention at once. In the first place, as has already been said, the specification in each case claims a process or processes and then a product or products by reference to the process or processes. There was in no. 6972 as lodged one claim for a product without reference to a process, but the claim in this form disappeared before grant. One would think that the product is the important thing, because Mr. Gallafent says that "when a chemical compound has been specified in terms of its chemical structure, it is within the skill of any competent chemist to formulate ways of making it by virtue of his general knowledge of organic chemical reactions". In tying, so to speak, the product to the process in each case, Mr. Gallafent may have had in mind s. 38A of the English Patents Act formerly in force, and s. 10 (1) (c) of the English Act of 1949. These provisions have never, so far as I know, been in force in Australia, but the formulation of the process as the primary thing was apparently thought to be a wise precaution. Be all this as it may, however, the process is in each case put as the primary thing, and the product is in fact tied to the process. And I think, as will be seen, that this fact assumes importance.

The second thing to be borne in mind is a technical and more complex matter. Each of the processes with which we are concerned has, of course, as its object and result the production of a substance by means of a chemical reaction. Two classes of chemical reaction are referred to or involved in the relevant specifications. The first is what is called a cyclisation reaction. With this we are only indirectly concerned. It takes place whenever a closed ring is formed from a chain - or, as we may perhaps say, whenever the second side of the appropriate chemical equation contains a ring formation which does not appear on the first side of the equation. A cyclisation reaction is referred to in the specifications of no. 6971 and no. 6972 both before and after amendment, and also in the specifications of no. 756 and no. 757. But it comes into the picture, so to speak, only at a preliminary stage and in the course of describing a method of obtaining the halogenophenthiazine compounds which are employed in the processes actually claimed. The substances referred to in the description of this method were referred to by counsel as the "basic" starting material. The method so described does not enter into any of the claims of the amended specifications of no. 6971 or no. 6972 on which the patents were granted. Nor does it enter into any of the claims of the specifications of no. 756 or no. 757.

The other relevant class of reaction is known as a condensation reaction, and is defined as "a union of like or unlike molecules, usually with the elimination of one or more molecules of water, hydrochloric acid (HCl) or alcohol". The commissioner's memorandum takes as a typical case the reaction which will take place (under appropriate conditions, of course) between methylamine and propyl chloride or chloropropane (as it is alternatively called). The reaction is represented as follows: It is seen that the methyl and propyl radicals have become linked to the nitrogen atom, while one of the hydrogen atoms has left the nitrogen atom and attached itself to the chlorine atom to form hydrochloric acid.

The reactions which take place in the processes claimed in the relevant specifications are all condensation reactions, but two different processes are involved. These have been distinguished by Mr. Gallafent as "Reaction Type A" and "Reaction Type B", and by Mr. Aickin in argument more shortly as "Reaction A" and "Reaction B". The difference between the two can hardly be understood without reference to one or other of the specifications, and I think it is convenient to take that of no. 6971. The structure of the molecule of the substance with which that specification is concerned is represented in "chemist's shorthand" thus: -

This formula, to which, for convenience, I will refer later as the formula for "Largactil," may be regarded as consisting of three parts or constituents - it would, of course, be wrong to say "elements". The first is what may be called the triple-ring structure at the top. Two benzene rings are connected by way of a sulphur atom and a nitrogen atom, the former being here divalent and the latter here trivalent. Two of the valence links of the nitrogen atom are linked to carbon atoms in the benzene rings. If the third had a hydrogen atom linked to it, the triple-ring structure would represent a phenthiazine compound known as chlorophenthiazine. The line connecting the chlorine atom with the righthand ring indicates that the chlorine atom may be substituted for a hydrogen atom at more than one position on the ring: in fact we are told later that it may be so substituted either in the 1-position or in the 3-position. The second constituent, which is linked to the nitrogen atom in the first, consists of a propyl group, which is divalent. The third, which is linked to the other valence link of the propyl group, is an amine group, consisting of a nitrogen atom and two methyl groups. In the process which involves Reaction A we begin with 1- or 3- chlorophenthiazine, which, as has been indicated, is represented by the formula - We then react this with a 3-dimethylamino-1-halogeno-propane, the formula of which is - The numbers indicate, if I have succeeded in understanding the matter, in the case of the chlorophenthiazine, the number of the hydrogen atom which is replaced on the benzene ring by the chlorine atom, and, in the case of the dimethylamino-halogeno-propane, the positions of the dimethylamine and the chlorine atom relatively to the members of the propyl chain. The reaction is shown by the following equation, which I take from Mr. Gallafent's affidavit.

