MAJOR, Circuit Judge.
This is an appeal from a judgment entered January 7, 1947, dismissing plaintiffs’ complaint after a trial on the merits. The complaint purports to state two causes of action, (1) for violation of an agreement entered into between the plaintiffs and the defendant, dated March 1, 1935, and (2) for the infringement of certain Letters Patent. Damages for violation of the agreement and an injunction to prevent [616]*616further violation thereof were sought. An injunction was also sought to prevent infringement of the Letters Patent and an accounting .therefor. The patents in suit are Stirling Patent No. 1,773,887, issued August 26, 1930, and Wilkinson Patent No. 2,155,458, issued April 25, 1939. Claims 7 and 8 of Stirling and claims 1, 3 to 8, inclusive, and 12 of Wilkinson are relied upon. Both patents relate to the art of printing, and specifically to what is known as the reverse half-tone intaglio process.
The plaintiff Oxford Corporation is a corporation of Michigan and owner of the legal title to the Stirling patent. The plaintiff Miehle Printing Press and Manufacturing Company is a corporation of Illinois and holds an exclusive license, with the right to grant sub-licenses, under the Stirling patent. The legal title to the Wilkinson patent is one-half in the plaintiff and paten-tee William J. Wilkinson, a citizen of California, and the other one-half in the plaintiff Miehle Company. The latter company is also the exclusive licensee, with the right to grant sub-licenses, under the one-half interest owned by Wilkinson. The defendant Publication Corporation is a corporation of New York, with a place of business in Chicago, Illinois, where it is engaged in rotary web intaglio printing. Speedry Gravure Corporation, the licensee and a party -to the agreement of March 1, 1935, was at that time a subsidiary of Alco-Gra-vure, Inc., and the latter was a subsidiary of the defendant. The agreement contained guarantees both by Alco-Gravure, Inc. and the defendant Publication Corporation for the performance of the obligations imposed upon Speedry Corporation. The defendant concedes that it has assumed and is responsible for the covenants and liabilities of Speedry.
The license agreement of March 1, 1935 was cancelled by the defendant, effective October 31, 1935. No question is raised as to the right of the defendant under the agreement to effect such cancellation. The •cause of action predicated upon the agreement arises from plaintiffs’ contention that certain liabilities imposed upon the defendant survived cancellation. More specifically, the contention is that the defendant became obligated, notwithstanding cancellation, to refrain from using the process and teachings of the claims in suit, and that it employs such process and teachings in violation of the agreement. Plaintiffs in their cause predicated upon the infringement of the claims in suit contend that the defendant is estopped by reason of the agreement of March 1, 1935, to contest the validity of such claims. Plaintiffs further contend that irrespective of this issue of estoppel, the claims in suit are valid and infringed by the defendant. The District Court, by its findings of fact, its conclusions of law and its judgment of dismissal determined all issues adversely to the plaintiffs.
Plaintiffs attach great importance to the effect to be ascribed to the agreement of March 1, 1935. Admittedly, it has its place in the lawsuit, but we are of the view that its importance has been overemphasized. The agreement specifically relates, to certain methods of printing and making printing plates, as described in Stirling Patent No. 1,773,887, and to certain methods of printing and making printing plates for which applications for Letters Patent were then pending, referred to as Serial No. 576.014, Serial No. 755,649, and Serial No. 1,487, and provides: “Miehle hereby grants unto Speedry an exclusive license to practice the method and to use the inventions disclosed and claimed in applications for United States Letters Patent Serial Nos. 576.014, 755,649, and 1487, and United States Letters Patent No. 1,773,887, for manufacturing plates and/or cylinders to be used on web intaglio printing machines for printing continuous webs of the following material [describing materials].”
