140 US 481 Clark Thread Co v. Willimantic Linen Co

140 U.S. 481

11 S.Ct. 846

35 L.Ed. 521


May 25, 1891.

Livingston Gifford and Edmund Wetmore, for appellant.

W. C. Miller, Clarence A. Seward, W. H. Kenyon, and B. F. Thurston, for appellees.



This is a suit brought by the appellees against the appellant on a patent issued to Hezekiah Conant for an improvement in machines for winding thread on spools. The patent was issued December 13, 1859, but antedated 22d June, 1859, and at its expiraion was renewed for seven years from 22d June, 1873, finally expiring in 1880. Its number was 26,415. The bill was filed in February, 1872, charging the defendants with infringement, and praying for injunction, damages, etc. The defendants, promptly answered, and the cause lay until February, 1874, when the complainants filed a supplemental bill, setting up the extension of the patent, and repeating the charge of infringement. The defendants answered, and the parties went into proofs. In March, 1879, the circuit court, held by District Judge NIXON, rendered a decree in favor of the validity of the patent, adjudged that the defendants had infringed the first and third claims thereof, granted an injunction, and ordered a reference to a master to take an account of profits and damages against the defendants. 4 Ban. & A. 133. After a long contest in the master's office a report was filed in October, 1884, awarding damages to the complainants in the sum of $159,035.22. The defendants filed exceptions, which were overruled by the court, and a final decree for the amount awarded was entered on the 17th of June, 1886, 27 Fed. Rep. 865. The present appeal was taken from that decree.


One of the principal points of controversy on which the case turns was indicated by the complainants themselves in their original bill. They say: 'And your orators further represent that the said defendants sometimes pretend that they have a right to make use of their said machines for winding thread on spools, and to make sale of like machines to others to be used, because they say that such machines are described in letters patent of Great Britain granted to William Weild on the 22d day of January, A. D. 1858, upon a specification filed July 22, 1858, and that the same invention was subsequently patented to the said Weild in the United States on the 2d of January, A. D. 1866; AND THEY FURTHER PRETEND THAT SAID BRITISh letters patent anticipate the invention of said Conant. But your orators aver the fact to be that the application of said Conant for letters patent for his invention was made and filed in the patent-office prior to the date of the sealing of said British letters patent to said Weild, and that the invention of said Conant, for which letters patent were granted to him as aforesaid, was made before the publication or date of sealing of said British letters patent of said Weild.' The allegation that Conant's application for his patent was made and filed in the patent-office prior to the date of the sealing of Weild's British patent is not correct. It is not proved, and the contrary appears to be the truth. Conant's application was first filed on the 5th or 6th of January, 1859, and was afterwards withdrawn and renewed on the last of April or first of May in the same year. The specification annexed to the patent is dated 11th of April, 1859; and the drawings are marked as received in the office and filed January 6, 1858, (an evident mistake for 1859,) and received and filed in new application May 2, 1859. There is a certified copy of the file-wrapper and contents in the record, which shows that the original application was filed in the office January 5, 1859, and was withdrawn, and a new application filed April 30, 1859, the papers being received in the examiner's office a day or two later in each case. But as this copy of the file-wrapper and contents was only introduced on an unsuccessful motion for a rehearing, and not in the principal case, it may not be proper to rely upon it in a matter affecting the merits. There is other evidence, however, sufficient to verify the same facts. The allegation that the invention of Conant, for which his said letters patent were granted, was made before the publication or sealing of Weild's patent requires more careful consideration. The defendants, in their answer, denied that they had infringed Conant's patent, and denied that he was the first inventor of what is claimed to be patented thereby, and averred 'that, on the contrary, the same, under the broad >construction thereof claimed for it by complainants, was, prior to any invention thereof by said Conant, described in and patented by letters patent granted by the government of Great Britain to Archibald Thomson, which were dated the 10th day of November, 1801, and numbered 25,053; also described in and patented by letters patent granted by the government of Great Britain to Wm. Young, which were dated the 2d day of December, 1848, and numbered 12,353; also described in and patented by letters patent granted by the government of Great Britain to Thomas Willis, which were dated the 1st day of June, 1852, and numbered 14,151; also described in and patented by letters patent granted by the government of Great Britain to John Wibberly, which were dated the 4th day of December, 1853, and numbered 2,901.' As to the Weild patent, referred to in the bill of complaint, the defendants answered as follows: 'And these defendants, further answering, say that letters patent of the United States for an invention in machines for winding thread upon spools were granted to William Weild, dated the 2d day of January, 1866, and that the only machine which these defendants have used for winding thread on spools, and those which they now have in use for that purpose, were purchased by them from said Weild under said patent, and were made in conformity therewith, and that they paid said Weild royalty for the use of the same; and these defendants, further answering on information and belief, say that said letters patent for said last-named invention were granted by the government of Great Britain to said Weild, dated January 22, 1858, and sealed April 30, 1858; and these defendants, further answering, say that they do not know and are not informed, save by said bill of complaint, when said Conant made his application for the letters patent upon which this suit is brought, or whether or not the same or the invention of said Conant was made prior to the sealing of the English patent to Weild, and leave the complainants to make such proof thereof as they may be advised is material; and these defendants, further answering on information and belief, deny that said Conant made his alleged invention before the date of said foreign letters patent to said Weild; and these defendants say that they are informed and believe that said Weild made the invention for which said patents were issued to him and put the same into public use prior to the time of said Conant's alleged invention.'


