The invention is particularly directed to an improved form of work arm on the frame or easing of a sewing machine commonly known as a “feed-off-the-arm” machine. In sewing tubular articles, it is customary to employ a cylindrical work arm lengthwise of which the fabrics advance as they *547are being stitched, instead of the usual flat work plate on which they are supported while they are sewed.
If the work travels up the arm (toward its supporting end), the mechanism is called a “feed-up-the-arm” machine; if it travels down the arm (away from its supporting end), it is called a “feed-off-the-arm” machine.
The specification of the patent in suit (at page 2, line 91) sets forth in detail the advantages of the invention as follows:
“In many machines heretofore designed and commercially used for analogous work, it has been common to use a work arm extending beneath and parallel to the bridge of the framework, the feed operating to advance the work onto the free end of and along said arm toward the end at which it is supported from one of the end standards of the machine. Therefore, in sewing up tubular work, such as stockings and the arms and legs of garments, for which such machines are especially designed, the work must accumulate on and about this arm until completed. Thereupon, in order to free and deliver the completed work from the machine, it is necessary to arrest its operation, elevate the pxesser foot, needles, etc., cut or break the threads, and then pass the work, as gradually extended and flattened, back along the work arm through the small space between the presser foot above and the throat plate, feed surface, etc., below; or in other words, the work must traverse in a reverse direction and with the machine stopped, the same path which it was caused to traverse, by the feed of the machine during the formation of the seam. It is obvious that this operation of removing the work necessarily involves the double loss of consuming much of the operator’s time and of putting the machine out of use for a material portion of each day’s work. Furthermore, removal of the work in this way frequently so disturbs the positions of the several threads that the correct formation of stitches does not proceed immediately upon again starting the machine, and imperfect stitches are formed at the beginning of the new seam.
“In the machine of the present invention, the losses and difficulties above referred to are obviated, as the feed operates to advance the completed work off the arm (i. e., longitudinally thereof and toward its free end). Furthermore, the peculiar shape and disposition of the work arm affords the desirable conditions of having the feed advance the work away from the operator, and, of affording ample space above and below the work arm, unobstructed by other parts of the machine, so that the operator may easily manipulate and arrange the work on or about the work arm, and may unobstructedly view the operation of the mechanism, without having to assume an uncomfortable position or reach around or behind other parts of the mechanism. In operating the machine herein shown, the operator may preferably sit with her right side toward the machine, in which position she will be able to look in the direction of the feed directly along its path and may comfortably extend her right hand and arm across the underside of the bridging portion of the work arm and readily handle the work and arrange it on the work-supporting portion of the arm to rapidly present it to the feed mechanism. * * * ”
The general advantages claimed for the invention in the above extract from the specification would seem to have been confirmed by later results. But, at the time Borton designed his work arm, “feed-off-the-arm” sewing machines involved in themselves no new inventive thought. The defendant had since 1900 filled special orders for a device in which it had reversed the feed and slightly changed some of the parts on its standard machine, so as to feed the fabric off the end of the work arm. The old French patent, No. 17,744 of December 18, 1856, to Journaux; British patent, No. 7,939, of 1887, to Baker; United States patent, No. 380,981, to Bollman, with its accompanying British patents; United Stales patent, No. 999,718, to Hawes; United States patents, Nos. 839,075, 839,076, and 839,077, to Onderdonk — all showed devices for feeding the fabric off the end of the work arm. Although the complainant was the first to manufacture a “feed-off-the-arm” machine on a large scale, it nevertheless came into a field where numerous patents for “feed-off-the-arm” machines had issued, and the idea was certainly an old one. Commercial development in respect to “feed-off-the-arm” machines had been slow, but the delay seems to have been due to disinclination to change the customary model. There has been no satisfactory showing that the problem in respect to such machines was unsolved, or that the designs already made wore not practically adapted to commercial feed-off-the-arm machines.
Certainly defendant’s auto-lap machine, which had a substantial sale, approximated the Onderdonk device, and its tandem inter*548lock machine was not far from the Onderdonk model, though having something' like the bridged work arm of the patent in suit. It is contended by complainant that the auto-lap machines never worked well, and that they have been supplanted by the Borton invention. Nevertheless, 693 of these auto-lap machines were placed on the market by the defendant between 1920 and 1927, and we cannot suppose that a design based on the Onderdonk patents, which had a continuous sale over a period of seven or eight years, was impracticable merely because it was superseded by machines which turned out to be preferable.
