This is a suit for the alleged infringement of three letters patent, granted to John G. Lewis, of St. Louis, Mo., for improvements in sewing machines used in blind stitching, wherein the stitches enter and emerge on the same side of the cloth without penetrating to or showing on the other side. Of these letters patent, No. 731,695 was granted June 23, 1903, on an applica¿ion filed October 18, 1897; No. 731,696 was granted at the same time on an application filed October 23, 1902, and No. 746,853 was granted December 15, 1903, on an application filed January 27, 1900. All were subsequently assigned to the complainant. The defendant is the user of a blind-stitch sewing machine, constructed in measurable accordance with the specification of letters patent No. 679,553, granted July 20, 1901, to Charles A. Dearborn of New York City on an application filed March 22, 1901; and the defense of the suit has been assumed by the Union Specialty Manufacturing Company, the present manufacturer of such machines. The defenses interposed are want of novelty and noniniringement. Upon a consideration of the proofs the circuit court dismissed the bill, and the complainant has appealed. A brief memorandum opinion, shown in the record, discloses that the circuit court was disposed, in a general way, to regard the Lewis improvements as possessing sufficient of novelty to sustain the patents therefor, but as coming so late in the art and being so slight as not to embrace the somewhat modified forms of construction in defendant’s machine, citing Railway v. Sayles, 97 U. S. 554, 24 L. Ed. 1053, and Sander v. Rose, 58 C. C. A. 171, 121 Fed. 835.
In the beginning the bill charged infringement of claims 1 to 4 and 10 to 22 of the first Eewis.patent, claims 1 and 2 of the second Lewis patent, and claims 2, 8, 9, and 14’ of the third Lewis patent, but claim 10 of the first patent and claim 14 of the third have since been withdrawn from consideration. The art of blind stitching by machinery did not originate with Lewis, but was disclosed in several prior letters patent, some antedating his first application about 20 years. Most of the earlier machines were adapted to sewing leather, felt, and the like, and were incapable of successfully blind stitching the more flexible and thinner materials from which clothing is made. One *952more nearly approaching success in that regard was the Borton and Wilcox machine, covered by letters patent Nos, 255,576 and 255,580. It was adapted to sewing welts or hems' upon hosiery, underwear, and other knit goods, but the stitching' done by it not infrequently penetrated to the other side of the goods, and it also produced such a groove, or drawing of the material, along the line of sewing as to make the seam quite noticeable on the fair side. These objections, ■although not of much importance as respects undergarments, were sufficient to prevent the use of the machine in making outer garments. Another machine, designed to be used in sewing flexible cloth, but not shown to have been ever actually so used, was that of Hoffman and Meyers, covered by letters patent No. 207,035. But it was not adapted to blind stitching garments in which there are occasional cross-seams, as in trousers’ bottoms; for its construction and adjustment were such that, when any such increased thickness in the material was encountered, the outer layer — that is, the one in which the stitching is to be blind — would move away from the path of the needle, and would not be caught by the stitches. Still another machine is said by the defendant to have been used by one Gammons in successfully blind stitching hosiery, trousers’ bottoms, and the like, before the date of Lewis’ invention, but the evidence thereof is of such a character that, when it is considered that Gammons, a year or so thereafter, applied for and obtained letters patent covering his machine, without disclosing or claiming its applicability to such work, we feel constrained to hold that this alleged prior use is not established with the requisite certainty. The Barbed Wire Patent, 143 U. S. 275, 12 Sup. Ct. 450, 36 L. Ed. 161; Deering v. Winona Harvester Works, 155 U. S. 286, 15 Sup. Ct. 118, 39 L. Ed. 153. Besides, the Gammons machine, like that of Hoffman and Meyers, belongs to the class in which the construction and adjustment are such that they will not do satisfactory work over cross-seams. While fully recognizing that the earlier machines and patents'show a steady and marked progress in the art of blind stitching by machinery, we are yet of opinion that Lewis was the first to devise a machine capable of doing satisfactory work upon the more flexible and thinner fabrics, and of successfully overcoming the obstacles presented by cross-seams and like inequalities in the thickness of the material operated upon. In point of result, these are the distinguishing features of the improvements covered by his first letters patent. He says in the specification:
“The object of my invention is to provide a blind stitching machine which will work properly on any kind of cloth on which blind stitching can be done by hand, and which shall be simple of construction, and not liable to get out of order. My invention consists in the combination, with a suitable stitch forming mechanism, of a normally stationary back guide — i. e., a guide on the back or opposite side of the doth from that which the needle enters — and means of holding the work up to said guide, and in various other novel features and details of construction all of which are described in the following specification and pointed out in the claims affixed hereto.”
And again:
“The (back) guide 43 is cylindrical in form and slightly tapering, so as to form a stretching device for the work. * * * The needle 85 is formed, as *953shown in Fig. 4, so that the sido of the needle next the back guide is substantially a straight line; that is, the point of the needle, instead of being in a line with the center of the needle, is on a line or nearly on a line with the side of the needle.”
