593 F. Supp. 208 | D. Del. | 1984
AUSTIN POWDER COMPANY, Plaintiff,
v.
ATLAS POWDER COMPANY, Defendant.
United States District Court, D. Delaware.
Robert K. Beste, Jr., Biggs & Battaglia, Wilmington, Del., Robert V. Vickers, Body, Vickers & Daniels, Daniel R. Cherry, Squire, Sanders & Dempsey, Cleveland, Ohio, for plaintiff.
Douglas E. Whitney, Morris, Nichols, Arsht & Tunnell, Wilmington, Del., Garland P. Andrews, Roy W. Hardin, David L. Hitchcock, Richards, Harris & Medlock, Dallas, Tex., for defendant.
OPINION
STAPLETON, Chief Judge.
Four issues remain in this four year old patent case pitting Austin Powder Company ("Austin") against Atlas Powder Company ("Atlas"):[1] (1) Are Claims 25 through 27 of the '621 reissue patent unenforceable *209 because of the failure of Austin's counsel to submit to the Patent and Trademark Office ("PTO") certain documents obtained from Atlas during discovery in this litigation? (2) If not, was Atlas' infringement of Claims 25 and 27 willful within the meaning of 35 U.S.C. § 284? (3) Is either side entitled to a determination that this is an "exceptional case" within the meaning of 35 U.S.C. § 285 and (4) Is Atlas entitled to intervening rights and, if so, under what circumstances may it continue to manufacture the Deckmaster? This opinion constitutes my findings of fact and conclusions of law relating to those issues.
I. THE "FRAUD" OR "INEQUITABLE CONDUCT" ISSUE
A. The Facts
Atlas' answer to the complaint in this action alleged that the patents in suit were invalid under Sections 102(a) and (g) and 103 of the Patent Act. The alleged prior art relied upon by Atlas consisted in part of work done by it prior to Austin's filing date of November 11, 1976. As a result, during discovery, Austin came into possession of all of Atlas' documents relating to this work and the development of the Deckmaster. When this occurred, Austin's reissue proceeding was still before the PTO and Atlas asserts that Austin's counsel committed fraud on that office by not bringing thirty-four of these documents (the "Paragraph 22 documents")[2] to the attention of the examiner in that proceeding. Austin acknowledges that the documents identified by Atlas were not submitted to the PTO, but maintains that its counsel had no duty to make such a submission.
1. Prosecution of Original Patent
On February 27, 1979, the PTO issued United States Letters Patent No. 4,141,296 ("the '296 patent") containing twenty-four claims directed to a delayed primer which was the subject matter of the prior liability trial decided on May 31, 1983. This original patent issued from an application prepared, filed and prosecuted by Robert V. Vickers ("Vickers"). Throughout the prosecution of the original patent application, David H. Brown was the examiner.
On October 20, 1977, Examiner Brown allowed some claims in the application and rejected others over Driscoll 3,709,149. Subsequently he rejected the remaining claims based on two patents issued on November 29, 1977, to Atlas, Postupack 4,060,033 ('033) and Bowman 4,060,034 ('034). In response to this new rejection, Vickers filed Rule 131 affidavits (37 C.F.R. § 1.131). At the same time (June 14, 1978), Vickers made several amendments to the existing claims and added two claims which ultimately became claims 23-24 of the original '296 patent. On August 11, 1978, Examiner Brown "considered and accepted" the Rule 131 affidavits and reported that "the previous rejection on these references is accordingly withdrawn."
The '296 patent then issued. There is no evidence that Vickers did anything improper during examination or allowance of this patent.
2. The '621 Reissue
On April 16, 1980, within two years of the date the '296 patent issued, Vickers prepared and filed an application to reissue the original '296 patent on the basis of application papers, including a Reissue Oath. This Reissue Oath was considered and accepted by the same Examiner Brown. Within the normal three months, Vickers submitted a "Statement Under Rule 97." This submission mentioned and forwarded copies of all patents cited against the original '296 patent, all "patents cited against Postupack 4,060,033 and Bowman 4,060,034," and the C.I.P. of Bowman 4,060,034, i.e., 4,165,691. In addition, the Reissue Oath specifically advised the PTO that "Atlas Powder Company" was "assignee of the Postupack and Bowman patents." The Reissue Oath further indicated that the Postupack and Bowman patents *210 were "withdrawn as references against the original patent under 37 C.F.R. § 1.131." The likelihood of a patent infringement suit by Austin against Atlas under either the original '296 patent or the reissue thereof was specifically mentioned by Vickers in the Reissue Oath. Before Examiner Brown acted upon the reissue, the present suit was filed on June 13, 1980.
