December 3, 1928. Rehearing Denied December 22, 1928.
Appeal from the District Court of the United States for the Southern District of New York.
Suit by the Peerless Roll Leaf Company, Inc., against the H. Griffin Sons Company. From a decree holding valid and infringed claims 19 and 20 of patent No. 1,526,209, issued to W.F. Grupe on February 10, 1925, defendant appeals. Affirmed as to claim 19; reversed as to claim 20.
The purpose of the patented machine is to feed metallized transfer strips of paper to a stamping press for use in printing titles upon book covers. There had been other such machines before, but the patent in suit was granted for the improvements disclosed by Grupe. As it is necessary to print several separate titles upon a book cover, as, for example, the title proper, the author's name, the publisher's, and often more, separate strips of paper are more convenient and economical for each title. Moreover, since the line printed for each title is of different length, the advance of each strip after an impression should vary with the length of the line. The art had already succeeded in accomplishing these results, but by different means from those disclosed here.
Grupe's machine has a pair of rolls for each strip, an idler and a feed roll, which grasp the strip between them and advance it over a length determined by the rotation of the feed roll. Each pair of rolls extends across the whole width of an open frame, which holds them, and all the feed rolls are actuated by a single shaft, and set in motion at the same moment by the operation of the press when the proper point in its cycle arrives. This is accomplished by a complicated mechanism, which it would be impossible to describe intelligibly without the use of figures, and whose detailed description is not in any case necessary to an understanding of the controversy. It is enough to say that the single shaft, just mentioned, controls a mechanism which gives a unidirectional rotation to each feed roll, though it is itself oscillated by means of a rack and gear; the rack being at the end of a pitman operated by the movement of the press. Part of the mechanism for transferring the oscillating movement of the shaft into unidirectional rotation of each feed roll is a ratchet clutch, which unites the roll and a concentric shaft when the shaft rotates in one direction, and sets it free when it rotates in the other. The extent of rotation of each feed rod is prescribed by a dial at one end of the roll, by means of a contrivance which throws it out of gear after the proper length of strip has been fed. Each roll starts from rest and rotates at the same velocity, but through a different arc.
The defendant's machine is made substantially after a patent to one Richard Lange, and departs in a number of respects from the patent in suit. It has, however, two rolls for each strip, extending across an open frame, one an idler, the other a feed roll, which grip between them a single strip and feed it for a predetermined length, different for each strip, and of course in one direction only. Each feed roll has a shaft within it so connected by a ratchet clutch, that, when the shaft rotates in the direction of the strip's advance, the roll locks with it. When the shaft is rotated in the opposite direction, the clutch unlocks and the roll is idle. In this it is similar to Grupe's feed roll. The movement of the shafts, is however, differently accomplished. A cord is wound upon a drum at the end of each, and when the cord is unwound the shaft rotates against the torsion of a spring within the feed roll. During this phase the shaft and feed roll are unlocked. When the cord is released, the spring rotates the shaft backwards, the clutch locks it to the feed roll, the strip is advanced, and the cord is rewound. A governor retards the rotation of the shaft on its return. The other end of the cord is fastened to a radial arm itself set upon a shaft, actuated through gears by a motor-driven shaft, which is driven always in one direction.
Claim 19 in suit reads as follows:
"A feeding mechanism for stamping and other impression machines provided with common die-holding means, comprising means for supporting a plurality of strips of impression material and means for feeding said plurality of strips respectively to said common die-holding means to regulatable extents, said feeding means comprising an open frame, a plurality of feeding members, said feeding members being rotatably mounted in said open frame, each of said feeding members being substantially cylindrical and extending substantially from one side of said open frame to the opposite side thereof and means for individually regulating the extents of feeding movement of said feeding members, said feeding means further comprising ratchet means respectively for said plurality of feeding members."
Claim 20 is the same, with the addition of the clause: "And common means driven from impression machine for actuating said feeding means."
The defendant offered various patents in evidence, but it is not necessary to describe them in detail. While they disclosed machines for advancing several strips for different predetermined lengths, in none was each strip moved by a separate set of rolls extending across the whole face of an open frame, each feeding means consisting in part of a ratchet. The defendant did not put them forward as anticipations of the claims in suit, nor are they mechanically near enough to Grupe's disclosure to be dangerous as anticipations.