The Y is the halogen atom which has been introduced as a constituent of the halogeno-propane. The H is the hydrogen atom which was originally in the chlorophenthiazine compound. The links at the top of the N are the links by which the nitrogen atom is attached to carbon atoms of the benzene rings of the phenthiazine compound, which, apart from the loss of the hydrogen atom, remains unchanged. The broken lines enclosing "H + Y" are merely to indicate where the HY of the product has come from.

I turn now to what has been called Reaction B. In the process which involves Reaction B we react dimethylamine ( ) with a chlorophenthiazine compound in which the hydrogen atom linked to the nitrogen atom has been replaced by the propyl group ((CH2)3), to the other valence link of which is linked a halogen atom. This chlorophenthiazine compound is represented by the formula -

The reaction of this compound with the dimethylamine is shown by the following equation (which again I take from Mr. Gallafent's affidavit): - The Y is the halogen atom, which figured in the phenthiazine compound with which we began. The H comes this time not from the phenthiazine compound with which we began, but from the dimethylamine. The links at the top of the N are the links by which the nitrogen atom is attached to carbon atoms of the benzene rings of the phenthiazine compound, which, apart from the loss of the halogen atom, remains unchanged. The broken lines enclosing "Y + H" are merely to indicate where the HY of the product has come from.

I have now to consider the four relevant specifications and the history of their passage through the Patent Office. I do not propose to go into more detail than seems absolutely necessary, and in the view which I have ultimately taken I think I can be fairly brief. I will begin with no. 6971. The specification accompanying this application began by giving what I have called the formula for "Largactil", which has been set out above. The ring positions are numbered in the conventional way, and it is stated that the chlorine substituent may be in the 1-position or in the 3-position. The document then proceeds: - "The compounds of the present invention are prepared from meta-chlorodiphenylamine by known methods for the conversion of a diphenylamine into an n-dialkylamino alkyl phenthiazine. By the expression "known method" is meant any method heretofore described in the chemical literature." So far the applicant has described a process which proceeds by "known methods" from a basic starting point. It goes on to give what it calls "a preferred process for preparing the new compounds". This "preferred" process, as I understand it, really begins at a later stage, after the necessary phenthiazine compound has been obtained. It proceeds by way of "Reaction A", which I have described above. After giving certain detailed laboratory directions, the specification then states: - The new compounds of this invention can also be obtained by the condensation of dimethylamine with a phenthiazine compound of the formula -

wherein the chlorine substituent is in the 1- or 3-position and Y represents a halogen atom. This is the other process which I have described above, and which works by way of "Reaction B". Directions are then given for preparing the chlorophenthiazine compounds "employed as starting material in one or other of the two abovementioned processes", i.e. the process by Reaction A and the process by Reaction B. Then follow four "examples". The first three involve Reaction A: the fourth involves Reaction B. The specification then claims: - "1. A process for the preparation of new therapeutically useful phenthiazine compounds of the formula" (which is then set out again) "wherein the chlorine substituent is in the 1- or 3-position, which comprises the conversion of meta-chloro-diphenylamine by known methods for the conversion of a diphenylamine into a n-dialkylamino alkyl phenthiazine. 2. A process as claimed in claim 1 which comprises condensing 1- or 3-chlorophenthiazine or a mixture thereof with a 3-dimethylamino-1-halogeno-propane." Claim 6 is the only other claim that need be mentioned. This claim claims a product as distinct from a process, and is for "phenthiazine compounds of the formula" (which is again set out) "wherein the chlorine substituent is in the 1- or 3-position when prepared by the process (sic) hereinbefore particularly described and ascertained." For reasons which I have given, it is, in my opinion, of no consequence whether this claim 6 should be here read as including a claim for the product when prepared by Reaction B. Claim 6 was never amended (except that it was re-numbered and became claim 5), and in the specification on which the patent was granted the singular - "process" - is, as will be seen, correct. I am disposed to think, however, that in the specification as lodged, the singular appears by reason of a clerical error, and that the word should be read as "processes". Two distinct processes have been clearly described in the body of the specification.

