Both these patents relate to the same type of coupler, and have the same fundamental features of form and operation. Patent No. 728,049, applied for March 17, 1902, is the basic patent, and patent No. 728,182, applied for May 29, 1902, covers certain detail improvements suggested while constructing patterns for the device shown in the former patent.
No commercial use was made of the exact device shown in patent No. 728,049. The coupler made and sold by the complainant, known as the “Climax coupler,” is that shown in such patent, with the improvements covered by patent No. 728,182. Two forms of the improved coupler have been marketed by the complainant, respectively known as high head and low head Climax couplers. The high head contained a lock identical with that shown in the later patent, and the low head — a reduced height of head and shortened shank — because of its reduced height, contains a shorter lock. Both these are functionally the same, and are covered by such patents and directly based on the claims in suit. The lock of the high head coupler is known in the trade as the “A” lock, and that of the low head as the “C” lock.
Modern car couplers conform to certain limitations fixed by the Master Car Builders’ Association. The design and coupling face contour is based upon the original Janney type. Briefly stated, the necessary parts are a hollow coupler casing, a drawhead, a swinging coupling hook or knuckle, and a locking device to lock the knuckle in coupled position. The contour of the drawhead and front end of the knuckle must be of the prescribed configuration, that all couplers of the same general type shall be capable of interlocking. In addition to these elemental parts, and as accessories thereto, means for lifting the lock are provided, usually a link or chain attached thereto, passing through an opening in the coupler wall, connected with a lever extended along the end of, and projected beyond the side of, the car, that the operator may throw the knuckle from its closed to its open *628position without passing between the cars. No limitations exist as to the configuration of the tail of the knuckle, the locking means, nor the inner contour of the chamber in which the knuckle tail and locking means operate. It is, however, essential that the knuckle, when locked, should not move out of its locked position while the cars are in motion, and that the coupler may be unlocked when desired. Since the adoption of this standard, the scope of invention in this type of coupler has been practically limited to the locking and unlocking features. While the main function of the lock itself is to lock and unlock, the trade also demands that it may be put and kept in unlocked position while the knuckle remains locked, that the cars may be uncoupled by their own movement. This is called lock-setting. The means for presenting the lock from moving or creeping up out of locked position, under draught or shock, is called the “anti-creep” or “lock to the lock.”
Complainant claims that with its lock — a single piece or block — the functions of locking and unlocking, knuckle throwing, lock-setting, and locking the lock, are performed, and that' this is a decided advance on the prior state of the art. Such a device necessarily coacts upon, and works in combination with, the knuckle tail, and, in the case of complainant’s coupler, with the interior configuration of the coupler head.
Of the claims of these patents, only four are specified in the bill as having been infringed, viz., claim 11 of patent No. 728,049 and claims 8, 9, and 10 of patent No. 728,182. These read as follows:
“Claim 11. A coupler locking and opening piece having its lower end inclined forwardly when in locking position, and having at its upper end an attachment for a lifting link, and a fulcrum-bearing, substantially as described.’.’
“Claim 8. A coupler opening and locking piece upwardly and rearwardly movable and having on its forward side projections h and g adapted to engage the knuckle successively, substantially as described.
“Claim 9. A coupler lock having on the lock a stop adapted to bear on the rear side of a lifting-link, and to support the same, substantially as described.
“Claim 10. A coupler lock having' a lifting-link pivoted in a slot in said lock, and a stop in the slot on the rear side of the link for supporting the same, substantially as described.”
These claims embrace the locking and unlocking features of the coupler. Claims 11 and 8 alone cover the lock, the construction and operation of which will' be presently described. There is no contention that defendant supplied the Climax coupler as a whole. It manufactures car couplers, but of á different type; and also parts of car couplers of many different types. It admits that it manufactured, advertised, and sold Climax locks of the “A” and “C” types, and Climax lifting links to be attached thereto. It furnished these attached to each other or separately, as desired, to purchasers and users of Climax couplers, to replace similar parts which originally had been furnished by the complainant as parts of a complete coupler.
The complainant charges that the supplying of these parts constitutes a direct infringement of such claims; that the Climax lock itself, was inventively a new thing, and is protected by such claims. The defendant contends that it supplied such parts only as repair parts to replace injured or defective parts, and, while these claims are drawn *629in form to cover the lock separately, the lock alone is wholly without function or utility, and not patentable as a separate thing; that it is useful only when associated with the peculiar form of the coupler head and knuckle shown in the patent; that such claims must be so read and construed as to embody such other necessary parts of the coupler; and that, when so read, the lock is but an element of a complete combination, and may be repaired by the owner and user of the complete device, so long as the identity of the original device was preserved. The other defense is invalidity, because: (1) Patents were anticipated by prior patents; and (2) that such locking features lack patentable novelty, and are faulty and inoperative.
