Opinion
Nos. 7223-7225, 8094-8096.
October 6, 1939. Supplemental Opinion December 13, 1939.
Appeals from the District Court of the United States for the Northern District of Ohio, Eastern Division; Paul Jones, Judge.
Suits by the Cleveland Trust Company and another against the Schriber-Schroth Company, against the Aberdeen Motor Supply Company, and against the F.E. Rowe Sales Company for infringement of Gulick patent No. 1,815,733, claims 1, 11-13, 15, 18, 30 and 33, Maynard patent No. 1,655,968, Jardine patent No. 1,763,523, Mooers patent No. 1,402,309, and Schmiedeknecht patent No. 1,256,265. The plaintiffs appealed from adverse decrees to the Circuit Court of Appeals, 92 F.2d 330, which set aside the decrees and remanded the causes with directions. The Supreme Court, 305 U.S. 47, 573, 59 S.Ct. 8, 83 L. Ed. 34, after denying certiorari and a petition for rehearing, received the cases upon a second rehearing petition and reversed with remand to the Circuit Court of Appeals. A second group of appeals by the Schriber-Schroth Company, the Aberdeen Motor Supply Company, and the F.E. Rowe Sales Company from the decrees is consolidated for consideration with the reconsideration required by the order of remand.
Reversed and remanded for further proceedings.
Arthur C. Denison and F.O. Richey, both of Cleveland, Ohio (Wm. C. McCoy, of Cleveland, Ohio, Milton Tibbetts, of Detroit, Mich., and Baker, Hostetler Patterson, Richey Watts and Evans McCoy, all of Cleveland, Ohio, on the briefs), for Cleveland Trust Co. and Chrysler Corporation.
John H. Sutherland and John H. Bruninga, both of St. Louis, Mo., for Schriber-Schroth, Aberdeen Motor Supply Co., and F.E. Rowe Sales Co.
Before HICKS, SIMONS, and ALLEN, Circuit Judges.
Our former opinion in this much litigated patent controversy involving automobile pistons is reported at 92 F.2d 330, under identical style, but including disposition of the companion cases brought by appellants against the Aberdeen Motor Supply Company, No. 7024, and the F.E. Rowe Sales Company, No. 7025. The Supreme Court, after denying certiorari and a petition for rehearing, received the cases upon a second rehearing petition and reversed with remand to us for consideration of questions presented and not determined, Schriber-Schroth Co. v. Cleveland Trust Co., 305 U.S. 47, 59 S.Ct. 8, 83 L.Ed. 34; directions amended, 305 U.S. 573, 59 S.Ct. 8, 83 L.Ed. 34; A second group of appeals (8094, 8095, 8096) from decrees against appellees in the first case, responsive to our decision therein, is now consolidated for consideration with the reconsideration required by the order of remand.
The Supreme Court granted certiorari after twice denying it on the ground that conflict was improbable because of the concentration of the automobile industry in the Sixth Circuit. It might be interesting to explore the availability of the Declaratory Judgment Act, 28 U.S.C.A. § 400, to dealers in other circuits to raise conflict.
The claims of five patents, and a number of infringing devices, were in the first instance presented for adjudication. In their consideration we were greatly impressed by the story disclosed by the record of the long search and sustained effort on the part of exceptionally competent automotive engineers to completely solve the problem of a successful aluminum piston operating freely with minute clearances in an iron cylinder. There had been advances, "foreshadowing," perhaps, that solution would not always defy inventive genius, though on the whole the aluminum piston as heretofore noted had received a "black eye" in the industry. Much had been expected of a piston invented by one Jardine, and it was about to be put into commercial production when a design disclosed by Maynard, which seemed an improvement thereon, was produced by the appellants' licensee, was accepted by major companies in the industry, and went into use to the number of many millions. In analyzing the Maynard invention it appeared to us that however meritorious certain structural improvements might be, it nevertheless embodied the basic concept of the patent to Gulick owned by the appellant and in suit. We held certain claims of Gulick and Maynard to be valid and infringed without passing upon the intermediate disclosure of Jardine, though recognizing that Jardine may have made substantial contribution to the art.