The specification of no. 6971 as lodged was amended in important respects before grant. In the body of the specification the references to "known methods" and to the process involving Reaction A as a "preferred method" were deleted, and for the word "from" in the relevant passage was substituted the word "by". The result was that the passage read: - "The compounds of the present invention are prepared by condensing a phenthiazine compound represented by the general formula" (which is set out, as before) "wherein the chlorine substituent is in the 1- or 3-position with a 3-dimethylamino-1-halogeno-propane." The passage which commenced, "The new compounds of this invention can also be obtained", and which described the process involving Reaction B, was also struck out. Example 4 was struck out. Claims 1 and 2 were amended so as to make one claim which read: - "A process for the preparation of new therapeutically useful phenthiazine compounds of the formula" (which is set out as before) "wherein the chlorine substituent is in the 1- or 3-position which comprises condensing 1- or 3-chlorophenthiazine or a mixture thereof with a 3-dimethylamino-1-halogeno-propane." The reference to the mixture is occasioned by the fact (already noted) that the two isomers have the same therapeutic properties, which properties belong also to a mixture of the two.

One important result of the amendments made was that the "preliminary" process by known methods from a basic starting point ceased to be part of what was claimed. The other important result was that all references to a process involving Reaction B were deleted. The patent granted on no. 6971 was for a process involving Reaction A and for compounds prepared by that process.

It is convenient to deal next with no. 6972, which was lodged at the same time as no. 6971. Again it is not necessary to follow in detail the objections and discussions which took place in the progress of the application through the Patent Office. There was confusion at one stage between nos. 6971 and 6972, which caused difficulty and misunderstanding, but I do not think that anything now turns on this.

The specification of no. 6972 as lodged began: "This invention relates to new phenthiazine derivatives having valuable therapeutic properties and to processes for the preparation of new phenthiazine derivatives." The specification then recites no. 6971 as a "co-pending application", setting out the Largactil formula, and proceeds: - As a result of further research, we have found that interesting therapeutic properties are also possessed by the other members of the class of phenthiazine compounds having the general formula:

wherein: R represents a hydrogen, chlorine or bromine atom or a methyl or methoxy group in the 6- or 8-position, X represents either a chlorine or a bromine atom in the 1- or 3-position, A represents a divalent, straight or branched aliphatic chain containing from 2 to 5 carbon atoms and R1 and R2 represent either individual methyl or ethyl groups or divalent groups which together with the adjacent nitrogen atom form a mono-nuclear heterocyclic ring. This formula is later referred to as "general formula II". Pausing here for a moment, it is obvious that the formula covers a very considerable number of combinations. One of the combinations included is the case where R represents a hydrogen atom (which means that there is no substituent in the left-hand ring), X represents a chlorine atom in the 1- or 3-position, A represents a straight aliphatic chain containing 3 carbon atoms, and R1 and R2 represent methyl groups. In this combination we have the very combination which is the subject of no. 6971. As I have said, however, it was realised by the applicant at the time of lodgment that no. 6971 and no. 6972 "overlapped", and that amendments would have to be made in one or both of the specifications before either could be the subject of a grant. It may be mentioned here also that the case where R1 and R2 represent groups which together with the adjacent nitrogen atom form a mono-nuclear heterocyclic ring was later made the subject of the separate application, no. 19702, which was lodged on 9th July 1953, and was ultimately granted with the same priority date as nos. 6971 and 6972 (17th December 1951).

The specification proceeds: - In a preferred form of the invention, A represents an alkylene group containing from 2 to 5 carbon atoms and R1 and R2 each represent a methyl or ethyl group or together represent the atoms necessary to complete a pyrrolidine, piperidine or morpholine nucleus. The compounds of the present invention are prepared from a meta-chloro-or-bromo diphenylamine (which may be substituted by a further chlorine or bromine atom or by a methyl or methoxy group) by known methods for the conversion of a diphenylamine into a N-dialkylaminoalkylphenthiazine. By the expression "known method" is meant any method heretofore described in the chemical literature. One preferred process for preparing the new compounds involves condensing a phenthiazine compound represented by the general formula:

