Opinion
No. 8010.
June 15, 1956.
Julius G. Maurer, J. Wesley Everett, and Wendell D. Allen, Baltimore, Md., for plaintiff.
George B. Finnegan, Jr., Granville M. Pine, and Morgan, Finnegan, Durham Pine, New York City, and Piper Marbury, Baltimore, Md., for defendants.
The principal issue in this case is whether a suture package, manufactured by defendant Ethicon during the years 1951 to 1954, infringes any of the claims in two patents issued to plaintiff in 1940., In brief, these claims embrace the combination in a suture package of (1) a conical bobbin, (2) a twisted suture strand wound helically in a single layer thereon, with each convolution around the bobbin in closely adjacent relationship, and (3) with either (a) an additional twist administered to the strand during winding, or (b) means at the base of the conical bobbin to receive and secure the strand, so that the suture is supported throughout in the form of a conical coil, which may be readily removed over the apex of the bobbin, or (c) a split in the conical bobbin, to make it radially yieldable under contraction and expansion of the strand coiled thereon. The accused package has a suture wound loosely about a pair of flat fibre tabs or arms extending diagonally upward from points near the ends of a flat fibre base, with no additional twist administered during the winding, with no means anywhere to secure the trailing end of the strand, and with no regular relationship between the successive coils and the tabs, or among themselves, as they lie loosely in a more or less random pile of elongated loops. For the reasons set out below, I find no infringement.
The word "suture" is used throughout this opinion to include ligature.
Plaintiff relies upon the three claims set out verbatim in this note. Their essential elements are itemized and discussed in the opinion.
Patent 2,225,833, claim 1: "In a suture package, a surgical suture and supporting bobbin therefor, the bobbin being generally of conical formation, and the suture comprising a twisted strand wound helically in a single layer thereon and twisted in the direction of the twist while being wound with each convolution about the bobbin in closely adjacent relationship and in which the twist in the suture about its longitudinal axis is uniform about the bobbin from end to end thereof as a result of said last mentioned twist and maintaining a continuous smooth surface throughout the strand."
Patent 2,225,833, claim 2: "In a suture package, a surgical suture and supporting bobbin therefor, the bobbin being generally of conical formation, and the suture comprising a twisted strand wound helically in a single layer on the conical body thereof with each convolution about the body in closely adjacent relationship and with a uniform twist in the suture strand from end to end thereof and a continuous smooth surface throughout, the bobbin having means at its base end to receive and secure said strand, whereby the suture is supported throughout in the form of a conical coil which may be readily removed over the apex of the bobbin without kinking or bending of the strand or the loss of twist therein."
Patent 2,225,834, claim 3: "In a suture package, a bobbin having a twisted suture strand wound helically thereon in a single layer with each convolution about the bobbin in closely adjacent relationship and with a uniform twist in the suture strand from end to end thereof and a continuous smooth surface throughout, said bobbin having a longitudinally divided conical strand supporting body, and said body being radially yieldable under contraction and expansion of said strand to maintain substantially uniform tension throughout the strand coiled thereon, whereby twist in said strand will not be caused to open while supported on or during removal from said bobbin.
Findings of Fact.
1. The suture package in general use from 1930 to the present consists of a twisted strand, varying in length from 12 to 60 inches, made of catgut, silk, or other suitable material, wound by hand either (a) in a concentrated coil, or (b) upon a relatively stiff reel, shaped like an elongated "H", with the crossbar almost as thick as the length of the side arms, submerged in a preserving solution in a sealed transparent container, usually a glass tube, which may be broken on a predetermined line to release the suture for use. In the case at bar, we are dealing primarily with strands of catgut, i.e. the dried and twisted intestines of sheep. Such a strand is customarily wound by hand onto a reel, which is inserted into a glass tube; it is then dried and sterilized with dry heat, which generally causes a contraction in the length but not the diameter of the twisted strand. A preserving solution, usually 90 to 95% alcohol, is then injected into the tube, which is sealed off to produce a package approximately four inches in length. The small amount of water in the preserving solution causes the strand to soften and swell transversely; this tends to open up the twist and thereby to weaken the strand. That tendency is accentuated if the strand has been bent sharply while being wound onto the reel, as often happens.