The defendant did, however, prove the prior use by the plaintiff of three machines alleged to embody the invention at a period more than two years earlier than March 31, 1924, the date of the application. The circumstances of this use were as follows. The plaintiff was a subsidiary company to the Boucher Cork Machine Company, as also were the Ebcoe Machine Company and the Sterling Company. Before March, 1922, the Ebcoe Company built three machines, which were turned over to the Sterling Company, which used them commercially during 1921 and 1922. This use is established by sufficient evidence, as the plaintiff does not dispute. They were not, however, sold, nor were they displayed to any persons who came to the factory. Only three employees were admitted to the room where they were kept, one Mangold, who was in charge of them, with the aid first of one, and later of two, assistants. The plaintiff was "right secretive about" them, "partitioned the machines off in a little room on the floor," and kept them secret "by isolating them." Access to them was possible only by crossing a bridge between two buildings. Moreover, the Sterling Company was an experimental department of the Boucher Company, and it was to test out the product that the Sterling Company used them. On the other hand, they were used to ornament paper boxes and the like, which were shown freely and sold in quantity. Customers for the product were not, however, shown the machines themselves.
Samuel E. Darby, Jr., and Darby Darby, all of New York City, for appellant.
Henry J. Lucke, of New York City, for appellee.
Before MANTON, L. HAND, and SWAN, Circuit Judges.
The most important issue, as we see it, is that of the public use by the plaintiff (for all the companies are to be taken as one for this purpose) before March 31, 1922. The machines then used were, we think, complete embodiments of the claims in suit. That question turns upon what is meant by the phrase, in claims 19 and 20, "ratchet means respectively for said plurality of feeding means." If this means the gear and rack, 29, 33, of the patent in suit, the three machines did not embody it, because they had a separate rack and gear for each feed roll, and thus dispensed with the elaborate mechanism by which the single shaft, 32, drove all the feed rolls together. But the claims do not admit of such an interpretation, because it would require elimination of the word "respectively," which can only mean that each roll has its own "ratchet means." If we do not understand the element referred to as the rack and gear, we must take it to be the clutch, and that we think it is, though the plaintiff does not apparently so understand it. While the specifications always speak of that member as a "clutch," Lange in his own specifications calls its exact equivalent a "ratchet clutch," and a ratchet in fact it is. While the word is used in mechanics interchangeably with "rack," when the teeth are upon a straight bar, it is strange if the member, 29, was intended, that it was not given the only name used for it in the specifications. Whatever doubt there might be is, however, laid by considering other claims not here in suit. It is clear that claims 19, 20, and 21 were intended in general to cover the same combination as claims 16, 17, and 18, which are precisely alike, except that "clutch" is used for "ratchet." The scrivener apparently feared that "clutch" might not be comprehensive enough, and wished to cover any "ratchet" which might not answer the strict definition. The ensuing duplication of claims, an abuse against which courts have again and again vainly protested, cannot disguise the intent of the words in their context, and we hold that the machines embodied claims 19 and 20, though we cannot see why claims 16 and 17, originally in suit, would not have done quite as well.
So it becomes necessary to decide whether the use was "public," within the meaning of Revised Statutes, § 4886 (35 USCA § 31). We think it reasonably clear that the machines were intended to be kept from the knowledge of the public in general; there could have been no other reason for isolating them in a separate room and allowing access to only three employees. It is true that these three were not expressly pledged to secrecy, and this has at times been regarded as important, Bliss Co. v. Southern Can Co. (D.C.) 251 F. 903, 909; and indeed it was doubted in Perkins v. Nashua, etc., Co. (C.C.) 2 F. 451, whether even an express pledge would serve. It seems to us that a pledge is only one circumstance, and that the question is to be decided by considering whether in fact the use was such as to impart knowledge to the trade at large. We held in Schrader v. Wein Sales Corporation (C.C.A.) 9 F.2d 306, that, when only the inventor and an assistant had access to the machine, no pledge need be proved; it was implicit in the situation. In Grasselli Chemical Co. v. National, etc., Co. (C.C.A.) 26 F.2d 306, 309, we declined to say whether a process carried on with obvious secrecy might not be public when a number of employees had knowledge of it, and showed that they did not feel bound to preserve the secret after they left the owner's employ. Brush v. Condit, 132 U.S. 39, 10 S. Ct. 1, 33 L. Ed. 251, concerned a use obviously public, like American Ballast Co. v. Davy, etc., Co., 220 F. 887 (C.C.A. 7), and Wendell v. American Laundry Machinery Co., 248 F. 698 (C.C.A. 3). In Smith, etc., Co. v. Sprague, 123 U.S. 249, 8 S. Ct. 122, 31 L. Ed. 141, not only workmen generally, but customers who bought the product, had access to the machine; there was no effort at secrecy, except as against those who "we thought were manufacturing buckles" — its product.
