Lewis Blind-Stitch Mach. Co. v. Arbetter Felling Mach. Co.

208 F. 992 | N.D. Ill. | 1913

SANBORN, District Judge.

Infringement suit on Patent No. 862,-830, issued August 6, 1907, to J. G. Lewis, relating to blind-stitch sewing machines for cloth fabrics.

The main question is whether there is infringement of two claims of the Lewis patent relating to the formation of stitches adjacent to and over the edge of a fold laid on the surface of a main fabric, like a cloak facing, or cuff attached to the end of a coat sleeve, by two rows of stitches made wholly on the outside of the cloth, and which are partly concealed under the edge of the facing or cuff, by a process called “felling.” Infringement of two other sets of- claims less important in character is also claimed. The argument mainly turns around two exhibit machines, called the Lewis old machine and the Arbetter machine; but the question of infringement depends also on many other considerations, based on facts appearing in evidence, and which relate to the respective objects Lewis and Arbetter were seeking to accomplish, and the means they were taking to reach these objects.

Defendant’s expert Robert P. Plains gives the following definitions and explanations necessary to understand any discussion of blind-stitcli machines:

“2. A ‘through and through stitch’ is one in which the needle enters at one surface ol: the material and passes out of the opposite surface and is there interlocked so that on withdrawal of the needle the thread will be found to extend through the fabric from surface to surface.

“3. A ‘blind stitch’ is one in v-hich the machine needle enters and emerges from the same surface of the material, the sewing thread not appearing- on the opposite surface.

“4. A ‘chain stitch,’ which may be either a through and through or blind, is formed by a single thread, such thread being interlocked or looped with itself at the point of emergence of the needle, so that on the surface where the inteiiooping, of such thread takes place the stitch will appear as a chain like structure.

‘‘5. A ‘lock stitch’ is one formed of two threads, the needle carrying one thread and at its point of emergence being inter looped or interlocked with another thread, usually termed the ‘bobbin’ thread.

“6. The chain stitch and the lock stitch may be either a through and through stitch or a blind stitch, according to whether the needle in making its penetrating stroke passes entirely through the fabric or enters and emerges from the same surface thereof.

“7. These varieties of stitches may lie used in the formation of a seam to unite two or more fabrics, or they may be made merely as ornamentations.

“8. If, in making a lock stitch, for instance, the needle makes its thrusting excursion always in the sanie plane, there will l.o produced a straight line of stitches. If, however, the needle should be moved bodily in a lateral direction after each stitch, there will be formed two lines of stitches, the threads of which are parallel. This last form of stitch formation may be termed ‘zigzag’ stitching, and the movement of the needle laterally is known in the art as ‘shogglng.’ ”

While the machines are complicated, they are not much more so than the 'old domestic sewing machines, with their feed mechanism, needle-control, and hook_’and bobbin devices. Blind-stitch machines existed for work on leather, felt, braid, matting, and cloth, as well as blind-stitch machines for making two rows of stitches connected by bobbin threads making zigzag stitches. The patents to Bosworth, Sleppy, ‘ Shea, Arbes, and Bevy all show blind-stitching mechanism, where the goods were bent around a back guide so as to present the surface to the stroke of a vertically descending or ascending needle which pierced the surface as it was descending around the edge of the back guide, and went through 'the upper layer and into but not through the lower, resulting in a stitch showing only on the upper side of the goods. The Lewis machine is of the same general type, where the goods are bent around the edge of a back guide adjacent to the needle.

Another class of prior blind-stitch machines fed the material horizontally, in which there was no bending over a back guide at the needle point, but the same result was secured by using a “bender,” which reciprocated vertically to push or bend the goods through a hole or slot in the bed plate, and thus get the cloth surfaces into the plane of a horizontally reciprocating needle, either straight or curved. Examples of this type are the Plummer, Henshall, Reece 1889, Reece 1890, Thomson, and Dearborn machines, and others. The Arbetter machine is of this general type, using the bender and horizontal feed at the needle point, instead of the vertically descending feed around the back guide edge, as in Lewis.

