DENISON, Circuit Judge
(after stating the facts as above). The question which defendant pleaded as one of utility and argues as one of operativeness seems also to be one of reduction to practice. It is clear on this record that the only structure ever built by Hotsapillar himself was a model which did not give to one skilled in the art any instructions or suggestions beyond those contained in the specification and drawings. The question is, therefore, of that constructive reduction to practice by filing of the specification and drawings which the established rule makes the equivalent of the actual building of a machine; and the first question we meet is as to the true time of such constructive reduction. After the filing of his application in 1892, it was prosecuted through various rejections and amendments. -In June, 1894, he filed a substitute specification and new drawings. These were accompanied by a new oath, and were accepted by the Patent Office. So far as the issues in this case are concerned, it makes no difference whether these substitute specifications and drawings are considered as a part of the original application or as a new application. We are therefore required to consider these 1894 drawings in determining whether Hotsapillar had made a complete invention.
The primary question is: To what extent must a device, built as shown in the drawings and specifications, be operative in order to support a conclusion that the conceived invention has been completely made? This question was quite fully considered by this court in an opinion hy Judge (now Mr. Justice) Eurton in Standard Cartridge Company v. Peters' Cartridge Company, 77 Fed. 630. 23 C. C. A. 367. He quotes from Doom Co. v. Higgins, 105 U. S. 586, 26 L. Ed. 1177, the illustration of an invention in some appurtenance of the steam engine whicli may be completely shown without describing other parts of the engine. He lays down the rules, also, as fully as contended for by the patentee here, that a completed invention is not negatived because the drawings are rude or imperfect, or because they are incomprehensible to one unacquainted with that class of machinery, or because they do not in all respects show the relation of the' novel features to the old device nor describe precisely the mode of attachment nor with scientific exactness show other details of the combination. He concludes that the controlling question is whether “the absent features are such as would be readily supplied by a mechanic familiar with the subject and without requiring further invention” (77 Fed. 647, 23 C. C. A. 383); and whether the mechan[512] ical questions left unsolved by the drawings were of such dignity as to require invention to carry into effect the idea indicated by the drawings (77 Fed. 652, 23 C. C. A. 388); and whether “that which remained imperfect in the sketches was remediable by the exercise of the technical knowledge of mechanics familiar with the construction and operation of the old machines” (77 Fed. 655, 23 C. C. A. 391). It is true Judge Lurton was considering a question really of conception, but he was adopting and applying the rule appropriate to questions of completion or reduction to practice.
We must then examine the nature of the alleged defects in Hotsapillar’s machine, and, from their inherent character and from testimony relating thereto, determine whether they are inconsistent with a completed invention. The operation of the scale is effected by the energy inchoate in the suspended article to be weighed and exercised through the force of gravity, which will normally bring the article into the same vertical plane as the point of suspension; and this principle operates freely in the old fashioned steelyards, or in any simple balance scales. In ordinary platform scales, it operates through pivoting the free end of the platform lever to the bottom of the suspending rod, and by permitting the rod to hang vertically from its suspension point. In every computing scale before Hotsapillar this principle had been preserved without impairment; and, under that situation, the only thing necessary in connecting the platform lever and the bottom of the vertical rod was to avoid friction. The specific form of the connection—whether by a short pivot or long hinge or mere hook and eye—was immaterial, except for the question of friction, and expedients for minimizing friction in this connection were common and well known. It follows.that, although in the claim of a patent for- an improvement in the value beam of a scale of this type, it would have been necessary to include, expressly or by implication, and in order to make an operative combination, the platform lever and the vertical rod and the connection between them, it would not have been necessary to show or describe anything more than a conventional connection, and such a claim would have covered any form of connection that might thereafter have been devised and used in the entire combination. This being the situation surrounding Hotsapillar and any invention he might make pertaining to the upper part of the device, we find that, by using his T-shaped connector, and sliding along its upper edge the link which suspends it from the value beam, he contemplated that his point of suspension should be-removed outside of the vertical plane of the lower end of the suspending rod. The extent of such removal would vary from