The cases nearly always also raise the question whether the invention is in process of experiment or testing, and, while this honestly continues, the publicity of the use is not important. Egbert v. Lippmann, 104 U.S. 333, 336, 26 L. Ed. 755; Elizabeth v. Pavement Co., 97 U.S. 126, 134, 24 L. Ed. 1000. It is probable that the use at bar was, at least in part, and perhaps altogether, of this kind, but we prefer not to strain that aspect of the evidence. The word "public" means more than that the use shall not be experimental; it is possible, even after the time of experiment has passed, to practice the invention without abandonment. An inventor working by himself presents a simple case, but we cannot suppose that his rights are greater than those of a company. There must be some way by which they can privately practice their inventions as well as he, and yet any practice involves the knowledge of some employees. When the number of these is limited to as few as are necessary to practice it at all, when customers and the public generally are excluded, and adequate precautions are taken to prevent dispersion of the knowledge until at least two years before application is made, it seems to us enough, whether a formal pledge of secrecy be exacted or not. Those workmen who will keep it do not need it; those who will not, it will not bind. We do not mean that, if in fact knowledge leaks out, the invention may not be lost; as to that we say nothing. But when, as here, adequate means are taken to confine all information as closely as is consistent with any exploitation at all, and when, so far as appears, they are successful, the knowledge of the necessary workmen not explicitly pledged to secrecy does not make the use public. The sale of the product was irrelevant, since no knowledge could possibly be acquired of the machine in that way. In this respect a machine differs from a process, Grasselli, etc., Co. v. National, etc., Co., 26 F.2d 306 (C.C.A. 2); or from any other invention necessarily contained in a product, Hall v. Macneale, 107 U.S. 90, 2 S. Ct. 73, 27 L. Ed. 367.
This issue being disposed of, the other points appear to us simple. The art had indeed made many efforts towards the same result, and they were successful, but it had not used the combination culled out in the claims in suit. The machines are very complicated, and it is idle to argue that their contrivance did not take a high degree of ingenuity; indeed, the defendant does not so argue. It relies for noninfringement upon the mechanical differences between its machine and that disclosed. But while there are, indeed, important differences, the infringement contains all the elements of claim 19, functionally as well as verbally; that is, if by the "ratchet means" the claim intends the clutch in the feed roll. This combination was new, and Lange saw the three machines which embodied it, while he was employed by the plaintiff; apparently he was not excluded. Indeed, he at one time supposed that the invention was his own. Whatever improvements he made were his; but he must submit to Grupe's priority in respect of what he did not invent.
We think, however, that claim 20 was not infringed. The defendant has no means driven by the press for actuating the feed rolls. That means is directly driven by an independent motor, and Lange gained nothing from the patent as respects this feature.
Decree affirmed as to claim 19; reversed as to claim 20.
On Petition for Rehearing.
The petition depends upon the supposed proof that the machines were shown to customers before March 31, 1922, the effective date. The defendant's only possible reliance is that part of Mangold's testimony with which the record closes, a matter so much discussed in the briefs and on the argument that it seemed impossible to us that counsel should have supposed we had overlooked it. The defendant is so insistent as to justify some further discussion. Grupe had sworn that the machines had been kept isolated by partitioning them off without any name on the door of the room where they were kept, and that Mangold and two assistants were the only workmen employed upon them; these steps were taken to keep them secret. Hayes' testimony is clearly valueless. The defendant then recalled Mangold, who swore that the sales manager of the plaintiff brought customers "to demonstrate and sell them" the machines during the year 1921, although there were no sales until 1925 according to Grupe. Upon cross-examination he added that the manager's purpose was to show them "roll leaf and machinery, * * * to show them that it could be done." That his only purpose was "to show the use of Peerless roll leaf, * * * and to show what it could do in making box tops."
We thought, and still think, that this single sentence of Mangold, which was certainly wrong in part, is too slight to overcome the general evidence that the machines were kept from the public. Indeed, just what he meant is not very clear; he may have meant that the machines themselves as well as their product were shown, but the matter was left in the air. Supposing he did, this was not enough to prove public use before March 31, 1922. The date at any rate rests wholly upon his faulty recollection, and he remained with the plaintiff until October, 1922. The defendant had the burden of a severe degree of proof, and at best went no further than to show that customers might have been shown the machines before the critical date. This, read with the testimony of Grupe, will not answer.