As the validity of the patents, particularly the claims relied on, must be passed, upon regardless of the question whether the parts supplied come under the head of repairs, this question will be first considered. Complainant contends that its lock was the first to give, in a single piece, the function of the “anti-creep” or “lock to the lock,” in addition to locking, unlocking, knuckle throwing, and lock-setting; and that the latter functions were performed in a new and desirable way. Tower, in his specification of patent No. 728,049 (basic patent), states the object of his invention and describes its several parts, their operation and functions, and novel features and advantage of his device, as follows:
“The object of my invention is to provide a coupler with a locking and opening piece of simpler construction and easier of operation than heretofore and which will enable the coupler to be made of great strength in proportion to the weight of the metal used therein. These characteristics which I obtain by my invention are of the greatest practical importance, for the reason that the couplers when in use are subjected to severe strains and often to careless .use. Safety and certainty of operation require them to be ■easily operated and to have as few parts as possible.
“In the drawings, 2 represents the coupler head, and 3 is the knuckle, whose tail, 3', extends rearwardly from the pivot-pin, 4, the front side of the tail or the end portion thereof being preferably approximately parallel with the central line of the drawbar when the knuckle is locked.
“The locking and opening piece, 5, in so far as its locking function is concerned, is in the nature of a pin, which extends on the front side of the tail ■of the knuckle and in locking has a bearing against the coupler head extended both above and below the knuckle. At its lower end is a guiding portion, b, which is inclined forwardly and is adapted to fit in a correspondingly-inclined hole, 6, in the floor of the coupler. Its upper end or head has a forward projection, c, or is otherwise suitably shaped for pivotal attachment to a lifting link, 7, and at the rear of the head is a bearing, d, which is adapted to engage a fulcrum, 8, at the top of the recess, 9, of the coupler head, in which the piece 5 is set. This recess is preferably nearly of the same width as the piece 5, so that it may guide and steady the latter. At the lower part of the recess, 9, is a shoulder, 10, which is adapted to receive the forward portion of the head of the piece 6 when it is in its locked position.
“The operation is as follows: The parts being in the position shown at the coupler B in Fig. 2, the knuckle is closed, and the locking and opening piece fits in front of it, with its guiding portion, b, in the hole, 6. The knuckle is thus held locked. If it be desired to unlock the knuckle and to swing it open, the operator lifts the link, 7, thus causing the locking and opening piece to rise. During its first motion the guiding portion, b, at its lower end, causes it to move rearwardly until it reaches the position indicated by the dotted lines, e, when it has passed back of the path of the tail of the knuckle and has left the latter free to swing open and its head has engaged the •fulcrum, 8, at the top of the coupler head. This fulcrum, 8, is inclined or *630beveled laterally, as shown in Fig. 4, so that, when the piece bears thereon, it will tip laterally in a direction transverse to .the length of the coupler head to a sufficient extent to bring the locking and opening piece somewhat to the rear of the toil of the knuckle, as shown by dotted lines, e, in Fig. 3 and by full lines in Fig. 4. The effect of this is to bring the piece 5 into knuckle-throwing position and also to free it from the path of stop 12, which is formed on the coupler head and is adapted to prevent throwing of the piece 5 until it is at the back of the knuckle. Continued lifting of the link, 7, will cause the piece 6 to tip forwardly on the fulcrum, 8, and to move the knuckle into open position, as shown by full lines at the coupler A of Fig. 3 and by dotted lines at the coupler B of Fig. 2. The operator, having thus opened the knuckle, can allow the locking and opening piece to drop, whereupon it will be restored to the position shown by full lines on the coupler B of Fig. 2. When the knuckle is next swung into closed position, its tail will engage the shank of the piece 5 and, pushing it rearwardly, its guiding portion b will cause it to rise and to move backwardly sufficiently to allow the tail of the knuckle to pass, whereupon it will drop again into the hole. 6, and will hold the knuckle in locked position.