Observing that many if not all of the elements of the Gulick combination were old, we held his precise combination to be novel and to have the quality of invention, and that its reproduction by the appellees in the specific form disclosed by Maynard was infringement of both. In a recital of the Gulick achievement in terms of elements, function and result, we assigned to the webs of Gulick the rigidity proclaimed by his specification in the dimension in which they furnished "a particularly strong support for the bosses," but a lateral flexibility in the dimension wherein rigidity appears not to have been required, where it would not serve the purpose announced, and where it would appear to interpose an obstacle to the expected functioning of a split skirt. There was assumption that since the specification was addressed to those skilled in the art, a purpose to provide a particularly strong support for the bosses was fully served by vertical rigidity, and made implicit in the original specification a lack of rigidity otherwise. With this view we found no difficulty in agreeing with the tribunals that sustained the Gulick amendment reciting lateral flexibility as mere clarification.
The Supreme Court, concluding that Gulick's failure to describe the flexible web element as a feature of his invention, and his failure in this respect to explain the principle of his machine so as to distinguish it from prior art, did not conform to R.S. § 4888, 35 U.S.C.A. § 33, held that his amendment added new matter beyond the scope of the device described in the application as filed, in consequence of which the appellant might take no benefit therefrom. Assuming that we sustained certain claims of his patent only by reading into them the flexible web element, it reversed our decision as to Gulick, and since Maynard neither claims nor describes flexible webs, it likewise reversed our decision as to its validity and infringement. Reciting, however, that since it had been thought by us that full recovery could be had by the appellants under the two patents adjudicated, we did not decide other questions involved in the appeals, it remanded the causes to us for further proceedings "with respect to such claims of the patents in suit as appellant below submitted to that court for adjudication." [ 305 U.S. 47, 59 S.Ct. 14, 83 L.Ed. 34.] Upon petition for rehearing and clarification, it added to the direction for judgment the following: "On the remand the Court of Appeals will be free to consider whether the amendments to the Gulick application rendered void the patent issued upon it, and to consider all questions affecting the validity and infringement of the claims in suit of the Gulick and Maynard patents, but without including web flexibility or laterally flexible webs as an element in the combinations patented by them."
Controversy has arisen over the scope of the remand, the appellees contending that we are precluded from consideration of the Jardine, Schmiedeknecht and Mooers patents since the bills stand dismissed as to them, albeit without prejudice, so that they may be brought to us for consideration only by new actions. The order of remand is, however, specific, requiring no interpretation. We are directed to take further proceedings with respect to such claims of the patents in suit as appellants submitted to us for adjudication, not merely in respect to those adjudicated. The phrasing of the amended direction is one of enlargement and clarification and not of limitation. That decrees have been entered below dismissing the bills as to the three patents not considered is of no moment. These decrees are brought into question in appeals 8094, 8095 and 8096 here involved. If they are in conflict with conclusions now reached they must be set aside and our mandates recalled, just as they must be set aside as in conflict with decision above. We hold that the remand operates to reopen the entire cause except insofar as questions necessary to decision have been decided by the Supreme Court and insofar as our own conclusions have not there been held erroneous and so constitute the law of the case.
The first question that confronts us is in respect to the validity of the claims of the Gulick and Maynard patents as circumscribed since we are now obliged to consider them without reading into them web flexibility or laterally flexible webs as an element, and in this connection it seems immaterial whether we read an element into them or merely construed an existing element as having in its environment a given characteristic. We recognized in our first opinion that Gulick was for a combination of elements, many if not all of which were old. Yet the combination appeared to us to be novel and to denote invention since a new result was secured in consequence of a nicely balanced flexibility of piston skirt, permitting successful operation at minute clearances without bending concentration at any point to impair ruggedness, and so distinguishing from Franquist and other pistons. In a case not free from doubt, since probably no one has seen an automobile piston reciprocating in its cylinder under operating conditions, the success of the commercial embodiment of Maynard, comprehending what was then conceived to be Gulick's basic concept, after so much failure, clearly appeared to denote that somewhere was invention. But the Gulick conception as we then considered it was of balanced flexibility derived from interplay of yielding skirt periphery and flexing supporting webs in response to the mechanical force of the cylinder wall and the thermal expansion of the piston head. The combination we are now obliged to consider is one with rigid webs opposing their rigidity to the resiliency of a skirt slotted to accommodate itself to forces to which practically it may not yield. This is not the combination we held to be valid. It is not the combination to which those most captious and skilled in the industry gave tribute, to the end that all our doubt as to validity was resolved. We are constrained to hold all of the claims in suit of Gulick invalid as marking no inventive step over prior art. It therefore becomes unnecessary to consider whether his amendments voided his patent since the direction in this respect is but permissive and not imperative. It likewise becomes unnecessary to consider the effect of the disclaimer filed since decision above.