The applications referred to in the agreement subsequently culminated in the Wilkinson Patent in suit, No. 2,155,458, which issued April 25, 1939. The process adopted by the defendant in-1939, which it is charged violates the agreement and infringes the claims in suit, is known as the Dultgen process. (It will be noted that the Dultgen process was adopted by the defendant some four years subsequent to the agreement of March 1, 1935, and also subsequent to the Wilkinson patent which issued on the applications referred to in the agreement.) Plaintiffs’ claim of a two-fold [617]*617cause of action arising from the use by the defendant of the Dultgen process — that is, that such use (1) violates the agreement and (2) is an infringement of the claims in suit —appears to be without substance. More specifically, we are unable to discern how the use of defendant’s process could violate the agreement, short of a showing that it was an infringement of the patent claims. In other words, it seems to us that a violation of the agreement and an infringement of the claims, in substance at least, amounts to one and the same thing. Plaintiffs’ contention in this respect, as we understand it, is that the practices and methods referred to in the agreement are broader in scope than the practices and methods described in the patents, and hence, plaintiffs are entitled to recover for a violation of the agreement, regardless of whether there has been an infringement of the claims. As a basis for this contention, it is pointed out that all inventions are not patentable and that the defendant was obligated to respect the inventions referred to in the agreement, despite their patentability. In our view, the inventions referred to in the agreement are those having a patentable status. Certainly that is true as to the disclosures contained in the Stirling patent then in existence, and we think it is also true as to the disclosures embodied in the applications for patents then pending, which were subsequently allowed and embodied in the Wilkinson patent. We think there was at least an implied representation on the part of the plaintiffs that the inventions referred to in the agreement were actual and patentable and that this is so irrespective of whether the Stirling patent referred to in the agreement and in the Wilkinson patent subsequently issued upon the disclosures contained in the applications also referred to in the agreement, were valid or invalid. The point is that at the time the agreement was entered into, the inventions and disclosures therein referred to were either actually embodied in the Stirling patent or in pending applications which subsequently culminated in the Wilkinson patent. Under these circumstances, we think the contention is without merit that the inventions and disclosures referred to in the agreement were different than the disclosures of the issued patent or the disclosures in the applications for patents then pending.
Consistent with the view thus expressed, it follows that the cause of action for violation of the agreement and that for infringement must rest upon the Same premise, that is, that the process employed by the defendant infringes the claims in suit. If the question of infringement be decided adversely to the plaintiffs, we are in doubt as to the necessity of considering plaintiffs’ contention that the defendant is estopped, by reason of the agreement, to deny the validity of the patents, or if the issue of estoppel be decided adversely to the plaintiffs, the necessity of deciding the validity of the patents over the prior art. On the other hand, if infringement be decided adversely to the defendant, the issue of estoppel and the validity of the patents must be decided, for obviously, there can be no infringement of an invalid patent. In the court below, both the plaintiffs and the defendant requested the court to pass upon the issue of validity, irrespective of its decision on other issues. In compliance with such request, the court, although deciding other issues adversely to the plaintiffs, also passed along the validity of the claims in suit and held that they were invalid. Inasmuch as that court has seen fit to dispose of all the issues presented, including that of the validity of the claims in suit, we have concluded to do likewise.
In view of what we have said, the suit is reduced to one for ordinary patent infringement, subject to the contention that the defendant is estopped to deny the validity of the claims by reason of the agreement of March 1, 1935. Therefore, we return to a consideration of the agreement as it pertains to this issue. The agreement contains twenty-one paragraphs. Under the terms of Par. 1, defendant was given the right of cancellation at the end of the ninety days (this time was subsequently extended by agreement of the parties), and this right was exercised by the defendant and the cancellation became effective October 31, 1935. The pertinent portion of this paragraph reads: “Speedry shall have Ninety (90) days from the execution of this agreement [618]*618in which to satisfy itself as to the quality of work produced, the time saved and other economies of the processes covered by the inventions of said patent and said applications, and as to the relation of the said inventions and their use to prior patents. In the event Speedry at the end of Ninety (90) days is not satisfied that the inventions and processes will accomplish the results desired, or is not satisfied as to the patent situation relative thereto, it shall have the right to cancel this agreement and be relieved of all liability thereunder by notifying Miehle in writing of its desire so to do, Subject, however, to the express agreement by Speedry that neither it nor any of its licensees, or Alco-Gravure, Inc., or Crowell Publishing Company, or Publication Corporation, will use any of the inventions disclosed in the patent or such patent applications as are finally granted, either singly or In conjunction with other inventions. In the event Speedry has not notified Miehle within Ninety (90) days from the execution of this agreement of its desire to cancel this agreement, Speedry shall be deemed to be satisfied that the inventions and processes will accomplish the results desired and shall be thereafter bound by all the terms of this agreement.”