Nothing in the prior art, however, covered blind-stitch felling, or fastening a folded edge to a bottom layer, so that that row of stitches made in such bottom layer could be concealed, and the bobbin thread only be visible over the edge of the fold. To “fell” means to flatten and sew down level with the cloth, as, to fell a seam. This was first accomplished by machinery by Lewis, and later by’Arbetter in a different way.

Lewis and Arbetter started out in nearly parallel paths, without knowledge of each other, to produce machines which would accomplish similar objects. Lewis started first, and was first in the field; but Arbetter wás not far behind, and in some respects has greatly outstripped his competitor, especially in having a large number of practical machines at work. Lewis also has machines at work, but to what extent or in just what form does not clearly appear. Both classes of machines, however, do remarkably good blind-stitch felling, producing results which look very much alike. Eor about seven years the Ar-better company has had a large number of machines in operation under patent licenses in many of the large cities of the country. This is mentioned, not as bearing directly on possible infringement, but to show the importance of the case to the parties, and the great care with which it should be examined.

Lewis secured this blind-stitch, concealed-effect felling result' by feeding the goods up to the needle point in an angular fashion, so that the needle would go in under the edge of the fold close to the bottom layer, and with a diagonal stroke come out on top of the folded edge, a little back of the edge. This is the needle to edge relationship so much referred to in the briefs and argument, and is the whole gist of Lewis’ discovery. This had never been done before, and his was the first machine to accomplish it. Arbetter soon after accomplished a like result in another way.

With these general observations, an attempt will now be made to describe the means and operations which Lewis and Arbetter used to secure their respective results; plaintiff claiming that means, operation, and result are equivalent in both forms, and defendant claiming that all three are distinct. In describing Lewis’ and Arbetter’s methods, the correctness of their respective contentions as to what each was trying to do, and did finally do, will be assumed, leaving to subsequent examination of the record to determine how far the respective claims are sustained by the proofs. In the ordinary sewing machine the needle, bearing its thread-loop, moves up and down in a vertical plane, and the stitch is made either by a revolving hook surrounding a bobbin carrying the second thread catching the loop and pulling it around the bobbin thread, or by a shuttle-bobbin shooting back and forth through the needle-loop, somewhat like the bobbin action in a loom. In these old machines the needle goes through and through the cloth, making a stitch visible on both sides. Now a blind stitch, visible only on one side, may be made in two ways. If made with a straight needle, the cloth must be bent over the edge of a guide, so that the needle may go in and come out on the same side; if made with a curved needle, or straight needle and bender, the cloth may lie flat as in the ordinary machine, and the surface of the cloth placed in such adjustment that it will be just inside the path of the straight needle or just inside the arc of the circle described by the curved needle, so that the needle point will catch the surface and come out close, to the place where the cloth is first pierced by the needle point. This operation may be aided by a device coming up through the bed plate either at each needle-thrust, or at every other one, to' bend up the cloth, and aid the needle in catching its surface. This device is called a bender. Either a straight or curved needle may be used, since tile bender pushes the cloth into the line of needle-thrust. Lewis uses the straight needle and Arbetter the curved. To make the Lewis method a little clearer, let the finger represent the machine feed guide or back guide. If two pieces of cloth are bent around the forefinger, the pieces may be blind-stitched together (the usual way in the household), by sewing only through the upper piece and the surface of the lower piece held over the finger, without letting the needle go through the thickness of the lower piece, a blind stitch can be made. The thread will show only on the top or edge, and not on the other side. This, as well as the use of the curved-needle, is an old method of machine sewing, as applied to leather, felt, cloth, and matting. This is known as “padding.”

Felling, however, or blind-stitch felling, is a more difficult operation, when done with a machine. Here the problem is to bury the needle thread, and allow only the bobbin thread to show over the folded edge, and at the same time do a good job, one which will be firm and neat, and nearly resemble hand felling. The edge of a fold must be fastened or hung to the main fabric by two rows of blind stitches (common in the old art), near to and over the edge of the fold. This can best be done by making the lines of one of these rows of stitches, or of both, at an angle to the edge line of the fold, and then using the fold t:o cover one line of stitches. Both Lewis and Arbetter accomplish this, with practically the same results.

