"In the Exhibit Bellis Machine the loopers take the supplemental thread, and push it up against the shanks of the dial needles, lifting the junctions of the loops into the path of the cylinder needles, which rise through them; and thereby two things are done, to wit; (1) to form within the angle of the two sets of needles the loops of supplemental thread which are to constitute the backing in the finished fabric, and (2) to bring the thread, w, over a (cylinder) needle to be interlaced with a mesh formed by such needle.' Bellis patent, p. 3, line 80. When later, the old loops of body thread on the cylinder needles are 'cast off,' the parts of the supplemental thread over the cylinder needles are cast off with them, resulting in the interlacing of the supplemental thread with the meshes formed by the cylinder needles, or, as the Fisher patent describes it, the extra thread is secured on one surface of the fabric, forming floating loops.' "In the defendants' machine the loopers take the supplemental thread, and bring it over the shanks of the dial needles, and press it down between those needles, and thereby the same two things above mentioned as in the Bellis machine are done, to wit: (1) to form within the angle of the two sets of needles the loops of supplemental thread which are to constitute the backing in the finished fabric, and (2) to bring the thread, w, over a (dial) needle to be interlaced with a mesh formed by such needle.' Bellis patent, p. 3, line 80. When, later, the old loops of body thread on the dial needles are 'cast off,' the parts of the supplemental thread over the dial needles are cast off with them, resulting in the interlacing of the supplemental thread with the meshes formed by the dial needles, or, as the Fisher patent describes it, the extra thread is secured on one-surface of the fabric, forming floating loops.' * * "In both complainants' and defendants' machines, it is the upper crest of the crinkled supplemental thread (in black in diagram) which is brought over and lies outside the needles, and is afterwards caught into or interlaced with the meshes of the fabric, and in both machines it is the lower or bottom wave of the crinkled supplemental thread which is inside the angle of the two sets of needles, and which afterwards becomes the backing or floating loop in the fabric. In one case the thread is caught into or interlaced with the meshes formed by the cylinder needles and in the other case with the meshes formed by the dial needles. In other words, the fabric from defendants' machine has the looped supplemental thread interlaced with the front wales of the fabric, while the fabric from the Exhibit Bellis Machine has the looped supplemental thread interlaced with the rear wales, that is, when you look at the fabric on its looped face. "Defendants say that Fisher's loopers cooperate with one set of ne dles only-the dial needles. Fisher's loopers cooperate also with his cylinder needles, because these loopers, after bringing the thread down over the dial needles to form the crinkle, as shown in his Fig. V, continue to hold down the thread in loops, not only until it has been cast off the dial needles with the loops of body thread, but the loopers still continue to hold down the extra thread loops after they have been brought down onto the loops of body thread of the cylinder needles, until tied in by the loops cast off by the latter, and furthermore the loopers are then further depressed a trifle to tighten the extra thread loops' (Fisher patent, p. 318. fines 33-34). As will be seen by the shape of the looper camn, in Fig. IV of Fisher patent (p. 326), the loopers are held down all the way round the cylinder, except where momentarily raised (by cam 31) to take the extra thread from guide eye 34. Fisher's loopers thus do co-operate with both sets of needles by first forming the extra thread loops over the dial needles, and by holding those loops onto the loops formed by the cylinder needles until the extra loops have been tied in. "In both the machine of the Bellis patent sued upon and defendants' machine, the shanks of the dial needles act in conjunction with the loopers to form the loops of the supplemental thread that are to be engaged with the knit fabric, and in both it requires the co-operation of both sets of needles and the loopers in order to form the said loops, and to cause them to be engaged with the fabric, so that the loops of supplemental thread will project from the back of the fabric to form the backing therefor, as is desired in the product of the machine." the Bellis loop is formed by having the extra thread pushed against the dial needles and out over the cylinder needles. The Fisher loop is formed by having the extra thread pushed against and over and down between the dial needles carrying a body thread. When the dial needle withdraws, the body and supplemental threads are cast off together, and thereby the two become interlaced. This operation is described in the Fisher patent as follows: "As the dial needles continue their outward `movement, the extra thread slips back over the ends of the latches onto the shanks of the needles, where it is left with just a trifle of slack, so that as the dial needles move in again, having engaged with the main thread, they draw it easily under the extra thread, which is cast off over the needle hooks at the same time as one loop of the main thread, and is