An analysis of this section shows that it deals with four subjects, (1) the purpose of the trial period, (2) defendant’s right of cancellation at the end of the trial period, (3) the surviving covenant .in the event of cancellation and (4) the effect of defendant’s failure to cancel. The important language of this paragraph is that the defendant upon cancellation of the agreement shall be “relieved of all liability thereunder,” subject to the agreement that it will not “use any of the inventions disclosed in the patent or such patent applications as are finally granted,” and that in the event that the agreement is not can-celled as provided for, the defendant shall be “thereafter” bound by all its terms. We have heretofore expressed the view that the inventions which the parties had in mind and which were referred to in the agreement were those either upon which the plaintiffs had or might obtain a valid patent. Certainly this paragraph contains no express language which would estop the defendant from contesting the validity of patents embodying such inventions, and neither do we think it can reasonably be implied. Moreover, it has been held that an estoppel will not be implied or inferred but that it must clearly appear from the language which the parties have employed. City of Chicago v. Joseph, 7 Cir., 95 F.2d 444, 446, certiorari denied 304 U.S. 578, 58 S.Ct. 1049, 82 L.Ed. 1542; Picard v. United Aircraft Corp., 2 Cir., 128 F.2d 632, 637; Measurements Corp. v. Ferris Instrument Corp., 3 Cir., 159 F.2d 590. We recognize, of course, the rule by which a patent licensee may be estopped during the term of the license to dispute the validity of the patent, but we are of the view that it has no application where the licensee effects a valid cancellation of the license agreement, as was done here.
Plaintiffs’ estoppel argument, however, rests largely upon language contained in Pars. 6 and 17 of the agreement. Par. 6 states: “Speedry admits the validity of United States Letters Patent [naming the patent and applications for patent referred to in the agreement] and agrees that at no time will it infringe or assist others to infringe or contest the validity of said Letters Patent either directly or indirectly.”
Par. 17 provides: “The cancellation of the license granted herein for any reason whatsoever * * * shall not relieve Speed-ry * * * from any of the covenants and conditions of this agreement excepting the agreement to pay royalty and the covenant to sue infringers.”
If these paragraphs survived cancellation, it appears that defendant by their express language is precluded from contesting validity, for it is obvious that they are in irreconcilable conflict with Par. 1, when applied to the same situation. At any rate, we are unable to discern how the defendant upon cancellation could be “relieved of all liability,” as provided in Par. 1 (subject to the condition therein stated), and at the same time be held to admit the validity of the patents as provided in Par. 6, or by continuing to be bound by the covenants and conditions of the agreement as provided in Par. 17. It [619]*619is unreasonable to think that the parties intended any such incongruous result.
The court below reconciled these conflicting provisions by holding, in substance, that Pars. 6 and 17 were of no binding effect after the cancellation of the agreement as provided in Par. 1. We agree with this construction of the agreement. It is rather plain, so we think, that the agreement was intended to cover two entirely different situations. Par. 1 provided for the right of cancellation, and upon its exercise the rights and liabilities of the parties were therein fixed. Under that situation, the remaining paragraphs, that is, Pars. 2 to 21, including 6 and 17, were no longer of any binding effect. By the act of cancellation their effectiveness was destroyed. On the other hand, if the defendant had failed to exercise its right of cancellation, it, and the plaintiffs as well, would have been bound by all of the provisions of the agreement. That the parties so intended is strongly indicated by the concluding sentence of Par. 1, which on failure of the defendant to cancel states: “* * * and shall be thereafter bound by all the terms of this agreement.” By this construction, all of the conflicting provisions of the agreement are reconcilable and when a situation arises to which they are applicable can all be given effect. See Cities Service Gas Co. v. Kelly-Dempsey & Co., Inc., 10 Cir., Ill F.2d 247, 249.
