delivered the opinion of the court:
The plaintiff seeks in these two suits to recover (1) royalties alleged to be due him under a contract and (2) compensation for the alleged unauthorized manufacture and use by the Government of inventions made and patented by the plaintiff. Both suits relate to the M-l or Garand semiautomatic rifle used by the United States Army in World War II.
*301The plaintiff has, since 1901, been a designer of guns. From 1903 to 1923 he designed guns for the Bemington Arms Co. and received a royalty on their manufacture. The United States Army Ordnance Department, being anxious to obtain for army use a satisfactory semiautomatic shoulder rifle, in 1923 made a contract with the plaintiff, who was then a well-known and highly competent designer of guns, and who had for some time been working on the design of such a rifle. In the contract the plaintiff agreed to disclose his prior studies and to work for the Department for one year beginning on June 1, 1923, in the completion of the weapon and the supervision of its construction. He was to receive $10,000, payable in monthly installments during the year. The Government was given the option to extend the contract for a period not to exceed a year, at the same monthly rate of pay. The contract provided that the plaintiff was to retain his patents, and any new ones taken out by him related to his work on the rifle, but that the Government should have the first option upon an exclusive license to manufacture, use, and sell the rifle produced, upon the payment to the plaintiff of a royalty of $1.25 per rifle up to the first 400,000 rifles, after which the Government should have a free, but nonexclusive license on the rifle. The option given the Government was to be exercised within six months after a complete and satisfactory test of the gun by the Ordnance Department, and not later than 12 months after the plaintiff’s services under the contract should terminate. The contract further provided that if the Government should exercise its option, but should not within 4 years enter upon the routine production of the gun, then it should have only a nonexclusive license, but should pay the same royalties as were provided for the exclusive license. The contract provided that the name Pedersen should appear on each rifle manufactured under the agreement.
The plaintiff went to work at the Springfield Armory, Springfield, Massachusetts, the Government providing all the necessary facilities. At the end of a year, the contract was extended for another year. By new contracts similar to the original one, the plaintiff’s services were continued down to January 3, 1930, on which date a final contract was *302made. This contract recited that, under the preceding contracts, a semiautomatic shoulder rifle had been designed and a number of such rifles had been made under the plaintiff’s supervision; that the 20 examples of that rifle delivered under the 1927 contract had successfully passed certain army tests; that certain changes in that gun were necessary or desirable, and were being worked on but had not been completed; that the plaintiff was willing, if the rifle already delivered or the improved„type still being worked on should be approved for adoption by the United States, to grant the United States a license to manufacture, use, and sell the rifle. Then followed the several articles of the contract, similar to those of the original contract of 1923, providing that the plaintiff was to retain his patents, the United States was to have an exclusive license to manufacture 400,000 rifles and pay a royalty of $1.25 each, and was thereafter to have a free nonexclusive license. The plaintiff was to be paid $1,000 upon his completion of the work begun “to meet such specifications as may arise or may have arisen as a result of the tests of the arm already delivered,” and upon his completion of the design and building of the improved type of rifle without further expense to the United States, and the delivery of one complete form of said rifle not later than December 31, 1930.
The Pedersen rifle, as such, was not adopted, or approved for adoption by the United States Army. Instead, some time prior to 1938 the United States began the manufacture of the M-l or Garand semiautomatic rifle, and it became standard equipment for the Army.
The contract of January 3, 1930, had in its Article VII provided that the license granted it by the contract should be automatically renewed from year to year unless the Government canceled it by written notice. By letters of October 20, 1938, and February 21, 1939, to the Chief of Ordnance, the plaintiff raised the question of his rights under his patents, and formally requested that he be paid royalties on all M-l (Garand) rifles which the Ordnance Department had made and that such payments be continued in the future. The Chief of Ordnance, apparently being reminded by the plaintiff’s letters of the contract of January *3033, 1930, .wrote a letter to the plaintiff on April 20, 1939, canceling the contract as of June 30, 1939. On April 26, 1939, the plaintiff wrote repeating his demand for royalties and the Chief of Ordnance responded that he had directed the Commanding Officer, Springfield Armory, to prepare Touchers for the plaintiff’s signature, which, when so signed, he would send to the General Accounting Office for settlement. He said that it was necessary to send the vouchers to the General Accounting Office for settlement because the funds had reverted to the Treasury. He further said that his action in submitting these vouchers to the General Accounting Office was not to be construed as an acknowledgment of the validity of the plaintiff’s patents. The vouchers were prepared, signed by the plaintiff, and sent to the General Accounting Office without any approval, on their face, by the Ordnance Department. They showed the manufacture, up to June 30,1939, of some 13,000 M-l rifles and called for a payment of $16,280, which was $1.25 per rifle. On January 10,1940, the General Accounting Office refused payment of the vouchers saying that the Army had not adopted the Pedersen Bifle but had adopted and was manufacturing the M-l (Garand) rifle, and that therefore no royalties were due under the contract.