nothing, when the link was at the center of the T connection, to (for example) six inches, when the link was at either end of the upper bar of the connector. It is apparent that, in case of such suspension from either end of this bar, the lower platform carrying end of the T connector and the suspension point at one end of its upper arm will tend to swing into the same vertical plane. This tendency must be resisted, in order to maintain parallelism between the upper horizontal bar of the T connector and the value beam when in balanced position; and in [513] Hotsapillar’s illustrated construction this tendency will be resisted, and the resistance will take the form of friction at all of the points of attachment to the general frame. There will be a side stress at the platform and a reverse or counteracting side stress at the value beam. The mechanical principle here involved can be illustrated by supposing that a five-pound weight is swinging freely by a cord from a fixed suspension point. It will cotne to rest directly below the suspension point, and will exert thereon, through gravity, its entire potential energy, viz., five pounds. If we now take a bar and push this weight to the left so that the cord takes an angle of thirty degrees from the vertical, it will require a constant exertion of force to maintain the weight in this position. In other words, the weight will be continually expending a portion of its potential energy, say one pound, in resisting this side stress, and, as its total energy cannot increase, it will have remaining only four pounds to manifest by downward pull at the point of suspension. It follows that the presence of these side strains, unbalanced or uncompensated, is inconsistent with the essential idea of a scale, viz., that it should accurately indicate the varying weights or values of the different articles placed upon the platform; and in Hotsapillar’s device the same article would indicate one weight, if the link was at the center of the connector, and it would indicate another weight, if the link was at one extremity of the horizontal arm; and there would be the same contradiction when the weight was translated into terms of value. Stated in another way, Hotsapillar approached a combination consisting of -five elements, and undertook to improve element No. 1. If he had accomplished this by a mere change in the form or attachment of element No. 1, his patent might have described and illustrated and claimed only this improved form of element No. 1, in combination with the other four well-known existing elements. However, he did not confine himself to this mere improvement. He made such a modification that he upset the law of the combination. Having modified one element so that it would not work in the pre-existing environment, he must modify the other elements so far as necessary to meet the new situation, and,_ until he had done so, he had made no invention, no matter how meritorious the idea which he had conceived.
Having seen that Hotsapillar's change at the upper end of the connecting rod introduced a new condition for the remainder of the combination and modified its operation, the next question is whether this modification was great enough to require the total rejection of the result, or whether it was so inconsiderable that it can be disregarded. Upon this subject, we may be informed by the testimony. Defendants constructed a scale after the 1894 drawings of Hotsapillar, retaining this inherent defect (and two others hereafter mentioned), but otherwise removing every imperfection indicated by the drawings and minimizing friction at every contact point in the best method of the scale-maker’s art. Complainant does not, by testimony or by argument, criticize this model in any particular. Its defective operation is manifested in two ways. When the suspension point is at one extreme of the connecting arm, the scale is “logy.” The value beam, when it should be in its true balance, may be put into its up[514] per tilted position and will stay there during a considerable lateral movement of the poise, or it may be put into its lower tilted position, and will stay there until after a correspondingly long continued movement of the poise. When the suspension point is at the other end of the horizontal connector rod arm, the value beam sticks at its central and apparently balanced position, in spite of moving the poise back and forth upon it.' The net result is that the device will weigh correctly when using the price unit above the center of the connector, that this accuracy at once changes into inaccuracy as the suspending link is moved in either direction, and that the inaccuracy increases with the'lateral progress of the suspending link and with the increase of weight upon the platform, until, with a heavy article, at the extreme high' or at the extreme low price, it will reach 10 per cent. Obviously such an error cannot be compensated by any system of markings on the value beam, and it is equally obvious that such device is not entitled to be called a computing- scale. The testimony is undisputed that such a scale would not be permitted to be used under the common, public regulation of weights and measures. Hotsapillar undertook to invent an improvement in computing scales, and, Unless what he produced could be applied to a scale so that the scale would compute, he had not accomplished his object.