We have thus adverted to the pleadings for the purpose of showing that the issue as to the priority of Weild's patent over the invention of Conant was raised by the complainants themselves in their bill of complaint, and was accepted by the defendants in their answer. This should settle all doubt as to the relevancy of that question in disposing of the case on its merits. Objection was made that the Weild patent was not duly proved in the case, but without foundation. It appears by the record that, at an examination of witnesses on the part of the defendants before W. C. Witter, examiner, by consent, continued from time to time from June 17, 1875 to April 8, 1876, one Boyd Eliot, being under examination, was asked, among other things: 'Question 2. Have you read and examined the copy—Weild patent, defendants' Exhibit William Weild? Answer. I have. Q. 4. Do you find therein described the defendants' machines, of which Exhibit 5 is a model in part? A. I do; substantially the same.' At the end of the depositions of the witnesses are the exhibits referred to therein, among which is the copy of the Weild patent in question, marked 'Defendants' Exhibit William Weild, W. C. W. Ex'r;' and at the end of the entire record is the certificate of the clerk of the circuit court verifying the same, under seal of the court, as a true transcript of all the proceedings in the cause on file and of record in his office at Trenton. The patent was referred to and used in the examination, was marked as an exhibit in the cause by the examiner, nd is actually found in the record, and returned and certified as a part thereof. Though the depositions contain no express minute that the patent was offered in evidence, we think that it must be received as so offered. Nearly the same question arose in Hoskin v. Fisher, 125 U. S. 217, 8 Sup. Ct. Rep. 834, in relation to a patent marked as an exhibit, and we held that it was sufficiently authenticated as a part of the evidence in the case. If the Weild patent was improperly inserted in the record, the complainants should have moved the court below to have it excluded before the transcript was sent to this court.