The plaintiff’s two suits are (1) for royalties for M-l rifles manufactured down to June 30,1939, when the royalty contract was canceled, and (2) for compensation for the use of his patented inventions after June 30, 1939, down to the date of the filing of the petition in the patent case, which date was May 6,1940. His theory in the royalty contract case is that the M-l or Garand rifle, is, in fact, the Pedersen rifle, apparently because, he claims, it includes certain of his patented inventions. As a consequence both the evidence and applicable legal doctrines in the two cases are substantially the same, hence we write only one set of findings of fact and one opinion for the two cases.
We have said that the plaintiff’s claim in the royalty contract case that the Garand rifle is in fact the Pedersen rifle is apparently based upon the incorporation in the Garand rifle of devices upon which the plaintiff claims he holds patents. It is probably true that the plaintiff would have *304been entitled to royalties under his contract if the Government had copied his rifle, even though his rifle had involved no inventions, or, if it did, he had taken out no patents upon them. But in appearance at least, and in the large working parts of the two guns, there is little resemblance, almost as little as there could be between two semiautomatic shoulder rifles. Perhaps the first thing to be decided in the design of a semiautomatic weapon is how to obtain the power to operate the mechanism. Pedersen obtained his power from the rearward push of the cartridge case when the powder exploded. In his gun patent he said that one advantage of this was that it saved the weight and bulk of a tube slung under the barrel to carry force from the compression in the barrel between the cartridge case and the bullet. Yet Garand used the tube device, with its added weight and bulk, after having worked, in earlier models, with a third method, that of taking power from the explosion of the primer. Both Pedersen’s method and that of Garand were old in the art, but they were completely different from each other, and each required different designing of other parts of the gun to cooperate with the method used. An apparent instance is the fact that Pedersen’s bolt could not be positively locked in position to hold the cartridge in the chamber, since it had to be free to yield to the rearward motion of the cartridge case as the bullet left it, while Garand’s bolt is positively so locked, and is only set free by a turning motion produced by the gas pressure created by the bullet the instant before it leaves the barrel. The Pedersen bolt works on a toggle which throws it almost vertical when it is in its rearward position. The Garand bolt slides straight back. The Ped-ersen bolt remains locked back when a clip of cartridges is exhausted and until a new clip is completely inserted, and then is released only by a separate manual pull backward. The Garand bolt is released as the inserted clip reaches its final position and is latched in and the operator’s thumb is pushed aside forcefully by the return of the bolt to its closed position. The Pedersen gun has an extensive covering of metal meshing around the barrel to cool it. The Garand gun has no such device.
*305The two guns are generally speaking not alike then, in either appearance, construction or operation, except in those particulars in which any two semiautomatic shoulder rifles would, it would seem, have to be alike. Would the fact that, somewhere inside the Garand rifle, some single device, or even more than one, on which Pedersen held a patent, make the Garand gun the Pedersen gun within the meaning of the royalty contract, and thus entitle the plaintiff to the stipulated royalty ? In view of what we have decided with regard to the plaintiff’s patents, it is not necessary for us to resolve this question.
We recognize that a license contract, under which a licensee agrees to pay a royalty for the privilege of manufacturing a patented article, may estop the licensee from questioning the validity of the patent held by the licensor. But the agreement here involved was an agreement to pay a specified royalty if the Army manufactured a gun developed by Pedersen. Article IV of the contract, as shown in our finding 8, provides for the payment of a royalty upon the manufacture of “this arm” which the context shows to mean the rifle developed by Pedersen, and not a rifle developed by someone else but incorporating certain features upon which Pedersen held a patent, but which in fact, as our findings show, were old in the art and just as available to the developer of the rival weapon as to Pedersen.