The gist of the Lewis invention in this part of his machine is the feeding of the cloth so that, when it is descending over and around the back edge of the guide, a folded edge is presented to the needle at a.n angle to the plane of needle movement, and this angular relation is maintained in both rows of stitches, that one made in the main fabric as well as the one made in the folded edge. While in the Arbetter machine the stitches made in'the main fabric are parallel to the folded edge, and only that line of stitches made in the folded edge are angular or diagonal to that edge. There is a theoretical advantage in the Arbetter method, because the line of stitches in the main fabric is parallel to the folded edge, and thus easier to cover up than a diagonal stitch, starting back from the edge and coming ‘out close under it. Practically there is little difference, because the stitches in the main fabric are very short. In both forms only the bobbin thread shows over the edge. A close observation of samples of work from both machines shows no discernible difference.

Referring again to hand work, hold the needle vertically in the right hand, with the point up, and bend the folded edge over the left finger so that such edge is parallel to the length of the needle. It will be seen that, when the needle is pushed up vertically through the edge of the fold, the points of entry and exit will each be the same distance from that edge, and the stitch will be straight and not diagonal. But if the cloth as it falls over on the near side of the finger is swung to the left without moving the folded edge away from the needle, the perpendicular portion drawn over the back of the finger will be at an angle to the vertical line of the needle-thrust, and, when the needle rises and is pushed through the perpendicular edge of the cloth, either the main fabric or folded edge, the stitch will be a diagonal one. This is-what Lewis means in claims 72 and 76 in counting on “means for presenting the goods to the needle with the edge adjacent to the needle inclined to the path of reciprocation of said needle,” and in claim 75, “inclined to both said paths of reciprocation of the needle.” The “means” referred to are the angular edge of the back guide over and around which the cloth is fed. So long as the cloth lies flat, away from the finger or edge of the guide, it may be swung either right or left clear ai'ound the circle without inclining the horizontal portion of the edge to the needle in any degree. It is always the perpendicular or descending part of the edge which' is so- inclined by such left or right movement. This point is not clearly brought out in the evidence, probably because it is an obvious one.

Applying this illustration to the Lewis machine making two rows of stitches, one in the main fabric and one in the folded edge, he provided for two parallel paths of needle movement. This he did by a cam to set over or “shog” the needle from side to side. Suppose the first needle-thrust was through the main fabric alone, on the next needle-stroke the needle is set back so as to make its diagonal stroke through the edge, as described; the two rows being locked together by the bobbin thread carried by the revolving hook, and by putting the thread under tension the folded edge is pulled over so as to bury the first row of stitches, and draw the bobbin thread firmly over the edge. The stitch made in the folded edge not being allowed to go into the underlying cloth, the edge is held by the stitch and the bobbin thread locked into it as if suspended from an inverted hanger, and when the bobbin thread is tightened from the side under the edge the latter is drawn down or felled over the first row of stitches, with the bobbin thread only visible across the edge o E the upper layer.

Mindful of the difficulty of clearly showing these operations, a description of Arbetter’s method of meeting the same problem will now be attempted, assuming as before that both parties were working to the same end. Arbetter gets a like result by rocking his circular needle. ITis cloth always lies horizontally and is fed under his curved needle, so that its surface is always parallel to the arc described by the needle point as it swings back and forth, except that the bender rises to the parallel needle-stroke so as to 'slightly push up the main fabric, and enable it to be the more readily caught by the needle point swinging parallel to the cloth surface. He makes the first line of stitches by keeping the needle-thrust parallel with the folded edge, and these stitches are made quite close to, but not touching, such edge; but the second line, by means of a rocking mechanism on his ne'edle-hanger, he makes in the edge itself by putting the needle through at an angle. On the first or straight thrust the curved needle picks up the surface of the main fabric over the bender, and on its second one its point is rocked over so as to pierce just under the folded edge, and a little forward of the first stitch, and go through such edge without touching the main fabric, diagonally to the line of the edge, coming out on top of the folded edge a little back of that edge. The thread-loops being caught by the hook and thus locked to the bobbin thread, both lines of stitches are locked and drawn together, and the first line in the main fabric more easily concealed, at least in theory, than in tlie Lewis operation. In both operations only the bobbin thread shows over the edge.

