switch situated under the car and operated, by gears more or less complicated, from the point of control."

The testimony of the inventor Davis negatives whatever presumption is afforded by Fig. 4 and by the specifications of an anticipatory invention of the features of construction in question. He testifies unequivocally that he is not, and does not claim to be, the originator or designer of the form of construction shown in Fig. 4 of his patent. His explanation of the use of this Fig. 4 is that he was employed by the Westinghouse Company at the time of his invention and that of Lange & Lamme, and that the latter were at that time doing the designing of all of the Westinghouse Company's controllers. He identifies the working drawings of the Lange & Lamme drum construction as made July 21, 1892, and October 20, 1892, and thus several months before the Davis application was filed; and he testifies that he knows from contact with both Lange & Lamme that they were at that time working together on the controllers in question, and it was thus his "understanding" that the latter jointly invented it. He testifies that the reversing switch, "which is my invention, was applied to a controller previously designed having this drum type of construction embodied in it, and in making up the drawings for the patent specifications a part of this controller drum was shown, to show the combination of this reversing switch and the controller drum"; and that he merely intended to illustrate in Fig. 4 the form of drum construction for the speed switch, which was not his invention, but the invention of others, as an example of the form of speed drum controllers. He also testifies that the Lange & Lamme construction had been embodied by the Westinghouse Company in commercial forms of controllers prior to the invention of patent No. 503,279. The charge contained in the answer of a surreptitious patenting by Lange & Lamme of a patent invented by Davis is thus rebutted; and while Davis' "understanding" may not have been competent evidence, standing by itself, that Lange & Lamme were the first inventors of the skeletonized metal drum construction, his testimony, especially in view of the lack of description in the patent specifications, to the effect that he was not the inventor of the device in question, and that it had been invented by others previous to his application, is clearly competent evidence to rebut whatever presumption of anticipation might be afforded by the unexplained representation of the drum construction in question in a figure illustrating the inventor's switch. The Davis patent may thus be disregarded.

The patent next presented as anticipating the Lange & Lamme invention is United States patent No. 309,167, issued December 2, 1884, to Frank J. Sprague upon an "adjustable resistance for electrical circuits." The object of the invention is declared to be "to provide an adjustable resistance of compact and convenient form." This object is accomplished by the use of two parallel wooden switch boards separated a short distance from each other, and connected by a shaft passing through both boards, each board carrying a series of contact plates attached directly thereto, a series of resistance coils being connected between these plates and with the circuit to be resisted, so that, by a circuit-controlling contact arm (carried by the shaft) upon each switch board, variations of resistance may be produced through differ

ent combinations of the resistance coils. Figs. 1 and 2 of the patent, which are reproduced below, represent a side and front elevation respectively of the switch board-the resistance coils below the boards. being here omitted from the figures accompanying the patent.

This Sprague device is set up as a "complete prototype of Lange & Lamme"; and it is urged that each element of the claims of the Lange & Lamme patent in suit can be read upon the Sprague invention. It is only in a narrow and limited way that this statement is even plausibly correct. There is no series of conducting sleeves carried by the axle, except that, as respects each of the two switch boards, there is one sleeve carried by the axle. There is no series of curved projecting strips, except as there is one strip which might be so characterized on each of the wooden switch boards. The invention, as described in the Sprague patent is neither in letter nor in spirit the device described in the Lange & Lamme patent. The former has in no sense the skeletonized metal sleeve construction of the latter. On the other hand, it has all the defects of the old wooden drum construction, including the exposure of a wooden switch board to electric arcking, a feature which the Lange & Lamme invention. was expressly designed to overcome. The patent claims contain no suggestion of the Lange & Lamme claims in suit, nor do the specifications or the claims suggest the object of overcoming or diminishing electric arcs. The sole object of the invention is stated to be "to provide an adjustable resistance of compact and convenient form, in which there shall be very little metal not in use at any time, variations of resistance being produced by the different combinations into which a small number of coils are thrown by the movement of the adjusting device."

There may be some remote analogy between the use to which Sprague put his resistance switch and the use to which Lange & Lamme put their current controller; but unless the uses are so analogous that the application of the old device to the use of the new would occur to a person of ordinary mechanical skill, inventive faculty may be involved in such new application. Potts v. Creager, supra.

If

there were otherwise any doubt that the disclosure of the Sprague invention would not naturally suggest the Lange & Lamme invention, such doubt would seem to be removed by the fact that when Sprague himself, the electrical inventor and railway expert, three years after the obtaining of his patent (and six years before the Lange & Lamme patent), was confronted with the problem of a suitable and practical railway controller, he disregarded his own resistance device, and used the wooden drum construction before referred to.

United States patent No. 510,596 to Germann, dated December 12, 1893, is also relied upon as an anticipation. This patent is upon a reversing switch for electric motors. The application was filed April 19, 1893, which is subsequent to the application for the Lange & Lamme patent. An attempt is made to carry the invention and public use of the Germann device back of Lange & Lamme's application. The Circuit Court was not satisfied that this alleged priority of invention and use had been proven with the requisite certainty. The testimony on this subject is somewhat confused. In the view we take of the patent, we do not find it necessary to decide the question of fact referred to. A prospective view of the Germann switch is given below.

