A-483-01

( T-913-95 )

2002 FCA 158

Canwell Enviro-Industries Ltd., Clive Titley and The City of Medicine Hat (Appellants) (Defendants)

v.

Baker Petrolite Corporation, Petrolite Holdings Inc., Baker Hughes Canada Company (Respondents) (Plaintiffs)

Indexed as: Baker Petrolite Corp. v. Canwell Enviro-Industries Ltd. (C.A.)

Court of Appeal, Strayer, Rothstein and Evans JJ.A.-- Toronto, March 20, 21; Ottawa, April 29, 2002.

Patents -- Anticipation -- Patent covering methods for removing hydrogen sulphide from natural gas -- Alleged anticipation based on prior sale of product -- 1996 Patent Act applicable retroactively according to s. 78.4(2) -- Under s. 28.2(1)(a), subject-matter defined by claims in patent must not have been disclosed more than one year before filing date -- Case law on anticipation reviewed -- Disclosure contemplated by s. 28.2(1)(a) proved -- Patent invalid by reason of anticipation.

Administrative Law -- Judicial Review -- Trial Judge making findings respecting anticipation of patent by prior use, sale -- Finding as to what constitutes invention question of law -- Standard of review on question of law correctness -- Trial Judge's conclusion appellant not proving anticipation mixed question of fact, law -- Trial Judge erred in law in reconstructing mixture W-3053, using test for anticipation -- Whether person skilled in art, using data, techniques available at relevant time could discover subject-matter of patent claims without use, application of inventive skill -- Application of correct legal test leading to conclusion such person would have been led inevitably to subject-matter of patent claims.

This was an appeal from a Trial Division decision finding that patent 2005946 (the 946 patent) of the respondents (Petrolite) had been infringed by the appellants Canwell Enviro-Industries Ltd. (Canwell) and the City of Medicine Hat. The 946 patent is a process patent covering methods for removing hydrogen sulphide from natural gas by sweetening or scavenging sour gas. The objective of the invention under the 946 patent was to reduce or eliminate undesirable side effects by use of a reaction product of an alkanolamine with an aldehyde. The triazine covered by the 946 patent had been known as a reaction product of monoethanolamine (MEA) and formaldehyde. Sales and deliveries of the MEA and formaldehyde mixture (now called W-3053) were made to Standard Oil and other customers in western Oklahoma in December 1987 and subsequent months. The sales were unconditional in that the product was not subject to any confidentiality agreement. The 946 patent application was filed in Canada on December 19, 1989. Canwell argued that the invention covered by Petrolite's 946 patent had been disclosed to the public with the sale of W-3053 commencing on December 10, 1987, more than one year before the filing date of the 946 patent. It also said that accurate reverse engineering of W-3053 would lead a person skilled in the art to the invention claimed in the 946 patent, that is, triazine, as a method to selectively reduce hydrogen sulphide in natural gas. Petrolite replied that the critical date was December 23, 1987, one year before the filing of the U.S. patent and that any reverse engineering would be kept confidential by anyone performing it and it would not be made public. The major issue on appeal was whether the 946 patent was invalid by reason of anticipation.

Held, the appeal should be allowed.

First, the Court had to determine which statute was applicable, the Patent Act brought into force on October 1, 1989 (the 1989 Act) or the Patent Act brought into force on October 1, 1996 (the 1996 Act). Under the latter, the critical date is December 19, 1989, the date on which the 946 patent application was filed in Canada. The Trial Judge correctly found that the 1996 Act was applicable to the facts of this case. Under subsection 78.4(2), any matter arising in respect of a patent issued on the basis of an application filed on or after October 1, 1989, but before October 1, 1996 is to be dealt with in accordance with the 1996 Act. The patent application was filed during the period referred to in subsection 78.4(2). The validity of the 946 patent must be dealt with and disposed of in accordance with the provisions of the 1996 Act.

According to paragraph 28.2(1)(a) of the 1996 Act, the subject-matter defined by the claims in the patent must not have been disclosed prior to December 19, 1988. The test for anticipation by publication set out by the Federal Court of Appeal in Beloit Can. Ltée/Ltd. v. Valmet Oy was cited with approval by the Supreme Court of Canada in Free World Trust v. Électro Santé Inc. The anticipation herein arose, not from prior publications or patents, but from the prior sale of Petrolite's W-3053 product. At a broad level, the principles relating to anticipation by prior publication are also applicable to anticipation by prior use or sale. Under paragraph 28.2(1)(a), it is necessary for the Court to have regard to the circumstances of prior use or sale in order to determine how a person skilled in the art might be led, without error, to the invention claimed. A number of principles were derived from United Kingdom and European decisions with respect to anticipation by prior use or sale. Those principles supplement, but do not alter the principles relating to anticipation as set out in Beloit or Free World Trust.

The Trial Judge made four findings respecting anticipation by prior use or sale. The first and second findings were purely factual. The appellate court may only intervene in conclusions of fact, including inferences of fact, in cases of palpable and overriding error. No palpable overriding error was demonstrated. The third finding, as to what constitutes the invention, involved construing the patent claims and that was a question of law. The standard of review on a question of law is correctness. As the Trial Judge did not construe the patent claims to identify the invention, the Court of Appeal did so. The invention in the claims at issue in the 946 patent was the use of a composition comprising the reaction product of a lower alkanolamine with a lower aldehyde or triazine in contact with natural gas streams to reduce hydrogen sulphide levels in the said streams. There was nothing inventive about the methods shown to bring the reaction product into contact with the sour gas streams. The fourth finding, one of mixed law and fact, involved applying a legal standard to a set of facts. Where on a question of mixed fact and law, it is possible to extricate the legal question from the factual one and determine that a legal error has been made, the standard of review will be correctness. Mischaracterizing the proper legal test results in the application of the correctness standard to the factual conclusions reached by the Trial Judge. The Trial Judge concluded that, based on all the evidence regarding attempts to successfully analyse and reconstruct W-3053, Canwell had not proven anticipation. The summary way in which he reached this conclusion respecting anticipation precluded the Court of Appeal from understanding his reasoning process. The Trial Judge made no findings of fact with respect to the expert evidence. The test for anticipation used by the Trial Judge consisted of analysing W-3053 to determine its composition or to reconstruct it. He erred in law in using this test. A defendant relying on the defence of anticipation by prior use or sale of a product does not have to prove that the product can be reproduced. A defendant relying on paragraph 28.2(1)(a) of the Patent Act need only prove that the subject-matter of the patent, the invention, has been disclosed. If triazine or its starting components were capable of being detected in the W-3053, the invention was disclosed by the sale to the public of W-3053 and the 946 patent was anticipated. The question was whether a person skilled in the art could have discovered triazine through analysis. The legal question, whether proof of anticipation required reproduction of the prior used or sold product or discovery of its composition on the one hand, or only discovery of the invention on the other, was extricable from the mixed fact and law conclusion of the Trial Judge that, based on all the evidence, Canwell did not prove anticipation. Such question is general in nature and not particular to the circumstances of this case. It is important that, in cases where anticipation by prior use or sale of a product is in issue, the evidence demonstrates that the analysis of the product discloses the invention and not necessarily an ability to reproduce the product.

The Trial Judge's conclusion that the evidence was ambivalent and totally unsatisfactory, and that Canwell had not met its burden, was premised on the wrong legal standard. Had the Trial Judge made findings of fact with respect to the expert evidence untainted by his error of law, such findings would have been reviewable only on a palpable and overriding error standard. At this stage, it was necessary that the evidence be re-examined in light of the application of the correct test for anticipation. The most important evidence was the analysis performed by Professor Keay. Based upon a blind test, the Keay analysis produced the correct answer, that is, that the major component of W-3053 was triazine. The conclusion reached by Dr. Keay provided evidence that a person skilled in the art, without the benefit of a priori knowledge and without taking inventive steps, could have successfully analysed W-3053 for the presence of triazine, for the MEA and formaldehyde starting compounds and for methanol. All analyses that were correctly performed led to the discovery of triazine or its starting materials. They were all remarkably consistent in leading the analysts to the invention. That left uncontradicted the Keay analysis, which was valid evidence of anticipation. To prove anticipation under paragraph 28.2(1)(a), it is not necessary to demonstrate that a particular purchaser did or would have conducted an analysis of the product; it is sufficient that the purchaser could have done so. The unconditional sale of the W-3053 made the product available to the public. Had he applied the correct legal test, the Trial Judge should have concluded that a person skilled in the art and using data and techniques available at the relevant time, and without the exercise of inventive skill, would have been led inevitably to the subject-matter of the patent claims, namely, the extraction of hydrogen sulphide from natural gas by contact with triazine. Canwell proved the disclosure contemplated by paragraph 28.2(1)(a). The claims in the 946 patent were invalid by reason of anticipation.

statutes and regulations judicially

considered

Convention on the Grant of European Patents (European Patent Convention), October 5, 1973, Art. 54.

Interpretation Act, R.S.C., 1985, c. I-21, ss. 3(1), 43(a),(b),(c).

North American Free Trade Agreement Between the Government of Canada, the Government of the United Mexican States and the Government of the United States of America, December 17, 1992, [1994] Can. T.S. No. 2, Art. 1709.

North American Free Trade Agreement Implementation Act, S.C. 1993, c. 44, s. 10.