It thus appears that Lewis’ discovery was that the folded edge must be kept in close angular relation to the path of reciprocation of the needle. In his device "for doing this a single necessary element additional to the old sewing machine is the angular edge or face of the back guide.. In claim 72, covering “means for presenting the goods to the needle with the edge adjacent to the needle inclined to the path of reciprocation of said needle,” the means are the angular edge. Tn claim 76 he adds a second element in the form of “an edge-guide for the goods.” This is convenient, but not essential. The whole gist of Lewds’ invention lies in the simple, small triangular piece of metal added to the face of the back guide, which gives him his whole needle to edge relation, and enables his machine to do good concealed effect felling.

The corresponding elements in the Arbetter machine are the rocking needle-hanger and the bender.

With "this attempt to show the operations and results of the two machines, it is now necessary to take up the Lewis patent, and other parts of the record, in order to try to discover just what Lewis was seeking, how he tried to get it, and what position his invention occupies, in view of the prior art, and any limitations imposed on him by the file wrapper contents, or self-imposed limitations, if any, appearing in his specifications and claims. Superficially the two operations are different, but the question is whether Arbetter’s machine is the equivalent of Lewis’, even though means and operation may appear to differ.

It is shown in the proofs that in February, 1901, Lewis installed the Lewis old machine, which contains his diagonal feed, in clothing factories in St. Louis, where it was used for several 'months doing, blind-stitch felling on the inturned hems of trouser bottoms. This was the first successful machine to do such felling ever produced, nor is there any other disclosure or machine in the prior art which anticipates the features of claim 72, relating to diagonal stitching.

The Lewis Me wrapper:' In the original specifications, filed August 2, 1902, there appear the following paragraphs relating to the subject-matter of claims 72 and 76:

“The object of my invention is to simplify the construction of blind-stitch sewing machines, and to provide mechanism whereby stitches can be made which are invisible on the opposite side of the goods to that on which the stitches are made, but which are also concealed, or nearly so on the side of the goods in which the stitches are made.”

“The needle 165 at one thrust passes along the line A (Fig. 22) entering and leaving the main layer 168 of goods at the same side, but does not enter the superimposed layer 169. At the next thrust the needle is moved laterally by the cam 86 so as to pass along the line B entering the edge or underside of the superimposed layer and emerging from the surface thus piercing the said superimposed layer transversely. By sufficiently tightening the tension the edge of the superimposed portion 169 is drawn under so that the stitch is concealed, or almost so on the side of the goods from which the stitch is done.”

There was originally no claim covering the subject-matter of patent claim 72, but this was added among the first amendments suggested by the patentee, under date of December 2, 1903, as follows:

“53. In a sewing machine the combination with stitch-forming mechanism' of means for presenting the goods to the needle with the edge inclined to the path of reeixmocation of the needle.”

This is the same as patent claim 72, except insertion of the words “adjacent to the needle,” after the word “edge,” which was made January 30, 1906. No explanation is given of the reason why this claim was not included at first, and it might well have been, since it is within the original description and drawings.

When the examiner came to deal with this and two other like claims, on December 9, 1903, -a rather unusual thing happened. Before Lewis filed his patent application, a patent was granted to Arbet-ter on an application filed August 13, 1901; the patent being dated January 7, 1902, No. 690,385. If this patent covered the Lewis invention, it was a good prima facie reference. The examiner thought it was, and so rejected claim 53 on this Arbetter patent. To this Lewis’. attorneys responded, under date of March 18, 1904, as follows:

“An affidavit has been submitted to carry applicant’s date of invention prior to the filing date of the Arbetter patent, and therefore said patent is not a good reference for present claim 5-1, formerly claim 53.