(wherein R and X are as hereinbefore defined) with a tertiary aminoalkyl halide of the formula: wherein Y represents a halogen atom and R1 and R2 have the significance hereinbefore defined and A1 represents a divalent straight or branched aliphatic chain containing from 2 to 5 carbon atoms which may be the same as the chain represented by A in general formula II or, in the case of a branched chain, an isomeric form thereof. It is to be noted that, as in the case of no. 6971, we get a reference to a process which proceeds by "known methods" from a basic starting point, followed by a description of a process which is called a "preferred process", but which really begins at a later stage, and works by way of Reaction A. The specification continues to follow the general lines of no. 6971. It proceeds to give certain detailed information for facilitating the desired reaction (including, by the way, a diagram which seems mistakenly to attribute trivalency to carbon). It then gives another method, as distinct from the "preferred process". It says: - The new compounds of this invention can also be obtained by the condensation of a secondary amine of the general formula: with a halogeno alkyl phenthiazine of the formula: wherein X, R1, R2, A1 and Y are as hereinbefore defined. This method is particularly suitable in the case where A1 has the formula - (CH2)n - , n having the value 4 or 5. This process involves Reaction B. Then follow no less than 24 "examples". I have not attempted to make any examination of these. Mr. Aickin said, and Mr. Tredinnick did not challenge the statement, that Example XXIII involved Reaction B, and that probably some others (which were later omitted by amendment) involved Reaction B, and I simply accept Mr. Aickin's statement.

The claims of no. 6972 as lodged need not be set out in detail. Claim 1 is for the process by "known methods" from a basic starting point which has been described in the opening part of the specification. Claim 2 is for the so-called "preferred" process, which involves Reaction A. Claim 6 is for A modification of the process claimed in claim 2 wherein a halogeno-alkyl phenthiazine of the type:

is condensed with a secondary amine of the type: This is a claim for the alternative process which has been described in the body of the specification, and which has been noted above as involving Reaction B. Claim 7 is for "A process as claimed in claim 1 wherein R1 and R2 complete a pyrrolidine, piperidine or morpholine ring." The remaining claims are claims for the products as distinct from the processes.

The amendments which were made before grant in the specification of no. 6972 were analogous to those which were made in the specification of no. 6971, and it is only necessary, I think, to state their effect in general terms. The printed copy which is before me of the amended specification on which the patent was granted contains a number of errors, and claim 1, as it stands, is unintelligible. The parties, however, are agreed as to the necessary corrections, and the applicant has undertaken to apply for, and the commissioner has intimated that he will allow, the amendments required to put the document in order.

In the body of the specification in its final form the compounds which are the subject of no. 6971 and no. 19702 are expressly and specifically excluded. The reference to preparation from a basic starting point by known methods is deleted. For the words "One preferred process for preparing the new compounds involves" are substituted the words "The compounds of the present invention are prepared by ". The effect of this is to restrict the invention described to the single process which is then set out, and which involves, as has been observed, Reaction A. The passage beginning "The new compounds of this invention can also be obtained" is deleted. The process described in the passage thus deleted was the process involving Reaction B. Several of the "examples", including Example XXIII, which was a process by way of Reaction B, are omitted. The claims in the final specification include only one process claim, which is claim 1. Counsel are agreed that this claim is intended to embody the process by way of Reaction A which is described in the body of the specification, and that it should be read as follows: - A process for the production of new therapeutically useful phenthiazine derivatives of the general formula:

wherein: R represents a hydrogen, chlorine or bromine atom or a methyl or methoxy group in the 6- or 8-position, X represents a chlorine or bromine atom in the 1- or 3-position, A represents a divalent, straight or branched aliphatic chain containing from 2 to 5 carbon atoms and R1 and R2 represent either individual methyl or ethyl groups or divalent groups which together with the adjacent nitrogen atom form a mononuclear saturated heterocyclic ring, with the exclusion of those compounds in which simultaneously R represents a hydrogen atom, X represents a chlorine atom, A represents n-propylene and R1 and R2 each represent methyl, and of those compounds in which simultaneously R represents a hydrogen atom and R1 and R2 together represent the residue of a pyrrolidyl or piperidyl ring, which process comprises condensing a phenthiazine of the type: where R represents a hydrogen chlorine or bromine atom or a methyl or methoxy group in the 6- or 8-position and X represents either a chlorine or a bromine atom in the 1- or 3- position with a tertiary amino alkyl halide of the type: where Y represents a halogen atom, R1 and R2 have the significance above defined and A1 represents a straight or branched aliphatic chain containing from 2 to 5 carbon atoms. Claims 2, 3 and 4 merely repeat claim 1 with additions desirable for ensuring a satisfactory reaction. The remaining claims are for products as distinct from processes. Claim 5 is a general claim, and claims 6-11 are for particular combinations. Of these seven claims it is sufficient to say that they are all tied to the processes claimed by claims 1-4, and that they all exclude, either by express words or by their very nature, the compounds claimed by nos. 6971 and 19702.