Patent Act, R.S.C., 1985, c. P-4, ss. 27(1)(c),(d) (as am. by R.S.C., 1985 (3rd Supp.), c. 33, s. 8), 28.2 (as enacted by S.C. 1993, c. 15, s. 33), 78.4 (as enacted by S.C. 1993, c. 15, s. 55).

Patents Act 1977 (U.K.), 1977, c. 37, s. 2.

cases judicially considered

applied:

Free World Trust v. Électro Santé Inc., [2000] 2 S.C.R. 1024; (2000), 194 D.L.R. (4th) 232; 9 C.P.R. (4th) 168; 263 N.R. 150; Beloit Can. Ltée/Ltd. v. Valmet Oy (1986), 7 C.I.P.R. 205; 8 C.P.R. (3d) 289; 64 N.R. 287 (F.C.A.); Housen v. Nikolaisen (2002), 211 D.L.R. (4th) 577; [2002] 7 W.W.R. 1; 10 C.C.L.T. (3d) 157; 30 M.P.L.R. (3d) 1; 286 N.R. 1; 219 Sask. R. 1 (S.C.C.); Hollis v. Dow Corning Corp., [1995] 4 S.C.R. 634; (1995), 129 D.L.R. (4th) 609; [1996] 2 W.W.R. 77; 14 B.C.L.R. (3d) 1; 26 B.L.R. (2d) 169; 27 C.C.L.T. (2d) 1; 190 N.R. 241.

considered:

Slattery (Trustee of) v. Slattery, [1993] 3 S.C.R. 430; (1993), 106 D.L.R. (4th) 212; 139 N.B.R. (2d) 246; 21 C.B.R. (3d) 161; [1993] 2 C.T.C. 243; 93 DTC 5443; 158 N.R. 324; Diversified Products Corp. v. Tye-Sil Corp. (1991), 35 C.P.R. (3d) 350; 125 N.R. 218 (F.C.A.); Merrell Dow Pharmaceuticals Inc. and Anr. v. H. N. Norton & Co. Ltd., [1996] R.P.C. 76 (H.L.); Lux Traffic Controls Ltd. v. Pike Signals Ltd. and Faronwise Ltd., [1993] R.P.C. 107; FISONS plc v. Packard Instrument B.V., EPO,T 0952/92--3.4.1, August 17, 1994; Bristol-Myers Co.'s Application, [1969] R.P.C. 146 (Q.B.D.).

referred to:

Nowegijick v. The Queen, [1983] 1 S.C.R. 29; (1983), 144 D.L.R. (3d) 193; [1983] 2 C.N.L.R. 89; [1983] CTC 20; 83 DTC 5041; 46 N.R. 41; Sarvanis v. Canada (2002), 210 D.L.R. (4th) 262; 284 N.R. 263 (S.C.C.); Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 S.C.R. 817; (1999), 174 D.L.R. (4th) 193; 14 Admin. L.R. (3d) 173; 1 Imm. L.R. (3d) 1; 243 N.R. 22; PLG Research Ltd. v. Ardon International Ltd., [1993] F.S.R. 197; Whirlpool Corp. v. Camco Inc., [2000] 2 S.C.R. 1067; (2000), 194 D.L.R. (4th) 193; 9 C.P.R. (4th) 129; 263 N.R. 88.

authors cited

Côté, Pierre-André. Interpretation of Legislation in Canada, 3rd ed. Scarborough, Ont.: Carswell, 2000.

Sullivan, Ruth. Driedger on the Construction of Statutes, 3rd ed. Toronto: Butterworths, 1994.

APPEAL from a Trial Division decision ([2002] 2 F.C. 3; (2001), 13 C.P.R. (4th) 193) that the respon-dents' process patent had been infringed. Appeal allowed on the ground that the patent was invalid by reason of anticipation.

appearances:

David W. Aitken for appellants (defendants).

Anthony G. Creber and Patrick S. Smith for respondents (plaintiffs).

solicitors of record:

Osler, Hoskin & Harcourt LLP, Ottawa, for appellants (defendants).

Gowling Lafleur Henderson LLP, Ottawa, for respondents (plaintiffs).

The following are the reasons for judgment rendered in English by

Rothstein J.A.:

INTRODUCTION

[1]This is an appeal from a decision of the Trial Division [[2002] 2 F.C. 3]. The Trial Division Judge found that patent 2005946 (the 946 patent) of the plaintiffs (Petrolite) had been infringed by the defendants Canwell Enviro-Industries Ltd. (Canwell) and the City of Medicine Hat (Medicine Hat). The major issue on appeal, and the basis for this decision, is whether the 946 patent is invalid by reason of anticipation. If the invention was disclosed in a manner and at a time provided in the relevant statutory provision, the invention will have been anticipated and the 946 patent will be invalid.

DECISION OF THE TRIAL JUDGE

[2]The Trial Judge's conclusions can be summarized as follows:

1. The 946 patent was not invalid by reason of anticipation, obviousness, insufficiency of disclosure or material misrepresentation in the patent.

2. Canwell and Medicine Hat infringed the 946 patent.

3. Canwell induced customers to infringe the 946 patent.

[3]The relief granted to Petrolite by the Trial Judge included:

1. Injunctions against Canwell and Medicine Hat from further infringing the 946 patent.

2. An order for delivery up of infringing products, apparatus and documents.

3. Judgment against Canwell and Medicine Hat for the profits flowing from infringement of the 946 patent.

4. Judgment against Clive Titley, president and sole shareholder of Canwell, as constructive trustee of Canwell's profits he received in the form of bonuses, subject to an inquiry to trace what assets in the possession of Titley represented the bonuses he received.

5. Interest.

6. Costs.

THE 946 PATENT

[4]The 946 patent is a process patent. It covers methods for removing hydrogen sulphide from natural gas. This is sometimes called sweetening or scavenging sour gas. Because hydrogen sulphide is corrosive in the presence of water and poisonous in very small concentrations, it must be removed almost completely from natural gas streams before use and before transport in a pipeline.

[5]Several methods for sweetening sour gas had been known. They involved bringing various chemicals into contact with natural gas at the wellhead or in a scrubber tower or absorber tower, or in other locations near the origin of the natural gas. However, use of these chemicals had undesirable side effects, e.g. health hazards, non-selective reaction resulting in removal of more than hydrogen sulphide, the production of precipitates that clog pipelines or poor performance at lower temperatures.

[6]The objective of the invention under the 946 patent is to reduce or eliminate these side effects by use of a reaction product of an alkanolamine with an aldehyde. The reaction product would be brought into contact with natural gas in a scrubber system, as a chemical solvent system, by injection into a pipeline or by other methods.

[7]The preferred embodiment of the invention is indicated to be monoethanolamine (MEA) as the alkanolamine and formaldehyde as the aldehyde. While the 946 patent claims cover a range of molar ratios of the alkanolamine and the aldehyde, from about 1:0.25 to 1:10, the preferable molar ratio is indicated to be from about 1:1 to about 1:1.5.

[8]The 946 patent states that the reaction product of MEA and formaldehyde is bisoxazolidine and triazine. However, the Trial Judge found that the expert testimony was unanimous that bisoxazolidine had never been conclusively identified as a reaction product of MEA and formaldehyde, and was equally unanimous that the predominant reaction product of MEA and formaldehyde was triazine.

[9]The triazine covered by the 946 patent had been known as a reaction product of MEA and formaldehyde. Accordingly, there is no product claim in the patent. The claims are limited to the reaction product used to selectively reduce the levels of hydrogen sulphide in natural gas.

FACTS

[10]The sequence of relevant events and facts, as found by the Trial Judge, can be summarized as follows:

1987

1.  August or September -- Quaker Petroleum Chemical Company began marketing, in western Oklahoma, a hydrogen sulphide scavenging product that was a formaldehyde solution. It was temperature-sensitive so that as the weather became colder, difficulties were encountered.

2. November -- Quaker experimented with mixing MEA and formaldehyde. The resulting product solved the temperature-sensitivity problem and showed a "pretty good" affinity for hydrogen sulphide removal.

3.  December 10 -- a 500-gallon batch of the MEA and formaldehyde mixture was delivered by Quaker to Standard Oil of Ohio in western Oklahoma.

4.  December and subsequent months -- sales and deliveries of the mixture, now called W-3053, were made to Standard Oil and other customers in western Oklahoma. The molar ratio of MEA and formaldehyde in W-3053 was 1:2.77. The sales were unconditional in that the product was not subject to any confidentiality agreement.

1988

5. December 23 -- Patent Application No. 4978512 was filed in the United States.

1989

6. December 19 -- The 946 patent application was filed in Canada.

1991

7. Canwell began selling Cansweet 300SX and, subsequently, CW1000. Both contain a reaction product of MEA and formaldehyde. In 300SX, the molar ratio is 1:1. In CW1000, it is 1:2.66.

1995

8. February 7 -- the 946 patent issued in Canada.

9. May 2 -- statement of claim filed giving rise to the action in the Trial Division.

2001

10. August 15--Trial Division judgment issued.

11. August 24--Interim stay granted.

12. October 2--Partial stay pending appeal granted on conditions and order for expedited hearing made.

[11]As often occurs, there have been corporate changes and assignments affecting the ownership of the 946 patent. There is no controversy about this and no need to address those details. It is sufficient to say that Petrolite and the other respondents were proper plaintiffs in the Trial Division seeking relief for infringement of the 946 patent.