This brings us to a consideration of what are perhaps the more important issues of the case — the asserted infringement of the claims in suit, as well as their validity. Prior thereto, however, we think it is pertinent as bearing upon both issues to further note the circumstances which led to the cancellation of the agreement of March 1, 1935, and the employment by the defendant of the Dultgen process. At the time of the execution of the agreement and prior thereto, defendant’s presses were all high speed, web fed, rotary intaglio presses, and printed from cylinders prepared by the “conventional gravure method.”1 The plaintiff Wilkinson interested the defendant’s president in using the Stirling-Wilkinson methods of preparing intaglio printing surfaces. Wilkinson disclosed certain prints made on a sheet fed press from plates made by his process. It is apparent that neither of the parties had too much confidence in the adaptability of plaintiffs’ process to the presses used by the defendant. Hence, the provision in the agreement for a trial period and the right by the defendant to cancel if the results were not satisfactory.
We think it is a fair inference from the record that both parties entered into this trial period test in the utmost good faith, each hoping that plaintiffs’ process could be made operable on the defendant’s presses. The test was carried on from March 1, 1935 (date of the agreement) until about October 31, 1935, when the agreement was cancelled. The tests were employed under Wilkinson’s direct supervision, aided by his own experts as well as those of the defendant. There is nothing in the record which is disclosed so convincingly as the fact that plaintiffs’ process, described in its then existing patent and its pending applications, was not operable on the character of presses used by the defendant. The utilization of plaintiffs’ process by the defendant was recognized by both parties as hopeless. We find nothing in the record to indicate that defendant cancelled the agreement for any other reason, and apparently the cancellation took place without a word of complaint from plaintiffs.
Much is said by the plaintiffs concerning the valuable secret information which the defendant acquired during the trial period of the agreement. The value of such information appears to have been more fanciful than real in view of the fact that with all the assistance which plaintiffs and their experts were capable of giving, the process was demonstrated as unsuitable for 1 defendant’s purpose. Furthermore, it would seem that the information necessary to employ the process of the Stirling patent would have been found in the patent. It is true, no doubt, that information relative to the application which resulted in the Wil[620]*620kinson patent was not then in the public domain, but, as heretofore noted, this patent issued prior to the employment by the defendant of the alleged infringing process, and its teachings were no longer secret.
In connection with the failure of the plaintiff Wilkinson (the patentee) and the defendant, together with their experts, to demonstrate the operability of plaintiffs’ process on defendant’s presses, it is also significant that for some four years after the cancellation of the agreement the defendant failed to utilize in its business any information which it had obtained from plaintiffs or any of plaintiffs’ processes. It was not until 1939 that it employed the alleged infringing process, and the record clearly discloses that this process was acquired by the defendant from the New York Daily News, where it had been independently developed.
The court below held all of the claims of both patents in suit to be invalid, and specifically found that such claims were anticipated by the prior art patents to Ives No. 495,341, Reckard No. 962,098 and Ballard No. 2,024,086. The court further found that Stirling’s claim 7 was anticipated by the prior art patent to Deeks No. 1,477,886, and that all the claims of the Wilkinson patent in suit were anticipated by the prior art patents to Stirling No. 1,773,887 (the Stirling patent in suit) and Ballard No. 2,024,087.
