DENISON; Circuit Judge.
Bill filed under R. S. § 4915 (Comp. St. § 9460), to procure for Clements the patent issued to Kirby May 23, 1916, No. 1,184,459. Clements and Kirby were contesting applicants in the Patent Office with reference to an improvement in vacuum cleaners, whereby the suction intake could be diverted from the regular nozzle to a more distant hose-connected nozzle. The primary elements of such a structure are the pump or fan chamber and the suction chamber. The two are divided by a partition having an opening. The suction chamber is provided with a suitable intake nozzle, and the pump chamber has a tangential outlet to the dust receptacle. By the suction effect of the pump, the dust is pulled through the intake nozzle suction chamber and the partition opening into the pump chamber, and is therefrom blown out to the dust receptacle. Each of the contesting applicants improved this old structure by making an opening in the outer shell of the suction chamber, directly opposite and registering with the partition opening, and inserting .through the outer shell opening and tightly into the partition opening an open-ended tube for a hose connection, whereby the suction pull was cut off entirely from the normal suction chamber, [577]*577and this function transferred to the suction nozzle at the outer end of the hose. This improvement has demonstrated its utility and has been largely adopted.
It is quite obvious that the possible presence of the inserted tube is an essential part of this invention; for, without it, the invention did not exist. How this conception should be formulated is a matter of arbitrary rule; and while such rules are convenient or even necessary in the Patent Office practice, their application'ought not to go to such an extent as to divert the patent from the true inventor to another. When the matter of formulation is considered in connection with the structure here described, it appears that the Patent Office procedure forbade uniting in the same claim two elements which do not simultaneously co-operate, but which have alternative and successive use. Accordingly, it was required that the inserted hose tube should not be named as a positive element of the claims, but that the inventive thought should be expressed in terms of the basic structure adapted to receive the inserted tube. Whatever language is selected for this purpose, confusion at once arises. If it has the broadest meaning, referring to any construction having oppositely disposed openings into which such a tube miglii have been inserted, without regard to whether the desired transfer of function was accomplished, it at once is apparent that there can be no valid patent to either contestant, because so broad a construction would cover earlier structures. If it is confined to a precise adaptability- — which may be called the moderately broad construction— to receive such a particular tube in such a manner as to operate successfully for the purpose now desired, there might or might not be pat-entability, depending (perhaps) on whether any earlier such precise adaptability was analogously purposeful or merely accidental. If it is given the narrow meaning, implying that a suitable tube is present ready for use and that provision for changing is the essence of the invention, patentability is not challenged upon this record — though it is difficult to comprehend how there is substantial distinction between the optional insertion of the tube thereby implied and the actual presence of the tube as a named element in the combination. It is not easy to see why a claim should not have been permitted including both an element like this removable tube, which has an active function when it is in place, and an element like the suction chamber which co-operates with the tube to furnish its setting, although the chief function of the suction chamber is temporarily dormant. However that might be (a subject later mentioned), it is this confusion — -which later became worse confounded — which has caused the trouble in this case.
Clements was the senior applicant. His attorneys conceived that the inserted tube was an element of his invention, and drew claims upon this theory. They were rejected, and the examiner said:
“Applicant can only cover, in this application, such a construction, com-pleto in itself without any additions to or substructions from it, as adapts it to any desired class of work or to such classes of work as may be done by it by the mere changes of adjustment of the parts, without removing any of the parts or adding other parts or substituting one part for another.”
[578]*578The applicant amended the claim without prejudice, insisting that the objections went only to the form, and not to the substance, of the invention. Eventually, the applicant and the examiner agreed upon tire proper form, as expressed in claims 1 and 2 quoted in the margin,1 and they became the only claims remaining in the application.
