This is an appeal from a final decree entered September 13, 1935, by the District Court for the Northern District of California, Southern Division, dismissing a bill of complaint charging infringement of two patents. The two patents are No. 1,875,-259, applied for November 1, 1921, and No. 1,875,260, applied for March 12, 1925, both issued August 30, 1932, in the name of George D. Parker. For convenience here, as in the trial court, these patents will hereinafter be referred to by only the last three digits, viz., ’259 and ’260 respectively, or as the Parker patents.

The original plaintiffs were Clara B. Parker, executrix of the last will of the patentee, George D. Parker, deceased; James M. Leaver, Jr., owner of a half interest in the ’260 patent; and Charles E. Evans; doing business under the name of Parker-Leaver-Evans Wire Tie. Under a contract made in October, 1922, each plaintiff had an equitable interest in the outcome of any suit brought on either Parker patent.

Before trial, Clara B. Parker and Charles E. Evans died, whereupon Donald Parker and Citizens National Trust & Savings Bank of Riverside, California, administrators of the Estate of George D. Parker, deceased; and Minnie Amanda Evans, executrix of the last will and testament of Charles M. Evans, deceased, were substituted as plaintiffs for Clara D. Parker and Charles E. Evans, respectively.

The original defendant was Pacific Box Corporation. At the trial the Eby Manufacturing Company, the manufacturer of the machine charged to infringe the patents in suit, intervenéd and defended the suit.' Unless otherwise herein noted, the original defendant and -the intervenor will be designated as Defendants, and all appellants will occasionally be referred to as Plaintiffs.

Subsequent to the entry of the decree for the defendants, notice was given to James M. Leaver, Jr., by the other plaintiffs to join in the contemplated appeal to this Court, but Leaver failed to join. The judge of the District Court made his order allowing appeal. by the remaining plaintiffs, in the following words: “ * * * it appearing that the remaining plaintiff named in the decree appealed from, James M. Leaver, Jr., has failed and refused to join in said appeal, Now Therefore * * * it is ordered that said appeal be, and the same is hereby allowed * * *

Appellees contend that the appeal should be dismissed, so far as patent ’260 is concerned, for the non-joinder of Leaver as party appellant. It is their contention that the finding of the judge of the District Court with reference to Leaver’s “refusal” to join is not justified; that all the record shows is the notice served on Leaver and his “failure” to join in the appeal; and that this is not the equivalent of summons and severance.

This Court has recognized the rule that on an appeal taken only by some of the persons against whom a joint judgment is rendered, there should be a summons and severance, or its equivalent, and that in the absence of such summons and severance or its equivalent, a motion to dismiss is well taken. Pflueger v. Sherman, 9 Cir., 75 F. 2d 84, 89 ; Mittry Bros. Const. Co. v. United States, 9 Cir., 75 F.2d 79, 81 . However, there appears to be no merit in appellees’ contention that the absence of the record to show a “refusal” as well as a “failure” of Leaver to join in the appeal destroys the effect of the notice given to Leaver and the court order based thereon as the equivalent of summons and severance.

The Supreme Court of the United States in the case of Masterson v. Howard, 10 Wall. 416, 418 , 19 L.Ed. 953, 954 , in holding that a case must be dismissed in the absence of summons and severance or its equivalent, said, “We do not attach importance to the technical mode of proceeding called summons and severance. We should have held this appeal good if it had appeared in any way by the record that Maverick had been notified in writing to appear, and that he had failed to appear, or, if appearing, had refused to join. * * * We think there should be a written notice and due service, or the record should show his appearance and refusal, and that the court on that ground granted an appeal to the party who prayed for it, as to his own interest.”

In Inglehart v. Stansbury, 151 U.S. 68, 72 , 14 S.Ct. 237, 238 , 38 L.Ed. 76, 77 , the Supreme Court said: “ * * * it is quite clear that Inglehart’s heirs could not appeal alone, without joining the other defendants as appellants, or showing a valid excuse for not joining them. This could only be shown by a summons and severance, or by some equivalent proceeding, such as a request to-the other defendants and their refusal to join in the appeal, or at least a notice to them to appear, and their failure to do so * *

Appellees next contend that the appeal should be dismissed because of the nonjoinder of Parker-Leaver-Evans Wire Tie; that the rule as to summons and severance cannot be invoked as to that concern because there is no showing of service of any notice at all on them.

This contention is likewise without merit, as the complaint shows on its face that the Parker-Leaver-Evans Wire Tie is not a legal entity. The original plaintiffs were designated as “Clara B. Parker, etc., James M. Leaver, Jr., and Charles E. Evans, doing business under the name of Parker-Leaver-Evans Wire Tie.”

After this appeal was taken, the Wire Tie Machinery Company, a California corporation, acquired the interests of the Parker administrators, and was permitted by this court to be substituted for the Parker administrators as party appellant. Ap- pellees complain that the Wire Tie Machinery Company has no position before this court, in view of certain provisions in the agreement between the original plaintiffs prohibiting assignment of their respective interests, without unanimous consent of the parties.

Suffice it to say that this court has already made its order substituting the Wire Tie Machinery Company as party appellant. We will not disturb that order at this time, in the absence of showing that the assignment of the interest of the Parker Estate to the Wire Tie Machinery Company was not made with the consent of the other parties involved.

_ The original complaint herein charged the Pacific Box Corporation as user of a bundle tying machine (hereinafter designated as the Eby machine), built accord-mg to the teaching of Dunn and Eldndge, with infringement of claims 43, 44, 45, 46, 47, 48, 49, 52, 55, 60, 67, 75 and 76 of the 259 patent and of claims 7, 12, 13, 65, 66, 69, 70, 76, 85, 86 and 87 of the 260 pat-en^-

The ’259 patent covers a machine with duplex revolving arms for the purpose of carrying two wires around a bundle simultaneously and tying both of the same m what is known as a fiat knot. The method of binding is what is known as the tight looping method, that is, laying the wire under tension completely around the bundle or ox'

Prior to the construction of the machine covered by this patent, there were being built various types of binding machines, One of such machines, covered by the Thompson patent, No. 1,152,670, issued September 7, 1915, also utilized a rotating arm to carry the wire around the bundle, That machine however, tied the ends of the wire in a corner knot, or what is known as a pigtail knot. The pigtail or corner knot left the twisted wire protruding from the bundle, and created a menace to stevedores handling the packages; furthermore, the twisting of the wire in this manner tended to crystallize the wire, and numerous breakages would occur. These disadvantages were eliminated by the use of the so-called flat knot. Another of such machines, covered by the McChesney patent, No. 1,357,883, issued November 2, 1920, did tie the wire in a flat knot, but that machine utilized what is known as the loose looping method of binding. In other words, that machine has a motor-driven feeding roller that pushes the wire through an open trough that curves around the bun-die. After the wire is thus pushed through the trough loosely around the bundle, a gripper that slides back and forth draws the excess wire back until the loop is pulled out of the trough and shrunk in size to t^le bundle.