In Maynard there was no emphasis on rigidity of web construction such as was found by the Supreme Court in Gulick, but we are compelled nevertheless to view the Maynard combination as one failing to disclose cooperation of flexible webs with vertical and longitudinal slotting to secure the balanced flexibility conceived to reside in Gulick, and constituting therein a basic achievement, and since it is apparent that details of construction recited by us as disclosing a piston lighter and more economical of manufacture than Gulick and more rugged and durable than Jardine made no impression upon the court as marking inventive steps over prior art, we are likewise constrained to hold the claims of Maynard in suit invalid.
In the conviction that the long search for a solution of the aluminum piston problem by engineers of great ability, their repeated failures, and the ultimate success of the Aluminum Company's commercial device point with persuasiveness to the fact that somewhere in the search one, with the spark of genius denied the skilled mechanic, contributed to the art the thought that brought achievement out of failure, we reconsider the Jardine patent, for though we are as sensitive as is the Court of Appeals of the Second Circuit [Buono v. Yankee Maid Dress Corp., 77 F.2d 274, 276], to "the high standard demanded for invention by the decisions of the Supreme Court in recent years," attested by many of our more recent decisions, yet we also share Judge Learned Hand's understanding that the Supreme Court has not departed from the principle that recourse must be had in doubtful cases to the history of the art for the interpretation of claims and the existence of invention.
The patent to Jardine, 1,763,523, was issued June 10, 1930, upon an application originally filed March 11, 1920, prior to the filing of the Gulick amendment in 1922, now held to be invalid. Jardine with great clarity and in precise detail explains the principle of operation of his machine and the flexibility of its webs. The knowledge that was not Gulick's or was by him concealed is clear to Jardine and by him proclaimed. "The webs 6 and guide segments preferably are so designed that this displacement of the segment 10, 10a, is permitted by virtue of a bending of the webs 6 at points remote from the guide segments. To this end, as shown in Fig. 4, the thickness of the guide segments is increased toward the webs 6 and the webs 6 are decreased in thickness from the guide portions inward toward the bosses to points in line with the inner ends of the slot 11a. This gives in effect a cantilever structure weakest at its support. * * * Thus the reaction force of the cylinder wall on the outer faces of the guide segments as the piston expands tends to cause bending of the webs 6 along said line 13. Due to the bending of the web sections 6 and the forcing together of each pair of segments, the guide part of the piston may undergo a considerable thermal expansion without a corresponding increase in the outer diameter thereof and thus a small initial clearance can be used without danger of scoring or seizure of the piston."
Jardine is also the first to refer specifically to an "aluminum alloy piston," and he states that one embodying his invention "can be fitted to the engine cylinder with a very small clearance, and that compensation for the relatively large thermal expansion of the guide section of the piston is provided by the displacement of the guide segments which results from the reaction force of the cylinder wall on the outer faces of the guides." It is true that the drawings of Jardine depict what is known as the slipper type of piston, illustrated also in the patent to Ricardo, No. 1,294,833, but his invention is not limited to that type, for he says: "It is to be understood that the invention is essentially applicable to other forms of construction."
While the specimen claims of Jardine 1, 8 and 11, though reciting the web as an element, do not incorporate therein the property of flexibility, yet in the light of the description they must be so interpreted, for it is clear that the Supreme Court in reversing our decision departs not at all from the rule that claims must be read in the light of the specification, since it is pointed out with particularity in respect to Gulick that his specifications point to no inadequacy in the structure or function of webs of the prior art which would be remedied by the webs specified, and to no function to be performed by them other than as a means of connecting and holding head and skirt so as to maintain the air gaps between them and to support the wrist pin bearings. Jardine repels this criticism of the Gulick application. It is also pointed out that Gulick failed to explain the principle of his machine so as to distinguish it from the prior art. Jardine points to prior art failure. "I have found," he says, "that these difficulties can be overcome by constructing a piston with its skirt or guide portion supported and slotted or divided in a manner to permit deformation and deflection of parts thereof without interfering with the performance of the essential functions of the respective parts."
To invalidate Jardine the appellees rely mainly upon the Long piston and Long's patent No. 1,872,772, upon an application prior to Jardine's, and emphasis is placed upon the observation of the court to the effect that the Long piston, with webs concededly flexible laterally, was in commercial use in 1922. This brings us to a consideration of the contention, vigorously urged by the appellants, that the Supreme Court having applied to the sufficiency of disclosures made by Gulick and Maynard the so-called strict rule of construction, an equally strict rule must be applied to prior art. To this the appellees reply that there is a distinction between the teachings of a prior art patent or publication and a prior use. They concede that the description in patent or publication must be complete when read in connection with the prior art as one skilled in the art would read it, but urge that when it comes to a prior use, and an article is in existence so that it may be duplicated, it is immaterial whether those skilled in the art knew or were unaware of its characteristics.