In the meantime, Kirby had come into the office, the structure of one of his drawings being substantially identical with Clements in the respects now involved. His claim which called for the suction chamber with its nozzle and the hose with its connection, was rejected as an aggregation. He then adopted, at the suggestion of the office, Clement’s claims 1 and 2, and an interference was declared with these two claims as the two counts of the issue. The parties both regarded the use of the substitutable hose connection, passing through the suction chamber and cutting it off, as the invention in controversy. Kirby’s preliminary statement (his pleading on the interference issues) claimed conception in May, 1911, followed by reduction to practice and filing of application on August 19, 1911. Clements claimed conception in October, 1910, reduction to practice in January, 1911, followed by application filed June 14, 1911. The proofs seem to show without question that Clements had an operative, experimental device in February, ordered castings in April, and made his first commercial machines as early as May, 1911. These embodied the alternative hose connection. Kirby testified that in September, 1910, he was manufacturing a vacuum cleaner which had the customary opening in the dividing wall between the suction chamber and pump chamber, and which had in the outer wall of the suction chamber and directly in line with tire partition opening, a removable cap; this removable cap being for the purpose of permitting access to the interior for cleaning purposes. Eater and about December, 1910, he made a hole in the cap and fastened therein a tube or suction hose and closed his suction chamber nozzle by a separate plug. He says:
“Then, later on, by extending this cap in far enough to reach the dividing wall, it automatically shut off the suction nozzle and connected the hose with the fan chamber. This idea of extending the cap in deeper was the thought that I had about the 1st of May, 1911.”
Obviously, upon the preliminary statements and this testimony and the conception of the meaning of the issue held by both parties, there was no escape from a decision in favor of Clements. Some testimony taken was said to have suggested to Kirby the question whether the issue ought not to be interpreted broadly enough to reach the construction made by him in September, 1910. Accordingly, he proved this [579]*579construction more fully in rebuttal, and on October 1, 1913, applied to the .Examiner of Interferences to'transmit the file to the primary examiner in order that he might amend the declaration of interference so as to remove the existing indefinitcness. The Examiner of Interferences reviewed the evidence and held that the primary examiner, in formulating the claims to be allowed and the issue, had considered the invention as that construction which included “provision for changing from one kind of cleaner to another”; that the claims of the issue were neither indefinite nor alternative; and that Kirby’s motion was unduly delayed. He therefore denied Kirby’s application; and the Commissioner, on appeal, affirmed the action of the Examiner of Interferences, referring to the obvious fact that if the issue should be amended, and Kirby’s statement dated back, Clements would have the right to take further testimony, since there would then be an issue upon which he had never been heard.
The interference itself then came on for hearing before the examiner. He first considered Kirby’s motion to amend and date back his preliminary statement, and denied it, both on account o f the delay and because the evidence which Kirby had taken was not sufficient to sustain his case, even upon the broad interpretation of the issue which he wished to have adopted. In substance, the examiner gave to the claim what we have called the moderately broad construction, and held that Kirby’s request to amend his statement was immaterial, since even with this relatively broader construction, Kirby must fail on the proofs — Kirby announcing that he did not care to take further proofs. Proceeding to the merits upon this moderately broad issue, the examiner awarded priority to Clements, upon jiroofs which seem to be too clear for controversy. On appeal, the Board took the third or narrowest view of the meaning of the issue, and observed that with the broadest view, which called merely for the existence of the two openings in general alignment, a patent must he refused, as the primary examiner had also held, because of earlier patents showing that general construction. The Board recommended that, after priority was determined, the successful applicant should be required to use a formulation including the removal tube as a positive element, in order that there might be no misunderstanding as to the .scope of the invention. Proceeding to the subject of priority, and with this view of the issue, the Board also affirmed the award of priority to Clements — as, indeed, it must.