The rotating arm of the machine covered by the Thompson patent operated in a uni-directional rotation. In order to provide for an overlapping of the ends of the wire necessary to the tying of a flat knot, Parker in his machine provided for the oscillation of the arm back and forth somewhat in excess of one full revolution,

Twomley, Parker’s technical emplovee, testified that early in 1920 Parker gave him the idea of making his drawing for the >259 machine. The drawings were started about july> i920, and completed in the early part of 1921. The patent was applied for jn November, 1921. The manufacture of the machine was commenced in the early part of 1921 and completed in the latter part of that year. Only one machine was manufactured under that patent,

D who testified for Eb tegtified that he went tQ work for Eb ¡n 0ctob im In DeCember of the same year he b t0 build a bindi maclline. In the summer Qr fall of 1921 he roduced an op_ erating maclline, wWch was completed to the satisfaction of Dunn in February, 1922. It was circularized to the trade in March, l922. That machine also utilized what is known as the tight binding method, but instead of a revolving arm to carry the wire around the bundle, it used a ring gear for that purpose. (Very loosely speaking, a “ring gear” as used in connection with binding machines, consists of a cogged ring, larger in circumference than the box or bundle to be tied. The wire is conducted by this rotating ring around the box or bundle.) The machine tied a corner, or pigtail knot, as distinguished from a flat knot. He applied for a patent for this machine on July 6, 1922. Two corporations ordered the machines, and in October, 1922, Dunn took the two machines to Honolulu to be demonstrated there. They were accepted and paid for. When Dunn was in Honolulu, he had a conference with White, manager of one of the purchasing corporations, who suggested that he preferred to have a flat knot rather than the corner knot as produced by the machine which he purchased. Upon his return from Honolulu, Dunn commenced his drawings to convert his machine into a machine tying a flat knot, Dunn testified that he completed these drawings in the summer of 1923.

There is much conflict in the testimony as to just when the construction of Dunn’s new flat knot machine was actually completed. The trial court found as a fact that “The first full-size flat knot type of Dunn and Eldridge machine was completed and successfully operated in June, 1923”. Plaintiffs contend, however, that the court was in error in this finding.

The following are excerpts from the record with reference to the date of completion of the Dunn and Eldridge flat-knot machine:

Guy A. Dunn testified:

On cross examination, Dunn testified: “I first saw a full automatic wire binding machine tying a flat knot at the Los An-geles Lumber Products Company. That was a Parker machine. That was not pri- or to the actual development on my part of the ®by flat knot machine.”

(Further evidence discloses that this was in August, 1923.)

From the above quoted excerpts from Dunn’s testimony on cross-examination, it seems apparent that Dunn admitted that he did not commence the construction of the Eby flat-knot machine until after his observation of the Parker machine in August, 1923.

Dunn further testified:

Dunn left Honolulu in December, 1922. Therefore, from the above testimony it seems apparent that the Eby flat-knot machine was not completed until September or October of 1923.

At the trial, in addition to the testimony taken in open court, portions of the testimony of Guy A. Dunn, of John D. Eby, deceased, president of the defendant Eby Machinery Co., of George D. Parker, deceased, the patentee of the Parker machines, an,d of other witnesses taken in various suits on other patents and in a Patent Office interference involving Dunn & Eld-ridge and Leaver, were introduced as exhibits. The following are excerpts from such testimony relative to the date of completion of the Eby flat-knot machine:

John D. Eby testified: “This last machine [flat knot machine] was completed in the spring of 1923 — no, it was sketched' — we commenced on this last machine in the spring of 1923, commenced to plan on this machine in the spring of 1923, and we sold one of these machines in December, 1923, and delivered it the next June.”

Guy A. Dunn testified in one of these previous suits: “I returned from Honolulu in the winter or late fall of 1922, and on arrival in San Francisco I discussed the matter with Mr. Eldridge and we immediately set out to build a flat knot machine. Some of the drawings of that machine were started immediately. They were completed in the early summer or spring of 1923. The construction of the first flat knot machine was held back for several months for the reason that we had quite a few corner knot machines in stock, and quite a bit of money invested, and Mr. Eby objected to proceeding with the building of this flat knot machine until these corner knot machines were sold. I think I said the construction of the first flat knot machine was started about the spring of 1923. I don’t remember exactly when the first flat knot machine was completed.”

The witness at this point was referring to the observation of the Parker machine in San Pedro in August, 1923. It should be noted that Dunn is making no claim at this time that the flat knot machine was completed in August,, 1923, but rather that he “had it thoroughly analyzed and knew exactly what (he) intended to do”.

■ Dunn later testified, “I have checked up on certain definite dates as to the completion of the manufacture and sale of these different machines on a card taken from our records. * * * The first flat knot machine of the type that is shown in the drawings Exhibit U, was built on June 1, 1923. That machine was placed on demonstration before several corporations, * * * but after some hold-back of the machine until we could get rid of the corner knot machine, that we had completed, as I stated yesterday. * * * The first machine of the flat knot type was sold on December 27, 1923, but it had been demonstrated before the trade months previous to that.”

This is the only place in the record where it is stated that June, 1923, was the date of the completion of the flat-knot machine, as distinguished from the drawings.

He later testified, “The flat knot machine and the corner tie machine had both been completed and constructed prior to December of 1923, according to the records.”

This testimony would indicate that Dunn is drawing a distinction between “comple tion” and “construction”, and that when he referred to the machine being “completed” he meant the drawings were completed.

Dunn testified in another of said previous suits, “Stating what I did subsequent to the completion of the design of the spiral twist machine in the line of designing a wire tying machine, as it became apparent that the market required a machine of higher speed and a knot that would be more desirable, I had the flat, so-called flat-knot in mind for some time, remembering it as the tie on cement barrels, etc., but up to this present time that you speak of I saw no practicable, feasible way of applying it to an automatic machine, but after doing some careful study the. development of the flat knot machine started. The idea came to me of how to apply the tie of a flat knot to an automatic machine in 1923. In my own mind I was perfectly satisfied with the operation of the corner knot machine, and it was a question of applying the flat knot to the other general principles. As near as I can remember at the present time, it took something like nine or ten months.”