“If it be desired to lift the piece 5 only far enough to unlock the knuckle without throwing it open and to leave the piece 5 in lock-set position, so that the knuckle can be moved open freely when the car to which it is attached is moved away from another car with which it is coupled, the operator simply lifts the piece 5 into the position shown by dotted lines e of the coupler B of Fig. 2 and the full lines in the coupler A of said figure, whereupon the inclination of the fulcrum, 8, as above explained, will cause the piece 5 to move laterally to a small extent. On releasing the lifting link the piece .5 will bear against the end of the tail of the knuckle, as shown in the coupler A of Fig. 2, and will be held against it by friction. There is, however, no locking engagement of the piece 5 with the coupler head, and the knuckle is free to move open when the cars are drawn apart.
“The floor of the coupler head is preferably provided with a longitudinal groove, h, in which the lower end of the piece 5 may rest and which serves to guide it into locked position when the knuckle is moved back.
“In Figs. 6 and'7, I show a modified construction of my invention, in which the lateral beveling of the fulcrum, 8, or other means for causing the piece 5 to move laterally, is rendered unnecessary by extending the end portion, b, of the piece 5 so that it shall have a lateral projection, f, which when the piece 5 is tipped forwardly after engagement with the fulcrum, 8, will engage the end of the tail of the knuckle and will start it on its outward motion and will thus constitute the means for imparting the initial engagement of the locking and opening member with the knuckle. In Fig. 7 I also show the shoulder, 10, provided with a recess or notch, 11, ‘ into which the forward portion of the head of the piece 5 can fit when it ,is in locked position, at which time it will prevent upward creeping of the piece 5, for any tendency to such creeping will simply cause the forward portion of the head to bear more firmly within the recess. At the same time, the recess presents no obstacle to the lifting of the piece 5 by the lifting-link. * * *
“The forwardly-iincllned guiding portion, b, of the piece 5, is desirable not only in performing the functions above stated, but also in preventing the tendency of the piece 5 to creep upwardly. It acts in this-way singly and also in co-operation with the recess, 11, when the latter is employed.
“One of the important and novel features of my coupler consists in the use of a locking and opening piece whose locking and opening member in locking fits in front of and locks the tail of the knuckle and when raised first frees and then by a continued movement opens the knuckle, the successive operations being functions of the same member, which preferably is also arranged to have the capacity of setting itself in unlocked position while the knuckle is still closed. The main member of the lock, therefore, has three functions, and the same part which locks the tail of the knuckle also operates to throw it open and, if desired, sets itself in unlocked position when the knuckle is still closed. The simplification of construction and strength which I thus secure will be appreciated by those skilled in the art
*631"I obtain a large locking surface of the piece 5 against the tail of the knuckle and against the coupler head. I can also employ a knuckle tail sufficiently long to prevent jamming of the knuckle when two couplers are brought together each with the knuckle in open position. Such jamming Is likely to occur in some couplers heretofore in use."
`rhe following are the drawings referred to:
It is conceded that this locking and opening device has three functions: That of locking or coupling automatically, of setting the lock piece in an unlocked position, and of throwing or kicking the knuckle into a complete open position. The second patent, No. 728,182, is an improvement on the coupler shown and described in the former patent. These improvements, so far as the locking piece and attachment are concerned, consist in a changed configuration of the lock *632piece and the addition of two separated projections on the front of the shank or locking part of the lock, and a' stop device on its link, connection to support the connecting link in upright position. Rigs. ¾, 4, and 5 of this patent are here inserted to show these changes and the coacting parts of the knuckle tail and interior contour of the coupler recess:
*633The general features and functions of this lock are the same as that of the earlier patent. Its special added features, adopting the Ian-guage of complainant’s brief, are as follows:
“On the middle forward surface of the lock shank are two projections, h and g, having a space between them which allows the lock to be Inserted in tile coupler cavity through the mouth of the coupler. These projections are adapted to perform the function of tipping off the lock from the lock-set position and of driving it rearwardly when the knuckle is closed. Thus, the knuckle tail strikes the lower projection, g, in pulling the knuckle open from lock-set position, while in closing, when the lock is resting on the floor of the coupler, it will strike first the projection g, and then the projection h, lift the lock and drive it rearwardly until it drops to locked position with the knuckle closed. If the lock is in lowermost position when the operation of closing the knuckle begins, the knuckle tail will first strike the upper projection, h, and then the lower projection, g, lifting and driving the lock rearwardly and upwardly out of the path of the knuckle tail, so that, when the knuckle has been closed, it will drop back to the locked or lowermost position.