Thereupon .Kirby, insisting that the broadest construction was the proper one and that the Board had held it unpatentable, asked that the matter be remanded to the primary examiner in order to redraft the issue so as to confine it to what was patentable. The Commissioner denied this application, for the reason that, when one construction of claims would make them unpatentable and another would make them valid, the latter construction is the one that should be adopted. Upon Kirby’s appeal from the decision of the Board denying the motion to amend and awarding priority to Clements, the Commissioner affirmed the Board. Pie held that the narrower construction confining the invention to a structure fitted for and intended for receiving the removable hose tube was the right construction; that the broadest one (and [580]*580probably he means the moderately broad one) would make the Invention unpatentable; and that, with this view, there was no occasion for Kirby to amend his primary statement, and the case against him on priority was clear. There was then an appeal to the Court of Appeals of the District. Error was assigned because Kirby should have been allowed to amend his preliminary statement and because priority should have been allowed to him. The Court of Appeals, on May 28, 1915, filed its opinion in which it gave its attention mainly to the distinction between the broad and narrow 'constructions of the issue, and held that it should have, for interference purposes, the broader construction, and concludes:
“It is apparent that appellant’s motion to amend his preliminary statem'ent would have been granted, but for the interpretation placed upon the claim, and we are of the opinion that it should have been. His evidence very clearly establishes the fact that, several months prior to the earliest date which may be given the appellee, he manufactured and sold cleaners answering the requirements of the counts as we have construed them. The decision is therefore reversed and the case returned to the Patent Office for appropriate action.”
Thereupon the various tribunals of the Patent Office took the view that the interference had been finally decided by the Court of Appeals, and that an award of priority to.Kirby had been made by that court, and that a patent must issue to him, and not to Clements, containing the claims which formed the issues in the interference.. Clements moved that the claims be amended, so as to specify the narrow construction, and that priority be thereupon awarded to him. The Commissioner denied this motion upon the ground that the Court of Appeals decision was conclusive, both as to what was decided and as to what might. have been decided. Clements then moved to dissolve the interference upon the ground that under the broad construction the issue was un-patentable. The Commissioner refused to accept this view, holding again that the Patent Office had no further authority to consider the question of priority, and that there was no longer any existing interference. Clements then amended his application, inserting claims positively including the hose connection as an element. These claims were rejected on the interference record. Clements appealed to the Board, and the Board affirmed. Clements appealed to the Commissioner, and the Commissioner affirmed. In the meantime, and while this last appeal was pending before the Board, and on the 23d day of May, 1916, the patent issued to Kirby, containing the claims which had been in interference and five others, which included the removable tube as a positive element; and Clements, on June 16, 1916, filed this bill in the court below, under Revised Statutes, § 4915 (U. S. C. S. § 9460).
The parties made a stipulation that a copy of the interference record should be received in evidence, and, together with copies of patents and other proceedings, should constitute the entire record in the case. It being thus ready for final hearing, Kirby made a motion to dismiss, upon two grounds: Eirst, that sucia a bill was maintainable only after the refusal of the patent by the Court of Appeals, which refusal the bill did not allege; and, second, that, if there had been such refusal, Clement’s proceeding in the Patent Office was abandoned under Revised [581]*581Statutes, § 4894 (U. S. C. S- § 9438), by failure to prosecute within a year after the last action. The bill was permitted to be amended to allege, in terms, a refusal by the Court of Appeals through the proceedings which had been recited, and the court postponed the motion for argument until the final hearing. Upon such hearing, the bill was dismissed, because not filed within a year after the action of the Court of Appeals, and with an opinion which did not indicate that the trial court had considered any other question.
The result reached by the Patent Office tribunals has worked a plain miscarriage of justice. It is beyond reasonable doubt that Clements, and not Kirby, was the first to use the inserted tube to shift the intake nozzle to the distant position, or to think of using it, or to make a device •which accomplished this result. It had been held by the primary examiner, by the Examiner of Interferences, by the Board, and by the Commissioner that this was the only patentable invention involved, and this conclusion was not criticized byr the Court oí Appeals; yet the patent therefor has been issued to Kirby and refused to Clements.