SUMMARY OF THE PARTIES' POSITIONS ON ANTICIPATION BY SALE OR USE OF W-3053

[12]Canwell says that the invention covered by Petrolite's 946 patent had been disclosed to the public with the sale of W-3053 commencing on December 10, 1987, and continuing thereafter. This was more than one year before the filing date of the 946 patent, namely December 19, 1989. Canwell says that accurate reverse engineering of W-3053 would lead a person skilled in the art to the invention claimed in the 946 patent, that is, triazine, as a method to selectively reduce hydrogen sulphide in natural gas. Therefore, the invention of the 946 patent was anticipated and the patent is invalid.

[13]Petrolite concedes that W-3053 was sold in western Oklahoma commencing on December 10, 1987. However, it says that the critical date was December 23, 1987, one year before the filing of the U.S. patent. It also says that the few sales made between December 10 and December 23 were not commercial sales and were to customers on their private property and, therefore, W-3053 was not available to the public. Further, it submits that any reverse engineering would be kept confidential by anyone performing it and it would not be made public.

[14]Although Petrolite does not rule out reverse engineering as a means of proving anticipation, it says that the evidence here was insufficient to meet the stringent test for anticipation in Beloit Can. Ltée/Ltd. v. Valmet Oy (1986), 7 C.I.P.R. 205 (F.C.A.), at page 214. Petrolite submits that the test is that a prior publication or, in this case, reverse engineering of a prior sold product must lead the person skilled in the art in every case and without the possibility of error, to the claimed invention. The evidence here does not meet the test. In any event, Petrolite says that there is no evidence that the methodology and data for accurate reverse engineering were available during the period from December 10 to December 23, 1987 or, indeed, from December 10, 1987, to December 19, 1988, if Canwell's critical date is accepted.

WHICH STATUTE APPLIES?

[15]The issue here is whether the Patent Act, R.S.C., 1985, c. P-4, as amended by R.S.C., 1985 (3rd Supp.), c. 33, brought into force on October 1, 1989 (the 1989 Act), or the Patent Act as amended by S.C. 1993, c. 15, brought into force on October 1, 1996 (the 1996 Act), is applicable. Petrolite argues that the 1989 Act is applicable. It says that under that Act, the critical date is the date on which the application for the United States patent was filed, December 23, 1988. The only time during which there could be anticipation by prior use or sale is between December 10, 1987, when W-3053 was first sold in western Oklahoma, and December 23, 1987, one year before the filing date of the United States patent. Petrolite argues that during this time the sales were experimental and, in any event, there was no evidence that the methodology and data were available to permit accurate reverse engineering to determine the contents of W-3053.

[16]Canwell says the 1996 Act is applicable. Under that Act, the critical date is December 19, 1989, the date on which the 946 patent application was filed in Canada. In Canwell's submission, the relevant period for anticipation by prior sale or use of the W-3053 product was December 10, 1987, to December 19, 1988, during which period sales were not experimental and the methodology and data for accurate reverse engineering were available.

[17]I have some difficulty accepting Petrolite's interpretation of the 1989 Act. However, it is not necessary to deal with that argument because, in my view, the 1996 Act is applicable. The 1996 Act contains transitional provisions including section 78.4 [as enacted by S.C. 1993, c. 15, s. 55]. Under subsection 78.4(1), applications for patents filed on or after October 1, 1989, are to be dealt with and disposed of in accordance with the 1996 Act. Under subsection 78.4(2), any matter arising in respect of a patent issued on the basis of an application filed on or after October 1, 1989, but before October 1, 1996, is to be dealt with in accordance with the 1996 Act. Section 78.4 of the 1996 Act provides:

78.4 (1) Applications for patents in Canada filed on or after October 1, 1989, but before this section came into force, shall be dealt with and disposed of in accordance with

(a) subsection 27(2) as it read immediately before this section came into force; and

(b) the provisions of this Act, including subsection 27(2), as they read immediately after this section came into force.

(2) Any matter arising in respect of a patent issued on the basis of an application filed on or after October 1, 1989, but before this section came into force, shall be dealt with and disposed of in accordance with

(a) subsection 27(2) as it read immediately before this section came into force; and

(b) the provisions of this Act, including subsection 27(2), as they read after this section came into force and as amended from time to time. [Restructured pursuant to S.C. 2001, c. 10, s. 4., but without substantive change.]

[18]The learned Trial Judge found that the 1996 Act was applicable to the facts of this case and I agree with him. The 946 patent application was filed on December 19, 1989, during the period referred to in subsection 78.4(2).

[19]Section 78.4 is a provision with retroactive effect. While there is a presumption against the retroactive application of legislation, the presumption can be rebutted by express words or by necessary implication. Such legislation is described by Sullivan, R., in Driedger on the Construction of Statutes, 3rd ed. (Toronto: Butterworths, 1994), at page 522:

The presumption against the retroactive application of the legislation can be rebutted by express words or by necessary implication. All that is required is some sufficient indication that the legislation is meant to apply not only to on-going and future facts but also to facts that are past. . . .

Retroactive legislation often states that it is deemed to come into force or to take effect on a date prior to the date of enactment. Or it may state that it applies to designated facts occurring from or before a particular date or time.

See also Côté, Interpretation of Legislation in Canada, 3rd ed., (Scarborough: Carswell, 2000), at page 150. Section 78.4 provides that the 1996 Act is to apply to matters occurring before a particular date and, specifically, before the coming into force of the 1996 Act. Accordingly, the 1996 Act has retroactive application.

[20]Here, the patent application was filed on December 19, 1989. Being filed after October 1, 1989 but before October 1, 1996, it is to be dealt with and disposed of in accordance with the 1996 Act. Likewise, any matter arising in respect of that patent is to be dealt with in accordance with the 1996 Act.

[21]While this conclusion seems straightforward, Petrolite makes a number of arguments to the contrary. First, it is argued that "any matter arising" must speak as of the date of the coming into force of the 1996 Act. It would apply, for example, to infringement occurring on or after October 1, 1996. However, it is said that in granting the 946 patent on February 7, 1995, the Patent Commissioner would not have had before him the 1996 Act. Since the question of validity goes to the appropriateness of the granting of the patent, questions of validity must be decided under the 1989 Act.

[22]I see no basis for this selective interpretation of section 78.4. The terms "any matter arising" and "in respect of" are words of the widest scope. In Slattery (Trustee of) v. Slattery, [1993] 3 S.C.R. 430, Iacobucci J., for the majority, refers to the words "in respect of" as they were previously considered by the Supreme Court in Nowegijick v. The Queen, [1983] 1 S.C.R. 29. He states at page 445:

The phrase "in respect of" was considered by this Court in Nowegijick v. The Queen, [1983] 1 S.C.R. 29, at p. 39:

The words "in respect of" are, in my opinion, words of the widest possible scope. They import such meanings as "in relation to", "with reference to" or "in connection with". The phrase "in respect of" is probably the widest of any expression intended to convey some connection between two related subject matters. [Emphasis added by Iacobucci J.]

The words "in respect of" as modified by the comprehensive "any matter arising" could not be more expansive. Parliament obviously contemplated that section 78.4 would be given wide scope. Nothing in the context of the legislation suggests a more restrictive interpretation (see Sarvanis v. Canada (2002), 210 D.L.R. (4th) 262 (S.C.C.), at paragraph 24). Petrolite's argument simply ignores the plain words of section 78.4.

[23]Petrolite relies on paragraphs 43(a), (b) and (c) of the Interpretation Act, R.S.C., 1985, c. I-21 which it says preclude the retroactive application of the 1996 Act:

43. Where an enactment is repealed in whole or in part, the repeal does not

(a) revive any enactment or anything not in force or existing at the time when the repeal takes effect,

(b) affect the previous operation of the enactment so repealed or anything duly done or suffered thereunder,

(c) affect any right, privilege, obligation or liability acquired, accrued, accruing or incurred under the enactment so repealed,    

However, Petrolite has had to acknowledge that these provisions are presumptions only and may be rebutted by express language in a statute, or by necessary implication. Subsection 3(1) of the Interpretation Act provides:

3. (1) Every provision of this Act applies, unless a contrary intention appears, to every enactment, whether enacted before or after the commencement of this Act. [Emphasis added.]

Section 78.4 expresses a contrary intention in so far as the 1996 Act is concerned and thus rebut the presumption in section 43.

[24]Petrolite also argues that section 78.4 of the 1996 Act must be read "in congruence with Article 1709(8)(a)" of the NAFTA [North American Free Trade Agreement Between the Government of Canada, the Government of the United Mexican States and the Government of the United States of America, December 17, 1992, [1994] Can. T.S. No. 2], and that doing so precludes retroactive application of the anticipation provisions of the 1996 Act. Article 1709(8)(a) provides:

Article 1709 . . .

8. A Party may revoke a patent only when:

(a) grounds exist that would have justified a refusal to grant the patent;

[25]I do not accept this argument for two reasons. First, Article 1709(8) is a provision of the NAFTA. The NAFTA has been approved by North American Free Trade Agreement Implementation Act, S.C. 1993, c. 44, s. 10. However, this does not give the provisions of the NAFTA themselves the force of an Act of Parliament. I accept that an international treaty may, where relevant, be used to assist in interpreting domestic legislation. See, for example, Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 S.C.R. 817, at paragraphs 69 and 70. However, the international treaty cannot be used to override the clear words used in a statute enacted by Parliament. Section 78.4 is plain and obvious. Petrolite, I think, is relying on Article 1709(8) of the NAFTA to give a restricted meaning to section 78.4 which its words cannot bear.