“In the top of the lock is a slot in which the lifting link, 7, is pivoted, and in the rear of this slot is a stop shoulder, m, which act to keep the link, 7, upright and to prevent its falling backward and wedging in the recess, 0, of the coupler head. This stop shoulder supports the link when the lock is in locked position, so that when the chain is operated the link is guided through the hole in the top of the coupler head.”
The essential characteristics of the locks of the two patents are summed up by complainant’s expert, Mr. Bently (Rec. p. 304), as follows:
“I find that in both of those patents there is a coupler having the same fundamental qualities, to wit, a locking piece characterized, as to form, by a forwardly-inclined toe at its lower end and a fulcrum and lifting-link attachment at its upper end, and characterized, as to operation, by a preliminary rearward swing produced by the cam action of the toe-hole upon the toe serving to release the knuckle tail and followed by a forward knuckle-opening swing produced by the lever action of the locking piece on its fulcrum; this to and fro swing of the lower end of the locking piece being produced by a continuous upward movement of the upper end of the piece by the pull of the lifting-link. I find that the device of the later improvement patent differs from that of the earlier patent in that the knuckle tail of the latter acts directly upon the curved shank of the locking piece, whereas in the improvement patent it acts upon two distinct lugs or projections on the front side of the locking piece; its engagement with the upper projection being on the under side of that projection and serving to lift the locking piece and free it, so that it hangs nicely balanced and easily pushed rearwardly by the engagement of the knuckle tail with the lower projection. In addition, the locking piece of the improvement patent is provided with a back-stop for the lifting-link, which serves to guide it through the operating hole in the top of the coupler head.”
The functions of locking automatically and lock-setting and knuckle throwing were old when the Tower patents were applied for. Step by step the art had advanced; the field of invention constantly narrowing. Many and varied were the patented devices by which all three of such functions were performed; nor was it new to perform all these functions with a single piece or lock.
Defendant cites a number of patents as anticipations and as showing the state of the art. As to the lock as shown in patent No. 728,049: *634In Tower reissue No. 11,477, Willison No. 643,581, and Springer No. 731,415, the locks are not only dissimilar in form, but in operation. Those locks swing in a transverse plane to the coupler head, whereas, with those of the complainant in suit, the swing is in a plane longitudinal therewith. The locks in the reissue and Willison have a vertical and a tilting motion, the latter on a fulcrum. The Springer lock is not moved bodily, but swings on a pivot. Tower No. 487,650, and White No. 543,071, though moving vertically, are not moved bodily, but swing on a fixed pivot. In White No. 514,296 like Tower reissue, the lock extends across the knuckle tail. Fabian & Widmark No. 717,663 is defendant’s chief reliance. It is claimed to be a complete anticipation of the patent in suit.
An interference was declared in the Patent Office between the Fabian & Widmark patent and Tower No. 728,049. This was dissolved on the ground that there, was no interference in fact, and that Fabian & Widmark had no right to make the claims constituting the issues involved. The rights to this Fabian & Widmark device were purchased by complainant about the time of this decision, while the applications for their patent and the patents in suit were pending. Defendant contends that the interference was not tried out on its merits; that, to avoid it, Tower had to restrict his invention to a locking and opening member extending in front of the tail; that this distinction was unsubstantial; that complainant disregarded it in its commercial coupler, in which this member does not lie in front of the tail, but distinctly beneath and in the rear thereof. Complainant’s attitude in the interference proceedings, its purchase of Fabian & Widmark’s rights, and subsequent change of construction in the knuckle opening member, are not without significance, but they fail to show first that the interference was improperly dissolved; second, anticipation by Fabian & Widmark; or, third, that, in changing the form of the toe piece of the locking and opening member, complainant departed substantially from its invention. They will have served their end by whetting the judicial mind to a keen scrutiny of the alleged interfering patents to ascertain whether anticipation exists.
The drawings of the Fabian & Widmark patent are not clear. The defendant’s experts found it impossible to give the movement of the lock with certainty, by reference to the drawings. See Rec. 129-130, 224. Figs. 2, 8, and 9 of this patent are here inserted, as showing the form of this lock, and its operation in connection with other parts of the coupler:
*635
There appears to be a vertical and bodily movement of this lock. There is, however, no pivotal or fulcruming action between the lock and the drawhead. Mr. Bently (complainant’s expert), in analyzing the respective forms and operations of the locks of the Fabian & Wid-mark, and the Tower patents here considered, said (Rec. p. 314):
“In particular, the Fabian & Widmark locking piece lacks entirely the preliminary rearwardly swing which characterizes the action of the locking piece of the Tower patents in suit, and it lacks the forwardly-inclined toe-portion engaging the corresponding toe-hole in the floor of the coupler to produce such rearward swing. It acts to release the knuckle tail by rising upward at its front end in contrast to the release of the knuckle tail in the Tower coupler, in controversy by the rearward swing of the lower end of a locking piece fulcrumed at its upper end. It contains no lifting link connected to its upper end, and no fulcrum at its upper end.”