“This affidavit also avoids this patent as a reference for present claims 55 and 50, former claims 54 and 55. It is not thought, however, that these claims are anticipated by the Arbetter patent, as in the Arbetter construction the goods are inclined only to one path of reciprocation.

“A reconsideration and allowance are respectfully solicited.”

Here we see the examiner taking a view in harmony with plaintiff’s present contention, and opposed to defendant’s contention, and Lewis (through his soliciting attorneys) arguing that the Arbetter construction was not an anticipation because the goods were inclined only to one path of needle reciprocation, which is inconsistent with plaintiff’s present claim.

The affidavit referred to satisfied the examiner, and the Arbetter patent was withdrawn as a reference, under date of April 5, 1904. The claim to the diagonal relation of the goods to the needle was, however, repeatedly rejected by the examiner on the Arbes patent, which did not have a diagonal feed, and on May 13, 1905, Lewis’ solicitors submitted the following argument:

“It is not seen that these claims are met by either of these references. The patent to Shea certainly has no bearing whatever upon the claims. Claim 54 calls for means of presenting the goods to the needle with the edge inclined to the path of reciprocation of the needle. This is not true of the Arbes patent. It is true that after the goods have passed around the back guide they pass down an inclined path, and it may be that the examiner had this in mind in rejecting the claim. An examination of the Arbes device, as shown in Fig. 6 of liis patent, will show, however, that at the time when the goods are presented to the needle as they are passing around the back guide, the edge is in a. line parallel with the path of reciprocation of the needle and not inclined thereto. The Arbes device will not accomplish the function which applicant obtains by presenting the edge of the goods inclined to the path of the needle, namely, that of being able to take a stitch through the edge of the goods passing in under the fold and out upon the face of the goods. It is therefore thought that the claim is clearly allowable.”

Again, the claim was rejected on the Arbes patent, with the following statement:

“The patent to Arbes, 595,090, of record, shows a back guide having a convex surface across which the goods are fed. This is the full equivalent of the back guide recited in claim 3, and said claim is rejected thereon.”

The claimant then varied his argument as follows:

“Present claims 53, 54, and 55 (formerly claims 54 , 55, and 56) are claarly not .anticipated by either of the references cited. In applicant’s last amendment he clearly pointed out his reasons for believing that these claims are not met. In addition, however, it may lie stated that the goods cannot be presented with the edge inclined to the path of the needle in either the Arbes or the Shea patent, as suggested in the examiner’s letter of July 18, 1904. If it is attempted to present the goods with the edge inclined to the path of the needle in either of these machines, the effect will be to feed the goods laterally away from the back guide and from the needle. This can be readily Illustrated by drawing a strip of cloth around a pencil or other cylindrical guide in a diagonal direction. In view of the above it is thought that these claims are clearly allowable.

“A reconsideration and allowance are respectfully solicited.”

Further rejection then followed, still on the Arbes patent, and the claimant, under date of June 21, 1906, again presented the same arguments as before:

“Present claims 57, 58, and 59 (former claims 5S, 59, and 60) are thought to clearly distinguish from the references. Present claim 57 calls for means for presenting the goods to the needle with the edge adjacent to the needle inclined to the path of reciprocation of the needle. This is not shown in either of the references. It is true that the patent to Arbes shows the edge inclined after it has left the needle. At the point where it is adjacent to the needle, however, it is parallel with the needle, and consequently the Arbes machine does not aeeonrplish the function which is accomplished by applicant’s machine. The patent to Shea is no nearer to applicant’s construction than the patent to Arbes. Present claim 56 calls for means for feeding the goods across the face of the guide adjacent to the needle in a path inclined to said face. This is not true of either of the references. In both the goods are fed straight across the face of the guide at the point adjacent to the needle, and therefore neither accomplishes the same function as applicant’s machine. Present claim 59 is similar to claim 57, except that it calls for a needle having two paths and means for lmesenting the goods with the edge adjacent to the needle inclined to both paths of reciprocation, and it is therefore more limited than the previous claim and is allowable for the same reasons.

“A reconsideration and allowance are respectfully solicited.”