[26]Second, I do not read the words of Article 1709(8) as precluding express retroactive legislation. It is true that the article provides that a patent may be revoked only when grounds exist that would have justified the refusal to grant the patent in the first place. However, the effect of the retroactive application of legislation is to make its provisions applicable as if they had existed at a prior time.

[27]It is true, as Petrolite argues, that some arbitrariness or anomalies may arise through the application of retroactive legislation. However, arbitrariness and anomalies are inherent in retroactive legislation. That is one reason why there is a presumption against interpreting legislation to have retroactive effect. However, the Court must take the statute as it finds it and interpret section 78.4 according to its words. The validity of the 946 patent is a matter that has arisen in respect of that patent. The patent application was filed in the period referred to in section 78.4. The validity of the 946 patent must be dealt with and disposed of in accordance with the provisions of the 1996 Act.

THE LAW RELATING TO ANTICIPATION

The Relevant Statutory Provision

[28]The application for the 946 patent was filed December 19, 1989. According to paragraph 28.2(1)(a) [as enacted by section 33] of the 1996 Act, the subject- matter defined by the claims in the patent must not have been disclosed prior to December 19, 1988. Paragraph 28.2(1)(a) provides:

28.2 (1) The subject-matter defined by a claim in an application for a patent in Canada (the "pending application") must not have been disclosed

(a) more than one year before the filing date by the applicant, or by a person who obtained knowledge, directly or indirectly, from the applicant, in such a manner that the subject-matter became available to the public in Canada or elsewhere;

Anticipation by Publication

[29]In Beloit, supra, Hugessen J.A. (as he then was) set out the test for anticipation by publication, at page 214:

It will be recalled that anticipation, or lack of novelty, asserts that the invention has been made known to the public prior to the relevant time. The inquiry is directed to the very invention in suit and not, as in the case of obviousness, to the state of the art and to common general knowledge. Also, as appears from the passage of the statute quoted above, anticipation must be found in a specific patent or other published document; it is not enough to pick bits and pieces from a variety of prior publications and to meld them together so as to come up with the claimed invention. One must, in effect, be able to look at a prior, single publication and find in it all the information which, for practical purposes, is needed to produce the claimed invention without the exercise of any inventive skill. The prior publication must contain so clear a direction that a skilled person reading and following it would in every case and without the possibility of error be led to the claimed invention.

The test for anticipation by publication in Beloit, supra, was cited with approval by the Supreme Court of Canada in Free World Trust v. Électro Santé Inc., [2000] 2 S.C.R. 1024, at paragraph 26. In Free World Trust, supra, Binnie J. explains, at paragraph 25, that anticipation by publication is a difficult test to meet because, after an invention has been disclosed, it is all too easy to find its antecedents in bits and pieces of earlier learning:

Anticipation by publication is a difficult defence to establish because courts recognize that it is all too easy after an invention has been disclosed to find its antecedents in bits and pieces of earlier learning. It takes little ingenuity to assemble a dossier of prior art with the benefit of 20-20 hindsight.

Anticipation by Prior Use or Sale

[30]The anticipation in the appeal at bar is about anticipation arising, not from prior publications or patents, but from the prior sale of Petrolite's W-3053 product. Anticipation by prior patents is also argued by Canwell, but it will not be necessary to deal with that argument.

[31]The Patent Act in force in 1986, when Beloit, supra, was decided, provided that anticipation could be proven by the public use or sale of the invention in Canada more than two years prior to the filing of a patent application. Before October 1, 1989, paragraph 27(1)(c) provided:

27. (1) Subject to this section, any inventor or legal representative of an inventor of an invention that was

. . .

(c) not in public use or on sale in Canada for more than two years prior to his application in Canada

may, on presentation to the Commissioner of a petition setting out the facts, in this Act termed the filing of the application, and on compliance with all other requirements of this Act, obtain a patent granting to him an exclusive property in the invention.

Under paragraph 27(1)(c), it appears that public use or sale of the invention was sufficient to prove anticipation even if the invention was not thereby disclosed; that is, even if it was not possible to determine the claimed invention, the invention would be anticipated simply by its use or sale. It was, therefore, not necessary, at that time, to prove more than use or sale of the invention in the relevant time period in order to prove anticipation.

[32]As a result of amendments made effective on October 1, 1989, and continued in the 1996 Act, evidence of use or sale of the invention was no longer sufficient in itself to prove anticipation. With the enactment of paragraph 27(1)(d) on October 1, 1989 [R.S.C., 1985 (3rd Supp.), c. 33, s. 8], which was replaced by paragraph 28.2(1)(a) on October 1, 1996 [S.C. 1996, c. 15, s. 33], the test for anticipation by any means became disclosure of "the subject-matter defined by a claim"(the invention) "in such a manner that the subject-matter became available to the public in Canada or elsewhere". The change to the legislation, inter alia:

1. reduced the relevant grace period prior to the filing of the patent application;

2. made disclosure anywhere, and not just in Canada, relevant;

3. eliminated use or sale of the invention per se as sufficient evidence of anticipation;

4. as proof of anticipation, required disclosure in such a manner that the invention became available to the public in Canada or elsewhere.

There is little Canadian jurisprudence interpreting paragraph 28.2(1)(a).

[33]At trial, Canwell referred to United Kingdom case law on the subject of anticipation by prior use or sale. The learned Trial Judge refused to engage in an analysis of that law. He was of the opinion that, although specifically referable to anticipation by prior publication, the principles in Beloit, supra, and Free World Trust, supra, were equally applicable to, and exhaustive of, anticipation by prior use or sale. In support of his view, he referred to the judgment of this Court in Diversified Products Corp. v. Tye-Sil Corp. (1991), 35 C.P.R. (3d) 350 (F.C.A.), at pages 360-361:

As was mentioned by Urie J. in Beecham . . . the defences of prior knowledge, prior use, prior publication and prior sale are "very much intermingled" and are referred to as "anticipation". I have noted throughout the cases that there does not appear to be any distinction in principle between these various defences and that what is said with respect, for example, to anticipation through prior knowledge is applicable, mutatis mutandis, to anticipation through prior publication. [Citation omitted; emphasis added in reasons of Trial Judge.]

[34]I agree with the learned Trial Judge that the United Kingdom authorities are not of assistance to the extent that they are contrary to the principles relating to anticipation articulated in Beloit, supra, and Free World Trust, supra. This Court is bound by Free World Trust and, to the extent it approved of Beloit, by Beloit. However, the U.K. authorities are useful to give content to the analysis of anticipation by prior use or sale and, to the extent they do not contradict the Canadian jurisprudence, guidance may be sought from such decisions for that purpose.

[35]I accept that, at a broad level, the principles in Beloit, supra, and Free World Trust, supra, relating to anticipation by prior publication are also applicable to anticipation by prior use or sale. For example, the evidence of anticipation by prior public use or sale, as well as by prior publication, should be subjected to close scrutiny. However, below a certain level of generality, the principles governing anticipation by prior publication may need to be tailored to fit the particular characteristics of anticipation by prior public use or sale. For example, the principle that the prior publication must contain so clear a direction that a skilled person reading and following it would be led, without error, to the invention claimed, applies to the specific context of prior publication. In the case of prior publication, the skilled person will read the publication. In the case of prior use or sale, reading may not be relevant. When faced with having to decide whether there has been anticipation by disclosure through prior use or sale under paragraph 28.2(1)(a), it is necessary for the Court to have regard to the circumstances of prior use or sale, in order to determine how a person skilled in the art might be led, without error, to the invention claimed. For example, was there an analytical method available at the relevant time to lead a skilled person to the invention? The United Kingdom authorities provide useful guidance in this respect.

[36]The United Kingdom authorities are also useful for another purpose. As indicated, although there is little Canadian jurisprudence interpreting paragraph 28.2(1)(a) of the 1996 Act, there is considerable United Kingdom jurisprudence interpreting provisions of the Patents Act 1977 (U.K.), 1977, c. 37, that are, in their substance, similar to paragraph 28.2(1)(a). Courts in the United Kingdom have had occasion to deal with cases of anticipation by prior use or sale under section 2 of the United Kingdom Act and those authorities will be instructive in assisting in the interpretation of paragraph 28.2(1)(a) of the Canadian Act.

Principles Derived from United Kingdom and European Authorities with respect to Anticipation by Prior Use or Sale

[37]I turn then to the United Kingdom authorities. I shall also have regard to decisions made under the European Patent Convention [Convention on the Grant of European Patents (European Patent Convention), October 5, 1973] because it contains provisions similar to the United Kingdom Patents Act 1977. Courts in the United Kingdom appear to be guided by decisions under the European Patent Convention and I see no reason why, where this Court finds the United Kingdom authorities of assistance and the European legislation is similar, it may not also seek guidance from the European decisions.

[38]The United Kingdom Patents Act 1977 contains provisions that are similar in effect to paragraph 28.2(1)(a) of the 1996 Act. Subsection 2(1) of the United Kingdom Patents Act 1977 provides that an invention is to be taken as novel if it does not form part of the state of the art. Subsection 2(2) defines the state of the art as comprising all matter which has been made available to the public in any way:

2.--(1) An invention shall be taken to be new if it does not form part of the state of the art.