On June 12, 1980, the Deputy Clerk of Courts for the Federal District of Delaware notified the PTO of the litigation involving the '296 patent and this statutory notice was placed in the file wrapper of the original '296 patent. On August 29, 1980, Vickers filed a Notice of Litigation with the Complaint, Answer and Reply from this litigation. The drawings from the '296 file wrapper were transferred to the '621 reissue application on October 24, 1980.
The reissue '621 issued on May 26, 1981.
3. Actions of Atlas in 1980
On October 30, 1980, Garland P. Andrews, attorney for Atlas, gained access to the publicly available '621 reissue application. At that time, Atlas was aware of everything that Vickers had or had not done in the reissue application. Vickers gave Atlas a copy of the '621 file history. Austin sued Atlas before the reissue prosecution was at the first action stage because Vickers wanted Atlas to have the opportunity to present anything it wanted to make of record in the '621 file history. Mr. Andrews knew that Atlas could give the PTO any fact or documents known to Atlas which would be relevant to the pending reissue application. Atlas chose to remain silent.
On November 3, 1980, Atlas submitted several thousand documents to Vickers. Most of the Paragraph 22 documents were included in this production. The letter covering this document production stressed that these "documents are confidential and will be subject to the protective order to be agreed upon by the parties." (DX 511). Additional documents including others of the Paragraph 22 documents were submitted after the final fee was paid to the PTO on December 19, 1980. It was not until 1982 that Atlas, in response to a Court Order, identified in amended Paragraph 22 the documents it claimed had been fraudulently withheld by Austin in 1980 and 1981.
4. Vickers' Actions From November 1980 to December 17, 1980
All except one of the Paragraph 22 documents consist of laboratory reports describing work at Atlas between May 1973 and March 1976. Their contents are summarized in my first opinion. Vickers received and studied most, if not all, of these documents during the pendency of the reissue proceeding. He became familiar with most of them prior to his meeting with Professor Kayton on December 17, 1980. His attention was not focused on these "particular documents," however, and his evaluation of them was made in the context of all the other documents and information available to him concerning Atlas' efforts to develop and market a non-electric delay primer.
It has been Vickers' view since November of 1980, based on all the information available to him that, while Atlas had decided to develop a non-electric delay primer well prior to November 11, 1976, it had not discovered a structure for reliably holding the proper relationship between the down-line and the delay unit until long after that date. Indeed, Vickers believed throughout the relevant period that Atlas' experience as a whole was strong evidence that Austin's invention was not obvious to one of ordinary skill in the art between 1973 and November 1976. Vickers was also aware that Atlas had abandoned work on a cast primer during the year following March 1976 in order to concentrate on developing a delay unit utilizing a "Kinepak" bottle. Finally, Vickers inferred from the fact that Atlas considered its laboratory reports confidential in 1980 and 1981, that it had not earlier published beyond its own personnel the facts there reported. Austin's counsel thus reached in November of 1980, substantially the same conclusions which I reached in my opinion of May 31, 1983.
*211 5. The Meeting with Professor Kayton
On December 17, 1980, Vickers conferred with Professor Irving Kayton in Washington for the purposes of evaluating the Reissue Oath and discussing the possible employment of Professor Kayton as a patent expert in this litigation. At this December 1980 meeting, Professor Kayton and Vickers primarily focused on the reissue application and specifically the Reissue Oath. The Atlas document production was also discussed, however.
Professor Kayton asked Vickers whether or not the documents of Atlas showed at any given time a conception or reduction to practice of the invention claimed in the '296 patent. Vickers replied that they did not. Vickers also mentioned that the Austin inventors were not aware of the Atlas work and that Atlas had insisted upon Vickers treating the Atlas documents as confidential. Professor Kayton did not attempt to make an independent study of the Atlas documents. He did advise Vickers of a recent decision of the CCPA entitled In re Clemens, 622 F.2d 1029 (1980).
