the hub is removed. In the Parshall patent the function of the dovetail grooves is only to receive the keys, by which the laminæ are locked to the spider, for a driving purpose. But when we turn to the Morrow patent we find that the functions of the dovetail connections between the spider and the laminæ are, first, to lock the laminæ to the spider for a driving purpose, and, second, to do so in such manner that they cannot possibly be driven from the spider by centrifugal force. In no patent of the prior art did the dovetail connection ever perform a function like the second one of the Morrow patent. I think, therefore, the Morrow patent was not anticipated by any of the patents of the prior art. But it appears in the proofs that the complainant shipped certain dynamo electric machines to Mexico on April 27, 1893, and also certain other dynamo electric machines to Providence, R. I., not later than May 1, 1893. This proof was stipulated into the case by the parties. Drawings of these machines, made as early as July, 1892, are in evidence, and the defendant's counsel contend that they show anticipation of the complainant's patent, the application for which was not filed until May 13, 1893. A comparison of the drawings of these machines with the drawings of the Morrow patent does disclose what appear to be striking similarities. But there is no testimony to explain the drawings thus stipulated into the case, and I feel sure I would be wholly unwarranted in concluding, in the absence of such testimony, that the drawings, which are quite intricate, show anticipation. The burden of proof as to anticipation is on the defendant, and must be established by preponderating and satisfying evidence. It has not been so established. As to infringement of the Morrow patent: The functions of the dovetail connections in the Morrow construction and in the defendant's device, are, first, to furnish such connections between the laminä and the spider that the laminæ will be forced to rotate with the spider, that is, to furnish a driving power, and, second, to hold the laminæ to the spider against the centrifugal force exerted by rapid rotation. The defendant insists that there are these differences between the two constructions: That in the defendant's construction there is no complete cylinder cast integrally with the arms of the spider, as in the Morrow construction, and that the dovetail connections are between the ends of the spider arms and the laminæ, and not, as in the Morrow construction, between the outer surface of a cylinder and the laminæ; and that in the defendant's construction the dovetail projections are on the ends of the spider arms and the dovetail grooves in the laminæ, while in the Morrow construction the dovetail projections are on the laminæ and the dovetail grooves in the surface of the cylinder. But, in my opinion, there is no material difference between the complete cylinder of the Morrow construction and the spider arms of the defendant's construction, the outer ends of which spread out laterally far enough to carry on the end of each arm several dovetail projections, though not far enough to form a complete cylinder. Nor do I think there is any material difference between the two constructions, in that in the Morrow construction the projections are on the laminæ, while in the defendant's construction the grooves are on the laminæ. In fact, the defendant's laminæ have alternate dovetail projections and grooves, just as the Morrow construction has. The projections on the lamina of the Morrow construction are narrower, and on the laminæ of the defendant's construction wider, than the grooves. The functions of the dovetail connections in the two constructions, however, as already observed, are precisely the same. The defendant must therefore be held to have infringed the second claim of the Morrow patent. The figures of the Reist patent are as follows: * In his specification Reist says: “My invention relates to dynamo electric armatures, and has for its object to provide a construction by which such armatures may be thoroughly ventilated, and also to so arrange the construction of the supporting structure or ‘spider,' as it is commonly called, that it may be cast or otherwise formed in a single piece without strains and without distortion. My invention further consists in the details of the construction by which I accomplish the purposes herein pointed out. In armature spiders, as commonly constructed heretofore, a number of radiating arms have been combined with a central body or hub and a rim connecting the arms. Where such structures have been cast in a single piece of large diameter, the rim and the portions adjacent thereto would cool first and take a permanent set, the arms would then cool, and later the hub, which is the largest mass of metal, would cool. The result of this has been that strains