The record in one of these previous suits also discloses the following proceedings at the trial: Counsel for plaintiffs objected to the pertinency of certain identification cards posted over the machines exhibited in evidence, on the ground that they were merely hearsay. The Court said, ■“The cards are merely here to identify the machines as exhibited. I cannot see any ■objection to them. It is understood that they are not evidence. They are simply for the purpose of identifying the machine. .1 think they should be read into the record."

The card was read into the record: '“This is the improved I)unn and Eldridge wire tying machine. First machine built June 1, 1923; first machine sold December 27, 1923; delivered June 30, d924.”

This latter excerpt from the record is .cited by defendants in their brief as testimony corroborating the finding of the trial court in this case- to the effect that the “first full-size flat knot type of Dunn and Eldridge machine was completed and successfully, operated in June, 1923”. However, as is evident from the above, said card cannot properly be regarded as evidence.

It should further be noted at this point that there is no evidence in the record to the effect that the Eby flat-knot machine was “successfully operated in June, 1923” as found by the trial court.

On March 17, 1924, Eby applied for a patent on his new flat knot machine.

Prior to this time, Twomley, working for Parker, started working on a new" machine, eventually covered by the ’260 patent involved in this suit. This machine makes use of a ring gear for winding the wire around the bundle, in the place of the revolving arms of the ’259 patent. It, as did the ’259 machine, ties the wire in a flat Jcnot.

Twomley testified that he started on his drawings for this ’260 machine in the latter part of 1922, and that he started work on the construction of the machine in the early part of 1923. He corroborated this by the introduction of shop orders for materials used in the machine. Twomley further testified that he completed and successfully operated the machine in June of 1923. This date he was unable to corroborate. In August, 1923, Twomley took the ’260 machine to the Los Angeles Lumber Products Company, at San Pedro, California. It was in operation in the neighborhood for thirty days.

At the same time, one of Eby’s machines (corner knot machine) was also being demonstrated at the Los Angeles Lumber Products Company. It is not clear from the record whether or not the Los Angeles Lumber Products Company was already under contract to purchase the Eby machine at the time Parker’s machine was demonstrated. Twomley testified that the purpose of taking the Parker machine over there was to put it in actual practice and to give it a test fun to develop any weakness that there might be in the machine. He further testified that he did not know whether or not the machine was put out there with the object of making a sale. Dunn, on the other hand, testified that the Eby Machinery Company was asked to demonstrate their machine in the plant of the Los Angeles Lumber Products Company, and when he arrived with the machine he found a Parker machine in competition. It is uncontradicted in the evidence, however, that the Los Angeles Lumber Products Company did purchase one of the Eby machines (corner knot machines), and that the Parker machine went back to Parker’s shop. Twomley testified that at the time of this demonstration the Parker ’260 machine was a “successful machine”, although some slight changes were made thereafter. He testified that those changes were merely safety devices.

The Parker ’260 machine was circularized in “The Timberman” in December, 1923, as appears from a copy of that publication introduced in evidence.

On March 12, 1925, the application for Parker patent ’260 was filed.

During the pendency of the Parker applications for the ’’259 and ’260 patents, various amendments were made by Parker, withdrawing certain claims and adding certain others. Appellees contend that the addition of new claims constituted “new matter” and that Eby and Dunn & Eldridge acquired what the law recognizes as “intervening rights”. This phase of the case will be discussed later in this opinion.

On March 23, 1932, Dunn requested the Patent Office to declare an interference between the Eby application of March 17, 1924 (his flat knot machine) with the Parker ’260 application. He copied for the purpose of this Interference some 32 claims from said Parker patent. Later the first Dunn and Eldridge application, filed July 6, 1922 (original corner knot machine) was added to the Interference on motion of Dunn, and Eldridge. Two more Parker claims were added, making 34 claims in all.

The two Parker patents, ’259 and ’260, were issued by the Patent Office on August 30, 1932.

The Interference above referred to was hotly contested in the Patent Office, and was still pending at the time of the trial of the principal case in the District Court.

Appellants have grouped their assignments of error into seven groups, and we will consider them herein in the same order.

Group one is headed “The Decree should have been confined to the issues.”

In the complaint filed by the plaintiffs . herein, defendants were charged with infringement of 13 o'f the 83 claims of the ’259 patent and of 11 of the 99 claims of the ’260 patent. The decree of the trial court, however, holds that all of the claims of both patents are void, for various reasons, and that Dunn and Eldridge were prior inventors of all that was patentable or that might be’infringed in either of said patents.

Claims of a patent are independent inventions. One may be infringed and others not; one may be valid and the rest invalid. The patent does not stand or fall as a unity. Leeds & Catlin Co. v. Victor Talking Machine Co., 213 U.S. 301, 319 , 29 S.Ct. 495 , 53 L.Ed. 805 . It is a fundamental concept of equity procedure that adjudication must be based upon, the issues created ' by the pleadings. Reynolds v. Stockton, 140 U.S. 254, 265 , 11 S.Ct. 773, 776 , 35 L.Ed. 464 ; McEwen et al. v. H. J. Grimes & Co. et al., 6 Cir., 90 F.2d 872, 874 .

In the principal case only the claims alleged in the complaint to have been infringed by defendant were in issue. The decree of the District Court, insofar as it went beyond the issues, should be reversed.

Appellants’ second group of errors is to the effect that the District Court erred in not finding invention in the subject matter of the claims in suit.

The District Court decreed that the device disclosed by the ’259 patent involves a mere unpatentable aggregation of parts found and disclosed in the prior art, and more particularly by the patents to Thompson and McChesney. The District Court also decreed that the device disclosed by the ’260 patent is a mere improvement over that disclosed in the Evans patent. (The Evans patent referred to is the Evans Reissue patent, No. 16,292, issued March 16, 1926.)

The ’259 patent is for a rotary arm binding machine, tying a flat knot. The Thompson patent referred to above is also a rotary arm binding machine. The machine does not make a flat knot, but rather a corner, or pigtail knot. The McChes-ney patent referred to is for a binding machine of what is known as the “loose looping” variety. It ties a flat knot.

The ’260 patent utilizes a ring gear in the place of the rotating arm, and also ties a flat knot. The Evans patent also uses the ring gear, but ties a corner, or pigtail knot.

It is the contention of the appellees, and in this they are supported by the findings of the trial court, that all that Parker did was to incorporate the flat knot of the Mc-Chesney patent into the binding machines' disclosed by the Thompson and Evans patents, and that hence patents ’259 and ’260 lack invention.

, Appellants contend that this is error, claiming that the flat knot' mechanism of the McChesney patent could not have been substituted .for the corner knot of the Thompson and Evans patents — that the arm of the binding machine of the Thompson patent and the ring gear of the Evans patent operated in a uni-djrectional rotation, and hence could not provide the proper overlapping of the ends of the wire necessary to make the flat knot. The Parker patents provide for an oscillating movement of the arm of ’259 and the ring gear of ’260.