B. The Atlas Position
Atlas' argument on the "fraud" issue begins with Section 1.56 of the Rules of the PTO as it existed in 1980 and 1981. 37 C.F.R. § 1.56 (1977). That section provided in relevant part:
§ 1.56 Duty of disclosure; striking of applications.
(a) A duty of candor and good faith toward the Patent and Trademark Office rests on the inventor, on each attorney or agent who prepares or prosecutes the application and on every other individual who is substantively involved in the preparation or prosecution of the application and who is associated with the inventor, with the assignee or with anyone to whom there is an obligation to assign the application. All such individuals have a duty to disclose to the Office information they are aware of which is material to the examination of the application. Such information is material where there is a substantial likelihood that a reasonable examiner would consider it important in deciding whether to allow the application to issue as a patent ...
Atlas stresses that the examiner had originally rejected Atlas' claims in the prosecution of the '296 patent based on Postupack and Bowman, two patents that reflected the work at Atlas prior to their filing date of March 9, 1976. Atlas further stresses that Vickers had successfully removed these patents from consideration as prior art by filing a Rule 131 affidavit in that prosecution indicating an Austin reduction to practice on March 5, 1976. Under these circumstances, Atlas asserts that a reasonable examiner in the reissue proceeding would have wanted to know about and consider any evidence tending to show a reduction to practice by Atlas prior to March 9, 1976. The Paragraph 22 documents, according to Atlas, were substantial evidence on this important issue and Vickers had a duty to draw the examiner's attention to this material information so he could decide that issue for himself. Since Vickers' evaluation of the evidence was either not made in good faith or from a grossly biased perspective and since he made a deliberate decision which had the effect of depriving the examiner of this opportunity, it is said that the '621 is unenforceable.
C. The Austin Position
Austin first correctly points out that the issue to be decided is not whether Vickers had a duty to submit to the PTO thirty-four documents hand picked by an adversary in litigation, but rather whether he had a duty to disclose all of the information he had learned having a bearing on whether there was an Atlas reduction to practice prior to March 5, 1976. If the purpose of Section 1.56 is to give the examiner a fair opportunity to apply the statutory criteria in light of the material facts, that objective would clearly not have been served here by the submission of the Paragraph 22 documents alone.
With respect to the issue of whether Vickers should have disclosed all of the Atlas documents, Austin asserts, inter alia, (1) that this Court's determination *212 that the Atlas work was not prior art means that these documents were not material as a matter of law, (2) that these documents are not material as a matter of fact, (3) that the documents as a whole are evidence of patentability which there was no duty to disclose, and (4) that, in any event, the '621 is valid and enforceable because Vickers' conduct was both objectively reasonable and undertaken in subjective good faith.
D. Analysis
In American Hoist & Derrick Co. v. Sowa & Sons, 725 F.2d 1350 (Fed.Cir.1984), the Court of Appeals for the Federal Circuit points out that there are degrees of materiality just as there are degrees of scienter and that any resolution of a fraud or inequitable conduct claim should be made with reference to where the case falls both on the continuum of materiality and the continuum of scienter. The Court observes that when the information withheld or misstated is highly material, less of a showing of culpability should be required to establish invalidity. The converse is said to be true as well; if there is a deliberate intent to deceive, for example, the information involved need not be as important to a resolution of the issues of patentability as would be required if only gross negligence were involved.
While the American Hoist opinion is carefully crafted to recognize the infinite variety of fact patterns which may be presented and to avoid prescribing rules which can be mechanically applied, the Court does suggest some helpful guidelines. It explicitly or implicitly recognizes five stations along the continuum of materiality: (1) information which is not relevant to the issues of patentability, (2) information which is relevant to those issues but cumulative, (3) information which is sufficiently relevant and non-cumulative that a reasonable examiner would want to know about it but which, upon consideration, he would not consider important in his decision making process,[3] (4) information which a reasonable examiner would consider important in his decision making process but which does not make the invention unpatentable, and (5) information which makes the invention unpatentable. Moreover, the Court suggests that when the information at issue falls below the first three of these stations on the continuum of materiality, the challenge to the validity of a patent should ordinarily be rejected without reference to the state of mind of the one charged with misconduct. When the information at issue falls in the range of the fourth station, the patent should ordinarily be declared invalid where there has been a deliberate intent to mislead, but ordinarily not otherwise.[4] Finally, the Court intimates that if the information is such that it would have caused rejection of one or more claims, even a showing of less culpability than gross negligence may suffice.[5]
*213 A number of the arguments advanced by the parties in this case seem to me inconsistent with the approach taken by the Court in American Hoist. First, Atlas' primary argument, as I understand it, is that Section 1.56 is designed to assure that patent examiners, rather than patent counsel, resolve issues of patentability and that whenever lawyers might disagree about the importance of certain information, counsel for an applicant must disclose it to the PTO and has no discretion to do otherwise. The approach of American Hoist, however, seems to me to assume that discretion will be exercised by those dealing with the PTO and to suggest that an exercise of discretion which is made in subjective good faith cannot result in invalidity except perhaps when the information withheld is of the utmost materiality.