2. Dr. Horine was concerned about this weakening, and conducted many experiments to determine the best method of overcoming it. Any winding of a twisted strand in coils or convolutions around a reel, either by passing the strand round and round, or by turning the reel round and round, tends to open up the twist in the strand on the outer surface of the strand in each of the loops. Moreover, if any strand is passed round and round a reel with a right throw, or with a left throw, a twist is necessarily applied to the strand if the ends of the strand are held firmly; this twist will tend to accentuate or to mitigate the opening of the twist in a twisted strand, depending upon the direction of the new throw or twist and the direction of the existing twist in the strand itself, if no means of offsetting the additional twist is provided. Different effects are produced if the reel itself is turned round and round, clockwise or counter-clockwise, and if the end of the strand is held tightly, loosely, or not at all.
3. Dr. Horine concluded that the best results can be obtained by winding twisted catgut strands helically in a single layer on a conical bobbin with each convolution about the bobbin in closely adjacent relationship, and by twisting the strand, while it is being wound, in the direction of the twist in the strand itself, which is to the right in commercially produced catgut. These means, in combination, prevent or tend to prevent the twist in the strand from opening up while it is supported on the bobbin or while it is being removed from the bobbin. Dr. Horine's later experiments indicated that it was desirable to split the conical bobbin to make it more resilient, so that it could offset the longitudinal shrinking and subsequent transverse swelling of the catgut in the tube. These means, in combination, are the best means now known to accomplish the desired purpose, but obviously require a relatively expensive bobbin and great care in the winding process.
4. The preparation of animal intestines, principally of sheep, first by the ancients for musical strings and later for surgical thread, is a very old art. The intestine is split into a number of flat strips, which are scraped, chemically treated, and twisted, either as single strips or two or more strips together, to form the desired strand. This strand is then polished and otherwise finished to achieve smoothness, strength and a uniform diameter.
5. Plaintiff's patents do not deal with any part of that process. Plaintiff's concern is to wind these twisted gut sutures in such a manner and on such a reel or bobbin as will eliminate the untwisting and the weakening of the strands which may be caused by sharp bends in the wound and packaged suture. This concept is not new; it was reviewed in detail in the patent to Themak, No. 2,081,389, issued in 1937. Themak chose to wrap his strand in a helical coil about a flat rectangular fibre holder, with the ends of the suture secured to the holder. Themak also taught that the direction of the turn of the coil should be the same as the direction of the twist of the suture itself.
6. A number of prior patents show the winding of suture strands about cylindrical surfaces in closely adjacent coils; e.g. Hollister No. 749,934, Johnson No. 799,166, and Pompilian French patent No. 475,953. The prior art also included: the patent to Kaysan No. 517,574 for a surgical ligature holder in the shape of a U, tapered toward the open end, with the strands wound in a figure 8 around the two arms; the patent to Davis, No. 1,123,497, which shows a suture strand wound in a spiral, circular coil about the extending arms of a flat, cardboard holder; and the patents to Lukens No. 1,234,051, to Hirsch No. 1,610,607, and to Randall No. 1,914,836, each of which shows the winding of sutures about the long axis of a flat, rectangular holder, notched at the ends to keep the strands from slipping off. Means for securing the end of the strand are shown by Davis, Themak and others.
7. The German patent No. 637,140, issued to Brinkmann on October 22, 1936, and the patent of addition, No. 638,953, issued to him on November 26, 1936, show a suture wound in helical coils side by side in a single layer on the surface of a cone. The patent of addition to Brinkmann and the early Kaysan patent (1894) show means of removing the suture over the apex of the conical or tapered bobbin.