It is a question'whether the description of this device is sufficient to meet the requirements of the patent law, or that it is operative; but, even if it can meet both of these tests, it, as well1 as the other locks mentioned, are so different in their movements and coaction with the other elements of their respective combinations that though they are all designed to lock and unlock the knuckle, and some have the lock-set feature, they all are substantially materially different both in form and operation from complainant’s lock, in effecting this purpose. In addition, none of these patents describe an “anti-creep” or the func*636tion of “lock to the lock” device, save Rabian & Widmark, and this-is not within the coupler cavity.
As to the lock of patent No. 728,182: The features of improvement of this patented device over that of the basic patent, their construction,, and mode of operation, have already been explained. All the functions of the lock of the basic patent are present in this one. Of the patents cited against this improvement, Curtis No. 852,352 and Mc-Conway reissue No. 11,546 show locking pins operating only in a vertical direction. Neither shows a knuckle throwing or lock to the lock device, or has any projection on forward side adapted to engage knuckle successively. Neither resembles the lock in question, or satisfies the terms of claim 8, viz.: First, a combined lock and opening piece; second, a piece upwardly and rearwardly movable, not having on its forward side projections, h and g, adapted to engage the knuckle-successively. Timms No. 685,802, Thurmond No. 313,386, Aabel No. 573,253, and Fabian- & Widmark No. 717,663, show lips, teeth,, or projections upon which the lock hangs from the knuckle tail when in lock-set position. These perform no function such as Tower projections, g and h, in successively contacting with the knuckle tail in driying the lock out of its path, as it moves to or from closed position; and Tower projections have no function in setting the lock, while that is the only function possessed by these lips, etc. Gifford No. 457,154, has two prejections on the forward face of the lock, but they have no function in lock-setting, locking, or opening. In fact, the lock has no lock-set, knuckle thrower, or lock to the lock function.
It is further contended that if the basic patent be valid the later one is not, as the alleged improvements do not amount to invention. The changed configuration of the lock, and the -coacting parts of the knuckle and inner contour of the coupler recess, do not change the fundamental' features of the form and operation of those of the basic patent. They undoubtedly make the automatic opening and closing movement after the lock has been put in lock-set position, and the “anti-creep” or “lock to the lock” function, more reliable and effective. The “anti-creep” is the essential in all locking devices. To insure it under all conditions of impact, jolt, or strain in coupling and moving cars is the chief end sought in all such inventions. Slight changes in; appearance may bring about radical changes in results. Invention is • not to be slighted because the changes are slight; neither is the inventor who makes his anti-creep feature more effective to be barred of his added invention because to his inventive brain the changes to bring about such result needed only to be few or slight in appearance. The doctrine of mechanical suggestion or obviousness loses much of its force when invoked in such a case. True, the earlier patent must be considered as part of the state of the art at the time this second patent was applied for, but that which was obvious to him as necessary to • improve his invention, and which now after its introduction is seemingly obvious to others, is not necessarily the suggestion which would come to the ordinarily skilled mechanic. The inventor is in a different class. To him the invention stands in a different relation. He mentally conceived it in form and operation before it was embodied in the. *637patented device. He approached the question of further improvement with a different mental vision from the mere mechanic, whatever his skill. The status of mechanical skill is not to be “measured in terms of the skill of the inventor to whom the patent issued.” Macomber, Fixed Raw of Patents, p. 48.
The border line between invention and mechanical skill is seldom clear and easy of demarkation. As was said by Justice Shiras in Krementz v. Cottle Co., 148 U. S. 556, 559, 13 Sup. Ct. 719, 720 (37 L. Ed. 558):
“It is not easy to draw the line that separates the ordinary skill of a mechanic, versed in his art, from the exercise of patentable invention, and the difficulty is specially great in the mechanic arts, where the successive steps in improvements are numerous, and where the changes and modifications are introduced by practical mechanics.”
When this doctrine, as well as that of “Equivalents,” is pressed against a later patent of an inventor, which is an improvement on his own invention, the equivalency of means or mechanical obviousness should be very clear before it is resolved against him.