The main net result of the amendments which I have attempted to describe was threefold. In the first place, the "preliminary" process by known methods from a basic starting point ceased to be part of what was claimed. In the second place, the subject matter of no. 6971 and the subject matter of no. 19702 were specifically excluded: there was no longer any "overlapping". And, in the third place, the only process claimed was a process by way of Reaction A. The patent granted on no. 6972 was for a process involving Reaction A and for compounds prepared by that process.

These results were, of course, desirable in themselves, but they left uncovered a part of the field which the applicant had desired to cover. They left the applicant unprotected in respect of "Largactil" (the subject matter of no. 6971) prepared by a process involving Reaction B, and they left it also unprotected in respect of the broader subject matter of no. 6972 prepared by a process involving Reaction B. It was for the purpose of filling the gap thus created, or about to be created, that nos. 756 and 757 were lodged on 7th June 1954. Each of these was accompanied by a complete specification which claimed for each of its claims the same priority date as nos. 6971 and 6972 (17th December 1951).

The specification of no. 756 gave as the general formula of its subject compounds the general formula given in the specification of no. 6972. It then proceeded: - "One method of making the said compounds is described and claimed in the said Application No. 6972/51 and the present Application is concerned with an alternative method of producing the said compounds. According to the present invention a secondary amine of the general formula shown in Fig. 2 of the drawings, is reacted with a halogen alkyl phenthiazine of the general formula shown in Fig. 3 of the drawing wherein R1, R2, X and A have the meanings assigned to them above and Y is a halogen atom." The figures referred to in this passage are the following: -

The process described in the passage quoted above proceeds by way of Reaction B. Claim 1 claimed that process, expressly excluding the processes covered by nos. 6971 and 19702. Claim 2 was for the corresponding product. It was tied to the process, and was subject to the same express exclusions. The specification was not amended in any respect now material. It was accepted on 2nd May 1956, and a patent in due course issued with 17th December 1951 as its priority date.

The specification of no. 757 gave as the formula of its subject compound the formula given in the specification of no. 6971. It then proceeded: - "One method of preparing the said compounds is described and claimed therein" (i.e. in the specification of no. 6971) The present invention is concerned with an alternative method of making the said compounds. According to the present invention the said compounds are obtained by condensing dimethylamine with a phenthiazine compound of the formula:

wherein the chlorine substituent is in the 1- or 3- position and Y represents a halogen atom. The process described in the passage quoted proceeds by way of Reaction B. Claim 1 claims that process, and claim 2 claims the corresponding product, tying it to the process. There was no need, of course, in this case for any express exclusion. The commissioner has refused to accept no. 757 and its accompanying specification, and the question for me is whether he was right in so refusing.

Before I approach this question it will clear the ground and make for simplicity if I say at once that, if no. 757 can properly be regarded as a "divisional" application at all, it seems to me that it will be correct to regard it as a divisional of no. 6971 rather than of no. 6972. It is true, of course, that the subject matter of no. 757 was included in the specification of no. 6972 as lodged, but so was the subject matter of the specification of no. 6971 itself. It makes no difference to the applicant whether we relate no. 757 to no. 6971 or to no. 6972. But it seems to me that a patent granted on the specification of no. 757 would naturally be regarded as the complement of the patent granted on no. 6971, just as the patent granted on no. 756 is naturally regarded as the complement of the patent granted on no. 6972.