8. Dr. Horine began his experiments in the late 1920's; he continued them until he became ill in 1934, and resumed them after his recovery in 1935. Dr. Horine testified during the trial proper that he determined as early as 1933 that material wound on a bobbin of general conical shape with progressive enlargement of the lengths was the best material, that he then "proved his point" by continued experiments, and that after he had proved his point and had an independent laboratory check his results, he applied for a patent. After the close of the evidence and after argument of counsel, he testified that in 1933 he wound the catgut on the cone, dried it in an oven, sterilized it at 325 degrees Fahrenheit, put it in ninety-five percent alcohol in a glass tube, sealed with glass or cork stopping to prevent the alcohol from evaporating. The witness Harne, a demonstrator technician at the University of Maryland Medical School, testified that he made cones for Dr. Horine in 1933; that he made the cones in his shop, and Dr. Horine took them home and split them. Dr. Horine, however, testified that no split cones were made until around 1938. Unfortunately, his records of various tests he had made could not be found, and the only reports of independent tests offered in evidence were those made by Penniman Browne, chemical engineers and analysts, which were dated 1939, after the patent applications were filed, but before plaintiff published his results in medical journals. These dates are important only on the question whether the German patents issued to Brinkmann on October 22, 1936, and November 26, 1936, may be considered on the question of the validity of plaintiff's first patent (solid cone bobbin), the application for which was filed on March 26, 1938.
9. Plaintiff filed his second application (split cone bobbin) on April 4, 1939. His original claims in both applications were broader than those finally allowed, and were rejected on the basis of a number of prior United States patents; the Brinkmann patents were not cited by the Patent Office. The two patents upon which plaintiff now relies were issued on December 24, 1940.
Claim 1 of issued patent 2,225,833 was claim 8 of the patent application, which originally read as follows: "In a suture package, a surgical suture and supporting bobbin therefor, the bobbin being generally of conical formation, and the suture comprising a twisted strand wound helically in a single layer thereon with each coil about the bobbin in closely adjacent relationship and in which the twist in the suture about its longitudinal axis is increased on each turn about the bobbin toward the apex thereof."
Claim 2 was claim 9 of the application, and supplanted claim 7, which, as last amended before cancellation, read: "In a suture package, a surgical suture and bobbin, the bobbin being tapered in the direction of its length, and the suture being wound helically thereon transversely to the direction of the taper with the consecutive turns closely adjacent, the suture being removable in the form of a spiral without unwinding, the bobbin having means for removably securing the ends of the suture."
A third patent No. 2,249,289, issued to plaintiff on March 17, 1939, was referred to in the complaint, and a suture package containing a paper reel manufactured by Ethicon was accused, but all charges based upon those packages have been abandoned.
10. Plaintiff relies upon three claims — Nos. 1 and 2 in Patent No. 2,225,833, his first patent, filed March 26, 1938, and No. 3 in Patent No. 2,225,834, his second patent, filed April 4, 1939. The parties are agreed that all three of these claims include, as necessary elements, the combination in a suture package of
(1) a conical bobbin, and
(2) a twisted strand wound (a) helically, (b) in a single layer thereon, (c) with each convolution about the bobbin in closely adjacent relationship; and in addition —
(3 — in claim 1 of the first patent) the strand being twisted in the direction of the twist while being wound, so that the twist in the suture about its longitudinal axis is uniform about the bobbin from end to end as a result of such additional twist;
(3 — in claim 2 of the first patent) no additional twist while winding, but (a) the bobbin having means at its base end to receive and secure said strand, whereby the suture is supported throughout in the form of a conical coil (b) which may be readily removed over the apex of the bobbin without kinking or binding of the strand or the loss of twist therein;
(3 — in claim 3 of the second patent) the body of the conical bobbin (a) being split longitudinally, and (b) being radially yieldable under contraction and expansion of the strand to maintain substantially uniform tension throughout the strand coiled thereon, whereby the twist in said strand will not be caused to open while supported on or during removal from the bobbin.