(2) The state of the art in the case of an invention shall be taken to comprise all matter (whether a product, a process, information about either, or anything else) which has at any time before the priority date of that invention been made available to the public (whether in the United Kingdom or elsewhere) by written or oral description, by use or in any other way. [Emphasis added.]     

The European Patent Convention contains similar provisions.

Article 54

(1) An invention shall be considered to be new if it does not form part of the state of the art.

(2) The state of the art shall be held to comprise everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application. [Emphasis added.]     

[39]The words employed in the United Kingdom and European legislation are similar, in effect, to the words in paragraph 28.2(1)(a) of the 1996 Act:

28.2 (1) The subject-matter . . . must not have been disclosed

(a) [. . .] in such a manner that the subject-matter became available to the public in Canada or elsewhere;

[40]While conceding that the United Kingdom and European legislation is similar to the 1996 Act on this issue, Petrolite argues that a significant difference is that the former (United Kingdom and European) use the term "made available to the public" while the latter (Canadian) uses the term "became available to the public". If I understand the argument, it is that "made available" implies "on the shelf" or, in other words, that the information could be available to the public. On the other hand, "became available" implies that the public was, in fact, in possession of the information. It is argued that the difference means that reverse engineering is permitted in the United Kingdom and Europe but not under Canadian law or, at least, that if reverse engineering is permitted under Canadian law, it must, in fact, have been performed more than one year before the filing date of the patent.

[41]I am unable to agree with Petrolite's argument. Whether information is made available or became available, it is available. Whether information is considered "available" if it is only ascertainable through reverse engineering is a question for determination. But that is so under either the "made available" or "became available" terminology. I see no material distinction between the terms "made available" or "became available" in the context of this case.

[42]For the purposes of analysing anticipation in the context of disclosure by prior sale or use under paragraph 28.2(1)(a) and without detracting from the general principles in Beloit, supra, and Free World Trust, supra, I deduce the following principles relevant to this appeal from the United Kingdom and European Patent Office jurisprudence:

1. Sale to the public or use by the public alone is insufficient to prove anticipation. Disclosure of the invention is required to constitute anticipation under paragraph 28.2(1)(a). In Merrell Dow Pharmaceuticals Inc. and Anr. v. H. N. Norton & Co. Ltd., [1996] R.P.C. 76 (H.L.), Lord Hoffmann found that use of a product makes an invention part of the state of the art, i.e. equivalent to disclosure, only so far as that use makes available the information which describes the invention. At page 86, he stated:

. . . to be part of the state of the art, the invention must have been made available to the public. An invention is a piece of information. Making matter available to the public within the meaning of section 2(2) therefore requires the communication of information. The use of a product makes the invention part of the state of the art only so far as that use makes available the necessary information. [Emphasis in italics in original; other emphasis added.]

2. For a prior sale or use to anticipate an invention, it must amount to "enabling disclosure". In Merrell Dow, supra, at pages 86-87, Lord Hoffmann quoted with approval the statement of Aldous J. in PLG Research Ltd. v. Ardon International Ltd., [1993] F.S.R. 197, at page 225:

Under the 1977 Act, patents may be granted for an invention covering a product that has been put on the market provided the product does not provide an enabling disclosure of the invention claimed. In most cases, prior sale of the product will make available information as to its contents and its method of manufacture, but it is possible to imagine circumstances where that will not happen. [Emphasis added.]

In Lux Traffic Controls Ltd. v. Pike Signals Ltd. and Faronwise Ltd., [1993] R.P.C. 107, Aldous J. stated at page 133:

It is settled law that to invalidate a patent a disclosure has to be what has been called an enabling disclosure. That is to say the disclosure has to be such as to enable the public to make or obtain the invention. [Emphasis added.]

3. The prior sale or use of a chemical product will constitute enabling disclosure to the public if its composition can be discovered through analysis of the product. The Board of Appeal of the European Patent Office in FISONS plc v. Packard Instrument B.V., E.P.O. case number T 0952/92 - 3.4.1, August 17, 1994, stated at page 21:

. . . in the Board's view it is the fact that direct and unambi-guous access to information concerning the composition or internal structure of a prior used product is possible, for example by means of analysis, which makes such composition or internal structure "available to the public" and thus part of the state of the art for the purpose of Article 54(2) EPC. [Emphasis added.]

4. The analysis must be able to be performed by a person skilled in the art in accordance with known analytical techniques available at the relevant time. In FISONS, supra, the Board of Appeal stated at page 21:

If such an analysis is possible in accordance with the known analytical techniques which were available for use by a skilled person before the relevant filing date, the composition or internal structure thereby is available to the public. [Emphasis added.]

This principle, in the context of enabling disclosure arising from prior use or sale, must be applied consistently with this Court's determination respecting prior publication in Beloit, supra. The person skilled in the art, using available analytical techniques, must be able to find the invention without the exercise of inventive skill.

5. In the context of patent anticipation under paragraph 28.2(1)(a), when reverse engineering is necessary and capable of discovering the invention, an invention becomes available to the public if a product containing the invention is sold to any member of the public who is free to use it as she or he pleases. In Bristol-Myers Co.'s Application, [1969] R.P.C. 146 (Q.B.D.), Lord Parker C.J. stated at page 155:

. . . if the information . . . has been communicated to a single member of the public without inhibiting fetter that is enough to amount to a making available to the public. . . . [Emphasis added.]

In Lux, supra, Aldous J. stated at page 134:

In the present case, a light system with a prototype controller was on a number of occasions made available to contractors over five months. Those contractors were free in law and equity to examine it. [Emphasis added.]

It makes sense that sale of a product to even one member of the public constitutes a making available to the public for purposes of paragraph 28.2(1)(a). The value of the patent that is sought lies in the secrecy of its subject-matter. Providing enabling disclosure to even one member of the public destroys this secrecy. The grant of a patent depends on the inventor giving to the public something it did not have before. If the public already has it, then the inventor gives nothing and is not entitled to anything in return, i.e. a monopoly for a period of years. (See Free World Trust, supra, at paragraph 13.)

6. It is not necessary to demonstrate that a member of the public actually analysed the product that was sold. In Lux, supra, Aldous J. stated at page 133:

Further it is settled law that there is no need to prove that anybody actually saw the disclosure provided the relevant disclosure was in public. Thus an anticipating description in a book will invalidate a patent if the book is on a shelf of a library open to the public, whether or not anybody read the book and whether or not it was situated in a dark and dusty corner of the library. If the book is available to the public, then the public have the right to make and use the information in the book without hindrance from a monopoly granted by the State. [Emphasis added.]

Although the comments of Aldous J. use the example of prior publication to make the point, Lux was a prior use case and the principle derived from his example is equally applicable to prior use or sale as well as prior publication.

7. The amount of time and work involved in conducting the analysis is not determinative of whether a skilled person could discover the invention. The relevant consideration, in this respect, is only whether inventive skill was required. In FISONS plc v. Packard Instruments B.V., supra, the Board of Appeal of the European Patent Office stated at pages 21-22:

The likelihood or otherwise of such a skilled person either reading such a written description, or analysing such a prior sold product, and the degree of burden (i.e. the amount of work and time) involved in such reading or in analysing, is in principle irrelevant to the determination of what constitutes the state of the art. [Emphasis added.]

There must be some evidence from which the use of inventive skill may be inferred. Complexity or time and work involved alone are insufficient.

8. It is not necessary that the product that is the subject of the analysis be capable of exact reproduction. It is the subject-matter of the patent claims (the invention) that must be disclosed through the analysis. Novelty of the claimed invention is destroyed if there is disclosure of an embodiment which falls within the claim. In FISONS, supra, the Board held at page 23:

According to the established jurisprudence of the Boards of Appeal, the novelty of a claimed invention is destroyed by the prior disclosure (by whatever means) of an embodiment which falls within the claim. Thus in the Board's view, the novelty of a claimed invention is destroyed by the prior use of a product, for example, sale of a product, if an analysis of a product using available analytical techniques is such as to inform the skilled person of an embodiment of the product which falls within the claim of the patent. The Board therefore does not accept the patent proprietor's submission to the effect that a complete analysis of a prior used product must be possible, so as to enable an exact reproduction of such product, in order to destroy the novelty of the claimed product. [Emphasis in bold in original; other emphasis added.]

[43]The foregoing eight principles (which are not intended to be exhaustive) are particularly relevant to disclosure by prior use or sale in the context of paragraph 28.2(1)(a) in this case, although some may also be applicable to disclosure by prior patent or publication. They supplement, but do not alter the principles relating to anticipation as set out in Beloit, supra, or Free World Trust, supra.

[44]Bearing in mind the above eight principles and those in Beloit, supra, and Free World Trust, supra, I turn to the relevant findings of the Trial Judge relating to anticipation by prior use or sale of Petrolite's W-3053 product.

THE FINDINGS OF THE TRIAL JUDGE RESPECTING ANTICIPATION BY PRIOR USE OR SALE

[45]The relevant findings of the learned Trial Judge were as follows:

1. W-3053 was in unconditional sale as early as December 1987.

2. A sample of W-3053 for the purposes of analysis could readily have been obtained as early as December 1987.

3. W-3053 was, for all intents and purposes, the invention described in the 946 patent.

4. Considering all the evidence respecting attempts to successfully analyse and reconstruct W-3053, Canwell had not discharged its burden of establishing that the invention disclosed by the 946 patent had been made available to the public. Accordingly, the invention was not anticipated.