Having found that the device, in the form shown, and constructed with the greatest skill compatible with that form, was inoperative, and that it was inoperative to a substantial degree, we come to the question whether 'the inherent defect could have been obviated or compensated or neutralized by the exercise of their expert knowledge by those skilled in the art, or whether such correction required the exercise of further invention before the originally planned and unitary invention was complete. This record indicates only one method known, even yet, for using the laterally adjustable suspending link at the upper end of the connector and avoiding the objectionable side strains. This consists in extending the foot of the connecting bar in a plane parallel with the upper horizontal arm and pivoting the platform lever thereto at two widely separated points. The record indicates that this plan was first adopted by Koehne in his application filed in 1896 (or perhaps by Culmer, by patent No. 552,278, issued December 31, 1895), and it has been employed both by complainant and defendant in every scale either one has ever sold. It is not clear to us that this construction will of itself substitute vertical lines of force for the oblique lines resulting in the Hotsapillar construction, but it obviously will tend to minimize the side strains, and there can be no doubt that it does either destroy them or minimize them so that they are -practically negligible. Complainant’s shop experts worked for several months modifying and improving the Hotsapillar device before they found it satisfactory, and all of these efforts apparently must have been directed toward the defect we have been considering. Koehne and Osborn fought out an interference in the Patent Office upon issues relating to this particular improvement, and a patent was granted- to Koehne; and we have no testimony indicating that those skilled in the art could tell how to do this thing without resorting to [515] their inventive faculty. Under these circumstances, we cannot say that the chauges in the lower part of the device necessary to make workable Hotsapillar’s improvements in the upper part were within the field of mechanical skill in that particular art in 1892 or in 1894. It follows that Hotsapillar is unable to supplement his imperfect disclosure by reference to the field of common knowledge in the art, and bis disclosure remains incomplete and imperfect.
We <lo not overlook that in the Hotsapillar model, constructed by defendant after his specification and drawings, there were retained two oilier defects besides the one considered, viz., the lower pivoted end of the rod had a slight lateral motion owing- to the fact that the end of the platform lever would swing vertically in the arc of a circle, and the pivot between the tare beam and the vertical rod was out of the proper vertical alignment. As to the first of these criticisms, the Patent Office held that its prejudicial effect would be negligible, and we see no reason to doubt this conclusion. As to the second, it may he that it would have been remediable by the ordinary know-ledge of the scale constructor, and that, if remedied, the model would have given a better performance; hut we think the burden was on complainant to show that the inoperativeness of the model was due to this defect, rather than to the other defect which we have found could not be remedied without further invention; and complainant has made no effort to do so.
[2] Complainant meets this situation by the .claim that the contract between the government and the patentee relates to the day of issue; that the patentee is, so far as the public is concerned, entitled to use, in any part of his machine, any equivalent known at the date the patent issues; and that, therefore, Hosapillar’s specification and drawings may be supplemented by reference either to the Koelme or the Osborn patent, both of which issued before Hotsapillar’s. This theory misapprehends the point at issue. The grant of letters patent is based upon a disclosure of an invention. The disclosure is made by the filing of the application. The invention is not made until there has been a reduction into practice, actual or constructive. This question must be determined by the situation existing when the application is filed. No subsequent invention by B. can have retroactive effect to transform an earlier mere conception by A. into a then completed invention. The rightful rejection of a patent application, because it does not disclose 1o one skilled in the art how to build and operate a machine, cannot become wrongful because before the application is abandoned some one else files an application or takes out a patent supplying the missing link. It is the date of the completed application, and not the issue dale, which must be the criterion in determining the fund of knowledge in the art which the patentee can, by implication, include in his disclosure.
[3 ] Complainant also relies upon the effect of the Patent Office decisions in the interference between Hotsapillar and Osborn, and we may, for the purpose of this opinion, assume that the present parties are affected as Hotsapillar and as Osborn would have been. An interference award, deciding a direct issue of fact, upon the question of priority, should be adopted by the courts in subsequent litigation [516] between the same parties, unless there is “thorough conviction” to the contrary (Morgan v. Daniels, 153 U. S., 124, 125, 14 Sup. Ct. 772, 38 L. Ed. 657); but the rule goes no further (Hildreth v. Curtis [C. C.] 157 Fed. 394, 395). We find here that Osborn contended, before the Examiner of Interferences, that Hotsapillar’s device was not operative. We "do not know whether he urged the same arguments and considerations which we have been considering. The examiner decided the issue in favor of Hotsapillar, treating this point as fairly one of reduction to practice, and so within his jurisdiction. Both the Board of Examiners in Chief and the Commissioner seemed to think the point was merely one of “right to make the claim,” and so was solely within the jurisdiction of the Primary Examiner and outside the power of the Examiner of Interferences to decide. For this reason, they refused to consider the merits of Osborn’s contention. It is therefore clear that the Patent Office has ' not, according to its own construction, authoritatively passed on the question of fact we have been considering, and no adjudication exists.
From these considerations, it follows that the patent in suit is invalid, and that the decree below must be affirmed, with costs.