Hemphill Company (hereinafter called Hemphill) filed, in the United States District Court for the Middle District of North Carolina, a civil action against Benjamin H. Jordan, doing business as Jordan Manufacturing Company, alleging infringement by Jordan of two patents owned by Hemphill. The District Court found for Hemphill on all the vital issues, held that the patents in suit were valid and infringed by Jordan, enjoined Jordan from further infringement, and referred the case to a Master to assess damages. Jordan has duly appealed to us.

The patents in suit are No. 1,872,760 (Claims 23, 24 and 25 being here involved) and No. 2,146,750 (Claims 18, 24, 25 and 43 being those with which we are here concerned). The subject matter of these patents relates to circular kniting machines for making striped hosiery. These machines are of the general type considered by our Court in the case of Scott and Williams v. Whisnant, 4 Cir., 126 F.2d 19. Both of the patents in suit have been held valid by the District Court below before the instant suit, in the case of Hemphill Company v. Martin and Smith, Civil Action No. 77. We shall consider separately each of these patents and the attacks thereon made by Jordan.

Claims 23, 24 and 25 of this patent read as follows:

"23. In a type of circular knitting machine adapted to introduce one (202 Figs. 6, 7, 36) or more (96 Fig. 22) lap threads each to a selected group (208, 209, 210 Fig. 35) of knitting needles of the machine in addition to the main yarn or yarns (A, B),

"yarn-engaging means (219) located in substantial proximity to the needle circle to act upon a retiring main yarn or yarns,

"a substantially semi-circular plate-like member (214) circumscribed by the needle circle,

"and having a crescent-shaped finger (216) extending in parallelism with substantially a semi-circle of the needles.

"and along which the lap threads move in the knitting operation,

"and from the end (217) of which finger they are released, substantially remote from the said yarn-engaging means (219).

"24. In a type of circular knitting machine adapted to introduce one (202) or more (96) lap threads each to a selected group (208, 209, 210 Fig. 35) of knitting needles of the machine in addition to the main yarn or yarns (A, B),

"yarn-engaging means (219) located in substantial proximity to the needle circle to act upon a retiring main yarn or yarns (A, B),

"a substantially semi-circular plate-like member (214) circumscribed by the needle circle,

"and having a crescent shaped finger (216) extending in parallelism with substantially a semi-circle of the needles,

"and along which the lap threads move in the knitting operation,

"and from the end (217) of which finger they are released and pass beneath the said plate-like member below the plane of the said yarn engaging means.

"25. In a type of circular knitting machine adapted to introduce one (202) or more (96) lap threads each to a selected group (208, 209, 210) of knitting needles of the machine in addition to the main yarn or yarns (A, B),

"yarn-engaging means (219) located in substantial proximity to the needle circle to act upon a retiring main yarn or yarns (A, B),

"a plate-like member (214) within the needle circle cut away to provide a substantially semi-circular crescent shaped guiding or receiving finger (216) for the lap threads along which the same slide to be released at the end (217) thereof substantially remote from the said yarn-engaging means (219)."

The italicized portions were inserted in the claims by amendment to overcome rejections.

Applications for patents No. 1,702,608 and No. 1,872,760 by Hemphill were copending in the Patent Office. The District Court, we think correctly, held that there was no merit in Jordan's contention that No. 1,872,760 was invalid on the score of double patenting, when compared with No. 1,702,608. This contention was fully aired and considered by the Patent Office, which decided in favor of Hemphill.

On this point, it is not necessary for us to add anything to what was said in the opinion of the District Judge below:

"Patent 1872760 is a continuation of patent 1702608 and the applications were copending in the Patent Office. '760 discloses an improvement over '608 in that the wrap horn and binder plate are both made substantially semi-circular. The end of the wrap horn is brought closely adjacent to the knitting point to give greater assurance that the wrap threads will not go below the latches of the selected needles or into other needles and at the same time maintains control of the tension on the wrap threads until they reach the knitting point and produce a more effective decoration. The binder plate and wrap horn being substantially semi-circular and the end of the horn being brought close to the binder plate facilitates the wrap threads passing beneath the binder plate, which is essential, as they pass off at the end of the wrap horn.