Taking these prior patents cited against both of the patents in suit, as a whole, while here and there there are resemblances between the means employed by the complainant’s device and those of some of the cited patents, that does not negative invention in the later combination device. Bates v. Coe, 98 U. S. 31, 25 L. Ed. 68; Columbus Watch Co. v. Robbins, 64 Fed. 384, 12 C. C. A. 174; Thomson-Houston Elect. Co. v. Black River Traction Co., 135 Fed. 759, 68 C. C. A. 461; Hinson Mfg. Co. v. Williams (C. C.) 86 Fed. 128. All of these cited patents are secondary in character, and none is entitled to a wide range of equivalents. At most they are entitled to no more than what they specifically cover. If some of these were of primary character, the question of anticipation might be serious. There are plain differences in means and operation between all of these inventions and that of complainant, sufficient to prevent anticipation and to sustain the Tower patents in suit, as showing some patentable advance in this limited field.
In addition, as already noted, none of the cited patents shows a lock having an “anti-creep” or “lock to the lock” function save Fabian & Widmark, and which, if operative, is radically different both in form and operation of its means from that of complainant. Complainant claims such a feature for its device, and if true, it being accomplished together with the other functions, in one piece, it is a long step forward in this art. The defendant contends that the complainant’s device does not effectively accomplish this function. The evidence indicates the contrary. But if it is ineffective as to this function, that will not avail the defendant, as the patent is valid without it. The defendant sells the lock to be used with complainant’s coupler, and it is intended to accomplish all the functions that it is capable of, and it does not lie in its mouth to deprecate its operativeness under such circumstances. In National Malleable Casting Co. v. Buckeye Malleable Iron & Coupler Co., 171 Fed. 847, at pages 852, 855, 96 C. C. A. 515, at page 520, where the patents in suit with relation to the Deitz pat*638ent, No. 576,094,. and Timms patent, were considered, Judge (now Justice) Rurton said:
“The Climax coupler, the coupler made by the complainant company, is made under the patents to Clinton Tower, issued in 1903, numbered 728,049 and 728,-182. That it embodies some of the principal features of Deitz is obvious, and, if the Deitz patent was one of primary character, the anticipation might be serious. There are plain differences, however, between the two inventions — quite enough to make the claim now made that the Climax is an embodiment of the Deitz quite unsatisfactory in view of the limitations which we must impose on Deitz to support his patent at all. * * *
“The art to which the patent in question belongs is so crowded with devices intended to accomplish the same purpose as to leave little room for the inventor. That there has been a slow, step by step, advance in the direction of the ultimate automatic coupler, and a final result, is as much as can be said. That the combination of Deitz involves some advance we fully concede; but that he has made such an improvement as to entitle him to any considerable range of equivalents we cannot concede. Each improver of which Deitz and Timms and Tower are the last to which our attention has been directed, has done something; but when we concede to each his own particular form of device, so long as it differs substantially from those which have gone before and does not include them, we shall have protected each as far as they are entitled to protection in a field already filled with the efforts of thousands struggling for the same result.”
The presumption of invention arising from the grant of the patent, whatever its vyeight, continues until evidence sufficient to overcome it is introduced. Simply raising a doubt as to whether a skilled mechanic conversant with the art involved would not have seen the means adopted does not rebut such presumption. After carefully examining and considering the evidence marshaled against this patent, and the arguments of counsel in this behalf, I am of the opinion that the device of patent No. 728,182, as well as that of the basic patent, shows patentable novelty.
As to the lack of utility: The defendant’s contention that the locking features of these patents are faulty and inoperative is not borne out by the facts. The large number — over 900,000 — of these couplers, embracing both high and low head types of lock, sold to more than 600 railroads and private car companies, sufficiently refutes such contention, and demonstrates the utility and commercial success of complainant’s device. If anything more were needed, it is had in the solicitation of the defendant for orders to supply these locks, its manufacture and sale of such parts, and its present strenuous insistence of its right to such trade. Rehnbeuter v. Holthaus, 105 U. S. 94, 26 L. Ed. 939; Goss Printing Co. v. Scott, 108 Fed. 253, 47 C. C. A. 302.
In my judgment Tower’s devices in both patents show invention, and, being patentably new and useful, both patents are valid.