The question chiefly considered by the court below, and to which the arguments of counsel are largely directed, is whether an applicant, who is refused a patent by the Commissioner, must exhaust any existing remedy by appeal to the Court of Appeals, before he may resort to his remedy under section 4915. It is claimed for Clements that a resort to appeal is optional by the very words oC the statute, “refused, either by tlic Commissioner or by the * * * court,” that the legislative history of the law confirms this view, and that the decisions to the contrary are obiter or ill-considered,- while it is contended for Kirby that we should apply what is called the distributive theory, and hold that, in those cases where there is no appeal from the action of the Commissioner to the. Court of Appeals, his final action should be treated as the refusal contemplated by section 4915, while, in those cases which may be appealed, this remedy must be followed before the other remedy can be invoked. See Smith v. Muller (C. C.) 75 Fed. 612; McKnight v. Metal Co. (C. C.) 128 Fed. 51. We do not find it necessary to decide this question, and for the purposes of this opinion only we accept the distributive theory; and we come to consider the character of the action which the Court of Appeals took in this case.
It is not dear, on the face of its opinion, whether the Court of Appeals intended to adopt what we have called the broadest or what we have called the moderately broad construction'of the issue. We have concluded that the latter was its true intent. It seems quite obvious that there was nothing patentable in merely locating the two openings opposite each other for convenient access to the inner through the outer. The Patent Office tribunals had all held that this was not new with either party, and that to give the issue this broadest construction would make it unpatentable, and the Commissioner had expressly held that, where a broad construction of an interference issue would make it unpatentable and the narrow one would make it patentable, the narrow should be taken as the true meaning. The Court of Appeals did not criticize any of these conclusions. The cases upon which it relied (Miel v. Young, 29 App. D. C. 481; Lindmark v. Hodgkinson, 31 App. D. C. [582]*582612), to the effect that the broader construction should be adopted, did not disclose any such situation as here existed, but pertained to situations where there was a possible choice between broad and narrow without destroying the whole subject-matter.
A claim to that precise adaptability, which would permit the insertion of the tube so as to make a fairly close joint at each opening, and thereby, of itself, cut off the suction from one nozzle and transfer it to the other, was not anticipated upon this record, and that limitation is fairly called for by the language of the issue. The two openings do not “register,” nor is the inner opening “adapted to receive * * * a hose tube extending into the inner opening,” nor are the two' “openings adapted to receive an inlet tube” in the sense which the context fairly requires, unless the inner opening is hot only in suitable alignment with, but is at least no larger than, the outer one; for, if it is larger, no tube which can pass through the outer opening can be received in or extended into the inner opening in such manner as to cut off the suction and accomplish the sole object of the invention. True, the fan suction might have been powerful enough, so that a leak around the outside of the tube when it entered the inner opening would have only weakened and not destroyed the suction pull at the distant nozzle; but substantially to cut off the primary nozzle was the whole object of the invention, and this required the close adaptation and registration. A consideration of all the circumstances, aided by the strong presumption that the Court of Appeals did not intend to do a vain thing, requires that this meaning should be given to its reference to “the broadest interpretation which they will reasonably support” and to the structure which “would readily permit the change from one kind of cleaner to another.”
[1] We must have grave doubt whether the action of the Court of Appeals was -intended to be an award of priority to Kirby upon an issue as to which Clements could persuasively claim he had never been heard, or to be anything more than a remand to permit an amendment of the pleadings and further proceedings thereon; but. we pass that question and assume that, as Kirby contends, the action of the court is rightly to be construed as an award of priority upon the two claims involved, and as final, in the sense that the Patent Office could only execute and not review. Was, then, the filing of this bill barred by the delay of more than one year between May 28, 1915, and June 16, 1916? There is no limitation statute which provides that a bill, under section 4915, based upon a refusal by the Commissioner or the Court of Appeals, must be filed within one year after such refusal — as counsel in argument seem more or less to assume. The only limitation of this character results from the decisions that to file a bill under section 4915 is, within the contemplation of section 4894, to “prosecute” the application for patent, as must be done “within one year after any action therein,” in order to avoid the inference that the application shall be “regarded as abandoned” — unless, even then, the delay may be found to have been unavoidable.