I agree with the commissioner that the real question in the case is whether the applicant can bring itself within the terms of s. 45 (5) of the Patents Act 1952-1955. As I have said, however, I do not agree with the commissioner's view that the sub-section cannot apply unless the invention which is the subject of the second or divisional application was specifically claimed in the original application as it stood before amendment. Also, as I have said, I am of opinion that the position is not affected by the fact that the specification of no. 6971 was published some time ago under s. 38A of the old Act. I have also expressed my opinion that the conjunction "so as to" in s. 45 (5) ought to be read as referring to consequence and not to purpose. These things having been said, I think that the only proper approach to the case is to look at s. 45 (5), and, taking the stated conditions of its application one by one, inquire whether that condition is fulfilled.

Up to a point the position seems clear enough. Clearly there was an application for a patent - no. 6971 - lodged under the Acts repealed by the Act of 1952. Clearly the commissioner required or allowed - it does not matter which - the applicant to amend the specification of no. 6971. Clearly the applicant has made an application - no. 757 - for a patent under the Act of 1952. And clearly, in my opinion, the claims in the specification of no. 757 are claims "fairly based on matter disclosed in the complete specification lodged" with no. 6971. The subject matter of no. 757 is described explicitly (even if not claimed) in the specification of no. 6971. But did the amendment of the specification of no. 6971 have the consequence of making that specification "apply to one invention only"? And is no. 757 "an application for an invention excluded" by that amendment?

The commissioner has answered both these questions in the negative, and his reasoning is not completely met by saying (as I have said) that s. 45 (5) may apply although the subject matter of the later application has not been actually claimed in the specification of the earlier application. The substance of the commissioner's view may, I think, fairly be stated as follows. He begins by saying that an amendment of the earlier specification cannot have the consequence of making that specification "apply to one invention only" unless that specification, before it was amended, at least disclosed - he would, of course, say "claimed" - more than one invention. He then says that neither no. 6971 nor no. 6972 originally disclosed more than one invention, and that no question of plurality of inventions was ever raised in connexion with the specification of either of those applications. He would say, I think, that the real and essential subject matter of each of those original specifications was a substance or compound, and that the mere fact that two alternative processes for obtaining the substance were disclosed does not mean that two inventions were disclosed. The conclusions follow, he says, that the amendment of the specification of no. 6971 did not make that specification apply to one invention only, and that no. 757 is not an application for an invention excluded by the amendment. He might perhaps seek support for his view from such cases as In re an Application for a Patent by G. & H. [1] , which were decided on s. 38A of the English Act formerly in force, but I do not think that these cases have any real bearing on the present case.

The question of what constitutes a single invention within the meaning of ss. 33 and 65 of the old Act (s. 35 of the new Act) may, I think, be a very difficult question, and there appears to be very little authority on it. This is perhaps not surprising, for, while one would suppose that the question must not seldom arise, it is a question which arises in the office and is probably generally settled there to the satisfaction of the applicant and of the commissioner: that a patent relates to more than one invention is not a ground for revocation after grant. The present case is curious in that it is the applicant who seeks to maintain that his specification claimed more than one invention, and the commissioner who opposes that view: it will usually, I should imagine, be the other way round. There are two cases in Griffin - Jones' Patent [2] and Re Hearson's Patent [3] . In the former of these the Law Officer (Herschell S.-G.) said: - "It seems to me that the general object of the invention is the test by which the question must be decided I should always allow alternative devices for producing a particular object as one invention." [2] Reference may also be made to In re Z's Application for a Patent [4] and In re J's Application [5] . In the former case the Law Officer (Evans S.-G.) said: - "It is impossible to give a definition of what constitutes "one invention" which would be applicable to every case The question cannot be decided upon any strict principle. A certain amount of discretion must be permitted to the Comptroller; and unless he exercises that discretion in a way which is clearly unreasonable I do not think that the Law Officer ought to interfere with it" [6] . This passage cannot, of course, be pressed too far. The matter is not simply one for the discretion of the comptroller (or commissioner). Ultimately it is a matter of the application of a legal standard. The passage is useful, however, in bringing out that a particular specification may quite reasonably be regarded either as disclosing one invention only or as disclosing two or more inventions.

1. (1885) Griff. 265. 2. (1885) Griff. 266. 3. (1885) Griff. 265. 4. (1910) 27 R.P.C. 285. 5. (1924) 42 R.P.C. 1. 6. (1910) 27 R.P.C., at pp. 286, 287.