Finally, on July 12, 1906, the examiner was satisfied, and no further rejection of the claim in question (72) was made, and the patent was issued, but more than a year later, on August 6, 1907.

No further description than that contained in the original specification was made. All the elements of claim 72 were contained in the first description, which shows that the “needle to edge” relation was a necessary part of the invention. Hence the first amendment to the claim, January 30, 1906, by inserting the words “adjacent to the needle,” added nothing to the previous description. It is evident all the way through that Lewis firmly insisted on the discovery of the diagonal stitch and needle to edge relation, with all that these imply. The patent is the measure of Lewis' grant, modified, of course, by the prior art, and by anything appearing in the file wrapper. There is nothing, however, in the latter which limits the grant, except possibly the admission of his attorneys that the Arbetter patent of 1902 did not anticipate, and could not therefore infringe. The patent in suit therefore stands substantially unimpeached, so far as the file wrapper is concerned. It is an advance on the prior art by a single step of considerable importance. While Lewis was not a pioneer, his patent, up to this point of our examination, is entitled to fair and reasonably liberal construction, because he was the first to do blind-stitch felling by machinery. Arbetter was only a few months behind, but Lewis was first.

The Lewis seam-application and interference: Much was said on the argument about another patent application made by Lewis, soon after his machine application was filed, called the seam-application. This was for a patent on the Lewis seam shown in the first application. Ar-better also had a seam-application, and the two were put into interference, which was finally decided in Arbetter’s favor, by the Court of Appeals of the District of Columbia, and a patent issued to him. Lewis, thus being denied a patent on his seam, brought suit under section 4915, Rev. St. U. S. (U. S. Comp. St. 1901, p. 3392), to compel the issue of a patent. This action is now pending and undetermined. Neither party to the interference took any proofs; each standing on the interference-record.

Both parties now claim some benefit from the seam-application and interference-records. Defendant claims that the Lewis seam-application shows indisputable evidence that Lewis’ real object and purpose were to make a diagonal or zigzag stitch in order to get greater binding effect and covering capacity, and not concealed effect felling, only claiming a partial concealment. This was decided by the District of Columbia Court of Appeals in the interference case; that court saying that “concealment of either thread was as foreign to his original design as it is impossible of accomplishment in his seam.” Plaintiff, on the other hand, claims that the interference case demonstrates that both seams are the same; that the Patent Office took four claims out of the Lewis application and gave them to Arbetter, and they are now among his patent claims. Defendant, having accepted a patent on its seam broad enough to include the seam of the Lewis patent in suit, with full knowledge of the seam-application and interference, on the theory that Arbetter had a prior right to make such claims drawn by Lewis, thus asserts that the subject-matter is identical. Defendant has taken out a patent on Arbetter’s claim to novelty and priority, and taken over Lewis’ discovery on the theory that it was later than Arbetter’s. This is a full recognition of the identity of the two seams. These are, in brief, the respective contentions.

[1] Ordinarily a patent is to be construed by its own contents, its file wrapper, and a consideration of what there was left previously undiscovered in the particular art. The patent is the measure of the grant; other applications not usually being material to its construction. But the rule of construction adopted respecting several deeds or contracts between the same parties, made about the same time, relating to the same subject-matter, and as parts of the same transactions, undoubtedly applies to patents and patent applications. All may be resorted to on an inquiry as to the meaning of any one. The seam-ap-plicatious, interference, and Lewis’ patent application, related to the same subject-matter, and were all pending during substantially the same period. Therefore all may be looked into to determine the proper construction of the Lewis patent.