The plaintiff relies upon three patents, one entitled “Magazine Rifle,” one entitled “Cartridge Clip,” and the third entitled “Cartridge Clip and Follower.” The first two were issued in 1929, the third in 1931. The rifle patent discloses and claims among other things a bolt lock mechanism functioning to hold a breech bolt in a rearward position after the last cartridge in the clip is fired and the clip ejected, and also discloses and claims mechanism pertaining to the automatic release of the empty clip and its ejection. Our findings 26 through 34 show in their intricate details the mechanisms disclosed and claimed in the Pedersen “Magazine Rifle” patent, and the construction and operation of the devices which perform the same functions in the Garand rifle. We have concluded in finding 33 that the Garand devices are covered by the language of some eight of the claims in the *306Pedersen patent. But in our findings 34 through 44, and 47, we have concluded that the devices covered by the claims in the Pedersen patent have been anticipated by prior patents in this country and in foreign countries.
The plaintiff urges that some of its claims describe an automatic operation in the discharge of the clip, and that that means that the whole operation must take place automatically and hence the bolt must be pushed rearward automatically, hence these claims cover only automatic or semiautomatic rifles. The preferred embodiment stated in the specifications of the patent is a gun of the semiautomatic type. But the claims relied on are in broad terms and refer only to “a gun” and hence the natural meaning of the language relating to the automatic discharge of the clip is that when the cartridges are all discharged from the clip it is automatically released and kicked out. Such a device would be an apparent infringement upon the patent claimed, if the patent were valid, and the plaintiff cannot give his claim an artificially narrow construction in order to escape the prior art. Hence the prior patents referred to in the findings anticipate the plaintiff’s claims, and render them invalid.
The plaintiff’s second patent here sued for is for a “Cartridge Clip.” His clip, as disclosed and claimed, is a U-shaped sheet metal clip constructed to hold a stack of cartridges in a tight, compact package, the clip to be loaded into the gun along with the cartridges, and not to be stripped off them as they go in. The clip is to be cheap enough so that it can be thrown away after use. The cartridges are to be in a double row in the clip and the clip is to hold them firmly in position under all usual conditions of handling. The details with regard to this patent, and with regard to the clip used in the Garand rifle are set out in our findings 48 through 52. In finding 53 we have concluded that the language of several claims in this patent is applicable to the clip used in the Garand gun. But here again we conclude, in our findings 54 through 63 that the Pedersen claims were anticipated by many prior inventions.
The third Pedersen patent here sued for is his “Clip and Follower” patent. Pedersen’s earlier patented clip, which fitted the Pedersen rifle which is in evidence and which was *307tested by the Army, was a nonreversible clip; that is, it had to be inserted into the rifle with a certain end up, and a certain side to the right. Cartridges were in a double but staggered arrangement in the clip, so that the lowest cartridge in one row was lower than the lowest one in the other row. In order to push them up evenly, it was thought necessary to construct the follower, that is, the device which pushed the cartridges up as the cartridges high in the clip were exhausted, so that it had two steps, one of which would push on the bottom cartridge in the low row, the other on the bottom cartridge in the other row. Since these steps on the follower were in a fixed arrangement, the clip had to be so constructed and so inserted in the gun that the two lowest cartridges would rest on these steps. When the Peder-sen gun with this clip and follower were tested at Fort Benning, Georgia, a generally very favorable report was made, but in the final report from the Infantry Board at that fort, dated May 8,1928, there was included a report from the Department of Experiment of the Infantry School there, suggesting certain improvements which would add to the efficiency of the gun. One of these suggestions was
e. The reversible clip. The adoption of such a clip would eliminate the only superiority of the service rifle clip over that of the Pedersen rifle. This fact has been brought to the attention of the inventor, Mr. J. D. Pedersen. Mechanical difficulties involved may outweigh the advantage which would accrue through its adoption.
The principal advantages of the reversible symmetrical clip are that it is easier to load the cartridges into the clip since no care need be taken to place one particular row in the higher position rather than the other row, and that it is easier for the soldier to insert in the gun, since he need take no care as to which end is up or which side is to the right.
We suppose that everybody concerned with the problem was aware that the crux of the problem of the reversible clip was the designing of a follower that would operate with such a clip. The designing of a symmetrical clip did not, in itself, present any difficulty. Mondragon, findings 60 and 75, and Frey, finding 61, had patented symmetrical clips.
