The subject-matter of these cross appeals is the Eickemeyer coil or winding for dynamo-electric machines covered by letters patent No. 377,996, dated February 14, 1888. Claims 1, 2, and 4, of the patent are alone in issue; and the only question which arises is whether the defendant’s coil infringes any of these claims. The circuit court held that claims 1 and 2 were not infringed, and that claim 4 was infringed. The form of the coil, with its axial line indicated, is illustrated in Fig. 3 of the drawings of the patent. Other coils shown in the drawings have more convolutions of wire, but all have essentially the same configuration.

*757The two parts of the coil containing the offsets are called the “ends.” They are the inactive portions outside of the magnetic field, and they lie alongside of or adjacent to the ends of the armature core. The two remaining parts opposite the axial line are called the “sides.” They are the active portions, which rest on the periphery of the armature core. It will be noticed that the axial line splits the offsets in the center. It will also be noticed that the coil is divided by its axial line into two unequal halves, and that the relative dimensions of the two halves are such that the smaller half may pass into and through the larger half, or, what is the same thing, the short side through the long side. In describing this structural feature, the patentee uses several synonymous forms of expression: The coil “at one side of what may be termed its axial line is of lesser external dimensions than the internal dimensions of the opposite portion.” Again: “Each [coil] having substantially one half thereof of lesser external dimensions than the internal dimensions of the other half.” Again:. “Thus making one side of the coil longer than the other side, so_ that the short side of any one coil may be passed into and through the long sides of other coils.” Again: “In each coil there is a long side, b, and a short side, b', and in each case the short side can be passed into or through the long side.”

The claims in issue are as follows.

“(1) A dynamo-electric armature coil or winding, which at one side of what may be termed ‘its axial line’ is of lesser external dimensions than the internal dimensions of the opposite portion, both of said portions being alike in contour, substantially as described. (2) In s dynamo-electric armature-winding, a series of coils which are counterparts in contour, each complete and separable from the others, and each having substantially one half thereof of lesser external dimensions than the interna! dimensions of the other half, substantially as described, whereby portions of each of said coils overlie and other portions underlie appropriate portions of other coils. (4) In a.dynamo-electric armature, a winding composed of detachable counterpart coils, each of which is placed in immediate contact with the periphery of tha armature core at one side only, substantially as described.”

The first claim is for the novel coil. The second claim is for a winding composed of a series of such coils. The fourth claim is for a winding in which the coils are placed in a particular* way on the periphery of the armature drum. These claims are carefully drawn. They are expressed in dear and unambiguous terms. The first two define with accuracy and exactness Eickemeyer’s main invention. Read in connection with the specification and drawings of the patent, their meaning is plain, unmistakable, and certain. Eickemeyer had a problem to solve, and he solved it, as is common with real inventors, by the application of a simple principle. The problem was the construction of a practical form-wound drum-armature winding for dynamo-electric machines, especially of the bipolar type, composed of detachable, interchangeable, counterpart coils. This problem he solved by the simple method of making the two halves of the coil of such unequal dimensions that the smaller half of one coil may pass into and through the larger halves of other coils. In this conception lay the Eickemeyer invention. Where this construction and mode of opera*758tion are present, there is present the Eickemeyer invention; and where this construction and mode of operation are absent, the Eickemeyer invention is absent. The lesser external and greater internal dimensions of the two halves of the Eickemeyer coil are made the sole test and criterion of the invention on every page of the drawings and specification of the patent, whether the coil be adapted to bipolar or multipolar machines, or to single or double layer windings. It is the one, fundamental, essential, fixed, and unvarying characteristic of the Eickemeyer coil. The first claim declares that the coil “at one side of what may be termed its ‘axial line’ is of lesser external dimensions than the internal dimensions of the opposite portion.” The second claim declares that each coil has “substantially one half thereof of lesser external dimensions than the internal dimensions of the other half.” Taking in the hand an Eickemeyer coil, we see at a glance the lesser external and greater internal dimensions of its two halves. Turning to the drawings of the patent, the same characteristic feature is always apparent, It is also stamped on every page of the specification, as sufficiently appears from the following extracts:

“Whether my colls or windings are adapted for use In bipolar or In multipolar machines, they are novel, in that each at one side of what may be termed Its ‘axial line’ is of lesser external dimensions than the internal dimensions of its opposite portion, and both of said portions are substantially alike in contour, so that they can be symmetrically assembled upon a drum or core, and enable at the ends of said core one portion of each coil to overlie and the other portion to underlie appropriate portions of other coils.” “In this armature [referring to Figs. 1 and 2] there are thirty-six counterpart coils, D, of conducting wire, and each coil at one side of its axial line (indicated in dotted lines in Fig. 3) is of lesser external dimensions than the internal dimensions of the opposite portion, and both of said portions are substantially alike in contour, and this characteristic feature is always maintained by me regardless of the number of windings in the coil and of variations in the form of the armature to be covered.” “In each coil there ls< a long side, b, and a short side, b', and in each case the short side can be passed into or through the long side, because for the first time a portion of each coil which is at one side of its axial line is of lesser external dimensions than the internal dimensions of the opposite portion of the coil, although both portions are substantially alike in contour.”

The inspection of an Eickemeyer coil, the examination of the drawings of the patent, and the reading of the specification, leave no room for doubt as to the meaning of the first two claims, and the novelty and scope of the invention therein described. The patentee expressly declares that his coil is “novel” in that one portion “at one side of what may be termed its ‘axial line’ is of lesser external dimensions than the internal dimensions of its opposite portion”; that “this characteristic feature is always maintained”; and that “for the first time, a portion of each coil which is at one side of its axial line is of lesser external dimensions than the internal dimensions of the opposite portion of the coil.”

The defendant’s coil, with its axial line corresponding to Fig. 3 of the Eickemeyer patent, is shown in the following cut:

*759

It is apparent that the two halves of this coil are equal and Its two sides equal, and that consequently one half cannot pass into and through the other half. Manifestly this coil has neither the structural form nor the mode oí operation of the Eickemeyer coil. It does not have the lesser external and greater internal dimensions of the two halves, nor the mode of operation whereby the smaller half of one coil may pass into and through the larger halves of other coils. There is absent from this coil the “novel” feature “always maintained” of the Eickemeyer coil, and hence the very essence oí the Eickemeyer invention.

To hold that defendant’s coil infringes claim I, we must, by construction, add to the claim the following words:

“This claim is not limited to a coil in which one portion at one side of its axial line is ol! lesser external dimensions than the Internal dimensions of the opposite portion, but includes a coil in which one portion, at one side of its axial line, is of the same dimensions as the opposite portion.31'

And to hold that defendant’s coil infringes claim 2, we must, by construction, add to the claim the following words:

“This claim is not limited to a series of colls ‘each having substantially one half thereof oí- lesser external dimensions than the internal dimensions of the other half,’ but includes a series of coils in which the dimensions ©£ each half of each coil are the same as the dimensions of the other bail'.’'

Further, to hold that defendant’s coil infringes these claims we must ignore the express language of the specification, wherein the patentee defines the “novel” feature of his coil, and declares that “this characteristic feature is always maintained by me.”

The. most liberal rule of construction known to the patent law will not sanction such an interpretation of the specification and claims of a patent, or such an expansion of the invention. Such a construction and enlargement of the patent are manifestly a contradiction of its specific terms, and an attempt to read into the patent a structure never contemplated by the patentee, and entirely outside of his invention.