were set up in the parts of the structure, mainly in the rim and in the arms, so that in some cases it would be distorted; and it has even been known to split, either in cooling or in the process of manufacture, when the outer skin would be turned off. This could be remedied by annealing, but this is an expensive and tedious process. I have therefore devised the construction herein set out. The arms, being unconnected at the outer ends, except by the laminæ of the armature, as pointed out herein, are free to move under strain set up in casting, and the position of the connecting means between the laminæ and the arms might be adjusted in the course of manufacture to conform to any trifling inequalities.” In describing the figures of the patent he says: "D is one of the arms of the spider. The arm, D, is expanded at its outer end into a head or face, E. The face of this head is cut away at Ei, and two enlargements, E2, E2, are shown, in each of which is a dovetail groove or mortise, F, serving to hold the laminæ of the armature in place. These laminæ are of the usual form except that they are provided with projecting tongues, G1, G1, of dovetail shape in the case illustrated, by means of which the arms of the spider are connected. The laminæ are also provided with the usual slots, i, for the coils of the armature. A sectional casting, H, is also used, and may be either made in several sections, or may be a split ring, according to the size of the armature. Bolts, K, K, passing through the rings, hold the laminæ together. As best shown in Fig. 3, verre tilating spaces, G2, are left between the several groups of the laminæ; this being a well-known construction." Figure 4 is referred to as showing “one of the laminæ,” with its projecting tongues, Gʻ. Figure 2 is also referred to for the purpose of showing, by the dotted and full lines, the manner in which the laminæ break joints. The construction thus described, says the inventor, “is exceedingly efficient, in that any inequalities occurring from heating in the armature are taken up without causing undue strain, and the parts are free to move within reasonable limits, as will be readily apparent.” The claims are three in number, each of which, the complainant charges, is infringed by the defendant's device. They are as follows: "1. In combination, an armature-spider having a central boss or hub, separate arms radiating therefrom, the arms being provided upon their outer ends with dovetail grooves, and laminæ having dovetail projections fitting in the grooves, thus connecting the arms. “2. In combination, an armature-spider having a central boss or hub and separate radiating arms, the arms having dovetail grooves upon their outer ends, laminæ arranged in groups and having dovetail projections fitting in the grooves, and clamping-rings for holding the laminæ in position. “3. As a new article of manufacture, an armature-spider for a dynamoelectric machine, comprising a central boss or hub, separate arms radiating therefrom, the arms having expanded faces, and dovetail grooves in the expanded faces." The defendant insists that all of these claims were anticipated in the prior art. The file wrapper shows that the application for the patent contained six claims. Of these, claims 1 and 4 were as follows: "1. In combination, an armature-spider having a central boss or hub, and arms radiating therefrom, with laminæ connecting the ends of the arms together." “4. As a new article of manufacture, an armature-spider for a dynamoelectric machine, consisting of a central boss or hub, separate arms radiating therefrom, and means for attaching the armature laminæ directly to the outer ends of the arms." These two claims were rejected, because they contained nothing patentable over Fitch, No. 293,441, Orton, No. 309,735, Keith, No. 353,310, Thomson, No. 400,973, and Collins, No. 451,894. Claim 6 was as follows: “6. As a new article of manufacture, a sheet metal lamina for the armature of a dynamo electric machine, comprising a body portion conforming to the curves in the armature, and projecting dovetail lugs or tongues for securing the lamina in place.” This claim was rejected upon the Morrow patent, now in suit. After the rejection of claims 1, 4, and 6 of the application, a substitute for claim 4 was filed, differing from that claim only in making the clause "separate arms radiating therefrom” read "separate arms radiating therefrom having expanded faces.” The substituted claim was also rejected, upon Lundell's two patents, No. 461,795, and No. 487,755. Reist submitted to the action of the Patent Office and authorized all of the rejected claims to be canceled. Claims 2, 3, and 5 of the application were the only ones allowed, and they became claims 1, 2, and 3 of the patent, which the defendant now insists are also anticipated. A comparison of the original claims, 1, 4, and 6 with