The question, therefore, is whether this application was abandoned by a failure to prosecute for more than a year. The mere recital of the [583]*583facts furnishes an indisputable answer. So far from being abandoned, the application was prosecuted with the utmost diligence and persistence. It is true that Clements made no further effort to get these two claims which had been in interference; but the application is and continues to be a unit, and it is plainly not abandoned unless it is wholly so. Even if it were true that Clements’ theory of the further action which should be taken was wholly erroneous, and he was asking action which he should have known he could not get, or even if the Patent Office had lost jurisdiction to do anything except issue the patent, the inference of abandonment would not follow. It is a familiar principle that, though a tribunal may have no power to grant the relief asked, it has jurisdiction to hear and determine the question of power; and we perceive no tenable theory upon which it may be said that Clements’ two applications to the Commissioner in June and July, 1916, each followed by “action” by the Commissioner, are consistent with a then maturing abandonment by him of his application. Even the language of Patent Office rule 171, “such proper action as the condition of the case may require,” cannot mean that a request is not “prosecution” unless it is grantable.
That inference of abandonment is as much without support in authority as in principle. In Gandy v. Marble, 122 U. S. 432, 7 Sup. Ct. 1290, 30 L. Ed. 1223, one year and four months more than the specified time had elapsed after the decision oí the appeal from the Commissioner, and no step of any kind had been taken. It was not denied that the application had been abandoned under section 4894, if that section applied at all to a suit under section 4915. In Westinghouse Co. v. Ohio Co. (C. C.) 186 Fed. 518, the facts in this respect are not stated, but it is assumed that the requisite time had elapsed without the taking of atiy step in prosecution — except the filing of a previous hill under section 4915, which was wholly ineffective because not filed against the right party; and manifestly whether certain action may properly'be characterized as prosecution of the claim presents materially and perhaps vitally different questions when it is taken in a tribunal with general jurisdiction gud when application is made in a forum which is open in one particular situation only.
In Colman v. American Co. (D. C.) 235 Fed. 531, the decision of the Court of Appeals was on June 3, 1913; the Supreme Court refused a certiorari on October 29, 1913; on February 24, 1914, the Commissioner gave to the applicant a formal notice, apparently to the effect that the case -was disposed of in accordance with the Court of Appeals decision; the bill, under section 4915, was, in effect, filed on February 1, 1915; and there was no suggestion that after October 29, 1913, the applicant had taken any step whatever in the prosecution of the application. The case holds, in effect, though not in words, that the formal notice by the Commissioner to the applicant in February, 1914, was not the “action therein” which section 4894 contemplates — a conclusion for which no reason is given, but whether right or wrong is immaterial here.
The present situation is not like a writ of error, which is followed in the wrong court until the statutory time for application to the right [584]*584court is gone. There the limitation runs from the day of the act to be reviewed; here it runs from the time when the plaintiff becomes acquiescent and begins to do nothing. From these considerations, it necessarily follows that, as to the two claims which were in interference, Clements’ application for patent had been refused within the meaning of section 4915, and that it had not been abandoned within the meaning of section 4894. As to the remaining claims issued to Kirby and sought in this bill by Clements, the situation is different. -
[2] The question as to the effect to be given in the Patent Office to a decision of the Court of Appeals on an interference is a modified question of res judicata. Usually, in declaring an interference, the examiner' includes all permissible claims, both generic and specific, readable upon the disclosure of both applicants, and these make the issues litigated and decided. If he fails to do so, the parties may have the omission corrected,, and such right of correction may stand for issues which might have been decided though they were not. Sometimes second thoughts suggest broadly generic claims which have been overlooked; but since the invention of the specific necessarily includes that of the generic, the issue on priority would often be the same. On the contrary, the finding of priority on a generic issue may or may not require the same finding as to the specific.