The so-called drum is constructed of two separate shafts, marked B and B', between which is an insulating disk marked C. Each of these shafts carries an arm supporting a contact plate; also an arm at the inner end of the shaft for connecting to the insulating disk. The outer end of the shaft on one side carries the main contact in the arm of a key, F, engaging clutches, G and G', respectively, as the reverser is set and the current made to flow in one direction or the other. In our opinion, the Germann reverser does not anticipate the Lange & Lamme

invention. It is obvious that the idea of a conducting metal sleeve drum is entirely absent. Germann's device employed a solid shaft (carrying but one strip), which shaft was itself a conductor. Had his switch contacts been carried by an insulated sleeve, his device as constructed would not have been operative. It is not, to our minds, an answer to this proposition to suggest that, had Germann been compelled to provide for a larger number of contact strips, he would or could have adopted a sleeve construction upon an insulated shaft; and that Lange & Lamm were forced to the sleeve construction because they had a large number of contact strips which it would be difficult to embody in one or two pieces. Laying speculation aside, the fact remains that Germann did not adopt the metal conducting sleeve device, and that Lange & Lamme did. While it is not necessarily invention to construct a shaft in one piece instead of two, or to use a sleeve construction instead of a solid shaft construction, where the function of the device is not changed (Standard Caster & Wheel Co. v. Caster Socket Co., 113 Fed. 162, 51 C. C. A. 109), as between Germann's and Lange & Lamme's inventions the functional difference is plain. While both devices are in a broad sense switches, the function of Germann's reverser is merely to change the direction of the current flowing through the motor, so as to govern the direction of its revolution. The function of Lange & Lamme's controller, on the other hand, is to manipulate a number of contacts, each carrying a live current, so as to increase or retard the speed of the motors. It is, to our minds, clear that Germann's reverser would not have suggested to a skilled mechanic the idea of the Lange & Lamme controller.

The Warner patent, No. 505,686, September 28, 1893, is not, in our opinion, an anticipation; nor does it suggest the principle of the Lange & Lamme invention. The patent is upon controller rollers for electric cars. For the purpose of obviating electric arcking between adjacent conducting strips, the inventor employed a series of slate disks placed alternately, large and small, upon the shafts, the contact strips being attached to the peripheries of the large disks. The design was that the air space thus formed between the disks carrying the contact strips would operate to prevent destructive sparking. The disks are not only sleeveless, but are nonconducting. The fundamental principle of the Lange & Lamme invention is entirely lacking.

The Beardslee patent, No. 264,230, dated September 12, 1882, remains to be considered. This patent is upon a commutator for dynamo-electric machines. Beardslee's commutator wheel is built up by so assembling four segments, each attached to an annular disk provided with an insulated bushing, as that when the disks are fitted side by side on the sleeve on which the commutator wheels are mounted, with their intervening insulated washers, the disks will be within the segments, and the peripheries of the segments will be in line with each other so as to form a complete and compact wheel against which the commutator brushes bear. There is in this construction no anticipation of the Lange & Lamme invention. While it is true that each segment of Beardslee's commutator wheel before assembling bears a resemblance to the Lange & Lamme sleeve, yet not only is the latter's patent not for a sleeve (but for a series of sleeves), but the complete

structure of Lange & Lamme differs radically from that of Beardslee not only in form but in function. The former is open-with widely separated parts; the latter, when the parts are assembled, forms "a complete and compact wheel," as it must be to fulfill its function. The former regulates the supply of current to the motor; the latter shifts the magnetic line of attraction. The object sought to be accomplished by the Lange & Lamme construction does not pertain to Beardslee's device. Devices in nonanalogous arts have no direct bearing upon questions of anticipation. Star Brass Works v. General Elec. Co., 111 Fed. 398, 49 C. C. A. 409. The fact that, for eleven years following Beardslee's commutator wheel patent, it does not seem to have occurred to any of those actively engaged in experiments to prevent destructive arcking in electric railway controllers to apply to the solution of the problem the segmented and insulated commutator wheel construction, is convincing evidence that the commutator art and the controller art are not so analogous as to preclude invention in the adoption by Lange & Lamme of their sleeve construction for the purpose of preventing electric arcking in controller drums. Hobbs v. Beach, 180 U. S. 383, 21 Sup. Ct. 409, 45 L. Ed. 586.

The objection that the claims are void for lack of sufficient description and disclosure is not, in our opinion, well taken. The specifications disclose "mounted sleeves, 20, carrying offsets as shown, curved to a cylindrical surface as shown, and acting as strips, as indicated by lettering agreeing with Fig. 1." They expressly state that "the various strip systems are insulated from each other by insulating washers, 21." Claim 13 expressly states that the sleeves carried by the axle are "conducting," and that they are insulated not only from each other but "from the axle." It is true that claim 14 is less specific than claim 13, in that it omits the statement that the sleeves are conducting and that they are insulated from each other and from the axle. But not only is the feature of insulation of the sleeves from each other covered by the express language of the specifications, but we think that insulation, in some form, of the sleeves from the axle, and thus their conducting nature, is necessarily implied by the specification that the sleeves be insulated from each other, as such insulation would be useless if the sleeves were not insulated from the axle, and the latter remained a conductor. We therefore construe claim 14 as requiring that the sleeves be conducting, and that they be insulated from the axle as well as from each other. Whether, as suggested by appellee, an insulation of the sleeves from the axle by means of air alone would satisfy the requirements of claim 14, we need not consider, as such a case is not presented.

That the defendant's structure does infringe complainant's patent is clear. Indeed, we understand it to be conceded that the lower part of defendant's structure, which its counsel styles "a reverse," does infringe the Lange & Lamme patent if valid; it being contended, however, that the upper part, styled a "flying helix." does not infringe. With this latter question we are not concerned. The structure in question is a unit, and contains an infringing device.

The decree of the Circuit Court will be affirmed.