The patents being valid, can the same be said of the claims in suit separately covering the locking features of the combination? It is obvious that the peculiar shape of complainant’s lock, whether that of the basic or improvement patent, is neither haphazard nor accidental, but that it is designedly so, to coact with parts of the knuckle tail and the inner contour of the coupler recess, which parts are likewise specially constructed that such coaction may be had, and through which the functions intended to’ be performed by the several movements of such lock are accomplished. The lock is not merely unique in *639configuration, but it embodies more of the idea of means to give practical effect to the inventor’s mental conception than any other or all the other elements of the combination. Restricted to the recess behind the knuckle head as the area of his operations, and to the knuckle tail, inner configuration of such recess and the locking piece as the parts, to secure the mechanism to make his mental vision physically operative, a peculiarly designed configuration of such parts to obtain a certain and effective coaction of all, was necessary. In producing such configuration, no doubt many modifications of these three parts took place till the desired coaction was obtained. It is not always possible to determine with certainty which of several elements is the vital one of the combination; but it would appear to be reasonably certain that within the mental' conception here worked out — a single piece to hold the knuckle in locked position, throw or kick it wide open, unlock without opening it, set in unlocked position to enable knuckle to automatically open and thereafter to automatically lock itself, that such single piece, the lock — would be the dominant element, and to^ which the others were necessarily subordinate. An inventor of a new and useful combination is not confined to his combination claims, unless all of the elements are old. If any of the elements are new and useful, and show invention, these may be claimed and patented. This may be done in a separate patent or by separate and distinct claims in the patent covering the combination, even though such parts are without utility, save in combination with the other parts of the device. 2 Rob. Pat. § 630; Walker, Pat. (4th Ed.) § 117; Roberts v. Nail Co. (C. C.) 63 Fed. 916; Holloway v. Dow (C. C.) 54 Fed. 511; Brammer v. Schroeder, 106 Fed. 918, 46 C. C. A. 41; Chambers, etc., Co. v. Faries (C. C.) 75 Fed. 662; Deering v. Winona, etc., 155 U. S. 286, 16 Sup. Ct. 118, 39 L. Ed. 153; Gill v. Wells, 22 Wall. 1, 22 L. Ed. 699; Miller v. Eagle Co., 151 U. S. 186, 14 Sup. Ct. 310, 38 L. Ed. 121; Adams v. Jones, Fed. Cas. No. 57.
The Climax locks being inventively a new and useful thing, claims 11 of patent No. 728,049, and 8 of patent 728,182, covering this element separately, are valid. Claims 9 and 10, which relate to the stop on the lock to bear on the rear side of a lifting link, and to support the same, are invalid as involving no invention. Claim 9 is the broader of the two, but it is not necessary to distinguish between them. The sole purpose of the stop here claimed is to maintain the link against a too great movement rearwardly, that it may not jam in the coupler cavity and prevent the movement of the lock. It is not necessary to consider the citations made against this expedient. The means —the particular stop in question, or its mechanical equivalent — would be obvious to one skilled in this particular art as soon as the need was made apparent. Hollister v. Benedict Mfg. Co., 113 U. S. 59-73, 5 Sup. Ct. 717, 28 L. Ed. 901. The protection of the patents extends over this stop only as an element of the combination. It cannot be made the basis of a separate claim. By itself it is not patentably new. Claims 11 and 8 being valid, can the defendant supply the locks thus separately covered, to purchasers of complainant’s complete coupler, to replace similar parts which have become broken or defective?
The question of the right to repair and what constitutes repair to *640a patented article, as distinguished from reconstruction, has been frequently dealt with by the courts, and, generally stated, the rule to be deduced from the cases is that, upon an absolute sale of a patented device, the purchaser has the right to the full enjoyment thereof, and an implied license to make repairs, so long as the identity of the original machine is preserved, and that such right of repair is not confined to the mere restoring of the part which is defective, but extends to replacing such part,- unless it is itself made the subject of a patent. The following are a few of the cases: Wilson v. Simpson, 50 U. S. 108, 13 L. Ed. 66; Mitchell v. Hawley, 83 U. S. 544, 21 L. Ed. 322; Cotton Tie Co. v. Simmons, 106 U. S. 89, 1 Sup. Ct. 52, 27 L. Ed. 79; Singer Mfg. Co. v. Springfield Foundry Co. (C. C.) 34 Fed. 393; Shickle v. St. Louis Car Coupler Co., 77 Fed. 739, 23 C. C. A. 433; Goodyear Shoe, etc., v. Jackson, 112 Fed. 146, 50 C. C. A. 159, 55 L. R. A. 692; Morrin v. Robert White Eng. Co. (C. C.) 138 Fed. 68; Gottfried v. Conrad Seipp Brew. Co. (C. C.) 8 Fed. 322; Aiken v. Manchester Print Works, Led. Cas. No. 113; Leeds & Catlin Co. v. Victor Talking Mach. Co., 154 Fed. 58, 83 C. C. A. 170, 23 L. R. A. (N. S.) 1027, affirmed 213 U. S. 325, 29 Sup. Ct. 503, 53 L. Ed. 816.