And still again:
“As the back guide is stationary during operation, work can be done on very thin cloth, and the machine can be run at a very high rate of speed, neither of which results can he accomplished by machines having the reciprocating or rotary guides heretofore used. The operation of the machine is also made more perfect by the form of the needle used. The point of the needle must pass through the center of the layer ol' goods next the guide, and If the ordinary form of needle with the point on the center line were used, the goods would be wedged between the guide and the rounded portion of the needle, with the result that thin goods would be cut through, so that the stitch would show.”
Claims 2, 4, 18, 19, and 20, in varying terms, specify, among other elements, a normally stationary back guide and means for holding the work in position around such guide. Several prior patents are cited against these claims, but we regard the combination specified in each as unanticipated and otherwise patentable. The essence of the invention is in the mechanism whereby the moving cloth is held closely around the back guide, and the outer layer, in which the stitching is to be blind, is held in such a fixed and continuous relation to the path of the needle that not only can such stitching be satisfactorily done on more flexible and thinner cloth than before, but the stitches can be made to penetrate the added material at cross-seams, and to catch properly in the outer layer as elsewhere.
Claim 1 differs from those, just mentioned, in that it omits the means for holding the work in position around the stationary hack guide, and for that reason we think it was without the requisite novelty. And we think; that claims 3, 17, 21, and 22 must also fail, for they omit the requirement that the hack guide be normally stationary.
Claims 11 to 17 call for a needle which is wholly or partially side-pointed, and this is relied upon as having given patentable novelty to them. But we feel constrained to hold otherwise. True, when the cloth is thin, there is an advantage in having the needle side-pointed, but such, needles were old in the sewing machine art, and their applicability to the present use was plainly suggested by the Wilcox machine, covered by letters patent No. 230,212, dated July 20, 1880. Tt was a machine for sewing covered wire upon hats and other articles, the pertinent portion of the specification being as follows:
“With machines using the ordinary needle, having the point in the cent ral line or axis, the stitching cannot be made as close to the wire as desirable, since the bevel near the point would, if the latter passed close to the wire, force it sidewise, tearing the covering of the wire, or breaking or bending the needle. In this invention a needle having the point in the plane of one of its sides (that next to the wire) is employed.”
Although that was a different use, we think it was so closely related to the present one as to come within the rule that, if the new use be so nearly analogous to the old one that the applicability of the devise to its new use would occur to a person of ordinary mechanical skill, it is only a case of double use. Mast, Foos & Co. v. Stover *954Manufacturing Co., 177 U. S. 485, 493, 20 Sup. Ct. 708, 44 L. Ed. 856.
. _ It is objected that no machine, conforming to the specification of the first patent, was ever used or placed upon the market by_ Lewis or the complainant, the fact being that complainant’s commercial machine embodies some of the features of each of several of Lewis’ patents and pending applications. But that this objection is not tenable is made perfectly plain in the recent decision in Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U. S. 405, 28 Sup. Ct. 748. 52 L. Ed. 1122, where it is held, following prior decisions, that, during the life of his monopoly, a patentee is under no obligation to use or place upon the market a device or machine embodying his invention.
The question of infringement turns upon the character of Lewis’ invention. We regard it as neither primary nor a slight improvement on the prior art, but as possessing enough of patentable novelty to command a reasonable range of equivalents. The defendant’s machine embodies every element or its equivalent, of the claims which we sustain, and accomplishes substantially the same result in substantially the same way. Probably there would be no infringement, if it fully conformed to the specification of the Dearborn patent No. 679,553, but it does not, and it is because of this that it accomplishes substantially the same result as the Lewis invention. The specification of the Dearborn patent does not call for mechanism adapted to so hold the moving cloth that the outer layer, in which the stitching is to be blind, will have a fixed and continuous relation to the needle path, but calls for a back guide yieldingly supported by a coil spring, so that it will automatically move backward to accommodate any increased thickness of the material incident to cross-seams. As this movement would permit the outer layer of cloth, which is next to the guide, to move away from the path of the needle, thereby preventing the stitches from catching therein, the manufacturer, as is shown by the machines produced in evidence, resorts to the expedient of making the supporting spring so strong and unyielding that the guide is not affected by cross-seams, but' remains normally stationary, any increased thickness of the material being accommodated by an opening in the presser foot, which, with that feature in its construction, is the equivalent of Lewis’ means of holding the work in position around the guide. There is also a marked difference in form between defendant’s back guide and that of Lewis; but as both operate to guide and stretch the goods preparatory to the stitching, and as defendant’s guide is appreciably rounded, both in the direction of the feed and transversely thereto, we think this difference is not enough to destroy the equivalence or to avoid infringement of such of Lewis’ claims as are limited to a guide having its face rounded transversely to the direction of the feed, or having a rounded tapering face.
