HAMLEY, Circuit Judge.

Seeking to defend its method of applying liquid rubber cement on tires, Elrick Rim Company brought this action for a declaratory judgment. The company sought a judicial declaration that a patent owned by Ralph R. Reading, covering a process for the application of liquid rubber cement, is invalid and not infringed by Elrick Rim Company. It also sought damages for unfair competition.

The defendants are Reading, to whom the patent (No. 2,721,148) was issued on October 18, 1955, and his exclusive licensee, Reading Tire Machinery Co., Inc. They denied the principal allegations of the complaint. They also counterclaimed for a judicial declaration that the patent is valid and infringed, and for treble damages for infringement. In addition defendants sought an award of reasonable attorney's fees.

After a trial, a judgment favorable to defendants was entered. Plaintiff was denied all relief. The patent was declared to be valid and infringed by plaintiff. Defendants were awarded single damages in an amount to be ascertained by a master. They were also awarded attoney's fees in the amount of $7,500.

Plaintiff appeals, contending that numerous findings of fact are clearly erroneous and that basic principles of patent law were disregarded or misapplied.

The retreading of an automobile tire is accomplished by attaching tread rubber to a tire carcass through the application of heat. The first step in doing this is to buff thoroughly the surface of the tire carcass. It is then necessary to apply an adhesive coating to the buffed surface before placing the tread rubber.

At the inception of the tire retreading industry and for many years thereafter this adhesive coating consisted of a thick, wet layer of rubber cement dissolved in an inflammable solvent having a petroleum base. A stippling brush was used in applying this coating. There was considerable difference of opinion in the industry as to the necessity for applying more than one such layer.

The viscosity and wetness as well as the thickness of the coating presented problems. It was necessary to remove the coated tire to a place of storage having adequate ventilation so that it could dry under fireproof conditions. It was also necessary that the drying room be dustproof. If particles of dust were permitted to settle on the surface of the drying cement, defective adhesion of the tread rubber to the tire would later result.

At times of high humidity droplets of water vapor would condense or settle upon the surface of the coating. This sometimes resulted in tire failures caused by the creation of vapor pockets under the tread. It was therefore necessary to suspend the described coating operations during periods of high humidity.

The application of this thick coating by the use of a stippling brush frequently forced the cement into surface irregularities and defects such as nail holes and cuts. Since rubber cement dries from the outside, the result often was that the inner wet cement formed gaseous pockets when heat was later applied during the retreading process. Defects of this kind, known as "tire blow holes," were a common hazard and sometimes resulted in the complete loss of the tread while the tire was in use on the highway.

In this prior practice the evaporation of the solvents in which the rubber cement was dissolved created a serious health hazard to the operator. The prior practice was also time consuming with respect to both the application of the coating and the drying period, which required from several hours to several days.

Use of the stippling-brush method of applying the adhesive coating made it necessary to stir the mixture of cement and solvent continuously. If this was not done, precipitation occurred and there was a consequent sludging and uneven application of the cement.

From about December 7, 1951, to December 1953, Reading worked almost continuously to perfect a method of applying an adequate adhesive coating to a tire carcass in a manner which would overcome the difficulties of the stippling-brush method. The central idea which Reading pursued was that of spraying on the liquid cement with a device similar to a conventional paint sprayer.

During this period Reading experimented with and tested numerous variations of the device, different combinations of air pressure within the container and in the bypass line leading to the nozzle, and diverse ratios of cement to solvent within the container. Most of the tires processed during the course of these experiments and tests proved satisfactory and were sold in the regular course of business. However, the number of retread failures and inability to completely overcome the major difficulties of the former art led Reading to continue his efforts.

Two features dominate the process as ultimately developed by Reading. One is so-called "emulsification" of air bubbles with the rubber cement in solution. The other is the forcing of this emulsion through the nozzle of the spraying device at a high pressure.

Emulsification, as Reading conceived it, is the entrainment of air bubbles in the liquid cement. The result is a stable mixture of cement solids and solvent. The need for constant stirring is avoided, and there is no precipitation or settling of cement solids as long as the container is kept under pressure.

It was discovered that the sprayed emulsion had less tendency to "cobweb" than where no emulsion effect was obtained. The elimination of "cobwebbing" made it possible to secure a more thorough and even coating of the tire. It was also observed that there is a greater degree of tackiness to cement which has been emulsified by means of the Reading process.

This process, as finally developed, teaches that the emulsifying effect may be produced by introducing compressed air into the container of the spray device through pinholes in an air inlet pipe near the bottom of the container. The desired effect can be obtained by an air pressure as low as five pounds per square inch. Reading believed, however, that the best results would be obtained by an application of forty pounds per square inch for several seconds, and then reducing the pressure to about fifteen pounds.

The high pressure in the independent air line, as taught by Reading, reaches a magnitude of from one hundred fifty to two hundred pounds per square inch. This makes possible the application of an exceedingly thin coating of rubber cement which dries quickly, thereby saving time and avoiding the dust, moisture, and health hazards associated with former methods. The utilization of high air pressure in the bypass line also aids in overcoming the "cobwebbing" effect to which reference has been made. In addition the high air pressure utilized in the Reading process causes the mixture to attain and retain an air volume above the flammable limits of the solvent. The need of extreme care to avoid the danger of fire or explosion is thus overcome.