As this patent, in our view, has an important bearing on the questions involved in the case, it is proper that we should examine with some care the allegation in the bill of complaint that Conant made his invention before the patent was recorded and published, which is conceded to have been on the 22d day of July, 1858, six months after it was granted and after its date. The question is important, because the law is that any person sued for infringement of an American patent may show in defense that the invention claimed was patented or described in some printed publication (not before the American patent was granted, nor before the application for it was filed, but) before the patentee's supposed invention or discovery thereof. Rev. St. U. S. § 4920. It is also important because the defendants proved that the machines used by them, and charged to be infringements of Conant's invention, were built in Manchester, England, and obtained from Weild himself, and constructed in accordance with his patent. It would seem to be very clear, therefore, that unless the invention of Conant was made and perfected before the 22d day of July, 1858, the time of publication of Weild's patent, the defendants had a perfect defense to the suit, either on the ground that the Weild patent anticipated Conant's invention, or that Conant's patent, in view of the state of the art, must be so construed and restricted as not to embrace any portion of the Weild patent, in which case the defendants could not be justly charged with infringement. The only evidence on the questions as to the time of Conant's invention is his own testimony, a species of evidence which, in cases of this kind, ought to be received with great caution. The following question was put to him by his counsel: 'Question 4. Please state when you made the invention which is described in the aforesaid letters patent granted to you. Give the date of its conception, and the successive stages of its development.' His answer was as follows: 'My attention was drawn to this point of the matter of windling thread automatically by machinery some time during the year 1857. I worked at intervals at making drawings and trying to develop different motions and devices, and, to the best of my recollection, I fixed upon a definite style of machine, which I proposed to build some time during that year, and prosecuted the thing so far as to have some of the patterns made, but was interrupted during the winter; and, on further reflection, in the spring of 1858 I decided upon another style of machine, which I thought would be more certain in its action, and which I completed during the summer of—and, to the best of my recollection, in July—1858. The part of my machine called the 'traverse changer' was the same in the first set of drawings as it is in present use. The various devices for accomplishing the change required to wind different lengths were carefully studied over, and I fixed on this present style of traverse changer, with different lengths of teeth or projections arranged upon the periphery of the wheel, as the most practical and best form in which it could be arranged, from the fact that the wheel is always in a position to wind the first course of thread upon the spool immediately after the last course on the previous spool is completed. The first machine, according to the best of my recollection, was deficint in that it had no way of adjustment by which I could wind spools of different lengths, but having the same number of courses, and which I considered defective, because in such case I would be obliged to have a different traverse changer for every different length of spool. So I made my traverse changer long enough for winding the coarsest number of thread in ordinary business, and then fixed the adjustment of the lips, so that, by spreading them apart, I could wind as short courses as was desirable, when by setting the lips close together, I could wind courses the full length of the traverse changer. It was also a point in my study to make the machine as simple as possible, so it could be readily operated by persons of ordinary ability. The first machine, as I mentioned, was completed and put in operation in the summer of 1858. After getting the machine completed, and testing it to my satisfaction, I went to work, and made a model myself. After that I made a set of patent drawings, and drew up my own specification, and made the first application for a patent myself. In January, 1859, I exhibited my machine at a meeting of the stockholders of the Willimantic Linen Company, in Willimantic, Connecticut, at which time they were so well pleased with the machine that they proposed to purchase one-half of the right; but they ascertained that I made the application for a patent myself, and, thinking perhaps, that it was not as perfectly done as it should be, employed Mr. Henry B. Renwick to redraft the specification. In the mean time a Mr. C. N. Spencer invented a self-acting winding machine, for which he made application for a patent, which the patent-office declared interfered with my application. I was obliged to take testimony, which delayed the granting of the patent, after which a patent was issued to me.' On a subsequent examination, being questioned with regard to the disposition of the machine constructed by him, he said: 'I made an exhibition of it in the month of January, 1859, at a meeting of the stockholders of the Willimantic Linen Company, and I run it the best part of a half day in their presence. Afterwards Iexhibited it to Mr. Harry B. Renwick, who, I suppose, was employed at that time as an expert for the Willimantic Linen Company, but after that it was laid away, and I don't know that it was ever again used.'


No person accustomed to weigh the credibility of human testimony can fail to perceive the stress under which this evidence was given. With the most favorable construction, the most that can be deduced from it is that the invention was not completed until July, 1858. The witness does not say in what part of July, although his interest strongly suggested that part of the month which preceded the 22d. The generality of the expression 'July, 1858,' excites attention, and in this case it is not applying too strict a rule to say that the evidence should be construed most strongly against the complainants; and this would necessarily lead to the inference that the invention was completed in the last part of July, subsequent to the publication of the English patent. We feel bound to put this strict construction upon the patentee's evidence, because such testimony, given for the purpose that this was, is necessarily subject to the gravest suspicion, however honest and well intentioned the witness may be. We say that Conant's testimony cannot be construed as showing that his invention was completed before the month of July, 1858. According to his statement it was not until then that he put it in visible form in the shape of a machine. He had conceived of a machine in 1857, and made some patterns, but in the spring of 1858 he decided upon another style of machine, which he completed during the summer,—to the best of his recollection in July, 1858. The part of his machine called the 'traverse changer' was the same in the first set of drawings as it is in present use. This is the substance of the testimony. It is evident that the invention was not completed until the construction of the machine. A conception of the mind is not an invention until represented in some physical form, and unsuccessful experiments or projects, abandoned by the inventor, are equally destitute of that character. These propositions have been so often reiterated as to be elementary.


