delivered the opinion of the Court:
This is an interference case from the Patent Office, wherein the controversy is as to the priority of right to an invention in voting machines. Sis counts have been formulated, which are as follows:
“1. In a voting machine, a registering mechanism and a pawl for actuating it for each candidate or ticket, arranged in rows or series, and an edgewise movable frame independently mounted having a series of bars one for each row or series of pawls, and adapted when moved to actuate every pawl that is in position to actuate the registering mechanism.
*138“2. In a voting machine, a stationary frame formed of bars or plates secured together at right angles, registering mechanisms mounted thereon, independent pawls movably mounted thereon for actuating the registering mechanism, and a frame independently mounted, made of bars or plates secured together at right angles and movable edgewise in said stationary frame for actuating said pawls.
“3. In a voting machine, a lever operated after voting, registering mechanisms, independent pawls to actuate the same, means for placing the pawls in position to actuate the registering mechanisms, an independently mounted frame that engages said pawls when in an actuating position, and means operated by the said lever for moving said frame in the same direction as said pawls, and thereby actuating the registering mechanisms.
“é. In a voting machine, registering mechanisms, independent pawls to actuate the same, means independently mounted and movable in the same direction as said pawls to cause said pawls to actuate the registering mechanisms, and means for preventing the actuation of some of said registering mechanisms.
“5. In a voting machine, registering mechanisms, independent pawls not in position to actuate said registering mechanisms, means for placing the pawls in position to actuate the registers, a common means independently mounted to actuate said pawls when in an actuating position and movable in the same direction, and means for preventing the actuation of some of said registering mechanisms.
“6. In a voting machine, registering mechanisms, independent locking mechanisms, and an edgewise movable frame independently mounted, which actuates the registers and returns the locking mechanisms to normal.”
There were originally three contestants and two interferences ; but one of the contestants has dropped out of the contest, and the two interferences have been consolidated into one. Of the two competitors who have remained, the appellant McKenzie is the senior by about three weeks. He filed his application in the Patent Office on February 12, 1900. The appellee Cummings filed his application on March 8, 1900.
*139McKenzie, in his preliminary statement, claimed conception of all the counts of the issue on November 1, 1894; disclosure on December 1, 1894; and reduction to practice on March 1, 1895, and again on December 1, 1897. Cummings alleged conception of the invention of counts 1, 2, and 6, as well as disclosure thereof in November, 1896, and reduction to practice in February, March, and April of 1897; and he alleged conception of the invention of counts 3, 4, and 5, and disclosure thereof in June of 1897, and reduction to practice in May, 1898.
The Examiner of Interferences found in favor of Cummings, the Board of Examiners-in-Chief in favor of McKenzie, and the Commissioner of Patents in favor of Cummings. From the Commissioner’s decision appeal has been taken to this court.
There is hopelessly irreconcilable antagonism in this case between the findings of the several tribunals of the Patent Office, not only as to the question of the conception and reduction to practice of the invention in controversy, but, what is most strange, as to how many and what counts of the issue were embodied in the several machines adduced in evidence by the contestants. Under these circumstances we feel ourselves justified in falling back upon two or three controlling circumstances to determine the controversy for us. But we may remark here that, while one of the machines of the appellant, McKenzie, was operated before us, there was no oral argument and no operation of any machine on behalf of the appellee, Cummings, in consequence of some misunderstanding on the part of counsel. Counsel for the appellee, however, has filed a brief which has been duly taken into consideration.
Leaving out of account, at least for the present, the question of conception of the invention, we find that, on the part of the appellee, Cummings, the determination of the question of his reduction of it to practice depends upon the force and efficacy to be given to two machines constructed by him or under his direction, — one made in the year 1897, and designated in the record and argument as the 1897 machine, which, as the Commissioner of Patents holds, embodies counts 1, 2, and 6 of the issue, — and the other made in the year 1898 and designated *140in the record as the 1898 machine, which, according to the Commissioner, embodies counts 3, 4, and 5 of the issue. There was another machine constructed by Cummings in 1899, which was offered in evidence, but which was not regarded by himself as satisfactory; and it plays no part in this case. There was still another machine exhibited by him, which was constructed in the year 1901; but, as this was long after the constructive reduction to practice by both parties by the filing of their respective applications, it need not be considered. The question, therefore, of reduction to practice by Cummings depends upon the operativeness of his two machines of 1897 and 1898. We do not think that either was operative in such sense as to be regarded as a reduction to practice, notwithstanding that they may have given hope of future fulfilment of the purpose for which they were designed.