the patents upon which those claims were rejected makes it clear, I think, that their rejection by the Patent Office was based on the conclusion that they were too broad. This conclusion, I also think, was correct. The patents upon which the rejection was made disclosed combinations which exactly fitted the broad and general descriptions of the rejected claims. That the patentee was himself convinced of this is shown by the fact that he authorized the cancellation of the claims. Bearing in mind the rule of the patent law that a claim allowed by the Patent Office cannot be so construed as to be the equivalent of a canceled claim, as well as the rule that a claim anticipated by an older invention is void, we find that, if the combination described in claim 1 of the patent possesses any novel feature whatever, it is only in the fact that the arms have at their outer ends “dovetail grooves,” and the laminæ “dovetail projections fitting in the grooves.' But I have held, in an opinion just filed in Westinghouse Electric & Manufacturing Company v. Prudential Insurance Company of America, 155 Fed. 1749, that there is no material difference between an armature spider which has a central hub with arms radiating therefrom connected at their outer ends by a rim or cylinder and an armature-spider which has a central hub with arms radiating therefrom not connected at their ends by a rim or cylinder. The Morrow patent, one of the two patents now in suit, discloses an armature-spider which has (1) a central hub; (2) arms radiating therefrom connected at their outer ends by a rim or cylinder; (3) dovetail grooves in the outer surface of the cylinder; and (4) laminæ having dovetail projections fitting in the grooves. The Reist patent discloses an armaturespider which has (1) a central hub; (2) arms radiating therefrom having at their outer ends expanded faces, but not sufficiently expanded to form a complete rim or cylinder; (3) dovetail grooves in the outer surfaces of the expanded faces; and (4) laminæ having dovetail projections fitting in the grooves. I can find no substantial difference between the combination disclosed by, though it is not claimed in, the Morrow patent, and the combination described in claim 1 of the Reist patent. Claim 1 of Reist is anticipated by Morrow. What has been said of claim 1 disposes also of claim 3. It differs from the spider disclosed by the Morrow patent only in the fact that the outer ends of the arms are not sufficiently expanded to form a complete rim or cylinder. Claim 2 requires further consideration. It is for a combination, the elements of which are (1) a central hub; (2) separate radiating arms; (3) dovetail grooves on the outer ends of the arms; (4) laminæ arranged in groups; (5) dovetail projections on the laminæ fitting in the grooves; and (6) clamping rings for holding the laminæ in position. By arranging the laminæ in groups, provision is made for the ventilating spaces, Gʻ, referred to in the specification. The separate radiating arms (that is, arms the outer ends of which are not connected by a continuous rim or cylinder) and the ventilating spaces between the groups of laminæ secure one of the objects of the combination stated in the specification, which is "a construction by which such armatures may be thoroughly ventilated.” If we consider the third and fifth elements of the combination-dovetail grooves on the outer ends of the arms and dovetail projections on the laminæ fitting in the grooves—as the equivalent of dovetail grooves on the laminæ with dovetail projections on the outer ends of the arms, as I think we should do, the six elements of the combination are all old. I am convinced that the combination is a very useful one; but, after careful consideration, I have reached the conclusion that it does not set forth any patentable novelty. Its language is so general that it can be read upon either of the prior Crompton patents. Each of these patents describes an armature-spider which has (1) a central hub; (2) separate radiating arms; (3) dovetail grooves in the laminæ (which are the equivalent of dovetail grooves in the outer ends of the arms of the claim now under consideration); (4) laminæ arranged in groups; (5) dovetail projections on the outer ends of the arms fitting in the grooves in the laminæ (which are the equivalent of dovetail projections on the laminæ of the claim under consideration); and (6) clamping rings (or “washers," as they are called by Crompton) to hold the laminæ in position. I have therefore been obliged to conclude that claim 2 of Reist is anticipated by Crompton. If I am in error in holding claim 2 of the Reist patent to be invalid, still the complainant has failed, I think, to show infringement of that claim. It is a combination claim. An infringing device must embody all the elements of the claim. I find nothing in the record explaining |