It must be admitted that s. 45 (5) is a difficult sub-section. But it is obviously intended for the protection of inventors, and I think that it should receive a liberal construction. If it were a vital question whether the original specification of no. 6971 related to one invention only, within the meaning of s. 33 of the old Act, I think that, having regard to what was said by the learned Solicitor-General in Jones' Patent [1] , I would agree with the commissioner that it did, and that no real question of plurality ever arose. But I do not think that this is a vital question. I cannot read s. 45 (5) as requiring that the original specification should, as a matter of law, claim or disclose a plurality of inventions. The relevant requirements of the sub-section are, in my opinion, fulfilled if what is eliminated by amendment (1) leaves the original specification applying to one invention only and (2) is capable by itself of being the subject of an application for a patent. In other words, it is enough if there is disclosed in the original specification an invention which could reasonably be regarded as severable, and it is in fact severed by amendment and made the subject of a separate application. If these conditions are fulfilled, it will be true to say that the original specification has been amended "so as to apply to one invention only", and that an application has been made for "an invention excluded by the amendment". This view cannot, I think, lead to any inconvenient result: the commissioner has full control of the amendment of specifications, and is in a position to prevent any unnecessary or undesirable "splitting up" of inventions.

The conditions which I have stated are fulfilled in the present case in relation to nos. 6971 and 757. It is important that the original application was primarily for a process. It is true that the specification of no. 6971 as lodged begins by saying that "This invention relates to new phenthiazine derivatives", and proceeds immediately to give the formula of the substance which has been called "Largactil". But it gives a little later a "preferred process for preparing the new compounds" - a process involving Reaction A, and it later describes an alternative process, which is a process involving Reaction B. The first five claims are for processes, the second being for a process specifically involving Reaction A. The sixth and last is for the product. But it is not for the product simpliciter: it is for the product only when prepared by the "process" which has been "described and ascertained". The important point is that there is no claim apart from a process, and two processes - the Reaction A process and the Reaction B process - have been clearly described in the specification. The applicant has tied himself to a process or processes. He has defined his invention by reference to his process or processes. When all reference to one of the two processes which he has disclosed is eliminated by amendment the result is, in my opinion, correctly described by saying that his specification has been made to "apply to one invention only". And, when he makes an application for the process which has been eliminated by amendment from his original application, he is, in my opinion, "making an application for an invention excluded by the amendment". The case thus falls, in my opinion, within s. 45 (5) of the Act, and the applicant is entitled to a grant of no. 757 with 17th December 1951 as its priority date. It may, of course, be (though I could not say so on the evidence) that what have been called Reaction A and Reaction B are no more than obvious chemical equivalents. But the relevant amendments of the specification of no. 6971 did not proceed on that basis. They eliminated all reference to the process by Reaction B, and left the applicant prima facie unprotected in respect of the product prepared by a process which he had disclosed.

It follows from what I have said that this appeal should, in my opinion, be allowed. There are one or two supplementary observations to be made.

In the first place, at the hearing before me a copy of an examiner's report on the specification of no. 6971 was tendered by the applicant, and counsel for the commissioner objected that it was inadmissible by reason of s. 55 of the Act. As the copy had been forwarded by the commissioner to the applicant in pursuance of s. 161, I ruled that it should be omitted in this particular proceeding. I have not, however, regarded the document as relevant. The only passage in it on which any reliance was placed seems to me to be of dubious import, and in any case this appeal must, I think, turn on what was in fact done and not on what the commissioner or one of his examiners thought.

In the second place, I have used such expressions as "the formula for Largactil" and "Largactil by Reaction B". It should be understood that I have used these expressions simply for the sake of brevity, as they were used in the argument before me. They may not be entirely accurate, but I do not think that they can lead to any confusion.

In the third place, counsel for the commissioner submitted that the costs of the commissioner should be paid by the applicant in any event. I think that the costs are entirely in my discretion under s. 149 (g) and I think, that the Commissioner of Patents, on such an appeal as this, is in the same position as the Commissioner of Taxation on an appeal to this Court against an assessment of income tax or estate duty. The appellant, having succeeded, should have an order for costs.