caught on the inside of the fabric by the loops of the main thread." The operation and result of the two devices thus differ in that in the one the supplemental thread is pushed from the inside outward over the cylinder needles, and thereby firmly interlaced in the body of the fabric; in the other the supplementary thread is pushed in from the outside, and interlaced, "secured," or "caught on the inside of the fabric," as the patentee says. In the light of the foregoing discussion, it is seen that not only was Bellis the first to produce a fleecing ribber and a fleece rib, but he was the first to produce a machine which interlaced loops on a double ribbed fabric. He was the first to conceive a practical method and means for obviating the mechanical difficulties attendant upon the introduction of the extra needles. It may, therefore, fairly be said that he fulfilled the requirements of a generic invention, in that he devised a machine which performed the function of looping on a ribbed fabric, a function never performed thereon by any earlier machine, and thereby produced a result never before produced, namely, a ribbed fabric capable of being fleeced, and that in doing this he exercised invention of a high order. And inasmuch as there is no prior art limiting Bellis as the first to produce a machine capable of providing loops on a ribbed fabric adapted to be fleeced, we think his invention is entitled to be regarded as of such primary character as to embrace defendants' machine, which performs the same functions of looping and engaging a thread in the fabric by a combination of means covered by the terms of the claims in suit, and operating on the same principle, with the same resulting fleecing capacity. It is argued by defendants, and found by the court below, that the Bellis invention is not a meritorious one because its machines are inoperative or impracticable for successful commercial operation. This argument rests upon the following contentions: Bellis, in his subsequent patent, No. 574,129, for an improvement, consisting in the addition of a ring to his prior machine, says as follows: "The upper peripheral edge, s. of the ring, S, also serves to direct the finished fabric, which is drawn by suitable takeups, downward, so that the newly formed backing loops do not come into the way of the jacks to be next actuated, which, as has been found, was the case where no such ring, S. was provided on the machine, the jacks on such machines generally catching the loop previously formed, and tearing on its upward motion the threads apart, destroying in this way the backing, but this is entirely obviated by the use of the said ring, S." All the Bellis machines, with perhaps one exception, are provided with this ring. Scott & Williams, the owners of one half of the patent in suit, have patented and make a different machine, and certain hostile. experts testify that in their opinion the Bellis machine is impracticable. for various reasons. This last contention is met by proof that the Bellis machines with the ring improvement have been in successful commercial operation for some eight years, and the theories advanced to show why the Bellis machines would not run successfully are denied by witnesses who claim to testify from actual observation. The Scott & Williams machines are different in some particulars from the specific construction of the patent, but the testimony of complainants' experts, not disproved by defendants, is to the effect that these machines are within the patented construction. The patented ring is an improvement on the original machine, and remedies a defect therein. But this is a mere detail improvement, and, in effect, merely amounts to an enlargement of the dial of the patent. We are unable to see, therefore, how it affects the meritorious character of the invention. It is frequently a characteristic of generic inventions that their first embodiments work imperfectly, and where the imperfections may be remedied, as in this case, by what amounts to a mere readjustment of relative sizes, such change does not affect the character of the underlying creative conception. In these circumstances, the majority of the court is of the opinion that the decree of the court below as to the defendant corporation must be reversed, with costs. Inasmuch as it does not appear that the defendant Kerman has any connection with the infringement, except as an officer of said corporation, the bill should be dismissed as to him, with costs. Hutter v. De. Q. Bottle Stopper Co., 128 Fed. 283, 62 C. C. A. 652. The cause is remanded to the Circuit Court, with instructions to enter a decree in accordance with this opinion. ROBINS CONVEYING BELT CO. v. AMERICAN ROAD MACH. CO. (Circuit Court of Appeals, Third Circuit. June 6, 1906.) No. 14. 1. PATENTS-CONSTRUCTION OF CLAIMS-REFERENCE TO DRAWINGS. SPECIFICATION AND Where the meaning of the language used in the claims of a patent is doubtful, or it is susceptible of two different constructions, the specification and drawings may properly be referred to for the purpose of ascertaining the true construction of the claims. [Ed. Note. For cases in point, see vol. 38, Cent. Dig. Patents, § 243.] 2. SAME--INFRINGEMENT-SUBSTITUTION OF EQUIVALENTS. A pulley revolving on a central shaft and one having trunions revolving in bearings at the ends are well known mechanical equivalents, and the substitution of one for the other, which is an element in a patented combination, does not avoid infringement. 