The Supreme Court in the case of Webster Loom Co. v. Higgins, 105 U.S. 580, 591 , 26 L.Ed. 1177 , gives its test as to what constitutes invention as follows r “It may be laid down as a general rule, though perhaps not an invariable one, that if a new combination and arrangement of known elements produce a new and beneficial result, never attained before, it is evidence of invention.”

• In the cited case Webster combined the advantages of Bigelow’s rigid lathe with Weild’s trough or wire bar for supporting the wire. By the combination the Webster loom could produce some' 50 yards a day, when previous looms could never produce more than 40 yards a day. The claim was made in that case that the mere combination of the prior patents did not constitute invention, but the Supreme Court in reply to that contention said, “But it is plain, from the evidence and from the very fact that it was not sooner adopted and used, that it did not, for years, occur in this light to even the most skillful persons. It may have been Under their very eyes, they may álmost be said to have stumbled over it; but they certainly failed to see it, to estimate its value and to bring it into notice”.

The court then gives the test as to what constitutes invention above quoted.

The Webster Loom case, supra, has-been cited by the courts many times in their decisions distinguishing between patentable combinations and mere aggregations.

It is well settled that what constitutes an invention, as distinguished from an aggregation, is a question of fact, and not o.f law. Reinharts, Inc., v. Caterpillar Tractor Co., 9 Cir., 1936, 85 F.2d 628, 630 . The court below found as a fact that the claims of the Parker patents in suit did not constitute invention. As a general rule this court will not overturn the findings of fact of the trial judge, since he has had an opportunity of hearing the witnesses testify and is in a better position than this court to judge their veracity. However, in a case such as the case at bar, the question of whether or not the subject matter constitutes invention does not turn upon the truth or falsity of the testimony of the witnesses. Their testimony of necessity is as to their opinions of whether it required more than mechanical skill to effect the combination of parts.

Applying the test laid down by the Supreme Court in the Webster Loom case, supra, we are of the opinion that the subject matter of the Parker patents did constitute invention. It is uncontroverted that never before had anyone produced a fully automatic wire binding machine that tied a flat knot. Dunn, “a skilled mechanic and designer”, who built the defendants’ machine, testified that at the time of his trip to Honolulu (where White suggested to him the desirability of having a binding machine that tied a flat knot) he knew “no practicable, feasible way of applying it (a flat knot) to an automatic machine” and that it took him something like nine or ten months to work it out for Eby.

One of the assignments of error in this group is that the said court erred in concluding as a matter of law, adjudging and decreeing that claim 69 of said patent ’260 is invalid as not readable upon the' dislosure of the patentee thereof in the application for said patent.

Claim 69 reads as follows: “In a wire tying machine, a wire guiding ring rotatable to describe a path encircling an object to be tied, means for supplying wire under tension to said ring, means for rotating the ring to effect encircling of the wire about the object so that one portion of the wire overlaps another portion thereof in a plane substantially parallel to the adjacent surface of the object, and means located within the periphery of said ring for intertwisting the overlapping portions to provide a substantially flat knot.”

The District Court in its Memorandum gives as its reason for such finding that “One portion of the wire does not ‘overlap another in a plane substantially parallel to the adjacent surface of the object.’”

In this connection, Roemer, testifying for the defendant, testified: “Claim 69 is a claim which I doubt reads on either the Parker machine or the Eby machine. This claim, which defines a ring gear type of machine, calls for means for rotating the ring to effect encircling of the wire about the object, so that one portion of the wire overlaps another portion thereof in a plane substantially parallel to the adjacent surface of the object. By that I understand that the slotted twister would have to be placed with its slot parallel to the face of the object at the time the wire was placed in the twister. By reference to the Parker patent 260 it can be seen that in the receiving position of the slotted pinion the slot is directed away from the adjacent surface of the bundle. This is likewise true in the Eby machine, as is clearly evidenced in the model, Defendants’ Exhibit B, wherein the slot is faced away from the bundle at the time it receives a wife. With this in view, I do not see how it is possible for one portion of the wire to overlap another portion thereof in a plane substantially parallel to the adjacent surface of the object. As a matter of fact, the wire in both Parker and Eby overlaps in a plane which is exactly perpendicular to the adjacent face of the bundle.”

We cannot agree with this view of claim 69. An examination of the patent will show that the wires do lie together in the slot of the twister pinion, and that that slot (or the bottom of it where the wires lie) is in a plane substantially parallel to the adjacent surface of the bundle. We think it reasonably accurate to say that one portion of the wire overlaps another portion thereof in a plane substantially parallel to the adjacent surface of the bundle.

Appellants’ third group of errors are to the effect that the District Court should have found that Parker — not Dunn and Eldridge — was the inventor of the subject matter of the claims in suit.

We think the record definitely establishes that Parker was the first inventor of a fully automatic binding machine that wound the wire around the bundle under tension and tied the ends of the wire in a flat knot. His first machine was the ’259, which was built in the latter part of 1921. The application for patent was filed November 1, 1921. Dunn testified that the Dunn and Eldridge first automatic machine was completed in February, 1922. That machine tied a corner knot, as distinguished from a flat knot.

The trial court decreed “That Guy A. Dunn and John Eldridge, joint applicants of application for patent, Serial Number 573,205, filed in the United States Patent Office on July 6, 1922, were prior inventors, with respect to the patentee of Patent No. 1.875.259, of all that was patentable or that might be herein infringed, and shown or disclosed and claimed in said Patent No. 1.875.259. ”

Such decree is erroneous. There is no testimony whatever establishing for Dunn and Eldridge a date prior to the date of the completion of the Parker ’259 machine —and it is further to be remembered that the first machine of Dunn and Eldridge did not tie a flat knot.

We shall now proceed to the question of priority as to patent ’260. Twomley testified that early in 1923 he started the construction of the Parker ’260 machine, and that he completed and successfully operated the machine in June, 1923. This latter date he was unable to corroborate. The machine was put on demonstration in San Pedro in August, 1923. Twomley produced shop records dated from December 14, 1922, to December, 1923, all marked “experimental”. I-n answer to the question as to how the machine could have been successfully operated in June, 1923, and parts purchased after that time, he explained that the parts purchased after June, 1923, were for minor changes that were made in the machine. The application for patent ’260 was not made until March, 1925.

Dunn testified that he started his preliminary drawings for his flat knot machine in December, 1922. We have noted above the conflict in the testimony as to just when this machine was completed. We think from the testimony quoted above, that it cannot properly be found that Dunn completed his flat knot machine before August, 1923, at which time the Parker ’260 machine was being demonstrated in San Pedro.