[46]The first and second of the Trial Judge's findings are purely factual. The third, what constitutes the invention, involves construing the patent claims and that is a question of law. See Whirlpool Corp. v. Camco Inc., [2000] 2 S.C.R. 1067, at paragraph 76. The fourth finding is one of mixed law and fact. It involves applying a legal standard to a set of facts--namely, did the facts establish anticipation under paragraph 28.2(1)(a).

THE STANDARD OF REVIEW

[47]The recent decision of the Supreme Court of Canada in Housen v. Nikolaisen (2002), 211 D.L.R. (4th) 577, provides a thorough analysis of the standard of review to be applied by appellate courts to decisions of trial judges. I summarize my understanding of the principles relevant to this case as follows.

[48]The standard of review on a question of law is correctness. At paragraph 9, Iacobucci and Major JJ., for the majority, state:

Thus, while the primary role of trial courts is to resolve individual disputes based on the facts before them and settled law, the primary role of appellate courts is to delineate and refine legal rules and ensure their universal application. In order to fulfill the above functions, appellate courts require a broad scope of review with respect to matters of law.

[49]The appellate court may only intervene in conclusions of fact, including inferences of fact, in cases of palpable and overriding error. Iacobucci and Major JJ. state at paragraph 25:

It is our view that the trial judge enjoys numerous advantages over appellate judges which bear on all conclusions of fact, and, even in the absence of these advantages, there are other compelling policy reasons supporting a deferential approach to inferences of fact. We conclude, therefore, by emphasizing that there is one, and only one, standard of review applicable to all factual conclusions made by the trial judge--that of palpable and overriding error.

[50]Finally, there are questions of mixed fact and law. Such questions involve the application of legal standards to a set of facts. Where there is no extricable error of law or where the facts are so particular that decisions about whether they satisfy legal tests have no precedential value, greater deference is mandated. At paragraph 28, Iacobucci and Major JJ. state:

However, where the error does not amount to an error of law, a higher standard is mandated. Where the trier of fact has considered all the evidence that the law requires him or her to consider, and still comes to the wrong conclusion, then this amounts to an error of mixed law and fact and is subject to a more stringent standard of review: Southam, supra, at paras. 41 and 45.

[51]Where, however, on a question of mixed fact and law it is possible to extricate the legal question from the factual and determine that a legal error has been made, the standard of review will be correctness. At paragraph 27 of Housen, supra, Iacobucci and Major JJ. state:

In Southam, supra, at para 39, this Court illustrated how an error on a question of mixed fact and law can amount to a pure error of law subject to the correctness standard

. . . if a decision-maker says that the correct test requires him or her to consider A, B, C, and D, but in fact the decision-maker considers only A, B, and C, then the outcome is as if he or she had applied a law that required consideration of only A, B, and C. If the correct test requires him or her to consider D as well, then the decision-maker has in effect applied the wrong law, and so has made an error of law.

[52]Mischaracterizing the proper legal test results in the application of the correctness standard to the factual conclusions reached by the trial judge. At paragraph 35, Iacobucci and Major JJ. state:

This mischaracterization of the proper legal test (the legal requirements to be a "directing mind") infected or tainted the lower courts' factual conclusion that Captain Kelch was part of the directing mind. As this erroneous finding can be traced to an error in law, less deference was required and the applicable standard was one of correctness.

As I understand the Supreme Court's reasons, it is the tainted factual conclusion to which less deference is owed, not specific facts found by the trial judge leading to that conclusion, unless, of course, they were also tainted by the legal error.

[53]Housen, supra, does not deal with the circumstance in which the Trial Judge, applying an incorrect legal standard, makes only conclusions but no findings of fact relevant to that conclusion--for example, where a trial judge makes no specific findings of fact, but simply says that he considered all the evidence and applied what is later determined to be the wrong legal test. The question remains as to how the appellate court is to review the application of the correct legal test in the absence of specific findings of fact by the trial judge. I deal with this circumstance later in these reasons.

REVIEW OF THE FINDINGS OF THE TRIAL JUDGE

Findings of Fact

[54]Since the first and second questions decided by the Trial Judge are purely factual, they are to be reviewed on a standard of palpable and overriding error. There is evidence to support these findings and no palpable or overriding error has been demonstrated with respect to them. This Court will not interfere with them.

Findings of Law

[55]The third finding, what constitutes the invention, is a legal question. The appropriate standard of review here is correctness. The Trial Judge found "there can be no doubt that W-3053 . . . was for all intents and purposes the invention described in the Patent in suit". In coming to this conclusion, the Trial Judge was making a number of implicit assumptions. As there is no product claim in the 946 patent, the Trial Judge must have been saying that the invention was contacting W-3053 with natural gas streams. In addition, it appears that his use of the words "for all intents and purposes" was meant to convey that he considered the use of W-3053 in contact with natural gas streams to capture the essence of the invention disclosed in the 946 patent, but was not precisely the invention in the patent claims. The Trial Judge's reasons do not reveal that he construed the patent claims in detail for the purpose of identifying the invention. I proceed to do so now.

[56]The 946 patent consists of 77 patent claims. Claim 1 is the most general and comprehensive claim. It discloses a method for selectively reducing hydrogen sulphide present in gaseous or liquid hydrocarbon streams or mixtures thereof (for purposes of this case, natural gas streams) by contacting the streams with a composition comprising the reaction product of a lower alkanolamine with a lower aldehyde in an amount and for a time sufficient to reduce the hydrogen sulphide in the gas stream:

A method for selectively reducing the levels of hydrogen sulphide in organic sulphides present in gaseous or liquid hydrocarbon streams or mixtures thereof comprising compacting said streams with a composition comprising the reaction product of (i) an alkanolamine comprising 1 to about 6 carbon atoms with (ii) an aldehyde comprising 1 to about 4 carbon atoms, in an amount and for a period of time sufficient to substantially reduce the levels of hydrogen sulphide in organic sulphides in said streams.

[57]Claim 1 is not a product claim. Indeed no claim is a product claim. In other words, there is nothing novel about a composition that comprises a reaction product of a lower alkanolamine and a lower aldehyde. Nor is it a novel process to reduce hydrogen sulphide in natural gas by contacting gas streams with a chemical product. The invention that is the subject of claim 1 is the contacting of the known, but not previously used for the purpose, reaction product disclosed in the claim with sour natural gas streams to reduce the level of hydrogen sulphide in the streams.

[58]Claim 2 provides that the molar ratio of the alkanolamine and aldehyde is from about 1:0.25 to 1:l0. This is a broad range of molar ratios. Presumably, within that range of molar ratios, the reaction product will be produced.

[59]Petrolite is not relying on all 77 claims in the 946 patent. According to the Trial Judge, Petrolite's counsel identified the other claims in issue as falling into four series:

I    --    Claims 44 to 50

II    --    Claims 19 to 21, 28, 29 and 34

III    --    Claims 3 to 6, 10, 13 to 16, 24, 25, 32 and 36

IV    --    Claims 57, 58, 66 and 67.

[60]Series I: Claim 44 is a particularization of claims 1 and 2. It covers a method for sweetening gas by contacting gas streams with a composition comprising the reaction product of monoethanolamine (MEA) (a lower alkanolamine) and formaldehyde (a lower aldehyde) when the molar ratio is from 1:0.25 to 1:1.15. Claim 45 limits the molar ratio from 1:1 to 1:1.5. Claim 46 provides that the reaction product is triazine. Claim 47 identifies the gas streams as comprising sour natural gas. Claim 48 specifies that the method of contacting the reaction product with the gas stream is in an H₂S scrubber tower.

[61]Series II: Claim 19 is also a particularization of claims 1 and 2. It covers a method for reducing hydrogen sulphide in gas streams using triazine. The other claims in this series also refer to triazine and otherwise do not affect the reaction product or the process of contacting it with sour gas streams.

[62]Series III: Claim 3 is also a particularization of claim 1. It limits the molar ratio of the alkanolamine and aldehyde to 1:0.25 to 1:.15. Claim 4 limits the molar ratio to 1:1 to 1:1.5. Claim 5 refers specifically to MEA and formaldehyde. Claim 6 refers to the reaction product as being triazine or bisoxazolidine. Claim 10 says that the reaction takes place at a temperature of not greater than about 200 degrees Fahrenheit. Claims 13-15 disclose different methods of contacting the reaction product with the sour gas. Claim 16 adds methanol to the composition that is brought into contact with natural gas. Claims 24, 25, 32 and 36 do not change the reaction product or the process for contacting it with sour gas streams.

[63]Series IV: Claims 57 and 58, again, are further particularizations of claim 1. They provide for contacting gas streams with a composition consisting of at least 70% triazine. Claims 66 and 67 do not change the reaction product or the process for contacting it with sour gas streams.

[64]As with many patents, there are numerous claims here. Many appear redundant. Be that as it may, it seems clear that the invention in the claims is contacting natural gas streams with a reaction product of a combination of a lower alkanolamine and a lower aldehyde. The Trial Judge found that the expert evidence was unanimous that bisoxazolidine had never been conclusively identified as a reaction product of MEA and formaldehyde, and that the testimony was equally unanimous that the predominant reaction product was triazine. The evidence was that when the molar ratio is about 1:1, the result is a higher concentration of triazine compounds. If the molar ratio is greater than 1:1, the result is triazine plus excess formaldehyde.