3. SAME IDLERS FOR BELT CONVEYORS. The Robins patent, No. 571,604, for a belt conveyor, claims 5 and 6 of which cover troughing idlers for supporting such conveyor-belts, were not anticipated, and, in view of the superior utility of the device over those in prior use and its immediate commercial success, must be held to disclose patentable invention. Also held infringed. Appeal from the Circuit Court of the United States for the Eastern District of Pennsylvania. For opinion below, see 142 Fed. 221. Frank Busser and George J. Harding, for appellant. Before DALLAS and GRAY, Circuit Judges, and CROSS, District Judge CROSS, District Judge. The patent in suit relates to new and useful improvements in conveyor-belt apparatus. It was issued to one Thomas Robins, Jr., November 17, 1896, is known as No. 571,604, and was duly assigned to the complainant on the 30th day of November, 1897. The device is intended for the transportation in bulk of coal, ore, minerals, grain, and other like substances. Conveyor-belts vary in width and length according to circumstances, and, to the end that they may convey the material without spilling, it is desirable that they should be concaved by turning up the edges, giving them a troughlike shape. It is also necessary that they should be supported at frequent intervals to prevent them from sagging, not only of their own weight, but more especially because of the load superimposed thereon. This support is afforded by rollers or pulleys actuated solely by the friction of the belt in passing over them, and they are therefore commonly known as "idlers." It is desirable that these idlers should not only support the belt as above indicated but should also permit it to move over them smoothly and with the least possible friction and disturbance. The patent in suit embraces the composition and construction of a conveyor-belt, and also the idlers or pulleys supporting the same. It is with the latter portion of the patent only that we are concerned. The claims involved in the suit are Nos. 5 and 6, and are as follows: "(5) The supporting pulleys, L, K, L, the hollow bearings, F. therefor, and the horizontal and turn-up hollow shafts secured in the said bearings, and the oil devices mounted on the ends of the turn-up shafts, substantially as set forth. "(6) In combination, the two brackets or castings suitably supported, the horizontal pulley mounted between them, the turn-up shafts secured in the said brackets or castings, and the pulleys L, loosely turning thereon, substantially as set forth." The bill of complaint is in the usual form, and charges the defendant with infringement of the above claims. The defendant denies infringement, and also denies the validity of the patent in suit. Appellant's brief, however, lays particular stress upon the defense of noninfringement. Conveyor-belts were originally supported by a single horizontal cylindrical pulley, upon which the belt rested in a flat position. Later the belt was supported and troughed by means of a spool-shaped pulley or idler. Both of these forms, however, proved to be objectionable and inadequate; the former because, as the belt was carried flat, there was a constant tendency to spill the load over its sides, and the latter, or spool-shaped pulley, because its varying diameter necessarily caused the belt in its progress to slip, with resultant friction, loss of power, and comparatively speedy destruction. Several patents have been cited as anticipations of the patent in suit, but reliance to show anticipation is chiefly placed upon the Healey patent, issued May 21, 1873, and the Creager patent, issued November 15, 1887. The Healey patent discloses a horizontal pulley in suitable bearings, with separate side pulleys set at an angle with the horizontal pulley to trough the belt. The side pulleys are, however, independently mounted, and do not lie in the same vertical plane with the horizontal pulley. The drawings and specifications clearly disclose an arrangement of the pulleys whereby the horizontal pulley is located to a greater or less degree forward of the angular or troughing pulleys. The result of this construction is that any given cross-section of the belt is not supported by the three pulleys at the same time. It is first supported at the center by the horizontal pulley, and then, as it advances, it is supported at the sides by the angular pulleys. The Healey device was to some extent a paper patent, since it never came into general or extensive use. It had obvious disadvantages. The testimony clearly demonstrates that in its use there is much more friction upon the belt than in the patent in suit. Furthermore, the belt is comparatively soon cracked by sagging between the lateral pulleys where it is unsupported by the horizontal pulley; then, too, the load is not borne so smoothly, nor can it be carried up so steep an incline, because of the constant shaking and joggling of the belt in passing over rollers adjusted in different vertical planes. The Healey patent, although issued about 18 years prior to the patent in suit, never seems to have suggested to any one a construction like that of the Robins patent, which was designed to, and does substantially, obviate all of the disadvantages just adverted to in the use of the Healey patent. The Creager patent is so unlike the complainant's that |