The complainant seeks to show infringement of these claims by two theories of construction, which are equally untenable. The first may be called the theory of the complainant’s experts; and the second, the theory of the complainant’s counsel, advanced during their closing argument in this court, and further supported by supplemental briefs. The first theory is founded upon the proposition that the two halves of the defendant’s coil are to be measured, not by their own dimensions, but by their position on the armature drum, and that, when so measured, there are found the lesser external and greater internal *760dimensions of the Eickemeyer patent. Mr. Bentley, complainant's expert, says:

“■When the coil is in place in the winding, one side lies close along the drum, and the other side is lifted up into an outer layer by means of the offset, to embrace the lower half of the next coil. The elevated half of the coil then has the larger dimensions, and the depressed portion the smaller dimensions, the former being on one side and the latter on the other side of the coil axis.”

The difficulty with this theory is that it is purely hypothetical. There is no suggestion of any such method of measurement in the Eickemeyer patent, and, consequently, there is no warrant for applying such a method of measurement to defendant’s coil. The halves of the Eickemeyer coil are measured by their own dimensions, and not by their dimensions when placed on the armature-core, or their dimensions relative to the axis of such core. The first claim of the patent is for a single coil of the specific form described. The second claim is for a series of these coils, each of the specific form described; and the only question is whether the defendant’s coil has substantially the same structural form. We are dealing with the form of specific things, and not with their position relative to something else. We cannot alter the form of a thing by placing it upon another thing. The Eickemeyer coil does not change its form either by laying it on a table or by the collection of a series upon a drum armature. Nor can changes in position alter the relative dimensions of the halves of such coils. They cannot make unequal halves equal, or more equal, or less equal; and the same is true of defendant’s coil. Its equal halves remain equal, and it is impossible to make them unequal by mere change of position. The measurement of the two halves of the defendant’s coil relative to their positions on the armature is an attempt to prove infringement by an assumed and indefensible method. The adoption of this theory of measurement would make the halves of the Eickemeyer coil equal in some of the various forms of its use described in the patent. This theory illustrates that, as soon as we depart from the plain, simple, and unmistakable description and drawings of the Eickemeyer patent, we become involved in a maze of irrational and contradictory conclusions. The complainant’s experts did not find the axial line of the Eickemeyer coil in any other position than as shown in Fig. 3 of the patent, or the axial line of defendant’s coil in any other position than as indicated in the cut shown above; the line in each case splitting the offsets of the coils in their center. They undertook, however, to prove that the defendant’s coil had, in fact, the unequal halves of the Eickemeyer coil, by measuring those halves with reference to their position on the armature core. This contention may be said to have been substantially abandoned at the argument, and the new theory advanced that the true axial line of the Eickemeyer coil is not the line indicated in Fig. 3, but another line which would include the whole of the offsets in the larger half or portion of the coil, and that the axial line of the defendant’s coil is not the line which divides the coil into equal halves, but is a line which includes the whole of the offsets in one half. This theory may be illustrated by the following figures, taken from complainant’s supplemental brief:

*761

*762Here again we have a pure theory based upon false premises. The first false premise is the assumption of an axial line for the Eickemeyer coil which does not exist in fact, and the second false premise is the assumption of an axial line for defendant’s coil which does not exist in fact. It can serve no good purpose to discuss the reasons why the axial line of the Eickemeyer coil should be or must be at one side of the offsets, or the consequences which flow from such an assumption, when the fact is otherwise. By arbitrarily changing the axial line of the Eickemeyer coil we can alter the relative dimensions of its two halves to suit our purpose, and by the same method we can make the two halves of the defendant’s coil to conform therewith. By the same method also, or by making another arbitrary change in the axial lines, we might bring other coils within the Eickemeyer patent. But the question of infringement in this case is not to be determined by assumptions and hypotheses. We must keep within the region of facts and things. An Eickemeyer coil, on its face, by virtue of its configuration, has unequal halves, and there is no theory or hypothesis or reasoning which can make those two halves equal. The defendant’s coil, on its face, by virtue of its configuration, has equal halves, and there is no theory or hypothesis or reasoning which can make these halves unequal. There is no justification in'the Eickemeyer patent for the assumption that the axial line is not exactly where it is located in Fig. 3. The specification says that its position is indicated in Fig. 3, and this is further confirmed by the following description in the specification respecting the curves and offsets of each convolution of wire in the coil:

“In order that the crossing of the wire In each coll may be obviated, the wire at each end in each convolution is curved- in evolute lines at both sides of what may be termed the ‘axial line’ of the coil, and at the center or inner ends of said evolute curves or bends the wire is offset and occupies lines which are parallel with the axis of the coil, thus making one side of the coil longer than the other side, so that the short side of any one coil may be passed into and through the long sides of other coils.”

Each wire is curved in evolute lines on both sides of the “axial line,” and at the center of the evolute curves the wire is offset and occupies lines parallel with the axis of the coil. This description locates the axial line exactly as shown in Fig. 3, where it splits the offsets in the center. The axial line of the Eickemeyer coil divides the coil into two unequal halves, so unequal that the lesser half must have the capacity of passing into and through the larger halves of other coils, for this is the function or mode of operation of this novel coil. The axial line in defendant’s coil divides the coil into equal halves, having no such function or mode of operation. It follows, of necessity, that the defendant’s coil is outside of the Eickemeyer invention disclosed in claims 1 and 2 of the patent, and therefore cannot infringe these claims.

The consideration of claim 4 remains:

“In a dynamo-electric armature, a winding composed of detachable counterpart coils, each of which is placed in immediate contact with the periphery of the armature core at one side only, substantially as described.”

Upon its face, this claim covers every winding for a dynamo-electric machine composed of any form of detachable counterpart coils placed upon the periphery of the armature core in a double-layer winding, *763having one side only of each coil in contact with the armature drum.

Claims i and 2 cover the novel Eickemeyer coil, whether the winding on the armature core be a singleJayer winding or a double-layer winding. Claim 4, in terms, is much broader, and includes coils in which-the Eickemeyer invention is absent; in fact all forms of coils, provided they are detachable and counterpart when assembled on the armature core in a particular type of double winding. Two features are embraced in claim 4, namely, a winding composed of detachable counterpart coils, and a method of double-layer winding. This broad claim, according to its literal reading, can only be sustained on one of three grounds: First. That the Eickemeyer invention set forth in the first two claims of the patent covers all forms of drum-armature coils which are detachable and counterpart. If this be true, then this claim may be a valid claim for the Eickemeyer invention when used in the type of winding described. Second. That the Eickemeyer invention disclosed in his patent resides in the, conception or discovery of the attaebahility and counterpartism oí armature coils, rather than in the means by which such coils are made detachable and counterpart. If this be true, then this claim may be a valid claim for such coils when placed on the armature core in a particular kind of double-layer winding. Third. That the Eickemeyer patent covers two distinct inventions, (1) a particular form of armature coils, and (2) a new method of double-layer winding, if this be true, this claim may be a valid claim for this method of winding where detachable counterpart coils are employed. This broad claim cannot be sustained on the first ground, because we have already held that the Eickemeyer invention is limited to a coil having the essential structural characteristics set out in claims 1 and 2, and therefore does not cover other detachable counterpart coils in which these essential structural characteristics are absent. This broad claim cannot be sustained on the second ground, because, at the date of the Eickemeyer invention, there was nothing new in the mere conception of detachable counterpart armature coils apart from the means by which such coils are made detachable and counterpart. The desirability of having form-wound coils detachable and counterpart had long been recognized in the art, and the invention of Eickemeyer resides wholly in the means by which these results are attained. This broad claim cannot be sustained on the third ground, because the Eickemeyer patent Is not for two distinct inventions, — a novel coil and a new method of double winding, — but is for a novel coil, or a series of such coils, which may be collected on the armature core in several types of winding. That the Eickemeyer patent is limited to the novel coil, and was not intended to cover, as a separate invention, and could not cover, if so intended, the method of double-layer winding described in claim 4, is' shown (1) by the patent itself, (2) by the history of the patent in the patent office, and (3) by the prior art. The Eickemeyer patent contains a full, dear, and comprehensive statement of the patentee’s invention, and a fairly comprehensive statement of the prior art; and in the consideration of the questions which arise in this case, it is a relief to turn from the conflicting and often insolvable opinions and explanations of experts respecting this patent and other patents in evidence, *764to the patent itself for light and guidance. As the meaning of claims i and 2 is made perfectly clear from the drawings and specification, so likewise is the meaning of claim 4. The general scheme of the patent is plain. The patentee points out his invention, and states its objects and advantages, and he then proceeds to describe, and illustrate by the drawings, the various arrangements in which his novel coils may be used. These show the capacity or adaptability of the patented coil, by virtue of its construction and mode of operation, for single-layer and double-layer windings. Most of the drawings of the patent illustrate single-layer windings, which the patentee evidently preferred. There are, however, two drawings which illustrate, as an alternative arrangement, the type of double-layer winding mentioned in claim 4. There is, however, no suggestion in the specification that Eickemeyer was the inventor of this method of double-layer winding. It only appears as one of the “arrangements” mentioned in the specification, and the patentee refers to it as follows :