"The disclosures of patent '608 are not prior art to the '760 patent in suit since the applications were copending. Claims 23, 24 and 25 of '760 are specifically directed to a substantially semi-circular horn and claims 23 and 24 to a substantially semi-circular binding plate neither of which is disclosed or claimed in '608; these claims do not read on '608 and there is not double patenting as between them.

"Patent 2146750 disclosed further improvements on the mechanism of patent '608. In circular knitting machines of the type involved here a latch or carrier ring forms a circular wall extending around the upper part of the needle circle, and serves to prevent accidental and unintentional closing of the latches of the knitting needles. Because of the eccentric mounting of the wrap head of the patents in suit and the necessity for locating this wrap hand near enough to the tops of the needles to insure the wrapping of the selected needles, it is necessary to cut away a portion of this latch ring wall. However, during reciprocatory knitting, at which time the wrap head is raised, some of the needles are positioned so that their latches may be closed at such cut-away portion. To prevent this, patent 2146750 in suit provides mechanism by which the raising of the wrap head is accompanied by the automatic moving of an arcuate member or gap closer into position to close the gap made by such cut-away portion in the latch ring, and for automatically removing the gap closer when the wrap head is lowered into operative position."

See, also, O'Reilly v. Morse, 15 How. 62, 122, 133, 134, 14 L. Ed. 601; Traitel Marble Co. v. U. T. Hungerford Brass & Copper Co., 2 Cir., 22 F.2d 259, 262; Thomson-Houston Electric Co. v. Ohio Brass Co., 6 Cir., 80 F. 712, 726, 727.

In an endeavor to show lack of patentable novelty under the prior art, Jordan relies heavily on the Nib-Jack machine, made and sold by Hemphill, as shown in Jones Patent No. 1,718,932. A careful reading of the testimony of Bryant, Hutton and Canavan discloses decided differences between the Nib-Jack machine and the patent in suit.

This becomes readily apparent upon a comparison, in both structure and function, between the guide or guard of Nib-Jack and the wrap finger of Patent No. 1,872,760. Jordan did not successfully refute Hemphill's contention that the purpose of the Nib-Jack guard was that of creating a slot for the reception of a reinforcing yarn when it is not being knit and is not for the purpose of guiding the yarn to the needles; so that this guide serves no function during the introduction or the knitting of the reinforcing yarn. The Nib-Jack machine could not, as the machine in the patent before us could, knit a plurality of true wrap stripes. There is undoubted force in the suggestion that the effect of attaching the accused devices of Jordan to old machines was to enable these machines, with these attachments, then to do what the Nib-Jack was utterly unable to accomplish.

The claims of the patent in suit disclose patentable novelty and disclose such an advance over the prior art as to justify the finding of the District Court in upholding the validity of the patent. These claims read on a mechanism for making true wrap designs (as distinguished from mock stripes) on socks.

This mechanism discloses a latch ring surrounding the needle circle and supporting a binder plate within this needle circle. The binder plate is cut away in order to provide a wrap horn along which, during the knitting operation, the wrap threads move. The wrap yarns pass through a wrap disc with eyelets to be wrapped around the selected needles. This wrap disc rotates once for each revolution of the needle cylinder. Further, the wrap disc is mounted eccentrically in overlapping relation to the needle cylinder, with the result that, on each revolution of the needle cylinder, each eyelet wraps its yarn around a given selected needle or group of needles.

Probably the clearest improvement over the machines in the prior art in the mechanism of the patent before us is found in the shape and location of the wrap horn, which extends around substantially one half of the needle circle, and which has its end positioned close to the knitting point but beyond the yarn feed fingers. Also semicircular in form is the solid portion of the binder plate, so that the end of the wrap horn comes closely adjacent to the edge of the binder plate. During the passage of a yarn through an eyelet, and the consequent wrapping of the yarn around the selected needle or needles, the continued movement of the needles serves to carry the yarn (thus wrapped around the needles) along the wrap horn until the yarn reaches the end of the wrap horn. Then, as the yarn moves off of the end of the wrap horn, the yarn passes beneath that portion of the binder plate on which are supported a binder and cutter. Thus is avoided any entanglement of the yarns with the binder and cutter.