In connection with the demonstration of the Parker ’260 machine in San Pedro, the trial court found that at that time “Parker was struggling with his [inven tion].in August, 1923”. This finding seems to be unwarranted from the evidence. The testimony of Twomley is clearly to the effect that at that time it operated successfully. It appears that certain minor safety devices were later installed, but we do not feel that that implies that the machine was not successfully operated at the time of the demonstration. Mechanical perfection is hot an essential of reduction to practice. Downs v. Andrews, 1928, 58 App.D.C. 91 , 25 F.2d 218, 224 . The witnesses for the defendant did not testify that the machine failed to accomplish its purpose, but rather that there were difficulties in the operation. Dunn at first testified that, “I would say that the Parker machine was not a successful machine at that time (the time of the demonstration). Parker had a great number of difficulties with the machine, in the operation of it”.

Later, on cross examination, he testified, “I could not answer the question positively as to whether or not while I was present and while the party Leaver was present during the operation of the Parker flat tie machine at San Pedro that said machine functioned successfully and'that the only difficulty encountered was that of crystallization of the shear pin at certain intervals after a large number of bundles had been bound.”

We therefore feel that the record •shows that the Parker ’260 machine was completed and successfully operated before the'Eby flat knot machine was completed. The finding of the trial court that “the Eby Machine antedates anything that Parker did” is in error.

To summarize; Parker completed his ’259 machine before Dunn produced even his corner tie machine, and he completed his ’260 machine before Dunn produced his flat knot machine. True, Dunn’s corner 'tie machine was the first of the four machines in question to utilize a ring gear to carry the wire around the package, but that ■machine tied a corner knot as contrasted with a flat knot. Parker’s '259 machine was the first of the four machines to provide for oscillation of the binding, member so as to cause the necessary overlapping of the ends of the wire to tie the flat knot. All of the claims here in suit are either for the flat knot feature, or the oscillation of the binding member, and pri- ■ ority as to all those claims should be awarded to Parker.

Appellants’ next group of errors relate to the finding of the District Court that the claims of the ’259 patent, or particularly claim 43 thereof, are invalid as being too indefinite or comprehensive or as not complying with United States Revised Statutes, Section 4888, 35 U.S.C.A. §33 .

Claim No. 43 is as follows: “In an apparatus of the character described, a bundle supporting means, rotatable means for tightly winding a tie member under tension around the bundle so that the meeting portions of the tie member are in relative parallelism with each other and in relative parallelism with the adjacent face of the bundle, and means for twisting upon each other the parallel portions of the tie member so as to form a joining parallel with the adjacent face of the bundle.”

The District Court in its memorandum opinion gives as its reason for such decision that the claims of the ’259 patent are invalid because of indefiniteness, the following:

“A bundle ‘supporting means’ of course, is any suitable support. * * * ‘Means for winding wire under tension’ includes the- human arm or any mechanical contrivance carrying wire around the bundle. * * * In addition the language of the claims is so general and vague as in the opinion of the court to teach nothing. Claim No. 44 is as follows:

The court bases its reasoning on the case of White v. Dunbar, 119 U.S. 47, 51 , 7 S.Ct. 72, 74 , 30 L.Ed. 303 , wherein the court said, “Some persons seem to suppose that a claim in a patent is like a nose of wax, which may be turned and twisted in any direction, by merely referring to the specification, so as to make it include something- more than, or something different from, what its words express. The context may, undoubtedly, be resorted to, and often is resorted to, for the purpose of better understanding the meaning of the claim; but not for the purpose of changing it, and making it different from what it is. The elaim is a statutory requirement, prescribed for the very purpose of making the pat-entee define precisely what his invention is; and it is unjust to the public, as well as an evasion of the law, to construe it in a manner different from the plain import of its terms. This has been so often expressed in the opinions of this court that it is unnecessary to pursue the subject further.”

An examination of the cited case, however, indicates that it is not in point in the principal case. In that case the Supreme Court was concerned with the validity of a reissue patent. The claim of the original patent had specified “textile fabric”. The claim of the reissue patent specified “an enveloping material”. The Supreme Court held that the claim of the reissue patent was an enlargement over that of the original patent — that the claim of the original patent could not be altered or enlarged by its specifications. It therefore held that the reissue patent should not have been granted.

In the case of Shull Perforating Co. v. Cavins, 9 Cir. 1938, 94 F.2d 357, 364 , this court held, “The patentee is entitled to have the claims of the patent construed with reference to the drawings and specifications. Where the means referred to in claims are clearly shown in the description of the patent, this description is sufficient to cover the means thus disclosed and its mechanical equivalents.”

We think that in the .case at bar the means referred to in the claims are clearly shown in the drawings and specifications. For instance, said drawings and specifications certainly exclude the possibility referred to by the trial court that “ ‘Means for winding wire under tension’ includes the human arm”.

Appellants next assign as error “The District Court should have found that the machine of the ’259 patent and the defendant’s Eby machine operate on the same principle.” This assignment of error ties in with the next group of errors assigned, which are headed “The District Court should have found the claims in suit to be entitled to a breadth of construction sufficient to include the defendants’ machine”, and we shall therefore consider these two groups of assignments of error together.

The theory of the trial court, as indicated in its memorandum opinion, is that the Parker patents are mere improvement patents, and that therefore Parker is not entitled to a broad interpretation of the doctrine of mechanical equivalents. That the Eby machine is not the mechanical equivalent of the Parker ’259 machine. Appellants, on the other hand, contend that the gear ring of the Eby machine is the mechanical equivalent of the revolving arm of Parker ’259. The language of the trial court is “There can be no claim of mechanical equivalents here”, but in view of the language of the case quoted by the court (Hardison v. Brinkman, 9 Cir., 156 F. 962 ) we do not think that the court intended to imply that there could be no doctrine of mechanical equivalents in connection with an improvement patent, but rather that the patentee of an improvement is not entitled ‘to a broad interpretation of the doctrine. That is in accord with the decisions on the point. In Reinharts, Inc., v. Caterpillar Tractor Co., supra, this court said :

“Appellant contends that, since each of the patents in suit is for a mere improvement, the claims thereof ‘must be strictly limited to substantially the identical construction described in the specification.’ That is not the law. It is the claims of a patent, not its specifications, which measure the invention. [Citing cases.]

“Decisions cited by appellant [citing] do not support its contention. They hold merely that, in the case of a combination patent for an improvement, where all the elements of the combination are old in the art, the patent should be limited to the specific combinations defined by the claims. They do not hold that such a patent should be limited to substantially the identical construction described in the specification. Indeed, one of the decisions cited (Thomas Day Co. v. Doble Laboratories [(C.C. A. 9) 42 F.2d 6, 8 ]) expressly recognizes and declares that, in such a case, ‘the monopoly granted by the patent is not to be limited to the identical devices exhibited in the drawings or prescribed in the specifications.’