“It is sometimes desirable that the wire at the sides of each coil should be in one layer, superimposed by the wires of another coil, to form additional layers, as illustrated In the coils B 2 of Figs. 12 and 13. * * * These coils have the same general characteristics of those previously described; but it will be seen that at each side of the core (indicated in dotted lines) each coil at its one side overlaps or overlies one side of another coil, and that at the opposite side this overlapping is reversed, thus placing all of the convolutions in both colls in a uniform position on the armature drum or core. This general arrangement can be carried out to any possibly desired extent.’^

The patentee states that “it is sometimes desirable” to make such a disposition of the two sides of the coil, and adds, “These coils have the same general characteristics of those previously described.” There is no doubt, then, that all the patentee intended to cover by claim 4 was an alternative arrangement of his novel coils in a particular type of double-layer winding; and, if the claim is to be read in connection with the specification, or any significance is to be given to the words “substantially as described,” it plainly must be limited to the coils of the patent. The proceedings in the patent office show that Eickemeyer attempted to claim this method of double winding and abandoned it. In the first application for his patent, there appears the following claim:

“7th. In a dynamo-electric armature, a ‘winding’ In whicn but one-half of the effective portion is placed in immediate contact with the armature core, substantially as described.”

This claim was rejected on reference to the Freeman patent of July 29, 1884, and the Weston patent of June 13, 1882. Both of these patents describe the same method of double-layer winding with hand-wound coils. The prior art shows that this method of winding was well known. Not only is it disclosed in prior patents, but in one or more instances the patents state that it is a well-known method of winding. In the Hering patent of February 2, 1886, this method is mentioned as “a method heretofore employed” and “well known to those skilled in the art.” The patentee further says: “Nor do I claim the disposition of the two halves of each coil alternately on the inner and outer layers of wire on the *765armature.” The Freeman patent of July 29, 1884, and the Weston patent of June 13, 1882, as already stated, exhibit the same method of winding. In the Edison patent of August ,22, 1882, there is found a similar type of winding, and the specification says: “The double winding is in effect a single winding with the alternate bars located in an outer layer.” In the Jehl patent, dated January xo, x888, one side of the coil is in one layer, and the other side in another layer. The Vincent English patent of May 18, 1882, describes two layers, with the sides of each coil in different layers. In the Hering, Freeman, and Weston patents, the coils on the armature core were hand-wound, or wound on the armature core by hand, as distinguished from form-wound, or wound on a former and then placed on the armature core. Assuming that Eickemeyer was the first to extend the use of this method to iormvound coils which are detachable and counterpart, this would not give him the right to a monopoly of all form-wound coils which are detachable and counterpart when placed on the armature core in this type of double-layer winding. But the prior art does not stop with the haiul wound coils. In the Jehl, Vincent, and Edison patents, form-wound detachable counterpart coils were arranged in this type of double winding on a, disk armature; and the Vincent patent seems also to suggest the same arrangement for a drum armature. Without going further, it is manifest that the method of double-layer winding set out in claim 4 was old in the art.