The wrap horn also has other functions. Thus, it confines the wrap yarns to the needles which have been selected, holding the yarns in the selected needles, thus preventing the yarns from getting below the latches of the selected needles and being thrown off. Also the yarns are kept under control upon their approaching the knitting point. Since the yarns are under constant tension, they must be prevented from sliding off the horn too soon, or the yarns might be embedded in the fabric so as to conceal partially the design. Nor can the yarns be pulled around the selected needles so tightly as to cause breaks either in the yarns or in the fabric.

The District Court held (Conclusion of Law No. 2) that the claims in suit were: "* * * infringed by defendant by making parts and converting existing non-wrap knitting machines into wrap stripe machines by applying such parts thereto and by making and selling such parts to others with the knowledge and intent that such parts were to be used by others in such conversion of existing knitting machines."

With this Conclusion we must concur. In support of this Conclusion of Law, we quote from the District Court's Findings of Fact:

"13. Prior to the commencement of the present action, defendant made and sold parts and combinations of parts by which a non-wrap knitting machine could be converted into a machine capable of producing wrap stripes, which such knitting machine could not do before such conversion. In some instances, the defendant had so converted knitting machines in his shop at High Point, North Carolina. Defendant had also supplied sets of such parts with the knowledge and intention that such parts would be used in such conversion of knitting machines.

"14. The Photographs PX's 1 and 2 show parts made and sold by defendant for such conversion of knitting machines into wrap stripe machines. PX-9 is a knitting machine which was converted from a non-wrap knitting machine to its present form of a wrap stripe knitting machine at the defendant's shop in High Point, North Carolina, by persons employed by defendant. The drawings PX-10 illustrate the accused wrap stripe attachment as made from such parts and installed on the machine PX-9.

"15. As illustrated by PX-10, the knitting machine PX-9 was converted into a wrap stripe knitting machine by means of the parts made by the defendant and has a vertical shaft eccentric to the needle cylinder. On this shaft is a disc or wrap head adjacent the needles when the head is in operative position. The shaft is geared one-to-one with the needle cylinder so that threads, fed through the wrap head, will wrap selected needles in identically the manner of the mechanism shown and claimed in the patents in suit. As shown in the photographs PX's 1 and 2 F, a new cutter and binder plate has been substituted for the binder plate originally on the machine before conversion. When installed, this new plate is cut away so that the binder plate and the horn are each substantially a semi-circle with the end of the horn being located close to the knitting point and closely adjacent the edge of the plate.

"16. To close the opening made in the latch ring to accommodate the wrap head disc when it is in operative position, defendant provides a ring member with an opening therein and mechanism by which this opening is automatically located at the opening in the latch ring when the wrap head is in operative position and whereby a solid part of the ring is automatically located at such opening when the wrap head is raised to inoperative position. * * *"

To this, we think little need be added. The claims of the patent in suit read very closely on the accused devices of the defendant.

Both the machine under the patent and the accused devices are "adapted to introduce one or more lap threads each to a selected group of knitting needles of the machine in addition to the main yarn or yarns." In both, we find "yarn-engaging means located in a substantial proximity to the needle circle to act upon a retiring main yarn or yarns," which is the binder located on the binder plate. Both also have "a substantially semi-circular plate-like member circumscribed by the needle circles," and "a crescent-shaped finger extending in parallelism with substantially a semi-circle of the needles." And true of both is it that along this finger "the lap threads move in the knitting operation and from the end of which finger they are released substantially remote from the said yarn engaging means."

We are not impressed by the statement in defendant's brief that charges of contributory infringement are obsolete. What Mr. Justice Douglas said in Mercoid Corporation v. Mid-Continent Investment Co., 320 U.S. 661, 669, 64 S. Ct. 268, 273, 88 L. Ed. 376, was: "The result of this decision, together with those which have preceded it, is to limit substantially the doctrine of contributory infringement. What residuum may be left we need not stop to consider."