Recognizing distinction between what may be disclosed in a patent and what may appear to the understanding of those skilled in the art when it reaches commercial embodiment and so may be physically reproduced, yet it would seem to us that nothing is prior art, at least in respect to function, unless its function is understood by those skilled in the art. Otherwise there would be little support in reason for the rule that novelty is not negatived by any prior accidental occurrence or production the character and function of which was not recognized until later than the date of the patented invention sought to be anticipated thereby, 1 Walker, 6th Ed. Sec. 106, or the further rule that novelty is not negatived by anything which was neither designed nor apparently adapted, nor actually used, to perform the function of the thing covered by the patent, Idem, Sec. 107. Nor is the observation of the court that a Long piston was in commercial use in 1922, the webs of which were concededly flexible, to be taken as a conclusion that the webs were flexible in use or that their flexibility, cooperating with other resilient elements, was then understood, however clear it may later have come to be. The Long patent did not disclose flexible webs. Since the argument that the webs of Gulick were laterally flexible to the understanding of those skilled in the art was rejected, so must we reject inference that prior to Jardine the art understood the webs of Long to be flexible in the absence of clearly convincing evidence that they were known to have that quality. Moreover, the Long patents recite that the separated portions of the skirt are rigidly held against bending to the forcing thrust of the piston. A Long piston was introduced to the industry by advertisement in The Motor World, December, 1921. It was installed in the Franklin automobile, and use at a clearance of .005 was set forth in a service bulletin. The evils following clearance so great are sufficiently recited in our former opinion, and the announced clearance negatives the thought either of web flexibility or of its understanding by the art.
It is said that pistons could not in an air cooled motor as was the Franklin, be machined to the minute clearance possible in water cooled motors. If true, it follows that Long contributed little to the solution of a problem whereby successful operation of an aluminum piston in an iron cylinder at clearances of .002 or less was achieved. It should, moreover, be noted that Long was a worker in the piston art for a considerable period. Three of his patents are in the record. It is conceded that No. 1,872,772 was his best effort, yet the structures urged to anticipate Jardine were devices prior to it, and while they may have had some commercial history, his latest patent still failed to describe or claim web flexibility.
Hartog, in his patent No. 1,842,022, prior to Jardine, claims internal yielding ribs, and means for yieldingly connecting skirt sections with the head. It seems clear from the Hartog description and drawings that his thought was of a springing resiliency in arcuate extensions of his skirt sections, and not of the bending or flexing of straight ribs so proportioned as to yield to the thermal expansion of the bosses on the cantilever principle, and that he did not disclose Jardine's mode of operation. There is little in the record to aid us to an understanding of Hartog. The experts did not discuss it, and the appellees appear to place no great reliance upon it. It is a mere paper patent without influence upon the art, and it is difficult, if not indeed impossible, to comprehend its mode of operation or whether it is operable at all, since he describes a piston skirt of greater diameter than the cylinder, mechanically compressible when cold, to fit tightly therein. Hartog taught nothing to Jardine.
The specimen claims 1, 8 and 11 of the Jardine patent are valid. Since our conclusion in this respect is based upon the concept of a combination with balanced skirt flexibility due to cooperation of longitudinal and vertical slotting with flexing webs supporting retracted bosses and connected to a skirt thereby made responsive to physical compression and thermal expansion so as to permit of minute clearances between piston and cylinder, a concept perceived in Gulick as amended and minus amendment no longer perceived, and since this combination with modifications which may improve but do not destroy it appears in Jardine, and since it is the combination that appears in the commercial piston of the Aluminum Company of America responding to Maynard's specific form and in such form reproduced by Sterling Products Co., in exhibit I, sold by the appellees, we hold that the latter infringes claims 1, 8 and 11 of Jardine. As we were unable in our first decision to find Gulick's complete concept, as we then understood it, in other accused structures, so are we now unable to find in them Jardine's concept, and we hold they do not infringe.
We are no more concerned than was the master below with the character of the monopoly sought by the appellants' licensee. Monopoly is the essence of the inventor's reward, and an assignable property right. Its control, abridgment or destruction lies wholly within the legislative province.