There can be no doubt that the Borton patent was not basic, but for a mere improvement, and must be limited to something other than any kind of work arm, cylindrical in form, off which fabrics may be fed. Indeed, the specification begins by ■describing the invention as for “improvements in sewing machines, ’ ’ and goes on to say that “the present invention is partly directed to an improved form of work arm on the frame or casing of the machine.” It recognizes that this “work arm” is a part of the machine different from the three-part frame or casing, which includes the main standard, the sewing head, and their connecting bridge — elements common to the frame work of sewing machines. As Arthur S. Browne, defendant’s expert, stated in his affidavit on the motion:
“This familiar form of machine, shown in the drawing of the Wilson sewing machine, thus has a framework and a work arm or plate, and the framework, as distinguished from the work arm or plate, comprises the three elements: (1) Sewing head; (2) vertical standard; and (3) main bridge connecting sewing head and the vertical standard.
“When the machines came to be built in which, instead of a flat work plate over which the fabric to be sewn passes, there is used a cylindrical work arm, so'that the work can pass along in tubular relation, such arm began to be called the ‘work arm’ or sometimes the ‘work cylinder.’ It is an ‘arm,’ because it projects from the main frame of the machine and its outer'end is free, but in all re"spects it is the equivalent of the older work plate, and its relation to the framework is the same.
“It is this distribution of the parts of the sewing machine into the framework on the one hand and the work arm on the other, which is set forth by Borton in his specification and adhered to in all his claims.”
Borton’s second improved form of “work arm,” to which claims 7 and 12 are specifically addressed, is thus described in his specification :
“Figs. 10, 11, and 12 show a somewhat modified form of the hollow work arm embodying the characteristics and principles of constructions above set forth. In this ease the horizontal bridging part 400 of the work arm extends forward, from the supporting end of the aim which joins the upper part of the front face of the standard 40 at an angle of less than 90 degrees to the main portion of the framework, to, an elbow at 401, and from the latter the part 402 depends in an approximately vertical direction to a second elbow 403. Prom the elbow 403, the approximately horizontal work-supporting end or part 404 of the work arm extends toward the left .and rearward (Pig. 10) and terminates beneath an overhanging portion of the head of the machine in substantially the same manner as the corresponding portion of the work arm of Pigs. 1 to 3. The open space 405 (Pig. 12) thus provided for the right am of the operative affords still further freedom of movement in manipulation, arranging and advancing the work along the part 404 of the am. * * * ”
The claims relied upon on this appeal are 7 and 12, which are as follows :
“7. In a sewing machine, the combination with stitch-forming mechanism including work-feeding means, of a work arm free at one end at which the stitch-forming mechanism operates and supported from its other end which joins the framework of the machine, said arm extending from its supporting end outward from the framework thence downward and thence inward toward the framework to its said free end, said work-feeding means operating to feed the work toward or off the free end of the arm.”
“12. A sewing machine having a work am extending outward from the framework of the machine, thence downward, and thence in an approximately horizontal direction to its outer free end, said work arm being supported from its end which joins the framework of the machine, thereby providing clearance beneath the arm for a hand or arm of an operator.”
Now it is quite clear that the Onderdonk patents described a device which would enable the operator to have free access to the work am and to keep the fabrics to be stitched in view at all times. It fed off the am, and if the arm were conveniently positioned in respeet to the'table on which the machine rested, it is hard to see why there *549would not be an accessibility and visibility equal to that in the patent in suit.
But the Onderdonk work aim involved a rectangular arrangement of the parts of the frame and the work arm of the machine. Both the tandem interlock and the auto-lap machines of the defendant, which resembled the Onderdonk device, had a four-directional work arm, and necessarily involved the inconvenience of numerous angles around which the feed mechanism had to go.