[65]In construing the patent claims, I am satisfied that the invention in the claims at issue in the 946 patent is the use of a composition comprising the reaction product of a lower alkanolamine with a lower aldehyde or triazine in contact with natural gas streams to reduce hydrogen sulphide levels in the said streams. The claims describe the reaction product either as triazine or as its components, combined according to molar ratios that will produce triazine or triazine with excess formaldehyde. There is nothing inventive about the methods shown to bring the reaction product into contact with the sour gas streams. Bisoxazolidine, referred to in claim 6, according to the Trial Judge, had never been conclusively identified as a reaction product of MEA and formaldehyde. The temperature limitation in claim 10 cannot be inventive because a triazine solution is aqueous, i.e. dissolved in water, and the temperature obviously could not go over the boiling point of water. Methanol, referred to only in claim 16, is an anti-freeze. I am not certain that the use of methanol is novel. However, I will assume for purposes of claim 16, that the invention was the use of triazine with the addition of methanol for reducing hydrogen sulphide in natural gas streams.

[66]Construing the patent claims as disclosing contacting sour gas streams with a composition comprising the reaction product of a lower alkanolamine with a lower aldehyde or triazine is consistent with the abstract of the disclosure in the 946 patent which states:

Methods are disclosed for selectively reducing the levels of hydrogen sulphide and organic sulphides from gaseous and/or liquid hydrocarbon streams, particular to natural gas streams, comprising contacting the streams with a composition comprising the reaction product of

(i) a lower alkanolamine with

(ii) a lower aldehyde.

[67]Having regard to all the patent claims in issue, I conclude that the invention described by the 946 patent is the use of a composition comprising the reaction product of a lower alkanolamine with a lower aldehyde, or triazine, in contact with sour gas streams to reduce the level of hydrogen sulphide in the streams. For purposes of claim 16, the invention includes methanol. Because the addition of methanol appears incidental at best, I will refer to the invention as the use of triazine or its starting components, for hydrogen sulphide removal except where it is necessary to deal with claim 16.

Findings of Mixed Fact and Law

[68]The main area of controversy in this Court was with respect to the conclusion of the Trial Judge that, based on all the evidence regarding attempts to successfully analyse and reconstruct W-3053, Canwell had not proven anticipation. This is a question of mixed fact and law. It involves whether the test for anticipation was met by the facts found by the Trial Judge from the expert evidence.

[69]Although his reasons consist of 197 paragraphs, the Trial Judge dealt with this issue in only five paragraphs. At paragraph 90, he stated:

I found the expert evidence as to efforts at analysis that have been made over the years, including efforts in the context of this litigation, to be ambivalent and, in totality, to be quite unsatisfactory.

[70]At paragraph 90, he quoted from the evidence of Professor Brian Hunter, one of Petrolite's experts, to the effect that a skilled chemist would not be able to determine the composition of W-3053 nor its starting materials. At paragraph 91, he quoted from the evidence of Professor John MacRae Mellor, an expert for Canwell, that a skilled chemist would have been able to identify triazine as the major component of W-3053 and that triazine could be produced as a reaction product of MEA and formaldehyde.

[71]At paragraph 92, the Trial Judge quoted from Décary J.A.'s reasons in Tye-Sil, supra, that evidence of anticipation should be subjected to the closest scrutiny and that the person alleging anticipation assumes a weighty burden. Then, at paragraph 93, the Trial Judge concluded that Canwell had not successfully discharged the burden on it of proving anticipation.

Against the foregoing, and considering all of the evidence before the Court regarding attempts to successfully analyse and reconstruct W-3053, I simply cannot conclude that the defendant Canwell has successfully discharged the burden on it to establish that the commercial sales of W-3053 that took place in western Oklahoma prior to the grace period preceding filing of the patent in suit, anticipated the invention disclosed by the patent in that they amounted to a making available to the public in Canada or elsewhere, the subject-matter of the patent.

[72]I readily recognize that there was much evidence in this case and many issues, and that this placed a heavy burden on the Trial Judge. Nonetheless, the summary way in which the Trial Judge reached his conclusion respecting anticipation precludes this Court from understanding his reasoning process. The Trial Judge made no findings of fact with respect to the expert evidence. He simply quoted two opposing expert opinions without commenting on them. While he concluded that the expert evidence was "ambivalent and, in totality . . . quite unsatisfactory", he provided no explanation for this conclusion. Nor does his final conclusion, at paragraph 93, provide any factual underpinning to give content to his reasoning.

[73]Be that as it may, the threshold question for this Court is whether, in determining a question of mixed fact and law, the Trial Judge made an error of law that is extricable from the mixed question. In my respectful opinion, he did.

[74]The Trial Judge's reasons indicate that the standard he was applying for proof of anticipation was the ability through reverse engineering to "successfully analyse and reconstruct W-3053". It is clear that the Trial Judge thought it necessary to reconstruct W-3053 and found the expert evidence wanting for this purpose. The Trial Judge also referred to analysing W-3053, but if the purpose of analysis in his mind was different from reconstructing W-3053, there is nothing in his reasons to indicate what that different purpose was. On the contrary, other portions of his reasons suggest that his reference to analysis was to enable W-3053 to be reconstructed, or, what is the same thing, to determine its composition. A test for reconstruction is consistent with the Trial Judge's previous determination that the invention was, to all intents and purposes, W-3053. I conclude that the test for anticipation used by the Trial Judge was successfully analysing W-3053 to determine its composition or, what is the same thing, to reconstruct W-3053. I am of the respectful opinion that the Trial Judge erred in law in using this test for anticipation.

[75]A defendant relying on the defence of anticipation by prior use or sale of a product does not have to prove that the product can be reproduced (see FISONS, supra, at page 23). The product will contain the invention but may also contain other ingredients. A defendant relying on paragraph 28.2(1)(a) need only prove that the subject-matter of the patent, i.e. the invention, has been disclosed. The invention in this case was contacting natural gas streams with triazine (or its starting components) to reduce levels of hydrogen sulphide in the streams. Because the contacting of W-3053 with sour natural gas streams was known to customers who purchased the product, it is only necessary to consider whether triazine or its starting components were capable of being detected in the W-3053. If so, then the invention was disclosed by the sale to the public of W-3053 and the 946 patent is anticipated. Thus, in the circumstances of this case, the question is whether a person skilled in the art could have discovered triazine (or its starting components) through analysis. It is not whether that skilled person could have reproduced W-3053.

[76]The legal question--whether proof of anticipation required reproduction of the prior used or sold product or discovery of its composition on the one hand, or only discovery of the invention on the other, is extricable from the mixed fact and law conclusion of the Trial Judge, that, based on all the evidence, Canwell did not prove anticipation.

[77]Further, the legal question here is general in nature and not particular to the circumstances of this case. It is important for precedential purposes that, in cases where anticipation by prior use or sale of a product is in issue, the evidence demonstrates that the analysis of the product discloses the invention and not necessarily an ability to reproduce the product. Of course, where the product is the invention, analysis might well have to show an ability to reproduce the product. But that is not the case here. Contacting sour natural gas with triazine (or its starting components) and perhaps methanol, was the only inventive elements of W-3053. Whether or not a skilled person could discover other non-inventive elements of W-3053 is irrelevant to the question of anticipation in this case.

[78]Because the error of law is extricable from the Trial Judge's conclusion of mixed law and fact, his conclusion is reviewable on a standard of correctness.

ANALYSIS OF THE EVIDENCE APPLYING THE CORRECT LEGAL TEST

[79]I turn then to the application of facts to the correct standard under paragraph 28.2(1)(a). As I have indicated, the Trial Judge made no findings of fact with respect to the expert evidence. His conclusion that the evidence was ambivalent and totally unsatisfactory, and that Canwell had not met its burden, was premised on the wrong legal standard.

[80]Had the Trial Judge made findings of fact with respect to the expert evidence untainted by his error of law, such findings would have been reviewable only on a palpable and overriding error standard. However, here, there are no express findings of fact at all. Even if it could be inferred from the Trial Judge's reference to basing his decision on "all the evidence" that he had made a type of overall finding of fact, any inference one might assume he made relative to any specific evidence was tainted by the application of the wrong legal test.

[81]In these circumstances, the only options are to return the matter to the Trial Division for a redetermination based on applying the correct legal test to findings of fact, or for this Court to substitute its decision for that of the Trial Judge. I have some initial reservations about this Court making findings of fact without seeing and hearing the witnesses first hand. However, the critical expert evidence was documentary.

[82]I am mindful that in Housen, supra, the Supreme Court has said that the absence of credibility issues is not justification for an appellate court to interfere with findings of fact by a trial judge (at paragraph 24). However, the issue now is not the standard of review. At this stage, it is necessary that the evidence be re-examined in light of the application of the correct test for anticipation. I think a relevant consideration in determining whether the appellate court should reassess the evidence or whether the case should be remitted to the Trial Division, is whether the credibility of witnesses is an issue and, if it is not, this Court may be less hesitant about conducting the review itself. Further, this matter has been outstanding for some seven years. The alleged infringement goes back some 11 years and the alleged anticipation some 14 years. It is also relevant that the evidence appears to be strong in one direction, that is, this is not a close case in which it might be preferable to have the evidence reassessed by the Trial Judge. Finally, an order was granted expediting the hearing of the appeal.