These references to the patent, ilie proceedings in the patent office, and the prior art, demonstrate that Kickemeyer is not entitled to a broad claim covering the application to a drum armature of this type of double-layer winding when composed of any form of coils which are detachable and counterpart; and it follows that claim 4 must be held invalid unless it is limited to the detachable counterpart coils of the Eickemeyer patent. Assuming claim 4 to be valid when so limited, the defendant’s coil does not infringe this claim, because, as we have already found, it lacks the essential structural characteristics of the Eickemeyer coil.

The complainant’s counsel has sought to impress upon the court the importance, value, and broad scope of the Kickemeyer invention. It seems true, notwithstanding the disclosure in the obscure Rapieff British patent of 1879, that Eickemeyer made an important practical improvement in windings for dynamo-electric armatures; but we should not for this reason magnify his invention. Eickemeyer did not invent form-wound coils to take the place of handmade coils, nor was he the inventor of detachable counterpart form-wound coils; but the most which he accomplished was the production of a form-wound detachable counterpart coil adapted for use on a drum armature; as distinct from a disk or ring armature. But, even as to drum armatures, the prior art admittedly shows form-wound detachable coils which were in a degree counterpart. Eickemeyer specially directed his attention to an improved winding for a bipolar drum armature. In a bipolar machine, the disposition of the ends of the coils is a serious problem, as the two sides of each coil occupy diametrically opposite positions on the periphery *766of the drum. This problem he solved by making a coil of such shape that the lesser half of one coil may pass into and through the larger halves of other coils. By this means, the ends were economically disposed of in a minimum space alongside the ends of the armature core, and other advantages obtained which are set out in the patent. In his specification Eickemeyer recognizes the state of the art and the scope of his invention. He says:

“In certain prior multipolar machines the armature coils have been capable of ready attachment to and removal from the armature drum or core, and several of the coils in each armature have been counterparts in size and form, but several different sizes and forms have been necessary in each machine; but I know of no prior multipolar or bipolar machines in which the several armature coils were counterparts in size and form, or any prior bipolar machine which has had coils capable of being applied to or removed from the drum or core without actually unwinding the wire, although, in certain prior machines having bar conductors, the bars could be readily applied to and detached from the drum or core by separating the bars from such disks or other radial conductors. * * * Begardless of the character of the armature with reference to its polar arrangement, I believe it to be broadly new to cover a drum or core with a series of coils which are counterparts in size and contour, and which are therefore interchangeable, one with another, with reference to their positions on the armature drum or core.”

The defendant contends that the prior art was in advance of this statement, in that the Alioth German patent of 1885 and the Vincent British patent of 1882 disclose an armature drum with a series of coils which were counterpart and interchangeable. But it is unnecessary to enter into this field of controversy, for, taking Eickemeyer’s statement of the prior art and of his invention as correct, we find that all he invented was an armature coil for drum armatures having a certain structural form and a mode of operation by virtue of such form; and, this being true, his invention cannot cover other armature coils of an essentially different form, and with a different mode of operation.

As all the coils used by 'the defendant (which we assume are like the exhibits in evidence) have equal halves', so that one half of one coil cannot pass into and through the halves of other coils, they do not contain the Eickemeyer invention, and therefore do not infringe either the first, second, or fourth claims of the patent in suit.

The decree of the circuit court is reversed, and the case is remanded to that court, with the direction to dismiss the bill of complaint with costs; and the costs of this court are awarded to the Webster & Dudley Street Railway Company.