(after stating the facts as above). The art is old; general service cars — that is, cars serviceable as flat or gondola cars and also as dump cars — have been in use a number of years; and the elements requisite to their operation ‘are very much older. Furthermore, various types of door-operating mechanism have been and are in successful use; indeed, most of the patents covering them are owned by appellant. Ditchfield’s construction is not in general use; only two cars have been made; whether this is because it is not as efficient as others, or because the business conditions during the past few years would not have justified the attempt, expensive even to appellant since it acquired the patent, and more so to the inventor before that time, to get railroad systems to substitute it for other equipment that was working well, is immaterial. The two cars demonstrated the operativeness of Ditchfield’s combination, despite some possible difficulties in releasing the creeping shaft and opening the doors under a load.

We shall consider the several objections in their order:

(1) Invalidity. As a subcombination without locking means, it is asserted by appellee to be invalid because—

(a) Inoperative;

(b) Anticipated;

(c) Not the result of inventive genius, but only of mechanical skill.

(2) Noninfringement.

1 (a). Inoperativeness. It is urged that, apart from any question of prior art, the claims now in issue, which on their face do not cover the entire structure, as they do not limit the shaft to one that creeps under the lever for the purpose of finally locking the closed door, describe ^a device that is practically inoperative, and therefore, unless the locking mechanism be read into them, the claims cannot be supported as covering a valid subcombination. If the locking device is necessarily implied in these claims, then there clearly is no infringement. Ditchfield locks by causing the shaft through a final winding to creep up under the lever, while Pullman pushes in a cam-locking bar behind the series of elbow levers.

It is quite true that, without some kind of a locking device, a service car built in accordance with these claims alone would not be adopted in practice; it does not follow, however, that the subcombination without the locking element is therefore invalid. If, in and of itself, that combination of elements has a distinct operative function of practical utility, it must be upheld, unless invalidated by the prior art. And, in our judgment, the testimony fully demonstrates that, even without the locking device, the combination as described in the claims in issue does effectuate a new and practical result. For, while it is intended both in the Ditchfield and Pullman cars that the doors, when closed, should be held locked, and that, when locked, the entire strain theretofore exerted on the chain by the weight of door and load should *125be released, neither locking mechanism is automatic. A human agency, with its defects, must intervene; carelessness and forgetfulness on the part of the employe will continue, as in the past, to let many a car go on its way unlocked. When this happens, the strain on the door-supporting mechanism, Ditchfield’s short chain and long lever, Pullman’s long chain and short elbow lever, caused by the weight of the door and its load, remains; but to the extent that a rigid element, the lever, replaces one or more links of the chain, the strain on the chain itself, the weak member of the mechanism, is reduced, and this is true, whether the levers have been made fully to perform their leverage function of pressing the doors shut, or whether because some doors are warped, or the chains have become of unequal length, all the doors are not absolutely closed. For even such a car is capable of service for some freight, and, while not intended so to be used, it nevertheless is actually used until the condition is discovered. In such a case, the door-holding mechanism would inevitably wear out faster and stretch much sooner under the strain, if it were a continuous chain, instead of, like Ditchfield and Pullman, part chain and part rigid lever. Thus, without the locking element, even if that be essential, not only to lock, but also to insure a complete closing of all the doors, the subcombination of these claims described a device, operative in actual service, and securing definite results, independently of the additionally desirable elements.

1 (b). Anticipation. In this connection it is to be noted that the lever has other functions than that just considered. It diminishes the strain on the chain, not only when holding and supporting the closed, but unlocked, or the incompletely closed doors in place, but also while the door is being raised into a horizontal position, and when the pull designed to effectuate a complete closing is exerted.

Again, the lever permits the use of locking devices which, after the doors are closed, will wholly relieve the chains from strain. Furthermore, it acts, not merely as a rigid substitute for chain links, but as a true lever in multiplying the power required to overcome the resistance in finally closing an ofttimes warped door.

Concededly, the Hart & Meissner device, patent No. 764,355, is the nearest in construction and conception to Ditchfield’s, and especially, in view of the size of the lever element, to Pullman’s. All of the elements are present: the projecting fingers of the D-shaped bars, to which the chain is attached, is designed to' catch the side door, almost closed by gravity, and, acting as a lever, to> exert and multiply the power necessary to close it. In thus functioning, it anticipates both Ditchfield and Pullman. But it is not intended to, and it obviously cannot, relieve any strain exerted on the chains by the weight of the door and the load in raising, closing, or supporting the doors, because there is no such weight, and consequently no such strain, on the Hart & Meissner chains. They do not pass around the doors ; they extend down between the doors to become attached to the levers. They do not raise the door, and only in the sense that they exert the final pull on the lever can the chains be called a part of the door-closing mechanism. The Hart & Meissner lever, therefore, acts only as a true lever. Structurally, it is no rigid substitute for chain links. It does not form with *126the chain attachment a door raising or holding mechanism. It cannot function as a strain-reducing element.. The Hart & Meissner construction, therefore, shows, no' anticipation.

Other, devices especially relied Upon are Dindstrom and Streib, No. 791,348, Christianson, No. 828,458, and Swanson No. 727,487. The first has no lever; the second, a sliding lever; and the third, a catch which defendant calls a lever. These levers, however, have no strain reducing function. Confessedly, all of these devices differ both structurally and operatively from Ditchfield’s and Pullman’s.

Anticipation has not been shown.

1 (c). Invention. Perhaps the best evidence that the conception of this subcombination, with this function for general service car having bottom drop doors, was not obvious to the skilled mechanic, but required the exercise of inventive genius, is the failure of Hart and •Meissner to apply it to their bottom doors. It will be noted that the Hart &’ Meissner cars have both side and bottom drops; but their bottom doors have only the old chain. No lever is attached thereto, though, for them, too, it would have -had the additional functions secured both by Ditchfield and by Pullman. While the Ditchfield improvement is concededly a comparatively slight one in an. old art, it is both useful and novel, and, in our judgment, the result of invention.

2. Infringement. Ditchfield’s claims in issue are not expressly limited to the long lever, as shown in his drawings; the}'- read as well on Pullman’s short lever. The former has certain advantages; while both act not merely as rigid substitutes. for chain links and as true levers, the long lever functions as such from the very beginning, of the door raising operation, whereas the short one comes into play only when the door is already raised to within an inch or two of the horizontal position. But, when it does operate, it, too, acts both as rigid substitute for chain links and as a lever. It relieves the strain on the chain both in the last moments of the door-raising and door-closing acts, and subsequently in holding the unlocked closed, or nearly closed, door. By tire pull of the chain, it functions as a lever, in closing, or nearly closing, the door. That neither Ditchfield nor Pullman can completely close all doors, if some are badly warped or if the chains have become uneven in length, and that, to accomplish this, the non-infringing cam bar latch, in Pullman, or the additional winding that carries the creeping, shaft up under the lever, in Ditchfield, each designed primarily for locking the doors, is brought into action, does not diminish or destroy the functional similarity of the two levers. Con-cededly, all of the other elements read on the Pullman device. It therefore includes all of the elements of Ditchfield’s claims in issue, with at least some of their functions, both new and old, and operating by practically like mechanical means. There is thus complete equivalency and infringement.

The decree must be reversed, and the cause remanded for further proceedings in accordance with the views herein expressed.