Much of the language employed in the long and verbose specifications of the patents in suit, as well as the language found in the claims, is of a highly technical nature. We have our doubts as to whether it could be understood by persons trained in the printing and photographic fields and wonder if the purpose of its use was for enlightenment or confusion. Most of the oral testimony was given by two expert and highly qualified witnesses, one for the plaintiffs and the other for the defendant, and their testimony is sharply contradictory on many of the vital issues involved. One or the other of these experts, and perhaps both, apparently did not understand either the teachings of the patents in suit or that of the prior art, and one or the other did not understand defendant’s process alleged to infringe. We have read, studied and attempted to comprehend the testimony of these expert witnesses, and we are inclined to the view that the claims in suit could be found invalid solely on the testimony of plaintiffs’ expert. On the other hand, we are convinced that the testimony of the defendant’s expert demonstrates that the claims are anticipated by the prior art and therefore invalid, but, if valid, his testimony clearly demonstrates that defendant’s process does not infringe. The court below in making its findings evidently, as it had a right to do, relied strongly upon the testimony of defendant’s expert. , Contrary to plaintiffs’ insistence that such findings are of little consequence, a study of the record, and especially the testimony of the expert witnesses, is convincing that Rule 52(a) of the Federal Rules of Civil Procedure, 28 U.S.C.A. following section 723c, is particularly applicable. Certainly we find no basis for thinking that the findings are “clearly erroneous.” We think they must be accepted.
Thus, no useful purpose could be served in any detailed discussion of either the conflicting testimony or in comparing the claims in suit with the disclosures of the prior art. Both patents in suit relate to half-range reverse half-tone intaglio printing plates. Stirling describes and claims a method of making a half-range plate in which “only substantially onp-half of the normal full tone scale dot (pocket) formation” represents the subject and in which all pockets are disconnected. According to plaintiffs, Wilkinson merely improved on Stirling by making a screen positive which retains certain tones next to white which plaintiffs say (denied by .defendant) are lost by Stirling. Thus, Wilkinson describes and claims a half-range positive and a process for making it in which the shadow areas of the subject are represented by half value checkerboard dots and all lesser tones at half value by dots of correspondingly lesser area.
Both Stirling and Wilkinson use a halftone screen (Stirling in making his negative and Wilkinson his positive) to make half-range screen positives. Both admittedly used knowledge then common to the art to control exposure and screen distance to re[621]*621produce the subject in the screen image by only half the normal full tone scale of dot formation. Each positive represents black at one-half value by checkerboard dots and all lesser tones at one-half value by dots of corresponding lesser areas.
Both patents use these half-range positives to control the etching of pockets into the plate which correspond in area with the dots on the positive. In both patents the pockets are etched to a uniform depth which is twice that normal for full range pockets. The added depth is relied upon to restore full tone value to the print. The requisite volume of ink applied to the paper for proper tone reproduction in every area of the plate is controlled solely by maintaining the area of the dots on the positive (and, hence, the area of the pockets which are all of the same depth) at one-half the area of a normal full scale dot (pocket) for each such tone.
The Ives prior art patent No. 495,341 discloses a screen positive, printing plate and a method quite similar to that now claimed by Stirling and Wilkinson. Ives discloses reverse half-tone intaglio printing plates made photographically and by etching; he uses a half-tone screen in making a screen positive photographically from a continuous tone negative. His disclosure includes the making of a screen negative directly from the subject, as Stirling prescribes, and it appears that the essential purpose of the Ives disclosure was to teach the art how to produce a screen positive (and an intaglio plate) of less than full range. While Ives does not specify the exact exposure or screen distance to be used, as do some of the claims in suit, it would seem that these essentials could readily be determined by a person skilled in the work. Both Wilkinson and Ives say that the screen distance and exposure shall be sufficiently less (than normal for a full-range half-tone), that the positive dots in the shadow areas will be unconnected. When an Ives positive, made by controlling exposure so as to leave the dots slightly open in the deepest shades of the picture, is used in the usual way to control the etching of the plate to form pockets therein which correspond to the dots of the positive, the resulting plate is the half-range plate described by Stirling in claim 7 and made by the process described by Stirling in claim 8. We see no difference in substance between the disclosure made by Ives and that contained in the claims of the patent in suit.
