(after stating the facts as above). The Kurda patent, No. 600,228, for a “polyphase-current transformer,” was granted March 8, 1898, and purchased by the appellant, General Electric Company, January 4, 1902; and these further facts are uncontroverted: Utility of the device as a unitary transformer of more than one alternating current in an electric system; extensive manufacture by the appellant of the patent transformers, with a profitable branch of business in their use by the trade; and invasion of the prima facie right of monopoly under the patent, by the defendants as charged in its bill. The sole defense tendered by the testimony, as frankly stated in the argument on behalf of the appellee, “is that the patent in suit is invalid in view of the prior art” — in substance for want of patentable invention — and the technical difficulties involved in the special subject-matter of “polyphase currents” tend to increase the usual difficulty arising under that issue. While the patent specifications refer to prior transformers for polyphase currents as of another type, using a group of single-phase transformers of the “core type” placed side by side, and that none were of the “ironclad type” adopted for the patent device, the evidence shows prior patents for transformers approximating so closely the means and function of this improvement that the margin is narrow for distinction therein between invention and exercise of the skill of an electrical engineer.

The patent transformer deals with the complicated problem of utilizing two or more alternating electric currents, as distinguished from a direct current, to transform the changing values of such currents into the desired electric energy. It is of the ironclad or shell type, as distinguished from the core type, and economizes material and space through its unitary structure. The specifications describe the invention as consisting “in arranging the electromagnets one above the other instead of placing them side by side, as heretofore, thus giving the electric magnets a common vertical axis.” In the prior Dobrowolsky patent, No. 422,746, granted March 4, 1890, for a transformer for polyphase currents, the structure is alike unitary, with the three cores (A, B, and C) shown in the patent drawings as placed radially, although other embodiments are in evidence arranging them directly “side by side”; and the appellant contends, in substance, for distinction therefrom: That “the transformer of the Kurda patent in suit is a specific improvement in unitary polyphase transformers embodying the Dobrowolsky principle, characterized by having ‘common mag*421netic circuits’ ”; that the Kurda electromagnets (primary and secondary coils) placed “one above the other,” and “mounted upon a common core or axis, are surrounded and separated by iron pieces” thus obtaining the common magnetic circuit; that such combination is novel and “the gist of Kurda’s invention, as covered by claim 1”; and that further invention appears in reversing the middle set of coils, covered by claim 2, whereby “marked economies are obtained,” as conceded by the appellee’s expert.

The above-mentioned Dobrowolsky patent was considered and upheld by this court, in an infringement suit brought by the present appellant, as owner thereof (Kuhlman Electric Co. v. General Electric Co., 147 Fed. 709, 78 C. C. A. 97), and the opinion in that case furnishes sufficient description and definition of the patent; and in reference to the appellees’ contention here, not only of complete anticipation of Kurda by Dobrowolsky, but that, such prior patent having expired in the hands of this appellant, it “now seeks to again obtain monopoly of polyphase transformers by means of the Kurda patent,” we remark that, monopoly in production and use being the legitimate purpose of the grant, the sole test of right thereto is the validity and scope of the patent. It plainly appears that Kurda succeeded in his quest for economy in material and form over prior devices, for a commercial transformer, and, if liis product discloses invention to that end, the purchaser is entitled to the protection sought in exclusive use.

In the argument to defeat the patent various propositions are urged, with reference to the prior art, which may be summarized as follows: (a) that Kurda’s transformer differs from Dobrowolsky’s only “in form of mechanical construction,” with “no different function and operation and secures no new or different result,” and “no advantages over the use of three single-phase transformers not previously obtained by Dobrowolsky”; (b) that after Dobrowolsky “the particular form of other polyphase transfonners became merely a matter of-design and not of invention”; (c) that Kurda’s transformer “is merely a natural union of single-phase shell-type transformers in view of polyphase transformers such as Dobrowolsky”; (d) that claim 1 of the patent “is anticipated in every respect by the earlier Ilutin & Leblanc ¡latent,” and as well by other patents; and (e) that “claim 2 involves no invention over the prior art.” We believe, however, that solution of the issue rests upon the inquiry above stated of patentable invention, as a commercial improvement over Dobrowolsky, in the light of the prior art; and that such inquiry does not involve reconsideration of the issue settled in Kuhlman Electric Co. v. General Electric Co., supra, that the “union of single-phase shell-type transformers” in a unitary transformer for polyphase currents was invention in Dobrowolsky’s transformer, and not therefore an obvious expedient. Until means were devised to obtain common magnetic circuits in a unitary structure, such union of single-phase transformers was necessarily inoperative. The question, therefore, is this: Are Kurda’s means to that end, for economic improvement, as stated in either'or both claims, differentiated from Dobrowolsky’s transformer, plus the prior art, so that invention is disclosed therein?