“It is true that, in construing the claims of a patent, a greater degree of liberality and a wider range of equivalents are permitted where the patent is of a primary character than when the invention is simply an improvement [citing cases] but it is not true that only primary patents are entitled to invoke the doctrine of equivalents. [Citing cases.] A patent for a meritorious improvement in an old art is entitled to liberal treatment. [Citing cases.]” [ 85 F.2d 635 .]

We therefore hold that, so far as the objection to their being improvement patents goes, the Parker patents are entitled to a fair and reasonable, but not broad, range of equivalents.

Another objection is raised, however, to allowing the doctrine of equivalents to Parker patent ’259 — and that is that the machine was never put into commercial use. ' The record is clear that only one of the machines was manufactured, and that none was ever sold.

In the case of Cocks v. Rip Van Winkle Wall Bed Co., 9 Cir. 1928, 28 F.2d 921, 922 , this court said, “Another reason why the appellants’ combination should not receive the construction due to a pioneer invention, but, on the other hand, should be strictly construed, is the fact that although the invention had been patented nearly ten years before the present suit was begun, it had not been utilized or placed upon the market but was still a paper patent. While the validity of a patent is not affected by-its nonuser, Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U.S. 405 , 28 S.Ct. 748 , 52 L.Ed. 1122 , the nonuser has a bearing on its construction, and, the courts are not disposed to give the same any broader scope than it is clearly required to be given, Westinghouse Elec. & Mfg. Co. v. Toledo, P. C. & L. Ry. Co. (C.C. A.) 172 F. 371 ; Wesel Mfg. Co. v. Printing Machinery Co. (D.C.) 218 F. 178 ; Kestner Evaporator Co. v. American Evaporator Co. (C.C.) 182 F. 844 .”

The trial court found that the Parker patent ’259 is shown not to have been utilized or placed on the market, but is still ■a “paper patent” and as such is limited to a narrow construction. In this we think that the court was correct.

We shall now proceed to a consideration of the question of whether or not the Eby machine infringed the claims in suit of either the ’259 or the ’260 machine of Parker.

Taking up first the claims of the ’259 machine, as pointed out above, they are limited to a narrow construction. A typical claim is No. 46, which reads as follows: “In a tying machine, means for effecting encircling of a binding wire about an object to be tied, said means including a wire guiding member mounted for movement in a path around the object, and means for securing the strip in a substantially flat joint.”

As we have heretofore stated, the ’259 machine utilized a revolving arm for the purpose of carrying the wire around the bundle to be tied. The Eby machine utilizes a ring gear for this purpose.

Appellants cite cases decided by the United States Supreme Court (Union Paper Bag Machinery Company v. Merrick Murphy, 97 U.S. 120, 125 , 24 L.Ed. 935, 936 , and Ross Winans v. Denmead, 56 U. S. 330, 341 , 15 How. 330, 341 , 14 L.Ed. 717, 722 ) laying down the test of what constitutes a mechanical equivalent. In the Union Paper Bag Machinery Co. case supra, the Supreme Court said, “Authorities concur that the substantial equivalent of a thing, in the sense of the patent law, is the same as the thing itself; so that if two devices do the same work in substantially the same way, and accomplish substantially the same result, they are the same, even though they differ in name, form, or shape. Curtis, Patents (4th Ed.) sect. 310.”

Without intimating what our decision might be were it not for the fact that the claims of ’259 must be given a narrow range of equivalents because of the nonuser of the patent, we feel that the ring gear of the Eby machine cannot be said to be the mechanical equivalent of the revolving arm of Parker ’259. We therefore hold that the Eby machine does not infringe any of the claims in suit of Parker ’259.

Turning now to the claims of the Parker ’260 paten-t, we have a different situation. That machine utilized a ring gear, and, furthermore, we are not here limited to a narrow construction of the claims of the patent. Parker ’260 machine has been manufactured and sold and is unquestionably more than a “paper patent”.

The claims of Parker ’260 alleged to have been infringed are numbers 7, 12, 13, 65, 66, 69, 70, 76, 85, 86 and 87.

The District Court found that none of said claims were infringed, giving a§ its reasons, “Eby does not halt the device as provided therein, nor does he place either wire through the twister during the operation. He does not discharge the twister toward the bundle; rather his twister dis-chárges parallel to the bundle. One portion of the wire does not ‘overlap another in a plane substantially parallel to the adjacent surface of the object.’ ”

As pointed out above, the ’260 patent is not limited to a narrow range of equivalents. We think that the Eby machine does do the same work as Parker ’260, in substantially the same way, and accomplishes substantially the same result, and therefore can be said to be the equivalent of ’260. Union Paper Bag Machinery Co. v. Merrick Murphy, supra.

For instance, claim 7 recites, inter alia, “means on the ring for passing the wire around the bundle and twice in the same direction through the twisting means”. The defendants contend that in the Eby machine the wire does not pass through the twisting means “twice in the same direction”. Examining the " operation of the Parker ’260 and the Eby machine, we find that in the Parker ’260 the ring and wire guide rotate in one direction somewhat in excess of one revolution, passing the wire through the twisting device twice, once at the commencement and once at the end of the operation. The ring and wire guide thereupon reverse themselves and repeat the operation for the next tie. In the Eby machine the shuttle first places an end of the wire in the clamping member and then the ring rotates to carry the wire around the package to be bound and through the slot of the twister pinion. The ultimate result is that the two ends of the wire are placed in the twister pinion preparatory to twisting the flat knot, as in Parker ’260.

With reference to claims 12, 13 and 87, the objection of defendants is that the Parker machine discharges the wire toward the bundle, while the Eby machine discharges the wire at an angle thereto. In ot-her words, in the Parker machine the wire is received into the twister pinion with the slot opening away from the bundle. The twister makes one and one-half revolutions, discharging the wire with the slot opening toward the bundle. The twister of the Eby machine, instead of making one and one-half revolutions, makes one and three-quarter revolutions, discharging the wire with the slot parallel to the bundle. We do not think that this constitutes a change in the mode of operation so as to avoid a charge of infringement of Parker ’260.

We have hereinbefore discussed the finding of the District Court regarding the alleged invalidity of count 69 of the ’260 patent, and found that the language of the claim is reasonably accurate as to the function of the ’260 machine. On the same reasoning we hold that the Eby machine infringes claim 69 of ’260.