Our own views on that subject were very clearly expressed by Chief Judge Parker, in Florence-Mayo Nuway Co. v. Hardy, 4 Cir., 168 F.2d 778, 785: "There is nothing to the contrary in the decision in Mercoid Corporation v. Mid-Continent Investment Co., 320 U.S. 661, 64 S. Ct. 268, 88 L. Ed. 376. That case merely applies the salutary rule that a combination patent may not be used to protect an unpatented part from competition. Nothing of that sort is involved here. What the defendants are doing is manufacturing and selling machinery with the knowledge, purpose and intent that it shall be used in a combination which will infringe. The case is one where plaintiff is using his patent not to monopolize the sale of what is not patented but to prevent defendants from aiding others to infringe what is patented. A clearer case of contributory infringement could not well be imagined than that presented by this record."

We affirm, then, the conclusions of the District Court that the claims in suit of Patent No. 1,872,760 are valid, are infringed by the accused devices of defendant and we uphold the reference to a Master for the assessment of damages. Since this patent has now expired, however, there is no need for an injunction to prevent defendant from further infringement.

A careful study of the record and exhibits convinces us that this patent is invalid on the score of prior public use more than two years before the filing of the application for the patent. 35 U.S.C.A. §§ 31, 69. True it is, as Hemphill contends, that to invalidate a patent on the score of prior public use, a high degree of proof is required. The Barbed Wire Patent (Washburn & Moen Mfg. Co. v. Beat 'Em All Barb-Wire Co.), 143 U.S. 275, 284, 12 S. Ct. 443, 36 L. Ed. 154; United Shoe Machinery Corporation v. Brooklyn Wood Heel Corporation, 2 Cir., F.2d 263, 264. But, when the defendant meets this standard of proof (as we think he has done here), the courts have not been slow to declare the patent invalid. Said Mr. Justice Woods, nearly seventy years ago, in Egbert v. Lippmann, 104 U.S. 333, 336, 26 L. Ed. 755: "We remark, secondly, that, whether the use of an invention is public or private does not necessarily depend upon the number of persons to whom its use is known. If an inventor, having made his device, gives or sells it to another, to be used by the donee or vendee, without limitation or restriction, or injunction of secrecy, and it is so used, such use is public, even though the use and knowledge of the use may be confined to one person."

See, also, Hall v. Macneale, 107 U.S. 90, 96, 2 S. Ct. 73, 27 L. Ed. 367; Thomson-Houston Electric Co. v. Lorain Steel Co., C. C., 110 F. 654. Significant in this connection is not only what did appear in the record and exhibits but also what failed to appear, particularly in connection with the testimony of Hutton, President of Hemphill, who had been associated with that company for thirty-two years.

Hutton testified that about 1927 Hemphill had a knitting machine with a latch ring containing a recess other than the throat and provided with a means for closing that gap. This gap-closer was a segment pivoted at one point and thrown in and out of action to protect the latches of the needles. The additional recess was for wrap-stripe knitting. There was also an auxiliary yarn feeder movable relatively to the latch rings. "A very limited number" of these machines were made and operated to produce fabrics or hosiery in 1927, according to Hutton, who characterized the machines as "experimental models." However, whether or not the use of a machine is or is not purely "experimental," is not to be determined solely on a mere characterization as such, particularly when this characterization comes from a deeply interested witness.

Hutton proceeded to say that in 1931 machines of this type were exhibited to customers "to sell them machines," that they were in a saleable condition, and that some were sold and shipped to customers in that year. He testified also that between ten and fifty of such machines were made during 1931.

The gap-closers as described by Hutton were similar (though slightly different) to those covered by the patent here in suit. We are here concerned with whether Hutton's statements as to the prior use of these machines with gap-closers can be corroborated by other evidence and the effect of the subsequent retraction by Hutton of his statements concerning the prior sales of these machines.