*422Claim 1 of the patent is for a unitary transformer “consisting of two or more sets of induction coils of different phase placed one upon another, in combination with laminated sheet-inductive material surrounding said sets of induction coils and separating them from each other”; and claim 2 is alike, with this additional element: _ “The electrical connection for the inner set being made in an opposite sense to that of the outer sets” — in other words, reversal of the middle coil. Thus the Kurda combination of claim 1 obtained the needful common magnetic circuit, by vertical arrangement of the several sets of coils on a single core, together with provision of the “E-shaped flat iron pieces” to surround and separate the coils — distinctively a unitary single core, embraced in iron, or “ironclad”- — while that of Dobrowolsky has several cores, with provision for the common circuit, by joining the several cores at their inner ends by a (so-called) pooling basin, with connection of their outer ends “by means of the -bars E.” The advantages over Dobrowolsky appear in compactness, saving of material, and (we believe as well) in convenience of handling and installation. It is contended, however, that these departures of Kurda were plainly within the teachings of the prior art, and the only other reference tending to support that view which impresses us to require mention is the German patent (No. IS,825) issued to Hutin & Leblanc, January 12, 1895. Two of the drawings of this patent are cited as showing “three shell-type single-phase transformers united end to end and utilized as a polyphase transformer,” alleged to be identical with Kurda’s claim 1, although not showing- the reversing of the middle coil provided in claim 2. On the face of these drawings the analogy is close, but we believe neither specifications nor drawings to be ■directed to the Kurda purpose of' a unitary apparatus “to transform an alternating current of one potential into an alternating current of another potential,” and that disclosure to that end is not an inevitable presumption therefrom. The sole purpose referred to was changing “an alternating current into a direct current,” for which, as specified, ■“the apparatus must have three separate transformers whose primary circuits must each be fed from one of the three-phase currents”; and,' although they “are mounted on the same framework,” it is conceded that they are not “placed one upon another” on a common core. No sugg-estion appears of the common magnetic circuit, on which Kurda’s operation depends; and it is questionable under the testimbny, to say the least, whether the butt joints shown between these separate transformers of Hutin-Leblanc would permit arrangement for such common circuit. The cogent fact, however, is undisputed that manufacturers of these supplies were constantly seeking, through skilled engineers, economies in material and form of unitary transformers which were thus obtained by Kurda’s combination, and we believe it to be fairly attributable to invention, rather than an obvious expedient of the engineer; that without great advance in the art his new combination is entitled to protection within its narrow scope, and claim 1 may justly be upheld accordingly.

Claim 2 provides the additional element of reversing the middle coil, with undisputed advantages not shown in prior polyphase trans*423formers; and it plainly appears that this economy is due to reversal of both primary and secondary coils of the middle set. As such provision is not disclosed by either of the patents cited as anticipations — 1 Steinmetz Nos. 533,2-1-8 and 561,735; Dobrowolsky No. 455,683 — the sole test of validity is whether invention was involved in thus reversing both for the benefit sought. We believe no dear suggestion thereof appears in the evidence of prior art, and that the device of this claim is well within the doctrine of invention.

We are of opinion, therefore, that the charge of infringement is established by the evidence, and that the decree of the Circuit Court, dismissing the bill for want of equity, is erroneous. The decree is reversed, accordingly, and the cause remanded, with direction to enter a decree in conformity with the foregoing opinion.