The Lewis seam-application was filed P'ebruary 2, 1903, was allowed November 24, 1906, but thrown into interference with Arbetter’s like application early in 1907. Figures 1 and 2 are about the same as figures 21 and 22 of his patent application. Two methods of forming the stitches are shown, one with the first needle-thrust in the bottom layer, and the next entering under the folded edge and coming out on its top surface, the other with the first needle-thrust as before but the next one both goes in and comes out on top of the upper layer containing the folded edge. In the original application he says nothing about greater binding effect or greater covering capacity, but recommends that the edge be folded “so there will be no raw edge exposed in the hem or seam”; nor anything about concealed effect or thread covering; although he mentions both of them later. In each of the four Lewis claims suggested to Arbetter in the interference, one needle-thrust is described as “entering under the edge” of the superimposed layer. All the claims were rejected on the first Arbetter machine patent, just as was done in the Lewis machine application,' although the drawings of both the Arbetter applications (seam and machine) show the second needle-thrust going in and coming out wholly on the top of the superimposed layer. Lewis at once called attention to this, and also filed an affidavit carrying the date of his invention back of the Arbetter machine application, and submitted a sample of his prior work. He also argued that the reference was not pertinent because one of the Arbetter rows of stitches is parallel to the edge. The examiner thereupon withdrew the reference and allowed two. of the Lewis claims, and rejected the other. This rejection raised the question of the function of the angular or inclined stitch, and in discussing this Lewis’ attorneys took the position referred to by defendant’s counsel, and by the District of Columbia Court of Appeals. But this was done after the other claims relating to the needle to edge relation had been allowed. The matter of the greater binding effect, preventing the pulling of the woof from the warp, and vice versa, in case of a raw edge, was argued back and forth, and all the claims rewritten, and somewhat amended to meet the examiner’s views and distinguish the references, until ten claims were finally allowed, and the application was about ready for patent when the interference was declared, and four of the Lewis claims given to Arbetter, as already stated.

This seam-application record shows simply that Lewis claimed the same things as to the diagonal stitch as he did in his patent application, but in order to emphasize its importance in reference to certain of his claims he brought in the arguments about greater binding effect, protecting the warp of woof and the raw edge; and in the explanation of July 6, 1906, he describes the concealed effect. There is nothing in this feature of the case which seems to me to be of any great importance, either in arriving at the true construction of Lewis’ patent, or as bearing on infringement.

As to the result accomplished by Lewis, and to what extent his invention advanced the art.of blind-stitch felling by machines, I think Lewis and Arbetter may be spoken of in nearly equal terms, because they both got the same result. Whether one has made a greater commercial success than the other is important in itself, but not in this connection; by fully operative devices they accomplish practically the same result in concealed effect felling. Nor is the record complete enough to justify any comparison as to commercial success, though it is apparent that defendant has a large business in manufacturing and licensing its machines.

[2] The question of infringement of claims 72 and 76: Lewis and Arbetter reached like results by different paths, each unaided by the other. This record is happily free from any suspicion of piracy on either side. Each party worked out his own ideas in his own way, and by different means and different operations got the same result. It is even possible that Arbetter gets a slightly improved result, though I think they are practically identical. Suppose Arbetter had been first in the field, and he were here suing Lewis for infringement. Would not the diagonal presentation of the goods to the needle, the bending them around the back guide, piercing them vertically instead of horizontally, making both needle strokes in the diagonal plane, ‘‘¿hogging’’ the needle instead of oscillating it, making the lower set of stitches nearer to the turned edge, and tlie absence of the bender in Lewis’, together be taken as pretty clearly showing different means and a different mode of operation? These inventors, after immense study and labor, have produced wonderful pieces of machinery; but they have, done this independently of each other. Rach has produced his own, and each machine seems to be quite distinct from the other. When a supposed infringer appropriates a valuable invention, and then tries to shield himself behind a paper disclosure found in some publication never heard of by the inventor, there is some satisfaction in being able to find that the paper does not really anticipate. But when two inventors have reached identical results from diverging paths out of sight of each other, there is also satisfaction in being able to find that they reached those results by means and operations essentially different, and thus give each that which should he his own. I think the Arbetter machine performs a substantially different function in a substantially different way to obtain substantially the same result, and that infringement does not exist.