Austin, on the other hand, argues, as I understand it, that whenever a reference withheld from the PTO is ultimately found not to invalidate the patent, the withholding cannot have adverse consequences. In a like vein it argues that a reference judicially determined not to be prior art, as a matter of law, cannot be material. With some justification Austin cites Preemption Devices, Inc. v. Minnesota Mining & Mfg. Co., 732 F.2d 903 (CAFC 1984) and Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693 (CAFC 1983), respectively in support of these two propositions. I conclude that neither case stands for the proposition for which it is cited, however, and that both propositions are inconsistent with the approach of American Hoist.
American Hoist endorses the definition set out in Section 1.56(a) as one form of materiality which can have legal consequences. That definition focuses on what a reasonable examiner would be likely to consider truly important in making his decisions on patentability. It does not limit materiality to information which would be outcome determinative of such decisions. Moreover, I am relatively confident that the American Hoist Court would not hesitate to invalidate a patent in a case where a reference would clearly anticipate the patent if prior art, where the issue of whether it is prior art is considered by the applicant's counsel to be a "toss up," where counsel withholds the reference to avoid the risk of loss, but where the reference is ultimately determined not to constitute prior art.
I assess the materiality of the Atlas documents as falling within the range of stations 3 and 4 on the materiality continuum discussed above. They may be viewed as falling within the neighborhood of station 3 because the documents and the Postupack and Bowman patents neither anticipate the claims in suit nor are they more pertinent than the prior art which was before the examiner. As I observed in my earlier opinion:
Atlas also suggests that its own early development work is prior art under Section 103 which, when combined with the other prior art patents, makes Austin's invention obvious. Atlas' work, however, does not augment the teachings of Driscoll, Kern, and Turnbull in any way relevant to the Austin invention. As noted earlier, the Atlas development work, for example DX-101, did not aid in the solution of the problem left open by those patents, namely, the problem of how to use a structure to hold reliably the proper relationships of downline to delay unit while still permitting both sliding and detonation of the delay unit.
Even if the Atlas documents and/or the patents they allegedly would have qualified as prior art are viewed as more than cumulative, however, Atlas has failed to show a substantial likelihood that a reasonable examiner would have considered them important to his decisions on patentability. While the Atlas documents might be considered sufficiently relevant that a reasonable examiner would wish to be aware of them, the issue of whether Atlas' work is *214 prior art is not at all a close one and I am confident that a reasonable examiner, upon considering the Atlas documents, would not consider them important in his decision making process. The Atlas documents as a whole record the failure of Atlas' research program prior to November of 1976. Even more important, they demonstrated beyond any doubt that Atlas realized the great market potential for a non-electric delay primer in early 1973, that by 1976 Atlas knew it was behind its competition in actually marketing a device, that development of a marketable device was made a high priority item by Atlas management in 1977, and that Atlas did not market a commercial product until after April of 1979. These undisputed facts speak eloquently of the fact that Atlas simply did not reduce to practice prior to November 11, 1976.
Even assuming arguendo, however, that a reasonable examiner would have considered the Atlas documents "important" within the meaning of Section 1.56, it still would not follow that the claims in suit are invalid or unenforceable. Atlas cannot seriously contend that Vickers believed the Atlas work to be prior art or even that he believed this to be a close issue. Its argument is that Vickers failed to remain sufficiently objective in his evaluation of the evidence. Thus, Atlas states in its Opening Brief,[6] "The simple truth ... is that he [Vickers] convinced himself that the documents were immaterial and that, therefore, he had no duty to disclose same to the Patent Office." In short, this is simply not a case involving a subjective intent to mislead.