The patent in suit eliminated one of these elbows and provided a bridge, under which the arm of the operator might pass in order to manipulate freely the fabric to be stitched on the work arm. The defendant in its so-called thirly-five-seven machine, which is the alleged infringing device, eliminated still another elbow, which the patent in suit shows at the intersection of the main frame and the horizontal bridging part numbered 400. The thirty-five-seven machine had a free space for the arm of the operator between the bracket on which' the device was supported and the easing, by positioning the casing at a sufficient angle away from the bracket. While it is possible to regard this structure as the mechanical equivalent of the Borton device, it is certainly not the Borton desig-n.
The Borton specification thus defines the scope of the “work arm” and the “frame” at page 2, line 26:
“Work Arm and Frame of Machine.— The framework of the machine comprises a main standard 40 of the hollow column or box type, a standard 41 of the ribbed column type, a bridge 42 rigidly connecting the standards and a work arm 40b.”
The second Borton structure, on which Complainant’s commercial machine is founded, brings the work arm from the top rather than from the bottom of the main standard 40, but Borton’s method, in both his structures for obtaining free working space for the arm of the operator, is by extending the work arm out from the frame, whereas in the thirty-five-seven machine the space is obtained by hanging the casing from a bracket.
There was a sharp conflict in the affidavits submitted on the motion for a preliminary injunction as to the relative advantages of the two structures. Defendant’s machine appears to be somewhat simpler than Borton’s, and more free from elbows that are said to give rise to difficulties. While wo find no anticipation of the patent in suit in the Brit-' ish patent to Baker No. 7,939 of 1887, or in any of the other patents cited, there is a sufficient body of prior art to limit the patent to a substantial conformity with the structure disclosed. United States patent, No. 380,981, to Bollman, is of importance. It shows a machine where the casing is hung on a bracket. It is true that the work arm shown is not cylindrical, but it would hardly involve invention to adapt the cylindrical work arm of Onderdonk to the Bollman machine. At page 1, line 66, of the Bollman patent, it is said that:
“The piece B”, which connects the table to the arm, is very narrow, * * * and by it the table bangs freely forward, so that the material to be sewed can be laid around it in the form of a tube, the open and not yet sewed edges or selvages passing one on each side of B". As they pass then farther on toward the needles on the farther end of the table, they are gradually laid one over the other, as is illustrated in Big. 3, which may be done by hand or by proper self-acting guides.”
It is hard to see why there is not a reasonable basis for defendant’s claim that the Bollman patent suggested the general form of structure adapted in its thirty-five-seven machine. The foregoing extract from the Boll-man patent shows that the space between the bracket and the frame was to enable the operator to manipulate the fabric to be sowed. Nor can it be said that the space is too narrow for this purpose. That depends on the scale of the drawing. It might easily be made ample and that it was intended to afford room enough ought to he assumed from the statement in the patent that the material may be placed “by band.”
To be sure the work arm in Bollman was underneath the frame, while the work arm both of Borton and the alleged infringing device extends forward of the frame. But claims for a work arm extending forward from the frame were canceled when the patent in suit was before the Patent Office as insufficient, in view of Bollman, and Borton made no effort to secure any claims that did-not require the work arm to be elbowed. The work arm in the thirty-five-seven machine extends out from the casing and is not elbowed in the way shown in the patent drawings.
Complainant, in January, 1915, sued defendant for infringement because of sales by the latter of its tandem interlock machine. That suit was discontinued in 1917 under a license agreement, whieh provided that it should not prejudice either party in any'suit for infringement after its termination. The license agreement expired in 1921, and thereafter the defendant put its auto-lap machine *550on the market without interference or protest from the complainant, and continued the sale of. that for seven or eight years. No further action was taken until defendant put out its thirty-five-seven machine in 1927. The present suit was then brought, and the motion for a preliminary injunction made. Judge Bryant denied the motion, and held that the questions of validity and infringement should await final hearing.
In view of the reasons presented for regarding the Borton patent as limited to substantially the structure disclosed, and of the proofs offered on the motion that the defendant’s thirty-five-seven machine is in some respects closer to the prior art than to Bor-ton, and is at any rate quite different from Borton in features thought to be of special merit, we cannot say that the holder of this unadjudieated patent is entitled to an injunction pendente lite.
The order is accordingly affirmed.