It is conceded by Cummings himself, in regard to these two machines, that they would not always operate, and that they were not entirely accurate. It was admitted that, as often as once in a hundred times, the registering mechanism would fail to register as desired; and it is not clear from the testimony whether such inaccuracy was the result of some fundamental defect in the mechanism or merely of crude mechanical workmanship. If the defect was due merely to,crude workmanship, it would seem that it would have been easy to show that fact. The sequel would appear to indicate that it was due rather to the absence of some necessary element; for accuracy was not attained until the construction of new and different mechanism in the successful machine of 1901.
Now, when it is remembered that the machines in question Avere intended for use as voting machines to receive and register votes at municipal and other elections, in which hundreds, and even thousands, of votes may be cast, and in which the primary and most essential requisite is that of unfailing accuracy, we cannot understand how a machine that fails to register as often as once in one hundred times, — in other words, that will misrepresent 1 vote in every 100 votes, — can be regarded as a successful reduction to practice. Such repeatedly recurring inac*141curacy necessarily defeats the whole purpose of the contrivance, and renders the machine useless. As was well remarked by the Board of Examiners-in-Chief in the Patent Office, no electorate could afford to commit the registration of its electoral will to such a machine.
It is argued on hehalf of the appellee that this is a smaller percentage of inaccuracy than occurs in the ordinary methods of registration. But even if this were true, — and the correctness of the statement is greatly to he questioned, — it does not aid the contention of the appellee. Eor the ordinary methods, even if liable to mistake, have their own remedies for correction of errors, while the inaccuracy of the machine is remediless. There can be no doubt of the truth of the statement that no electorate could afford to commit the registration of its electoral will to such a machine; and, this being so, the inference is unmistakable that the machine is not an operative machine.
The case is cited of Coffee v. Guerrant, 3 App. D. C. 497, to the effect that, when a machine accomplishes its purpose in a manner reasonably satisfactory to practical men and fairly promising good results, although not then constructed with thorough mechanical skill in all its parts, it may be regarded as a successful reduction to practice: and this is undoubtedly a just and reasonable construction of the law. But the machine in that case was a tobacco-stemming machine, which did fairly good work, but occasionally permitted some of the tobacco leaves to pass the blades without having the stems properly removed. This inaccuracy, or this imperfection, as we may call it, would not necessarily have prevented the machine from coming into universal use, even though it had never been perfected so as to permit no leaf to pass unstemmed. The escaping leaves might well be gathered and thrown back; and the.machine might have been found remarkably useful, notwithstanding the imperfection. Bnt the imperfection in the present case wholly destroys tiie usefulness of the machine. In order to be operative at all, absolute accuracy is here required. It was apparent in the case of Coffee v. Guerrant that the defect was due mainly, if not exclusively, to crude workmanship. It is not apparent in the present case that it is not due to some fundamental fault in the *142constriiction of the machine. The apparent periodicity of the recurrence of the defect would indicate fundamental and radical inefficiency, beyond the scope of mere crudity of workmanship.
We find no sufficient proof in the record to show that the appellee, Cummings, reduced the invention of any of the counts of the issue to practice before the constructive reduction to practice involved in the filing of his application of March 8, 1900.
On behalf of McKenzie numerous machines were adduced in evidence to show reduction to practice by him. Six of these have been segregated from the others as most strongly illustrating the invention. One of these, designated as “Exhibit No. 18,” which is shown to have been finished on or before May 15, 1898, is stated by the Board of Examiners-in-Chief in the Patent Office to contain all the counts of the issue, and to have been satisfactorily operated in their presence with constant and unvarying accuracy. By the Examiner of Interferences, however, and by the Commissioner of Patents, this and other machines offered in evidence on behalf of the appellant McKenzie, were regarded merely as sections of machines, and not as full operative devices. We deem it unnecessary to express any opinion, either on the question of their operativeness, or on that of the extent of the devices in controversy which they embodied. Eor, if we assume that there was no actual reduction to practice by McKenzie any more than there was by Cummings before the filing of their respective applications, yet, inasmuch as McKenzie was the first to come into the office with a constructive reduction to practice, which he did on February 12, 1900, although by less than one month before his competitor, he is entitled thereby to an award of priority of invention.
. This conclusion necessitates a reversal of the Commissioner’s decision, and an adjudication of priority of invention in favor of the appellant McKenzie.
The clerk will certify this opinion and the proceedings in this court in the premises to the Commissioner of Patents, according to law. Reversed.
A motion by the appellee for a rehearing was denied August 2, 1904.