[83]The observations of La Forest J. in Hollis v. Dow Corning Corp., [1995] 4 S.C.R. 634, for the majority, are instructive for the circumstances here. At paragraph 33, he stated:

It is well established that appellate courts have the jurisdiction to make a fresh assessment of the evidence on the record where they deem such an assessment to be in the interests of justice and feasible on a practical level; see Prudential Trust Co. v. Forseth, [1960] S.C.R. 210, at pp. 216-17. In this case, there was sufficient evidence on the record to allow the Court of Appeal to make a full and proper re-assessment of the duty to warn issue without sending the case back to trial. While appellate courts are generally, and justifiably, wary of making findings of fact without having the advantage of seeing and hearing testimony first-hand, I do not believe that such concerns arise in this case because the bulk of the critical evidence adduced at trial was documentary, not testimonial. In light of the fact that Ms. Hollis has now waited close to seven years for the final resolution of her claim, and the high costs already created by the unusual length of this appeal process, I believe the Court of Appeal followed the proper course in weighing and assessing this evidence in order to achieve a measure of finality in this case; see, e.g., Davie Shipbuilding Ltd. v. The Queen, [1984] 1 F.C. 461, at p. 464 (C.A.); Nova, An Alberta Corporation v. Guelph Engineering Co. (1989), 70 Alta. L.R. (2d) 97 (Alta. C.A.), at pp. 110-12.

Having regard to the circumstances of this case and the comments of La Forest J. in Hollis, supra, I think the appropriate disposition is for this Court to consider the evidence and apply the correct legal standard and I proceed to do that now.

[84]The most important evidence was the analysis performed by Professor Keay. The Trial Judge did not comment on the Keay analysis in his reasons.

[85]Professor Keay was provided with a blind sample of W-3053 prepared by Dr. Bachelor, one of Petrolite's expert witnesses. Professor Keay analysed the sample and identified the major component of W-3053 as triazine. He also found methanol. He further proposed a method of synthesizing the triazine by reacting MEA and formaldehyde in a 1:1 molar ratio, which is the preferred method described in the 946 patent.

[86]In cross-examination, Petrolite's expert Dr. Hunter, agreed that, based upon a blind test, the Keay analysis produced the correct answer, that is, that the major component of W-3053 was triazine. Petrolite's expert, Dr. Bachelor in cross-examination, also agreed that the Keay analysis produced the correct answer.

[87]In conducting his analysis, Professor Keay used a mass spectrometry computer database to identify the presence of triazine in W-3053. However, that database had been updated in 1994, well after the relevant period, i.e. December 10, 1987, to December 19, 1988. It would not have been available to a person skilled in the art at the relevant time. Nonetheless, the evidence of Dr. Buchannon, one of Canwell's witnesses, was that he had in his possession in January 1988, a database that would also have identified the presence of triazine in W-3053.

[88]The conclusion reached by Dr. Keay, based on his blind analysis, provided evidence that a person skilled in the art, without the benefit of a priori knowledge and without taking inventive steps, could have successfully analysed W-3053 for the presence of triazine, for the MEA and formaldehyde starting compounds and for methanol. Having regard to Dr. Buchannon's evidence, the technology and data necessary for the analysis were available at the relevant time, that is, prior to December 19, 1988.

[89]Petrolite says that Professor Keay was advised to refrigerate the W-3053 sample that he had been given to analyse. They say that, prior to December 19, 1988, a person skilled in the art would not have had this information. However, the evidence is that Professor Keay analys ed the sample over the course of one week. There was no evidence that, refrigerated or not, the W-3053 would have degraded over that period. The advice to refrigerate does not undermine the Keay analysis as evidence of enabling disclosure.

[90]There was evidence of a number of other analyses of W-3053 and Canwell's CW1000 product. In most cases, these tests were performed under conditions in which either the starting materials were known or a particular component was being tested for. Therefore, these are not valid tests for anticipation because the analysts were provided with some information that would not have been available to the person skilled in the art conducting the analysis prior to December 19, 1988. A purchaser of W-3053 at that time would not have known that its starting materials were MEA and formaldehyde or that triazine was the reaction product in the mixture.

[91]Be that as it may, in all but one of the analyses, either triazine or the starting components, an alkanolamine and an aldehyde or, more specifically, MEA and formaldehyde were detected. In the one analysis that failed to discover triazine or its starting materials, the analysis was incorrectly performed. Chloroform, rather than deuterated water, was used as a solvent causing the triazine to disappear from the sample. However, Professor Keay avoided this problem by conducting a simple solubility test in his analysis, leading him to use deuterated water, rather than chloroform, as the solvent.

[92]Just as it is implicit under the Beloit test, that a person skilled in the art will, correctly and without error, read and follow the directions in a prior patent or publication, the same implication must be applicable in the case of reverse engineering. A flawed procedure leading to the wrong conclusion is not evidence that a person skilled in the art employing a procedure and correctly using data and methodology available at the relevant time, would not be led to the claimed invention.

[93]In the result, all analyses that were correctly performed led to the discovery of triazine or its starting materials. As I have stated, other than for Professor Keay's analysis, these other analyses are not in themselves sufficient evidence of anticipation because information not available to a purchaser of W-3053 at the relevant time, was known by the analysts performing these tests. However, all these analyses are remarkably consistent in leading the analysts to the invention. That leaves uncontradicted the Keay analysis, which is valid evidence of anticipation.

[94]Petrolite argued that Professor Keay did not reproduce W-3053 in its exact molar ratio. However, as indicated, reproduction of W-3053 is not the test for anticipation. Professor Keay found the invention--that triazine was the reaction product of W-3053. He found that triazine was optimally produced at a molar ratio of 1:1. This molar ratio falls within the range of molar ratios in all the claims at issue. He also found methanol referred to in claim 16 of the 946 patent.

[95]Petrolite has raised other issues relating to the reverse engineering analyses, none of which I find persuasive. It submits that W-3053 is of an extremely complex nature and is, therefore, difficult to analyse. This argument is not well founded. In law, complexity of a product or the time or work involved in the analysis are not themselves relevant considerations. The question is whether a person skilled in the art and using data and techniques available at the relevant time could discover the invention without the use or application of inventive skill. See Beloit, supra, at page 214, and FISONS, supra, at page 23. Complexity or time and work involved alone are insufficient to invalidate an analysis. The question is whether there is evidence from which the use of inventive skill may be inferred. There is none here.

[96]Petrolite says that the deliveries of W-3053 in western Oklahoma were to the private property of the purchasers and W-3053 was, therefore, not available to the public. This argument fails to appreciate that the purchaser is a member of the public, and that the Trial Judge found that sales were made unconditionally. To prove anticipation under paragraph 28.2(1)(a), it is not necessary to demonstrate that a particular purchaser did or would have conducted an analysis of the product. It is sufficient that the purchaser could have done so (see Lux, supra, at page 133).

[97]The same may be said about Petrolite's argument that the results of a reverse engineering analysis would be kept confidential by anyone performing the analysis and would, therefore, not be available to the public. It is the unconditional sale of the W-3053 to the purchaser that makes the product available to the public. If the purchaser is free to perform reverse engineering, that is sufficient. How a purchaser chooses to treat the analysis, i.e to keep it confidential or not, is not a relevant consideration.

[98]Petrolite argued that reverse engineering involves adopting a mosaic approach which is not permitted to prove anticipation. The mosaic rule applies to prior publications or patents and says that it is improper to assemble a dossier of prior art with the benefit of hindsight (see Free World Trust, supra, at paragraph 24). However, this is not what occurred here. It was W-3053 only that was the subject of analysis, and disclosure of triazine and methanol came from reverse engineering of it. The analysis of a previously used or sold product employing available techniques and data is equivalent to the reading by a skilled person of a single prior publication and following it to the claimed invention (see Beloit, supra, at page 214, and FISONS, supra, at page 21).

CONCLUSION

[99]I have carefully considered the law of anticipation articulated in Beloit, supra and Free World Trust, supra. I have also taken into account principles relating to disclosure through prior use or sale and paragraph 28.2(1)(a) of the 1996 Patent Act. I have also deferred to those findings of fact of the Trial Judge untainted by legal error. Based upon his analysis of the facts, I agree with him that W-3053 was sold unconditionally prior to December 19, 1988. Had he applied the correct legal test, the Trial Judge should have concluded that a person skilled in the art and using data and techniques available at the relevant time, and without the exercise of inventive skill, would have been led inevitably to the subject-matter of the patent claims, namely, the extraction of hydrogen sulphide from natural gas by contact with triazine (or its starting components).

[100]Canwell has proven the disclosure contemplated by paragraph 28.2(1)(a). Having found the invention, described by claims 1 and 2 and the series I, II, III and IV claims of the 946 patent, to be anticipated, I conclude that the claims relied upon by Petrolite in the case at bar are invalid. I do not find it necessary to address the other issues raised in this appeal.

[101]The appeal should be allowed and the decision of the Trial Judge should be set aside. The claims in issue in the 946 patent should be found to be invalid by reason of anticipation. The respondents' claim against the appellants should be dismissed with costs here and in the Trial Division.

Strayer J.A.: I concur.

Evans J.A.: I concur.