The locks being specifically covered by valid claims, the supplying of such parts by the defendant, without the complainant’s consent, whatever the purpose of their manufacture and sale or the use made of them by the purchaser, constitutes a direct infringement. But even if the locks were not protected by separate claims, and the patentee’s rights therein were limited to them as elements of his combination device, as the defendant contends, the supplying of such locks by defendant could not, on the facts of this case, be justified on the ground of repairs. The parts supplied by the defendant are not strictly repairs. The defendant does not actually repair a defective part; it supplies new p'arts to railroad companies in advance of the actual need thereof. It is said that railroad economy requires that, at different points along the railways, supplies of the several parts of the different types of coupler in use be kept to replace, as occasion shall require, those that may become broken, defective, or lost.
Car couplers are subjected to severe treatment, and, though made with a view of sustaining heavy shocks and strains, are bound to be injured in some of their parts. The brunt of the shock in coupling is undoubtedly borne by the knuckle, and, in the jolting and pulling which takes place in the continued movement of the cars, it is subjected to the greater strain. This is larger ■ and heavier than the lock, which must be made small enough to enter and operate in the chamber or recess back of the knuckle head, and in which the knuckle tail operates. The lock is less likely to become defective than the knuckle, and, save for the eye where the link is attached, is practically unbreakable. The chain or link attached to such locks, used for lifting them, is the lighter and more fragile, and is exposed to external injuries, and as a necessary result is the more easily injured. The injury to thb -links is the. more frequent cause of the replacement of the lock, for the reason that it serves the interests of the railroads better, at the time a substitution is necessary, to substitute for the good! lock with a defective link a new chain with an attached new *641lock. As a result, many new locks are put in operation while the old ones Which they replace are in perfect serviceable condition. Such a replacement is not a repair of the lock. These locks with the defective links are subsequently sent to the repair shop, and such defective parts repaired or replaced.
It is obvious that in the case of substituted locks and links, where the links only have been injured, no repairs are necessary to the locks, and that, when the defective links have been repaired or replaced, the company has an additional lock for each one purchased! of the complainant; whereas, in case of repair or replacement authorized by the implied license,. no greater number than that furnished by complainant would belong to the user.
If this method of furnishing supplies should be extended to other parts of the coupler, a duplication of complete couplers could be obtained without resorting to the patentee. Coupler heads and knuckles, as well as locks and lifting links, could be purchased from others than the complainant under the guise of having repair parts in stock, and to the extent that the replaced parts were subject to repair,-the railroads, by assembling such repaired parts would have such duplicated car couplers to put in service. Such' á result is manifestly in opposition to the principle upon which the license to repair rests. This right is for the purpose of maintaining the patented article in the service intended by its purchase, and not to- obtain duplicates. The right of one, other than the patentee, furnishing repair parts of a patented combination, can be no greater than that of the user, and he is bound to see that no other use of such parts is made than that authorized by the user’s license. Thomson-Houston Elect. Co. v. Ohio Brass Co. et al., 80 Fed. 712, 26 C. C. A. 107. If it is necessary to facilitate the movement of cars that the railroads be permitted to so substitute locks with attached links, though only the link is defective, and for such purpose to keep on hand at different points a stock of such locks with attached links — and this-would seem to be necessary — such locks should be obtained from the complainant. This requirement would protect the patentee in his monopoly without injury to the purchasers’ license to repair, for the latter would still have the right to repair the parts thus substituted.
Whether we measure the rights of the- parties in respect to supplying the Climax locks by the claims which cover this feature separately, or by considering such lock merely as an element of the combination, device, the result is infringement. By the former the infringement is direct; by the latter, contributory. The direct infringement'is the more serious, and calls for a more far-reaching remedy, as that exists regardless of the purpose of the manufacture and sale of such locks or the use made of them.
The complainant is therefore entitled to the usual decree- for injunction and account as to claim 11 of patent 728,049 andj claim 8 of patent 728,182.