11. In 1938-39 plaintiff sent models of his suture package to Johnson Johnson in an attempt to interest that company in its manufacture. His offer was refused and the models returned. Defendant Ethicon is a wholly owned subsidiary of Johnson Johnson. Plaintiff also sent models to Davis Geck, a large suture manufacturer, but they were not interested. In 1941 plaintiff entered into a license agreement with Bauer Black covering all three of his patents. That firm made and distributed on a test basis a number of such suture packages, but finally decided not to put the package into commercial production, and cancelled the license agreement.
12. Until 1951, all sutures were wound by hand. Defendant Ethicon, however, was anxious to develop a winding machine, and engineers employed by it or by its parent company engaged in a number of experiments. They found it was impractical to develop a machine to wind sutures on the conventional H-reel, and in 1951 Schaefer, an employee of Johnson Johnson, developed a reel suitable for use on the winding head of a machine. Schaefer's reel is a flat piece of fibre board with a pair of flat, inclined arms extending diagonally upward from points near the ends of the base.
13. The original winding head was developed by Schaefer, Bilane, and Barry for use on a winding machine which they developed and patented. See Patent No. 2,727,699, issued in 1955. The original winding head was made of metal, about 2¼" long, 5/8-¾" high, and 9/32" thick. At both ends the vertical sides were broken 5/16" from the top, and thereafter ran diagonally 9/16" at an angle of about 30° from the horizontal, so that the horizontal base was reduced to 1 17/64". The reel was placed immediately below the winding head, with the tabs imbedded in the sloping sides thereof, so that the strand was wound around those sloping sides and not against the tabs. A small metal cylinder extended outward from each side of the winding head, so the strand was wound in elliptical loops. Sixteen loops in all were wound. After the twelfth loop, the strand was cut to the estimated length of the last four loops; and after the last loop was wound, the machine pushed the reel off the winding head and placed it in a glass tube.
14. Although it probably was intended that the strands should be wound around the sloping sides of the winding head in loops of progressively increasing length, in practice the loops tended to pile up on each other or to slide down into the crotch formed by the sloping sides of the winding head and the base of the reel. This prevented any accurate determination of the length of the last four loops to be wound, which was important because the estimated length of the last four loops determined where the thread should be cut. Although the first four loops would, on the average, be shorter than the last four loops, the length of any particular loop and its relationship to its predecessor and successor were matters of chance. This uncertainty caused many practical difficulties in the operation of the machine, and in 1954 a young engineer was assigned by Ethicon to the task of making the machine operate more efficiently. He conceived the idea of reducing the length of the sloping vertical sides of the winding head and causing the bottom 5/32 of an inch of each side to be approximately vertical. This change enabled him to estimate more accurately the probable average length of the last four loops, although the exact length of any loop and its relationship to its immediate predecessor and successor remained a matter of chance. Further experiments indicated that these "vertical" lower portions of the sides should be tapered inwardly five degrees, to reduce the friction between the suture and the winding head when the reel was ejected, and that their height should vary slightly depending upon the thickness of the strand to be wound.
15. This machine and these winding heads were developed for the purpose of mechanizing the winding operation theretofore performed by hand, and not for the purpose of utilizing plaintiff's invention. The men who developed the machine and the reel had never heard of plaintiff's invention, and it was not brought to the attention of any of the engineers involved in the design, testing, or operation of the machine.
16. In Ethicon's sealed and completed package, the loops wound on either the old or the new type of winding head lie loosely on the reel in elongated loops, having a random relationship to each other, and not pressing against the tabs in any regular fashion, if indeed they press against the tabs at all.
17. Suture packages wound by Ethicon on the old style winding head were sold by the other defendants in Maryland during the years 1951 to 1955 or 1956.
18. Plaintiff originally contended that all packages containing reels wound on either type of winding head infringe the claims of his patent. However, during the closing argument, his counsel conceded that the packages wound on the new type of winding head do not infringe any of plaintiff's claims.
Conclusions of Law.