I should add that, in addition to his other testimony concerning the basis for his view of the prior art issue, I credit Vickers' testimony that he inferred the Atlas work was done on a confidential basis, that he knew the Austin inventors had had no actual knowledge about Atlas' work and that he understood from Professor Kayton and In re Clemens that, under these circumstances, the Atlas work was not prior art.
Vickers could not have thought he could sweep the Atlas work under the rug. He was involved in litigation in which the significance of that work would have to be determined. He was in a reissue proceeding into which Atlas could inject any documents it wished at any time. Realizing these facts, Atlas argues that Vickers, in withholding the documents, was motivated by a desire to avoid any delay in the reissue proceedings so he could rely on the new claims and the presumption of their validity in these proceedings. I conclude, however, that the Atlas documents were not submitted to the PTO by Vickers because he made a good faith and objectively reasonable judgment that the Atlas work was not prior art.
E. Conclusion
For the foregoing reasons, I hold that the claims in suit are not invalid or unenforceable because of fraud or inequitable conduct before the PTO.
II. THE WILLFUL INFRINGEMENT ISSUE
I have previously determined that there was no infringement by Atlas, willful or otherwise, prior to issuance of the reissue on May 26, 1981. By that date, Atlas had been familiar with the reissue proceeding for seven months and this litigation was already over eleven months old. In this litigation Atlas vigorously contended throughout this proceeding that Claims 25 and 27 of the '621 were not infringed by the Deckmaster, that those claims were invalid under the recapture doctrine and for other reasons, and that Atlas was entitled to continue production of the Deckmaster under 35 U.S.C. § 252 even if those claims were determined to be valid and infringed. Austin has failed to persuade me that Atlas' management lacked a good faith belief in the merits of these positions. It necessarily follows that Austin's claim under 35 U.S.C. § 284 must be denied.
*215 III. THE ATTORNEY'S FEE ISSUE
Atlas and Austin each accuse the other of litigating this case in bad faith. Each, accordingly, seeks an award of counsel fees. While I believe that each side of this hard fought fight lost its objectivity from time to time and that the Atlas camp did so more frequently than did the Austin camp, I am unwilling to say that either side litigated this case in bad faith.
IV. THE INTERVENING RIGHTS ISSUE
Atlas correctly points out that the only claims it has been held to have infringed are reissue claims which did not appear in the '296 patent and that the Deckmaster had been on the commercial market slightly over two years when Austin succeeded in securing a reissue to cover that competing primer. Atlas maintains that, under these circumstances, it must be permitted to continue to manufacture and sell the Deckmaster.
The Court of Appeals for the Federal Circuit recently passed upon a similar contention in Seattle Box Company, Inc. v. Industrial Crating & Packing, Inc., 731 F.2d 818 (Fed.Cir.1984). The teachings of that case are helpful here:
The second paragraph of section 252 ... [is intended] to protect intervening rights. The second paragraph provides, in pertinent part:
No reissued patent shall abridge or affect the right of any person * * * who made * * * or used prior to the grant of a reissue anything patented by the reissued patent, to continue the use of * * * the specific thing so made * * * or used, unless the making [or] using * * * of such thing infringes a valid claim of the reissued patent which was in the original patent. * *
The statute sets forth a single straightforward test for determining whether the doctrine of intervening rights protects an alleged infringer. The only question to ask under this test is whether claims of the original patent which are repeated in the reissue patent are infringed....
We have ... held ... that the claims appearing in Seattle Box's reissued patent are substantively different than those in the original patent.... Industrial, therefore, may properly raise a defense of intervening rights....
When the doctrine of intervening rights is properly raised, the court must consider whether to use its broad equity powers to fashion an appropriate remedy. The second paragraph of section 252 states:
[The court] may provide for the continued manufacture, use or sale of the thing made * * * or used as specified, or for the manufacture, use or sale of which substantial preparation was made before the grant of the reissue, and it may also provide for the continued practice of any process patented by the reissue, practiced * * * prior to the grant of the reissue, to the extent and under such terms as the court deems equitable for the protection of investments made or business commenced before the grant of the reissue.