As appears by the statement made on behalf of the Commissioner in Re Curtiss, 46 App. D. C. 183, at page 188,' it had been the practice of the Patent Office to permit second interferences between the same parties in special cases where the character of the claims before in issue had prevented a complete decision of the whole controversy. The decision of the Court of Appeals in Blackford v. Wilder, 28 App. D. C. 535, followed by Horine v. Wende, 29 App. D. C. 415, and Carroll v. Hallwood, 31 App. D. C. 165, had caused the Patent Office to change its practice and apply in all cases the sweeping rule that the Court of Appeals decision should be considered an adjudication as to eyery issue which might have been made upon the two° applications as well as to those which were made. These three cases were followed by the Patent Office as a rule of decision in its action in declining to consider all of Clements’ later applications, and were relied upon by the court below and in this court in support of this conclusion. They were reviewed and modified by the decision in Re Curtiss, and the earlier rule of the Patent Office was, in effect, approved. Paraphrasing that opinion to fit the facts of this case, it is held:
“The real test here is whether or not the granting of these narrow claims later asked by Clements would dominate the subject-matter upon which Kirby prevailed in the interference. * * * If Clements should be founds to be entitled to priority on these narrow claims, it would in no respect affect the prior rights of Kirby, acquired through the former interference.”
. It should be noted that this opinion was not rendered until March, 1917, but it nevertheless is applicable to the rightfulness of action which had been taken in 1916, and both by its reasoning and the force of its authority makes clear that the Patent Office ought to have entertained Clements’ later narrow claims, and, if necessary, declared an[585]*585other interference.2 If there could otherwise be doubt of this result, it would have followed from the peculiar situation existing in this case, where the Patent Oflice had refused to put the narrow claims into the issue because it held they were already there.
We think it right to say that these later narrow claims were “refused by the Commissioner” before this bill was filed, although it is trite that the action of the Commissioner, in person, did not occur until a later date. After the Commissioner had twice decided that the whole matter was concluded, and that the Patent Office had no power to consider narrow claims which Clements might ask, the later presentation of them by Clements to the examiner was a formality, and so with the appeal to the Board and with the appeal to the Commissioner. It was held in McKnight v. Metal Co., supra, 128 Fed. at page 54, that even before the application had been formally refused, or the patent granted to the other contestant, there was a sufficient refusal to justify resort to section 4915, when it became apparent that the controlling question had been several times decided in the Patent Office. We, think that, under the circumstances of this case, Clements was at liberty, after the Commissioner had decided the substantial question and the primary examiner had decided the formal ones, to desist from further proceedings which the Patent Office officials were likely to consider as vexatious, and consider his application as refused. We have the less hesitation in reaching this result, because, in any event, the filing of a supplemental bill, after the Commissioner did act, would have cured this defect and demonstrated that it was in this case an unsubstantial formality.
[3 J There remains, of procedural questions, the one whether, under the above-described distributive theory, there should have been an appeal to the Court of Appeals, from the refusal of the narrow claims, before this bill could be filed as to them. Again we assume, without deciding, that if there could be an appeal, there must be one; but it seems to be clear that under Ex parte Frasch, 192 U. S. 566, 24 Sup. Ct. 424, 48 L. Ed. 564, and Steinmetz v. Allen, 192 U. S. 543, 24 Sup. Ct. 416, 48 L. Ed. 555, an appeal would not lie, because the substantial holding of the Commissioner was that he had no jurisdiction, and appeals cannot be taken on such questions. The claims must have; been “twice rejected,” under 'Revised Statutes, § 4909 (Comp. St. § 9454), before starting the chain of appeals which now ends in the Court of Appeals (formerly the District Supreme Court, under R. S- § 4911). We do not overlook that in Re Curtiss the Court of Appeals entertained an ,appeal upon a similar point; but the question of jurisdiction was not raised, and the existence of this right of appeal is, at best, too uncertain, in this case to compel a resort to it, where the necessity of such resort in any case is only inferential.