Appellants head their next assignment of error “The District Court should not have found the claims in suit from the Parker ’260 patent void by reason of public use and sale of the Eby machine for more than two years before the application for the ’260 patent was filed”.

35 U.S.C.A. § 31 provides, so far as pertinent here: “Any person who has invented or discovered any new and useful art, machine, manufacture, or composition of matter, or any new and useful improvements thereof, * * * not in public use or on sale in this country for more than two years prior to his application, * * * may * * * obtain a patent therefor.”

The application for Parker ’260 patent was filed on March 12, 1925. We have found above that the Eby flat knot machine was not completed prior to August, 1923. The first sale of an Eby flat knot machine took place in February, 1924. The finding of the District Court that the ’260 patent was void by reason of prior public use and sale of the Eby machine is in error.

Appellees insist that even though the Parker patents be valid, the appellants have failed to sustain the burden of proof of infringement. They contend that the appellants failed to present any testimony to describe the defendants’ structure, or to compare the defendants’ structure with the patents in suit. We do not agree with this contention. Twomley, testifying for the plaintiffs, identified graphic charts of the various types of wire tying machines, including the Parker ’259 and ’260 models, and the Eby machine involved in this suit. Furthermore, he testified as to the mechanical operation of the various machines. He also identified photographs of the various machines which had been prepared in his presence. Even if the testimony introduced by the plaintiffs failed to sufficiently describe the defendants’ machine, the defendants as part of their case introduced drawings and specifications of their machine. These drawings, whether introduced by the plaintiffs or the defendants, may be considered by the court in determining the question of infringement.

We next come to the question raised by the appellees of whether or not the appel-lees have acquired “intervening rights” which would prevent a recovery by the appellants in this action. We need consider this defense only as to the ’260 patent, since we hold that the claims in suit of ’259 are not infringed.

The ’260 patent was applied for on. March 12, 1925, and the application was pending over seven .years, the patent having been issued on August 30, 1932. On July 9, 1932,- the application was amended by the addition of claims, including some of the claims in suit. This was after the Eby machine had been on the market for some eight years.

The doctrine that an applicant might be barred from -obtaining a patent by lapse of time was first introduced as to reissues in the case of Miller v. Bridgeport Brass Co., - 104 U.S. 350, 352 , 26 L.Ed. 783 . In that case the Supreme Court held that an omission of claims from a reissue constitutes a dedication to the public of that which is not claimed. The Court then states, “If two years public enjoyment of an invention with the consent and allowance of the inventor, is evidence of abandonment, and a bar to application for a patent, a public disclaimer in the patent itself should be construed equally favorable to the public. Nothing but a clear mistake or inadvertence, and a speedy application for its correction, is admissible when it is sought merely to enlarge the claim.”

The Supreme Court has also held that the rule announced in Miller v. Bridgeport Brass Co., supra, is. not qualified by any question relevant to the presence or absence of intervening rights of other inventors, accruing between the date of the original patent, and the date of the application for the reissue thereof. White v. Dunbar, 119 U.S. 47, 52 , 7 S.Ct. 72 , 30 L.Ed. 303 .

The Supreme Court later extended the rule, with qualifications, to cases of divisional applications. In the case of Chapman v. Wintroath, 252 U.S. 126 , 40 S.Ct. 234 , 64 L.Ed. 491 , the Court held that one whose patent application discloses, but does not claim, an invention which conflicts with that of a later patent, has two years in which to file a divisional application, in order to have the question of priority of invention determined, notwithstanding section 4894 ( 35 U.S.C.A. § 37 ) as to abandonment of the application by failure to prosecute for one year. In the cited case the contention was made that since the divisional application was filed more than one year after the filing of the parent application, the applicant was barred.

In the case of Webster Electric Co. v. Splitdorf Electric Co., 264 U.S. 463 , 44 S. Ct. 342 , 68 L.Ed. 792 , the Supreme Court, construing the Wintroath case, supra, held that the rule that a reissue patent, expanding the patentees original claims, will be invalidated, by a delay of two years in applying for it unless special circumstances be proven justifying a longer delay, is applicable also to patents issued on divisional applications. It held that there was not a hard and fast limit of two years in every case of a divisional application.

In the case of Crown Cork & Seal Co. v. Gutmann & Co., 86 F,2d 698, the Circuit Court of Appeals for the Second Circuit held a divisional patent invalid because of laches in the filing of the application on which it was granted. The Circuit Court of Appeals cited the Splitdorf case, supra, and held that prima facie the two year limit applies to divisional applications, and that an applicant who waits longer before claiming the invention disclosed in his application must justify his delay by proof of some excuse. On appeal to tire United States Supreme Court the Court reversed the Circuit Court of Appeals ( 304 U.S. 159 , 58 S.Ct. 842 , 82 L.Ed. 1265 ) holding that in the absence of intervening rights or abandonment the Splitdorf case does not mean that an excuse must be shown for a lapse of time for more than two years in presenting the divisional application. The Court pointed out that the District Court had found no adverse use of the patent pri- or to the filing date of the divisional application, and that the Circuit Court of Appeals had not disturbed that-finding. ,

The result of the above cited decisions of the Supreme Court is that in the case of reissues, the right to obtain a broadened reissue is lost by a lapse of two years from the date of the original, in the absence of special circumstances justifying a longer delay, whether or not there have been intervening rights. In the case of divisional applications, if there have been intervening rights, the rule relative to reissues will be applied. In the absence of intervening rights or abandonment, the rule does not apply.

The Supreme Court has never passed on the question of whether or not the rule will apply to amendments to the original application.

In the case of Dwight & Lloyd Sinter-ing Co., Inc., v. Greenawalt, 2 Cir., 1928, 27 F.2d 823 , the court in construing 35 U. S.C.A § 38 held that a division is not a new application, but is an amendment of the original. It is argued by the appellees that since a divisional application is in reality an amendment to the parent application, the rules above outlined relative to divisional applications should apply to amendments generally.

Appellants, on the other hand, argue that even assuming the defense of- intervening rights may be raised in connection with amendments, the same is not applicable in the case at bar, since the amendments complained of added narrowed claims and not “new matter”. We agree that running all through the rules above set forth is the first assumption that the claims have been broadened by the division or reissue, as the case may be. This court in the case of Carson v. American Smelting & Refining Co., 9 Cir., 4 F.2d 463, 470 , in refusing to apply the doctrine of the Splitdorf case, supra, said, “From that case [Splitdorf] the case at bar differs materially. Here the applicant asserted in his original application, in an unscientific way it is true, but in language sufficiently intelligible, the invention which is the substance of the divisional patent. * * * The case is far from being that of a patentee, who waits until others have produced new forms of improvement and then, with the light thus acquired, applies for an enlargement of his claims to embrace those forms.”