We now come to the second patent. It is later than the Dearborn patent, in measurable accordance with which defendant’s machine is made, but that fact is immaterial, because in respect of the matter *955now to be considered, defendant’s machine docs not conform to the Dearborn patent. Two claims are said to be infringed, but they are so nearly alike that we need refer only to the one, which reads:
“2. In a sewing machine a back guide, around which the goods are adapted to be held, a needle co-operating with said back guide, means for reciprocating said needle longitudinally and means for causing said needle to recede laterally from said back guide after its point has entered the material.”
It is not questioned that defendant’s machine embodies all the elements of this combination, save the last, but it is insisted that that element is omitted. Of course, if that be true, there is no infringement. The defendant’s needle is slightly inclined, and the path in which its point reciprocates past the back guide is nearer thereto than is the like path of its shaft. As the needle passes the guide, the distance between them slightly increases from the point to the other end of the needle; the purpose in this being to avoid any wedging or cutting of the goods between the guide and the body of the needle. The reciprocation of the needle is longitudinal, no lateral motion being imparted to it in any way. So tile question arises, does such a machine embody means for causing the needle to recede laterally from the guide, within the meaning of the claim? We think the answer must be in the negative, and for these reasons: First, it requires some straining of terms to speak of defendant’s needle as receding laterally from the guide, the natural import of such words being that the needle is moved sidewise, which is not the case with defendant’s needle. Second, the connection in which the word “means” is used in describing this element of the combination, and also the preceding one, indicates that it has a similar meaning in both, and refers to some mechanism other than the needle, the latter being separately specified as one element. Third, the specification, to which reference may be had for the purpose of ascertaining the true meaning of the claim (O. H. Jewell Filter Co. v. Jackson, 72 C. C. A. 304, 140 Fed. 340), shows beyond any question that the words “recede laterally” and “means” are used with the meanings just attributed to them; that is, they refer to a sidewise movement of the needle and to operating mechanism whereby that movement is produced. It follows that defendant’s machine wholly omits one element of the patented combination, and hence the charge of infringement fails.
We come then to the third patent, three claims in which are said to be infringed. One of them reads:
“2. In a sewing mac-bine for blind stitching, a guide for the goods and stitch forming mechanism, said stitch forming mechanism being provided with a needle inclined to the path of its reciprocation.”
And in the specification it is said of the inclined needle in this combination:
“The needle 49, instead of being set straight in the needle bar 29 in the usual manner, is inclined forward (that is, slightly toward the back guide -around which the cloth is held), as shown in Figs. 1 and 15. This inclination of the needle enables the point of the needle to be brought close to the hack guide, so as to pierce very thin goods, and at the same time prevents the shaft *956of the needle from chafing against the back guide, which would tend to cut through the goods and allow the stitches to show.”
The defendant does not question the validity of this claim, but only that its needle is so inclined as to sustain the charge of infringement. Its needle is curved and reciprocates in circular paths, while the one shown in Lewis’ drawings reciprocates in straight lines, but' that a curved needle may be given such an inclination as to enable its point to be brought close to the back guide, so as to pierce very thin goods, and at the same time to prevent its shaft from chafing against the guide and cutting through the goods, is plainly shown in a drawing having the inclination somewhat exaggerated for purposes of illustration, which we here reproduce from one of the briefs:
There is some difference of opinion among the witnesses as to whether or not defendant’s needle is so inclined, but a careful consideration of the evidence bearing thereon fully satisfies us that it is, and also that its inclination is not accidental, or the result of straining in use, as suggésted by defendant, but is produced by an intentional construction and adjustment of the lower part of the arm carrying the needle, as is shown by other machines and needle arms of the same manufacture produced *in evidence. We hold, therefore, that infringement of this claim is established.
Claims 8 and 9 of this patent relate to an entirely different subject-matter, and we deem it sufficient to say in respect of them that, if they can be regarded as possessing patentable novelty, the interpretation which must be placed upon them, in view of the prior state of the art, is such that the defendant’s machine cannot be held to infringe either of them.
The decree of the circuit court is affirmed as to claims 1, 3, 11 to 17, inclusive, 31 and 33 of the first patent, claims 1 and 3 of the second patent, and claims 8 and 9 of the third patent, and is reversed as to claims 3, 4, 18, 19, and 30 of the first patent, and 'claim 3 of *957the third patent, with directions to enter a decree for complainant upon the claims as to which the former decree is reversed, the complainant to recover one-half of its costs in this court, and the decree in the circuit court to be without costs in that court to either party up to the time of such decree. See Ide v. Trorlicht, etc., Co., 53 C. C. A. 341, 115 Fed. 137; Fairbanks, Morse & Co. v. Stickney, 59 C. C. A. 209, 123 Fed. 79; Johnson v. Foos Mfg. Co., 72 C. C. A. 105, 141 Fed. 73.