It is true, as urged, that the fundamental principle of both machines is the relative relation of folded edge to needle path, and that their absolute positions are wholly nonessenlial. I think it is also true, as argued by plaintiff’s counsel, that it is immaterial to the infringement of claim 72 that in defendant’s machine the stitches in the base layer are parallel with the edge, whereas in the Lewis machine the stitches in both the base layer and the superimposed layer are diagonal to the edge; the only essential thing in either case being that the stitches in the superimposed layer shall be diagonal to the edge. But giving these concessions all the logical force ( am able to do, and giving as full effect as I can to Machine Co. v. Murphy, 97 U. S. 120, 125, 24 L. Ed. 935, Hoyt v. Horne, 145 U. S. 302, 308, 309, 12 Sup. Ct. 922, 36 L. Ed. 713, the Button-Hole Machine Case, 61 Fed. 958, 10 C. C. A. 194, and the decisions there fully reviewed by Judge Putnam, I still think infringement is not shown. While I think the needle to edge relation, which is the gist of the Lewis invention, is present in both inventions, and that the products of both are alike, these results arc obtained in such a different way that infringement is avoided. As already stated, the only device essential to produce the angular presentation of the goods in the Lewis machine and discovery is the angled back guide, while in Arbetter’s machine the corresponding means are a rocking hanger and a bender. These are so completely distinct as to avoid infringement, even though Lewis he considered a primary inventor. Moreover, the two modes of operation are different. If the needle-shogging in Lewis be taken as equivalent to the needle-rocking in Ar-better, still the presentation of the goods to the needle is made at right angles in Arbetter, together with the betiding operation, and obliquely in Lewis. If, on the other hand, the needle-rocking in Arbetter be taken as the equivalent of Lewis’ diagonal presentation of the goods, there is still the additional bending operation, which should, I think, be regarded as part of the function of presenting the goods, and not as an additional element to Lewis’ combination which of itself would not avoid infringement.

The question of infringement of the needle-guide claims: As to these claims I think it is sufficient to say that defendant does not use the Lewis needle guide, but merely the old form found in a number of earlier patents. If defendant’s needle is put under tension, it is a matter of accident, and does not result from the construction or hanging of the needle, or from the way the guide is made or adjusted. The patent construction is novel, and the claims valid, but not infringed.

The “inclined-hook” claims: Infringement of two claims is charged. The claims are these:

“Claim 42. In a sewing machine, the combination with a reciprocating needle, of a rotary hook containing a bobbin and inclined to the path of reciprocation of the said needle, the rear face of the hook co-operating with said needle to take the loop.”

“Claim 71. In a sewing «machine for blind stitching, the combination with a needle, of a rotary hook inclined to the path of reciprocation of said needle, said hook having an inclined cast-off surface and an offset for guiding the thread onto said cast-off surface.”

The Lewis old machine, put in use in February, 1901, does not have an inclined hook, but it was claimed in original count 36 of his machine application of August 2, 1902. I shall assume the validity of the inclined hook claims, and consider only the question of infringement. The opposing experts disagree in their respective descriptions of the operation of the Lewis and Arbetter hooks, though I can see no reason why they should. It is perfectly clear that the Lewis claims count on improved cast-off of the needle thread over the face of the hook, and the inclination claimed must necessarily be a leaning over of the face of the hook towards the bed plate of the machine, so that the angle between the plane of needle reciprocation and the face of the hook increases from nothing up to 20 or 30 degrees when measured back from the hook-face, and 150 to 160 degrees measured the other way. That is, Lewis inclines the hook downward, so that gravity may assist in the cast-off, so far as gravity can do so in a rapidly flying movement. Arbetter, on the other hand, inclines the hook upward, so as to get a more difficult cast-off than he would if the hook were left at right angles to the path of needle reciprocation. This is substantially admitted by Mr. Browne, complainant’s expert, is expressly testified to by Mr. Hains, defendant’s expert, and is perfectly clear from the wooden models used on the argument, as well as from a comparison of the Lewis patent with the Arbetter machine. Defendant does not use the improved cast-off; hence there is no infringement.

Every angle of this complicated case was covered by the four oral arguments and the printed briefs in the clearest and most satisfactory wajr, so that the work of the court has been comparatively easy.

There should be a decree dismissing the bill because infringement does not appear.