As to corroboration, one important question to be considered is whether machine #86062, which Hemphill demonstrated to the Holeproof Hosiery Company prior to April 20, 1931, had a gap-closer and was thus one of the machines referred to by Hutton. The machine was designated in Hemphill's letters to Holeproof as a "latest model wrap stripe machine * * * fitted with 18 wrap yarn feeds and 3 color clocking spindle" and "other improvements." An advance sample was sent to Holeproof on June 4, 1931, and the machine described as a "latest improved machine HS & DS color feeds" was shipped to Holeproof on July 25, 1931.

Hemphill claims that this machine had no gap-closer. It relies on a letter of March 11, 1932, from Hemphill to Holeproof, which stated that Hemphill had shipped parts, among them being a carrier ring assembly (which included a gap closer), to Holeproof, to bring machine 86062 up to date, and an answering letter to the effect that old parts removed from the machine were being returned. Hemphill claims that the gap-closer found on the machine was the one thus referred to in these letters, and that previously no gap-closer was on the machine. But there is no other evidence that any changes had been made in the carrier ring (containing the gap-closer) and no evidence that the old carrier ring was actually returned.

Hemphill maintains that the first machines with a gap-closer sold by it were the machines sold to the Huntley-Jackson Company on an order dated March 21, 1932. Hutton testified that the gap-closer of the 1931 machines had a slightly different shape from that of patent No. 2,146,750. The Holeproof machine 86062 was found to have also a slightly different gap-closer from the Hemphill standard mechanism. Drawings made in 1931 show a gap-closer slightly different from the later machines, such as one sold to Huntley-Jackson in March, 1932. This would indicate that the alleged original gap-closer was still on the machine 86062. The drawing made in 1931 showed a gap-closer with the words "Reinf. vert. stripe mach." on it, and on the order for Holeproof machine 86062 Hemphill wrote "Style machine — Reinforced striper;" while on the Huntley-Jackson order of March, 1932, was the notation "Style machine — latest type" and on the accompanying drawing "Wrap stripe machine part * * *." From the similarity in these descriptions, it would seem fair for us to deduce that the drawing of 1931 was of the machine shipped to Holeproof.

Ward, who installed machines for Huntley-Jackson, testified that prior to 1932 he had never seen a Hemphill wrap-stripe machine with a gap-closer. Hemphill had received an order for a 12-color feed machine (described in the same general language, for the most part, as 86062) four days after the receipt of Holeproof's order, and the Huntley-Jackson machine was shipped one day before the Holeproof machine. This Huntley-Jackson machine of 1931 probably had no gap-closer; but it was not specified as a "latest improved machine" and it was 12-color instead of 18-color.

Hemphill also claims that there is no proof that the "18-color" machines required gap-closers — that the necessity for a gap-closer does not depend on the number of colors but on the depth of the recess in the latch ring. Jordan claims the opposite. Canavan testified that the deeper the latch recess, the greater is the need for a gap-closer but he did not mention colors. Hutton was unable to say how many color feeds were provided on machines that did not have gap-closers; he testified, however, that a deeper groove was necessary with a larger number of feeds and that this necessitated a gap-closer. Thus the testimony is not inconsistent. It is worthy of note that 18-color machines have more feeds than any other machines referred to in the record. Hutton testified further that all of the 18-color machines made prior to 1934 were of the type shown by PX-11 and shipped to Huntley-Jackson in March, 1932. These machines contained gap-closers. There is no justification in the record for Hemphill's argument that Hutton did not mean all of the "18-color machines" prior to 1934 required gap-closers but that merely some of them required gap-closers.

On January 30, 1932, Hemphill sent Holeproof six attachments for converting "8 color wrap machines to 18 color wrap." It would appear from Hemphill's Assembly Sketch of January 30, 1932, and from testimony previously referred to, that these assemblies contained gap-closers.

On the order were pencilled notations of parts list numbers. One of these lists included a gap-closer. But the components of these lists were sometimes changed, though the same list number was retained, and the lists were not dated. Further, it seems probable that the parts were not assembled prior to February 3, 1932, within two years of the application for Patent No. 2,146,750.