Another prior art patent which plaintiffs are hard-put to escape is that of Reckard No. 962,098, which discloses an intaglio plate made photographically and by etching, having a non-printing surface of even height in which the subject is carried by a large number of depressed, entirely disconnected pockets formed therein of varying size, the largest pockets corresponding in area to the darkest tones of the subject. He makes a continuous tone negative of the subject, and from this a screen positive which carries the subject by disconnected dots which vary in size with the tones of the subject. He uses either a dotted or a half-tone screen. In either case, he obtains disconnected dots on his positive which will produce disconnected pockets in the etched plates. According to defendant’s expert, Reckard’s normal use of a half-tone screen to obtain disconnected dots will form Wilkinson’s half-range positive, with checkerboard dots in the shadows, with other tones represented by dots of corresponding lesser area, as shown by Wilkinson’s process. From such a positive, Reckard uses conventional steps to produce Stirling’s plate, with disconnected pockets corresponding to the dots of the positive by Stirling’s process.
It is unnecessary to make further reference to the prior art. We think there is substantial support for the findings of the lower court th_t the patents in suit are anticipated thereby and, in any event, that they do not constitute invention over such art.
The lower court also found that even though the claims in suit be held valid, they were not infringed by the defendant’s use of the Dultgen process. Again we think that the findings of the District Court are substantially supported and must be accepted. As to the Dultgen process, the court found:
“The defendant’s process was independently developed in the plant of the New [622]*622York Daily News, and its originator, Dult-gen, derived no knowledge from the test work of March 1, 1935 to October 31, 1935. The process now used by defendant was in commercial operation at the plant of the New York Daily News before defendant or anyone connected with the defendant had any knowledge of it. Defendant first learned of such process when it was called to the attention of the defendant by the General Manager of the News in 1938. Such process was thereafter installed in defendant’s plant under Dultgen’s supervision and has since been used in defendant’s plant in accordance with instructions then given by Dultgen and as previously practiced in the plant of the New York Daily News.
“The defendant’s process is widely used in the gravure printing industry by twelve licensees of defendant’s licensor, the New York Daily News, and such process did not result from, and is entirely independent of, the test work conducted in 1935 under the contract in suit.
“The accused process used by defendant is based upon principles other than those underlying the processes disclosed to Aleo Gravure, Inc., during the 1935 test period.”
Plaintiffs concede that the Dultgen process is an improvement over that disclosed by the patents. This appears evident from the fact, as heretofore noted, that the plaintiffs’ process, after much effort, could not be made operable on defendant’s high speed web fed presses, while the Dultgen process produced a satisfactory result. One important feature of the Dultgen process which distinguishes it from that of the patents is that Dultgen calls for two positives to form the pockets in the printing plates. In practicing this process, defendant uses these two positives to form in its printing plates pockets of varying depths, having areas in the lighter tones greatly exceeding areas of full tone dots. This plate is, therefore, of much less than half-range, in contrast to the requirement of Wilkinson and Stirling that the pockets (dots) shall all be of equal depth and of areas exactly one-half those of full tone dots, i.e., half-range. While all three plates are full range plates as to the volumes of the pockets, for each applies enough ink to regain full range in the print, Stirling and Wilkinson do so by etching the half-size pockets twice as deep while defendant secures the required volume by increasing the areas while decreasing the depth of the pockets in the lighter tones.
Plaintiffs in their brief assert that Stirling in his printing plate used “only substantially one-half of the normal full tone scale of dot formation to reproduce the full tone range of the original subject.” The Dultgen process, however, does not have pockets in any tone except black, which are “substantially one-half the normal full tone scale,” in area, and its plate is much less than “one-half the normal full tone scale,” as is required by Stirling.
The Wilkinson patent requires that the screen positive must represent all tones of the subject by dots which are one-half the areas of normal full range half-tone dots for such tones. There is testimony that departure from this requirement will sacrifice true tone fidelity and deviate from the explicit directions of his patent specification and claim requirements. On the other hand, there is testimony that tone reproduction as taught by the Dultgen process excludes the possibility of using Wilkinson’s positive or process. Numerous other distinctions between the Dultgen process and that of the patents are relied upon by the defendant, which we think unnecessary to relate or discuss.
The judgment is affirmed.