*308On February 14, 1929, Pedersen filed an application for a patent on a combination of a reversible clip and a cooperating follower. On May 12,1931, a patent was issued pursuant to that application, and that is the third patent here sued upon. The specification of the patent says “Accordingly, the follower has a step for engaging the bottom of the double row stack, which step may be shifted automatically to a right or left hand position so as to accommodate itself properly to the arrangement of the staggered rows.” The patent drawings are reproduced in our finding 67. Figures 11 and 12 there show that on top of the regular flat follower plate is placed a piece of metal whose cross section is bullet shaped, which piece extends along the top of the follower, not rigidly, but on bearings, so that the top of it can be pushed either to the right or the left. When there are no cartridges pushing on it, it is held upright by small springs at its ends. When a clip of cartridges is pushed down against it, whichever cartridge is low pushes it sidewise to a position where its edge touches, extends along and supports the bottom cartridge of the other row, along a line lying straight below the line of the center of the cartridge. In the invention, in the proportions shown in the drawings, the pictured springs are necessary to make the device work, since if the piece were not restored to its central upright position when the clip was empty, it would, unless the clip happened to be put in the gun with the cartridges high on the side where the piece was lying, prevent the clip from reaching its place in the gun, and the gun from operating.
As we have said, the patent drawings show the springs. The specifications refer to them and say that “the movable step 25 is normally and yieldingly held centrally of the follower 19 as shown in figure 10 by small fiat springs * * And in two of the six claims, those numbered 3 and 4, claim 3 refers to “a movable step normally held centrally thereof” and claim 4 refers to “a step pivotally mounted thereon and yieldingly held in central position thereof.” Neither of these claims is here sued upon, since the alleged infringing device in the Garand rifle is not normally nor yieldingly held in a central position on the follower. We have found, in finding 73, that the claims in suit are anticipated by the *309Lewis patent in connection with the Mondragon patent. As we have said, Pedersen did not invent the symmetrical or reversible clip. Neither did he invent the separate follower, not built into the clip but being a plate or platform pushed up through the clip by a spring device. Hence the reversible clip.with the separate follower pushing on the lower row of cartridges was available to a designer. The Mondragon follower plate, finding 75, had concave surfaces to support firmly the bottom cartridge. Lewis had patented the spring device shown in finding 73 which was shiftable sidewise and thereby adaptable to the bottom cartridge, whichever side of the clip it was in. Surely the adjustment of the Lewis spring device so that it would support the bottom cartridge in the upper row, rather than in the lower row, would not have been an invention. And putting this device on a separate platform follower, which itself supported the bottom cartridge in the lower row, would produce the shifting step follower supporting both rows, which is the basis of the plaintiff’s claims. The idea, therefore, of a movable, sidewise adaptable follower was as old as the Lewis patent. Whether the particular mechanism embodying the idea should be leaf springs, as in Lewis, or a shaped bar turning on pivots, as in Pedersen, or a shaped bar sliding in a slot, as in Garand, the idea of meeting the problem by a shifting element was in all three, and was first in the Lewis patent. Pedersen did not use leaf springs, but pivots. Garand did not use either leaf springs or pivots, but a slot. The shiftable follower was not Pedersen’s invention, and his clip and follower patent is invalid for lack of invention.
As to the moralities of the situation, apart from any question of the validity of patents, we cannot see that those who designed the follower of the Garand gun got any benefit whatever from the disclosures of the Pedersen patent. They designed a follower slide shiftable from side to side by moving laterally in a slot, with no tipping, no pivots, and no springs. We cannot see how any part of the idea of such a device could have been gained from studying the Pedersen patent, when it must already have been well known that some sort of a moving part on the follower would be necessary if *310the upper row of cartridges in a reversible clip was to be supported.
There is a serious question in the case as to whether the Garand follower slide, which admittedly performs several functions not related to the claims of the plaintiff’s patent, if it were valid, performs at all the patented function of supporting and lifting the higher row of cartridges which is the function which the plaintiff’s patent purports to cover. In view of our conclusion that the relevant claims of the patent are invalid, we do not decide this question.
We conclude that the Pedersen rifle was not adopted by the Army; that the M-l or Garand rifle which was adopted was not the Pedersen rifle in appearance, structure, operation, or in any validly patented detail. It follows that the plaintiff is not entitled to recover either royalties under his contract, or compensation for the use of his patented inventions.
We think that the evidence shows that the plaintiff designed and constructed an excellent rifle which probably would have been adopted by the Army if the Garand rifle had not appeared at about the same time and seemed preferable to the Army. But since it did seem to the Army to be a preferable weapon, it was the duty of the Army, and no violation of the plaintiff’s rights, to adopt and use it.
The plaintiff’s petitions will be dismissed.
It is so ordered.
Whitaker, Judge/ LittletoN, Judge/ and Jokes, Chief Justice, concur.