A letter from Hemphill to Holeproof, dated January 30, 1932, explained the changes made in the new assemblies but no reference was made to gap-closers which, if a new development unfamiliar to Holeproof, would certainly have been mentioned. And a sketch of a gap-closer and carrier ring, also dated January 30, 1932, was found. The letter stated that a blue-print was enclosed. This is strong evidence that these assemblies had gap-closers. This sketch discloses that the gap-closers were of the old type as shown in the 1931 sketch and as described by Hutton.

In its letter of March 4, 1932, Holeproof indicated its familiarity with gap-closers, referring to the "new type of machines you are providing with a gap-closer." In Hemphill's letter to Holeproof dated January 30, 1932, we find the statement: "You will note the carrier ring construction and the general principal of this mechanism is more compact and of better construction than the machine that you have at your plant (86062)." Machine 86062 has a "spacing block * * * in connection with the gap ring or latch ring" which is not found on Hemphill's standard mechanism (as covered by Patent No. 2,146,750). The omission of this block would make the mechanism "more compact and of better construction."

We now consider the extent of Hutton's recantation of his original testimony as heretofore set forth. Hutton was asked if he had found anything in the Hemphill Company's records to show that a machine such as PX-11 (the later type as sold to Huntley-Jackson in March, 1932) was sold by Hemphill prior to March 31, 1932, and he answered in the negative. Then he was asked if he had any knowledge of the demonstration of a machine similar to that represented by PX-11 prior to February 3, 1932, and he replied: "No, sir, there was none prior to that date." But on cross-examination, Hutton stated that the testimony given earlier was true and correct except as he had attempted to modify it. He testified further that the testimony was based on the Company's records.

There was evidence that Hemphill destroyed its records at periodic intervals. Hutton instructed his secretary to search the records but requested her to give him only those records of equipment shipped to Huntley-Jackson during the years 1931 and 1932. Thus his testimony does not show that no gap-closers were shipped by Hemphill in 1931. As to his lack of knowledge of the demonstration of these machines, there is a fairly justifiable inference, in the light of the previous questioning, that Hutton meant that he had found nothing in the fragmentary search of the records of the Company to show any such demonstration. Also at this time Hutton stated that he had searched the records without finding any drawings of these parts (gap-closers), yet subsequently such drawings were found, after the Holeproof machine 86062 was located by Jordan. And this was after Hutton had been expressly asked to produce the records as to any shipment of such machines to Holeproof. Hemphill produced no drawing until after the existence of the Holeproof machine was discovered. Hutton, although a skilled draftsman, refused to make a sketch showing the differences between the 1931 gap-closers and the later type of gap-closers.

Hutton was questioned, on his second deposition, only about the type of machine shipped to Huntley-Jackson in 1932 and represented by PX-11. It would appear that Hutton had become purposely evasive and was seeking to limit his answers as much as possible. Hutton also testified that the machines sold Huntley-Jackson were Model 9, whereas the first machine which had gap-closers, he stated, was a Model 12 or 14.

From all the evidence it definitely appears that Hemphill made and sold machines with gap-closers prior to 1931, that the machine 86062 originally had a gap-closer, and that Hutton's first damaging admission was not effectively repudiated. It seems highly improbable that Hemphill, having perfected the gap-closer device in 1927, would have waited until 1932 to use it commercially, for the evidence clearly indicates that this device was very important and quite necessary for the efficient operation of the machine.

We therefore conclude that the defence of prior public use as to this patent was sustained and that the District Court erroneously held otherwise. Accordingly, this patent must be declared invalid. Our holding that prior public use invalidated this patent renders it unnecessary for us to discuss, as to this patent, any other contentions.

The decree of the District Court as to Patent No. 1,872,760 is affirmed in so far as it holds this patent valid and infringed by defendant and refers the case to a Master for the determination of damages; but, since this patent has now expired, there is no necessity for an injunction against further infringement. As to Patent No. 2,146,750, the decree of the District Court is reversed and the court below is instructed to enter a decree declaring this patent invalid on the score of prior public use.

Affirmed in part; reversed in part.