The Circuit Court of Appeals for the Second Circuit came to the same conclusion in Ball & Roller Bearing Co. v. F. C. Sanford Mfg. Co., 297 F. 163, 165 , wherein it said, “With the court below we are satisfied that this is a narrowed reissue. * * * As for intervening rights, it is difficult to see how that doctrine can find place in litigation over a narrowed reissue.”

We shall therefore examine the proceedings at the Patent office to determine whether the added claims were in fact broadened claims.

At the outset it should be noted that of the claims of the ’260 patent in suit, claims 7, 12 and 13 have been present in the application from the time of its filing on March 7, 1925. It is with reference to the remaining claims, 65, 66, 69, 70, 76, 85, 86 and 87 that the question of intervening rights arises. These claims were added by the amendment of July 9, 1932, after the Eby machine was on the market.

At the time of the filing of the original ’260 application, on March 7, 1925, it contained 55 claims. Subsequently 18 of these claims were cancelled, leaving 37 of the original claims at.the time of the July 9, 1932 amendment. It is to these 37 claims that we must look to determine whether the added claims were broader or narrower.

The added claims read as follows:

“65 — In a tying machine, means for effecting encircling of a binding wire about an object to be tied, said means including a rotatable wire guiding ring movable in a path around the object, and means for securing the wire in a substantially flat joint.

“66 — In a tying machine, a rotatable annular member mounted for movement to describe a path around an object to be tied, means for supplying a binding wire to said member, means for clamping the wire adjacent the object, means for rotating the member to effect encircling of the clamped wire about the object, and means for securing the wire in a substantially flat joint.

“69 — In a wire tying machine, a wire guiding ring rotatable to describe a path encircling an object to be tied, means for supplying wire under tension to said ring, means for rotating the ring to effect encircling of the wire about the object so that one portion of the wire overlaps another portion thereof in a plane substantially parallel to the adjacent surface of the object, and means located within the periphery of said ring for intertwisting the overlapped portions to provide a substantially flat knot.

“70 — In a wire tying machine, a rotatable ring, means for supplying wire to said ring, said ring having a wire lead-off member movable orbitally about an object to be tied upon rotation of the ring, means for rotating the ring to effect encircling of the wire about the object, and means for in-tertwisting adjacent portions of the wire into a substantially flat knot.

“76 — In a wire tying machine, means for forming a knot in wire wound around an object, a pair of wire clamping members between which said means is located, a rotatable ring for effecting winding of wire about an object, means for rotating the ring in opposite directions, and means for correlating the operation of the clamping members in accordance with the direction of the ring.

“85 — In a tying machine, means for effecting encircling of a tensioned binding wire about an object to be bound and tied, and a member having means to receive por tions of said wire in one position, said member being movable to intertwist said portions and place them for discharge in another position.

“86 — In a tying machine, means for supporting an object to be bound and tied, means for effecting encircling of a binding wire about the object with portions of the wire overlapping; and a slotted twister for receiving the overlapped portions in one position, said twister being movable to in-tertwist said portions and place them for discharge in another position. .

“87 — In a tying machine, means for supporting an object to be bound and tied, means for effecting encircling of a binding wire about the object with portions of the wire overlapping, a rotatable twister having a slot opening away from the object to receive the overlapped portions, and power operated means for rotating the twister to intertwist said portions and to position said slot toward the object for discharge of the twisted portions.”

Defendants contend that it was not until the amendment of July 9, 1932, that the flat knot claims got into the case, and that the flat knot claims constitute new matter. We cannot agree with that contention. For instance, claim 1 is one of the original claims filed -in 1925, and reads, “1 — In a bundle tying machine, means for supporting a bundle in a predetermined location, means for passing a tensioned wire around the located bundle so that the wire tightly encircles the bundle with one portion of the wire overlapping another portion of the wire in a plane substantially parallel to the plane of a side of the bundle, and meatis for twisting together the overlapping portions of the wire " (Italics supplied.)

In the first specifications, filed in 1925, we find the statement, which has remained unchanged, “Another object of the invention is to provide, in a wire tying machine, means for forming a flat knot of the wire, which will lie flat against one of the sides of the bundle.”

The specifications have also contained from the outset a detailed description of the operation of the twisting device.

Another illustration of the broad claims which have been in the application during the entire pendency thereof, is claim 7, which is one of the claims involved in this suit, and which we-. find infringed by defendants: “7 — In a bundle tying machine, a frame .adapted .to support the .bundle to be tied, a ring rotatably mounted in said frame and adapted to surround the bundle, a wire twisting means arranged on said frame contiguous to one side of the bundle, means on the ring for passing a wire around the bundle and twice in the same direction through the twisting means and means for rotating the twisting means to twist the two portions of the wire therein together.”

The' defense of intervening rights raised by the defendants, is denied.

The above necessarily disposes of the contention made by the appellees that the failure of Parker to verify his applications by supplemental oath invalidated the patent issued thereon. The requirements as to supplemental oath have only to do with “new matter” inserted in the application. We quote from Westinghouse Electric & Mfg. Co. v. Metropolitan Electric Mfg. Co., 2 Cir. 1923, 290 F. 661, 664 ,

■ “That statute [R.S. 4892 ; 35 U.S.C.A. § 35 ] requires that one shall swear to his-invention, and all of it; if he only swears to a part, and his attorney puts in the rest, it is exactly like any other yielding to the temptation of improving an affidavit after execution, by inserting additional allegations of fact. A claim is not a disclosure; but every claim must be justified by the disclosure. If, therefore, a new claim needs for its justification or validity an additional or new disclosure, such additional allegations of fact assuredly need a supplemental oath, not because they are to be used for an amended or new claim, but because they tell some fact of invention not told, before.

“Changes of language, not changing the substantial meaning as it stood before amendment, and even changes of meaning, narrowing the scope of the invention described, do not infringe the statute. * * *

The decree of the District Court is reversed as to the claims of the ’259 patent, insofar as it went beyond the issues, namely, the claims in suit. Sustained as to the-claims in suit, on the ground that said claims are not infringed by defendant.

As to the ’260 patent, the decree of the' District Court is reversed. We find that the claims in suit of said ’260 patent are valid and infringed by defendants.

The case is remanded to the District Court for further proceedings in accordance herewith. Costs ,are awarded to appellants.

On November 20, 1936, the Examiner of Interferences awarded priority to Barker of all claims limited to a machine producing flat knots. Dunn and Eldridge appealed from that decision to the Patent Office Board of Appeals, and on January 28, 1938 the Board of Appeals affirmed the decision of the Examiner of Interferences.