Plaintiff's patents are for suture packages, embracing specific features of construction. Their claims are limited in scope, and were deliberately drawn to those specific features. They are "paper patent[s], in a crowded field, which [have] never been used commercially. Thus, the claims, in the light of the specifications, must be rather narrowly construed." Yates v. Jones, 4 Cir., 176 F.2d 794, 795.
Claim 1 of plaintiff's first patent embraces a number of elements. See Paragraph 10 of the Findings of Fact, above. All of these elements, in combination, are essential to the claim. The same is true of claim 2 of the first patent and claim 3 of the second patent. So construed, as narrow combination claims, each of the claims is valid.
But if any of the claims should be construed so broadly as to be infringed by the Ethicon package, whether wound on the old or on the new winding head, that claim would be invalid and void because anticipated by prior United States patents, set out in Paragraphs 5 and 6 of the Findings, as well as by the Brinkmann German patents.
The Brinkmann patents are proper references against plaintiff's second patent under the law applicable thereto, because they were issued more than two years before plaintiff's application for that patent; they are also proper references against plaintiff's first patent unless he can prove a date of invention prior to the "ausgegeben" dates of the Brinkmann patents, October 22, 1936, and November 26, 1936, respectively. Plaintiff's evidence on this point must meet a very severe test. "Where one seeks to carry the date of invention back of the date of an anticipating patent, he assumes the burden of proof, and must establish an earlier date, `by evidence so cogent as to leave no reasonable doubt in the mind of the court, that the transaction occurred substantially as stated.' Deering v. [Winona] Harvester Works, 155 U.S. 286-301, 15 S.Ct. 118, 39 L.Ed. 153; Columbus Chain Co. v. Standard Chain Co., 6 Cir., 148 F. 622, 78 C.C.A. 394; Barbed Wire Patent, 143 U.S. 275, 12 S.Ct. 443, 450, 36 L.Ed. 154; Coffin v. Ogden, 18 Wall. 120-124, 21 L. Ed. 821." Moline Plow Co. v. Rock Island Plow Co., 7 Cir., 212 F. 727, 732. "When an inventor's date is to be carried back beyond his application, courts regard the effort with great jealousy, and must be persuaded with a certainty which is seldom demanded elsewhere; quite as absolute as in a criminal case, in practice perhaps even more so." Judge Learned Hand in United Shoe Machinery Corp. v. Brooklyn Wood Heel Corp., 2 Cir., 77 F.2d 263, 264. Plaintiff's evidence, set out in Paragraph 8 of the Findings of Fact, does not meet this test. The Brinkmann patents must be considered in determining the validity and scope of the claims in plaintiff's first patent; they confirm the conclusion that those claims must be narrowly construed.
That conclusion is also supported by the file histories of the Horine patents, which show that plaintiff reduced the scope of his original claims by cancellation or by amendment, in order to overcome rejections by the patent examiner. He is therefore now estopped to ignore the limiting language in the issued claims when he asserts those claims against others. Baker-Cammack Hosiery Mills v. Davis Co., 4 Cir., 181 F.2d 550, 563; Smith v. Florence-Mayo Nuway Co., 4 Cir., 182 F.2d 507, 509, and cases cited therein.
The essential elements of the claims in suit are not found in the accused packages.
(1) Ethicon's reel is not conical, or of a generally conical formation. It does not provide a conical surface on which the suture may lie.
(2) The suture is not wound helically, in a single layer, with each convolution about the bobbin in closely adjacent relationship. Ethicon's packages do not show the sutures in this condition. Moreover, the evidence does not show that Ethicon's sutures were wound in this manner, on either the old or the new type of winding head. The evidence shows that when the old winding head was used the strand did not wind helically, in a single layer, either about the winding head or about the reel, but that the loops either piled up on themselves, or slid down into the crotch, or both, and thus prevented any regular relationship between the successive loops. Plaintiff's demonstrations and other evidence did not show (a) the progressive enlargement of the coils claimed by plaintiff as distinguished from (b) the admitted tendency of the loops to increase in length, so that the average length of the last four loops would be greater than the average length of the first four, but with the relationship between any loop and its predecessor and successor a matter of chance. Defendants' demonstrations and exhibits indicate that there was no progressive enlargement of the coils wound on either winding head. But even if the evidence had proved progressive enlargement of the coils, that would not be the equivalent of the carefully drawn features of the claims in the patents issued to plaintiff.