[4J Kirby’s counsel say that, because Clements could have procured a writ of mandamus, under the cases just cited, to compel the Commissioner to receive and consider these claims, therefore there is no remedy under section 4915. The natural inference is to the contrary. Man[586]*586damus is an extraordinary remedy, awarded only when no other is adequate. It could, at the most, only direct the Commissioner to act, but could not control his action. The remedy under section 4915 is far more adequate, where the matter has progressed to the point of refusing the patent, and it would seem that if in either the Steinmetz or the Frasch Case there had been any right to get a full review of all questions by application to a court with adequate powers, the mandamus would have been refused because unnecessary. Kirby’s contention requires us to write into the statute an additional condition precedent before invoking section 4915 so as to make it say:
“Whenever a patent is refused either by the Commissioner of Patents or by the [Court of Appeals] of the District upon appeal, or by any other court having any jurisdiction over the Commissioner,” etc.
.We conclude that the dismissal of the bill because the right to file it had never accrued or had been lost, was error. The court below did not consider the merits, but that does not affect our duty to do so. The record was complete and there was a final hearing. The settled policy of the law is against piecemeal appeals, and they are approved only in special cases. We see no reason why the case should go back for further hearing. It is true that the physical exhibits used in the interference were not shown to the court below, and those for Kirby have not been sent to this court; but the interference record makes it clear that they are cumulative and of no controlling importance.
[5] It is well established that the court appealed to under section 4915 has full jurisdiction to re-examine all matters determined by the Court of Appeals. The latter becomes, for this purpose, one of the tribunals of the Patent Office, and its decision is in no sense a final decree, which cannot be thus collaterally attacked, though it will be followed unless the contrary is established by convincing testimony. Morgan v. Daniels, 153 U. S. 120, 14 Sup. Ct. 772, 38 L. Ed. 657, Courson v. O’Connor (C. C. A. 7) 227 Fed. 890, 894, 142 C. C. A. 414, and see Baldwin Co. v. Howard Co. (April 11, 1921) 255 U. S. -, 41 Sup. Ct. 405, 65 L. Ed. —.
[6] It is clear to us that, upon the moderately broad issue as we have defined it, priority should have been awarded to Clements, and that the record permits.no other result. Kirby himself testifies that in his structure of September, 1910, the inner opening was so much larger than the outer one (five-sixteenths of an inch in diameter) that any tube which could have been inserted through the latter into the former would not have cut off the suction chamber, and that, about the 1st of May, 1911, he first thought of accomplishing the object of the invention in controversy, and he did so by extending the tube across the suction chamber and by enlarging the outer opening so as to make it the same size as the inner one, so that the tube would fit both. Thus he, for the first time, conceived and reduced to practice the “registering” and the “adaptability” which are called for by the claims, even with the moderately broad construction which we think the Court of Appeals gave to them. Clements’ priority in this respect was' found by all the Patent Office tribunals and is hardly in dispute. It seems that the Court of Appeals overlooked the insufficiency of Kirby’s construe[587]*587tion of September, 1910, in the respect to which we have referred; and counsel for Clements say, in effect, that they did not call the court’s attention to it because not believing that it could become material.
If the patent had not issued to Kirby, there might be embarrassment about granting relief to Clements as to his narrower claitns, because they woidd not have been subjected to that revision and modification which ought to precede the action of the court, and mandamus might be necessary; but Clements is asking precisely and only the claims which the Patent Office has found properly formulated and allowable, as against all prior rights, and has included in the patent issued to Kirby. They are seven in number. The first two are those which were in interference. The remaining five are different formulations 'of the narrower conception, which include the hose connection or tube as a positive element. The formal objections of the Patent Office on the ground of aggregation have been overcome, either by a change of opinion on its part or by some particular wording which avoids the objections — it is not material to determine which.
[7j The decree should be reversed, and the record remanded, with instructions to enter a decree in favor of Clements pursuant to this opinion. The costs of this court and of the court below should go against Kirby. We think that the provision of section 4915, that “all the expenses of the proceeding' shall be paid by the applicant whether the final decision is in his favor or not,” applies only to those cases which had just been mentioned, viz. “where there is no opposing party” and the papers are served on the Commissioner only.