The Mooers patent 1,402,309 was not greatly stressed in the earlier briefs or argument. It is not strongly urged now. It is contended that Mooers disclosed the scheme for converting heat into mechanical energy and diverting expansion through this energy into relieved spaces limiting expansion in those unrelieved. But if this is the teaching of Mooers it must be derived from inferences not permitted to be drawn from Gulick and Maynard, and without such interpretive approach the conclusion must be that the Mooers disclosure is incomplete. We hold its specimen claims to be invalid.
Schmiedeknecht patent 1,256,265 has now expired. It discloses no web flexibility cooperating with other elements of resiliency to achieve the balanced flexibility perceived in Jardine on the basis of which alone the latter is thought to be valid, and we sense no advance in Schmiedeknecht that distinguishes it from the prior art in aspects beyond the skill of the art. It does not invalidate Jardine and is itself invalid.
Appeals 7223, 7224 and 7225 are from decrees in the same causes determined by our previous opinion, and responsive to our mandates therein. This unusual situation resulted from the denial of a petition for certiorari to our first judgment and denial of a petition for rehearing thereon. It was doubtless assumed below that the cases had reached finality. The decrees adjudicated certain claims of Gulick and Maynard valid and infringed by exhibit I, and granted the usual injunction and order for accounting. The issue as to them is now moot, and our mandates must be recalled and the decrees set aside.
The appellants, however, contend that the decrees should not have provided injunctive relief to the appellee Trust Company on the ground that it is the bare title holder of the patents in suit, with the Aluminum Company of America as exclusive licensee the only party entitled to seek it. They thereby deny our power to grant restraint in respect to claims now found valid and infringed. The master, however, concluded that the Trust Company holds the entire right, title and interest to all patents in suit except an undivided half of the Maynard patent, that it might sue in its own name without joining with it the beneficiaries under the trust, and that the Aluminum Company is a nonexclusive licensee and not a necessary party to the suit. No exception was taken to his findings or conclusions in this respect, and they were not challenged when the case was first before this court. The right of appellees to injunction was no longer in issue when we gave direction for mandate. It is now the law of the case that appellees are entitled to an injunction restraining infringements of any claims of any of the patents in suit now adjudicated as valid and infringed.
A second objection to the decrees, and so impliedly to similar substituted decrees, is that they were not specific. It is urged that they should have recited with particularity the exhibits held to infringe, and those held not to infringe. There is merit to the first phase of the contention since an injunctional order should be as explicit as the case permits. There is no virtue in the second, since decision carries its own vindication of unsuccessfully assailed structures and no affirmative relief by appellants was prayed.
The decrees below will be set aside upon the recall of the mandates. A new decree will be entered in each case adjudicating claims 1, 8 and 11 of Jardine patent valid and infringed by exhibit I. It will provide for an injunction restraining manufacture, use and sale of exhibit I, and all substantial equivalents thereof, and grant the usual accounting for past infringement by exhibit I or its equivalents. It will deny relief as to every other claim of any of the patents in suit, and as against all other accused structures. Costs of the appeals will be divided.
Reversed and remanded for further proceedings in conformity herewith.
Supplemental Opinion.
It appearing that the direction for mandate in our opinion in the above cause, announced at the October session of this court, was, upon petition of the appellees, amended so as to make specific the assailed structures held not to infringe, in accordance with which we amended the ordering clause to read:
"That the structures exemplified by exhibits 8, 9, BBB, CCC, 3-J, and 3-O, do not infringe any of the claims of any of the patents in suit."
And it now being called to our attention that not all of the claims of each of the patents adjudicated were in issue, and
It having been clearly our intention under command of the order of remand made by the Supreme Court, to determine all questions in the case not adjudicated, but to do no more than that; and
It being our view, upon consideration of the response of the appellees to the petition, that the res judicata effect of the present decision upon other litigation now pending should more properly be adjudicated in such pending suits, and that any effort to predetermine the effect of presently ordered decrees upon such suits without consideration of the issues there involved is entirely too speculative for appropriate present adjudication,
Now, therefore, it is hereby ordered that a supplemental mandate be issued in the above cause directed to the District Court, which supplemental mandate shall order that the original mandate heretofore issued, is amended so that the ordering clause therein shall provide in place of what is above recited, the following:
"That the structures exemplified by exhibits 8, 9, BBB, CCC, 3-J, and 3-O, do not infringe any of the claims of any of the patents in suit which they were charged to infringe."