(3-1) Ethicon's strands are not "twisted in the direction of the twist while being wound with each convolution about the bobbin in closely adjacent relationship and in which the twist in the suture about its longitudinal axis is uniform about the bobbin from end to end thereof as a result of said last mentioned twist", as required by claim 1 of plaintiff's first patent. The specifications and plaintiff's testimony indicate that such a twist would be a twist to the right. Plaintiff admits that Ethicon does not do this, but contends that holding the strand tight while winding to the left, is an equivalent. Plaintiff argues that "it makes no difference, in preserving the twist, whether the strand is wound to the right or to the left, provided, that (1) when winding is done to the right the strand is rotated on its long axis to the right and has imparted to it an additional twist, and that (2) when winding is done to the left, the strand is held against rotation on its long axis, thusly imparting an additional twist". But the evidence does not show that the strand is held so firmly in Ethicon's machine that the claimed effect is achieved. If it is achieved, it is not as a result of anything plaintiff teaches in his patent. The evidence does not show that what Ethicon does is the equivalent of what is set out in the claims of plaintiff's patents. As in Smith v. General Foundry Mach. Co., 4 Cir., 174 F.2d 147, 150, "Defendant's arrangement is not intended to and does not effect the purpose which the [plaintiff's] patent structure was expressly designed to accomplish." The proof of the pudding is in the eating. In Ethicon's package the twist in the suture about its longitudinal axis is not uniform about the bobbin from end to end as a result of any additional twist. Plaintiff himself testified that a suture wound by Ethicon on the old type winding head tends to develop four weak spots in each loop, at points where it does not rest against the reel at all. Defendant's evidence showed that the final condition of the suture, so far as its twist and strength are concerned, is substantially the same in a package containing a machine-wound, Schaefer-type reel, as in a package containing a hand-wound, conventional H-reel.
Finally, the doctrine of equivalents is not available to plaintiff to cover what he neither disclosed nor claimed. Having avoided the prior art by claiming his invention with precision, he cannot now discard that precision to establish infringement. Mackay Radio Telegraph Co. v. Radio Corporation of America, 1939, 306 U.S. 86, 618, 59 S.Ct. 427, 83 L.Ed. 506. What he gave up by amendment in the Patent Office, he cannot recapture by recourse to the doctrine of equivalents. Exhibit Supply Co. v. Ace Patents Corp., 1942, 315 U.S. 126, 62 S.Ct. 513, 86 L.Ed. 736. See also Schmitt v. Norcor Mfg. Co., 7 Cir., 152 F.2d 519, and Stelos Co. v. Hosiery Motor-Mend Corp., 2 Cir., 72 F.2d 405.
(3-2) Ethicon's package does not meet the following requirement of claim 2 of the first patent: "the bobbin having means at its base end to receive and secure said strand, whereby the suture is supported throughout in the form of a conical coil which may be readily removed over the apex of the bobbin without kinking or bending of the strand or the loss of twist therein".
(3-3) Nor does Ethicon's package meet the following requirement of claim 3 of the second patent: "said bobbin having a longitudinally divided conical strand supporting body, and said body being radially yieldable under contraction and expansion of said strand to maintain substantially uniform tension throughout the strand coiled thereon". Aside from the fact that there is no conical bobbin, much less a conical bobbin split longitudinally, plaintiff's own testimony with respect to the weak spots in defendant's sutures shows that his claimed uniform tension is not found in Ethicon's package.
Plaintiff's patents were not infringed, and the complaint must be dismissed, with costs. It is so ordered.