The court is given the discretion to fashion a remedy from a wide range of options available to it. The court may, for example, (1) confine Industrial to the use of those double-concave blocks already in existence, (2) permit Industrial to continue in business under conditions which limit the amount, type, or geographical location of its activities, or (3) permit Industrial to continue in business unconditionally.
Id. at 829-30 (emphasis added).
As I read Section 252 and Seattle Box, an infringer of a reissue claim not found in the original patent who made or used its infringing device prior to the reissue is, ordinarily, entitled to some relief from the Court. While the Court may take into account all the surrounding circumstances in fashioning an appropriate remedy, the infringer is normally entitled to continue its use or sale of the "specific thing" made or *216 used before the reissue "to the extent and under such terms as the court deems equitable for the protection of [its] investments made or [its] business commenced before the grant of the reissue." While a court of equity could perhaps deny intervening rights where the defendant has unclean hands or its preexisting investment and business were insubstantial, the infringer is ordinarily entitled to have the court fashion a remedy calculated to protect its preexisting investment and business. It is not necessary, for example, for the infringer to show reliance on the limited scope of the claims of the original patent.
Although I do not believe additional fact findings are necessary to support Atlas' entitlement to protection of its intervening rights, I note that Atlas' development of the Deckmaster was independent of Austin's work. While it is true that Atlas personnel were aware of the commercial Austin product before designing the Deckmaster, they did not utilize the design features of that product in that process.
As previously noted, Atlas had been marketing the Deckmaster for more than two years prior to the reissue. While the evidence is scant, the fair inference is that Atlas' Deckmaster business was substantial on May 21, 1981 when the '621 issued.[7] Based upon that fact and the absence of any countervailing ones, I conclude that Atlas is entitled to intervening rights under 35 U.S.C. § 252. This conclusion provides the answer for each of the questions regarding intervening rights set forth in the Pretrial Order for the initial trial of the liability issues.[8] The remaining issue, which was reserved for the remedy phase of these proceedings, is how Atlas' business as of May 21, 1981 should equitably be protected. Further proceedings on the issue will be scheduled promptly.
NOTES
[1] An opinion describing the patents-in-suit, the prior art and the accused device and resolving other issues in this litigation is reported at 568 F. Supp. 1294 (D.Del.1983).
[2] So named by the parties because they were ultimately described in Paragraph 22 of Atlas' answer.
[3] also emphasize that the pertinent inquiry is not whether a reasonable examiner would want to be aware of a particular thing, but whether, after he was aware of it, he would "consider it important" in deciding whether to reject one or more claims.
725 F.2d at 1363 n. 2.
[4] it is demonstrated that a reasonable examiner would merely have considered particular information to be important but not crucial to his decision not to reject, a showing of facts which would indicate something more than gross negligence or recklessness may be required, and good faith judgment or honest mistake might well be a sufficient defense.
[5] an objective "but for" inquiry is satisfied under the appropriate standard of proof, and although one is not necessarily grossly negligent in failing to anticipate judicial resolution of validity, a lesser showing of facts from which intent can be inferred may be sufficient to justify holding the patent invalid or unenforceable, in whole or in part....
For an analysis of American Hoist which considers a showing of gross negligence or recklessness a threshold requirement, see Allen M. Sokal, Impact of Decisions of the Court of Appeals for the Federal Circuit on the Duty of Disclosure, 66 Journal Pat.Off.Soc. 29 (Jan. 1984).
Where the objective "but for" standard is met, the patent can be declared invalid for failure to meet a statutory criterion without reference to the claim that the patentee breached its duty of candor to the PTO. The presence or absence of such a breach may nevertheless be important. Where there has been such a breach, for example, the case may be ruled an "exceptional one" under 35 U.S.C. § 285 or licensees may be held to be entitled to refunds from the date of the license. Troxel Manufacturing Co. v. Schwinn Bicycle Co., 465 F.2d 1253 (6th Cir.1972) (dictum).
[6] Defendant's Opening Brief p. 60.
[7] We know, for example, that Atlas was trying to catch up with Austin when it entered the market in April of 1979, that Atlas has been manufacturing and selling the Deckmaster continuously since that time, and that Atlas, at the time of trial in September of 1982, was estimating its then current annual volume at approximately 750,000 units per year.
[8] Docket No. 147: Austin Questions of Law 89-95, pp. 79-81; Atlas Question of Law 22, p. 85.