If the result to which we have come as to the date of Conant's invention is correct, it really determines the controversy, without any inquiry as to the similarity of Conant's invention to that shown in Weild's patent; for it is proved beyond all controversy that the defendants only used Weild's machine; and if they only used that which antedated Conant's invention they could not be guilty of violating his rights. The only part of Conant's testimony on which any plausible argument can be raised to show that his invention was prior to the 22d day of July, 1858, is that in which he states that the part of his machine called the 'traverse changer' was the same in the first set of drawings as it is in present use; but this is so vague and uncertain that no satisfactory inference can be drawn from it. The traverse changer, by itself, was not claimed in the patent as the invention of Conant, but it was claimed in combination with other things, which may or may not have been exhibited in his drawings; and traverse changers had been in use long before Conant had thought of the subject at all. To make the matter more manifest, it may be proper to give a brief description of the machine and the invention which related to it. In winding thread upon a spool it is necessary to wind it in layers, one above or around the other, until the spool is filled flush with the two ends. To do this with a continuous and unbroken thread, the first layer is commenced at one end of the spool and carried to the other as fast as the thickness of the thread wound in a continuous coil will admit. From thence the next layer is wound in a reverse direction back to the end at which the first layer commenced; and so on alternately until the spool is filled. The thread is fed on to the spool by means of a slender finger called a 'thread guide,' through and over which the thread is brought close to the spool; and in order to wind the alternate layers as above described this thread-guide has to be moved alternately backward and forward just as fast as the thread is wound, and just as far as the length of the spool. It receives this alternate movement by being attached to a slide-bar or traverse-rod, which is carried backward and forward by means of a parallel shaft furnished on one-half its length with a right-hand screw and on the other half of its length with a left-hand screw. Two half nuts, attached by arms to the slide-bar, are so arranged that by the action of the machine one of them may be engaged with the right-hand screw for a limited time, and then the other with the left-hand screw for an equal length of time. The result is that the slide-bar will be forced first in one direction and then in the other, according as the right or left-hand screw is engaged with its corresponding nut. The change of motion is effected by a change of engagement of the nuts, so that when one nut is withdrawn from the screw the other may be set up against it. For a constant and uniform extent or length of alternate movement this may be very easily effected by ordinary cam or stop devices. But a spool has not an invariable length from its center or barrel to its circumference. The inner surface of the two ends flare out a little, so that the outside layers of thread are longer than those beneath them. This renders it necessary that the length of the alternate movement of the thread-guide and slide-bar should be slightly and continuously increased from the first or inside layer to the last or outside layer. The device for effecting this change of extent of lateral movement is the thing which requires most inventive skill in the construction, of the machine. Conant evidently supposed that he was the first to cotrive a device of this kind, which, in his patent, he calls a 'pattern cam,' or 'traverse changer.' But in this he was mistaken; spool-winders had long been in use in England, with the right and left hand screw device for giving the thread-guide an alternate backward and forward movement, both in hand machines and in automatic machines driven by power; and in two of the latter Wibberly's, patented in 1853, and Young's, patented in 1848 traverse changers were used for the same purpose as in Conant's machine. Besides the regulation of the alternate movement of the thread-guide to correspond with the variable length of the layers of thread on the spool, it was also necessary, or at least desirable, that there should be some device for stopping the machine, or the winding of the thread, when the spool was full. This was effected in a certain way by Conant in his machine, and had been effected in different ways by Wibberly and Young in their machines. It is unnecessary here to go into a description of the various devices used. Enough has been said for a general understanding of the claims made in Conant's patent.


The court below held that the defendants had infringed the first and third claims of the patent, and it cannot be seriously contended that any other claims were infringed. The first and third claims of the patent were as follows: 'First. The combination, substantially in the manner hereinbefore set forth, of a traverse changer, with right and left hand screws, and with nuts which are alternately in gear with such screws, the combination operating as a whole substantially in the manner and for the purpose herein described.' 'Third. I claim a stop motion, substantially such as described herein, for causing the machine to come to rest when a spool is filled, in combination with automatic apparatus, substantially such as set forth, for regulating the length of motion and change of direction of a motion of a guide through which thread is delivered onto a bobbin or spool.' These claims are for combinations. The first claim is for a combination of a traverse changer, right and left hand screws, and nuts alternately in gear with such screws, the combination operating substantially in the manner and for the purpose described. This combination has three elements, and a specific mode of operation. The third claim is for the stop motion in combination with the apparatus described and combined in the first claim. From this review it is apparent how uncertain and unsatisfactory is the statement made by Conant in his testimony that 'the part of my inachine called the 'traverse changer' was the same in the first set of drawings as it is in present use.' It fails entirely to show that prior to the 22d of July, 1858, he ever had or ever exhibited any drawings of the invention described in the first or third claim of the patent. We conclude, therefore, that there is no proof on which reliance can be placed that Conant made his alleged invention before the publication of Weild's patent in England. After Weild's patent was introduced into the case, showing with certainty the date of its publication, and such date anterior to the issue of Conant's patent, it was incumbent on the plaintiffs, in rebuttal, to show, if not with equal certainty, yet to the satisfaction of the court, that Conant's invention preceded that date. Plow-Works v. Starling, ante, 803, (decided this term.) It is also clear that Conant was not a pioneer in this department of invention, and that he must be held strictly to the terms of his patent, and was entitled only to the specific form of device described and claimed therein; and the evidence is very clear to the effect that the defendants never used such device. It is unnecessary for us to enter into an examination of the evidence on this subject. We are satisfied that the complainants had no case on which to ground a decree, and that the bill of complaint ought to have been dismissed. The decree is reversed, and the cause remanded wit di rections to enter a decree dismissing the bill of complaint, and taking such further proceedings as may be in conformity with this opinion.