There are at least three paint-spraying devices, patented prior to Reading, which with certain adjustments or minor changes could have been utilized to carry out the Reading process. One of these is the Shelburne apparatus, patent No. 1,710,435. Another is the Gradolph device, patent No. 1,318,863. The third is the McLean sprayer, patent No. 1,395,965. These three patents claim devices and not methods. In each case the principal purpose of such device is to spray paint, but other liquids are also mentioned.

In paint spraying the application of air pressure in the independent air line in the 150 to 200-pound range is unnecessary and in fact undesirable. Likewise, there is no advantage in obtaining an emulsion effect where paint is to be sprayed. In these two particulars Reading teaches a process not contemplated by Shelburne, Gradolph or McLean.

In January 1953, which was towards the end of Reading's experimental period, one W.S. Cahill developed a method of spraying rubber cement on tire carcasses. After a little over a month of testing, Cahill began the manufacture and sale of a spray device to be used for this purpose. He sold five such appliances between February 7 and July 23, 1953. One D.S. Hartman assisted Cahill at the outset, and was later given Cahill's original machine. Hartman used this machine in the regular course of his tire retreading business from February 7 to July 1, 1953, during which period he processed 4,658 tires.

The Cahill method utilizes a paint-spray type of device but without provision for agitation or emulsion of the liquid in the tank. While compressed air is forced into the tank, this is done at a point above the liquid and only for the purpose of forcing the liquid into the spray gun.

Cahill employs an independent air line connected to the spray gun, utilized, as in the case of Reading, to atomize the liquid cement as it is applied. The patent issued to Cahill does not disclose any particular air pressure, although references are made to the "high pressure" air supply. In his recommendations to others using his device, Cahill advised the application of forty pounds pressure in the independent air line and five pounds in the tank.

The Cahill process contemplates that brushes will be applied to the cement after it is coated on the tire. Most of the appliances manufactured by Cahill were actually without brushes. Hartman, however, continued to use the brushes on his Cahill device and considered them to be a valuable part of the machine. This feature of Cahill has no counterpart in Reading.

The Cahill method of spraying liquid cement has a tendency to cause "cobwebbing" of the cement on the tire. The result is that tiny filaments and strands of dried cement would float in the atmosphere, creating a health, fire, and dust hazard. As before noted, this "cobwebbing" effect was avoided in the Reading process.

Reading made the first public disclosure of his concept on January 23, 1952. Cahill applied for his patent on June 17, 1953. Reading applied for his patent on July 23, 1954. The Reading patent was issued first — on October 18, 1955. The Cahill patent was issued on August 7, 1956.

The application for the Reading patent as originally filed was intended to cover both the device and the process. Thereafter the application was divided and the patent issued on October 18, 1955, unlike Cahill, is on a process and not an apparatus. The part of the Reading application pertaining to an apparatus is still pending in the patent office.

Appellant Elrick Rim Company has been in the tire-retreading machinery business since 1951. M.C. Elrick, a member of the original partnership, first met appellee Reading at a tire dealers' convention in 1954. Before meeting Reading, however, Elrick had heard of the Reading method and believed that it was excellent. When the two men met, Elrick asked Reading if any jobbing arrangement was available for northern California. Reading referred Elrick to Reading's northern California representative who informed Elrick that no distributorship was available.

Several days after the convention Elrick entered into a contract with an Oakland manufacturer for the production of a spray device. Before doing so Elrick had made an examination of the ultimate results of the Reading device and method, and had concluded that Reading was not entitled to patent protection.

The Elrick device as thereafter manufactured for and marketed by him utilized a conventional pressure paint pot. In use, the cement and solvent are put in the tank, which is then closed, and air compressed to ten pounds per square inch is allowed to bubble through the liquid for a few minutes. The liquid cement is then released by pressing the trigger of the spray gun. An independent source of compressed air mixes with the cement at the nozzle as in any spray device.

The instructions which Elrick issued to purchasers of his device describe in some detail the operating method to be employed. Elrick's instructions for air pressure in the independent line are the same as in the specifications of the Reading patent. Elrick's instructions for air pressure within the liquid cement receptacle, however, are for ten pounds per square inch, as compared to Reading's recommended initial pressure of forty pounds thereafter reduced to fifteen pounds. The Elrick instructions do not mention "emulsion," but use the term "air agitation."

The Elrick process, however, actually does produce an "emulsion," as Reading uses the term, though it may not be accomplished as quickly or as completely as by Reading's process. In view of this fact and the similarity of methods as to atomization, the Elrick practice is equivalent to the method disclosed by Reading. Whether Elrick's action in adopting this equivalent practice was willful, deliberate, and intentional is a fact question which will be dealt with at a later point in this opinion.

The facts summarized above incorporate the substance of the findings of fact, but with considerable amplification. Appellant challenges many of the individual findings of fact as clearly erroneous.

There are instances in these findings of fact, as with those in almost any case, where an individual finding, read alone, is incomplete or inaccurate. The limitations of language are such that it is often impossible to state a fact with all of its attendant qualifications and exceptions in a single sentence or even a single paragraph of reasonable length. The accuracy and completeness of findings of fact must therefore be judged by reading them as a whole and not by considering individual findings in isolation.

The findings of fact here in question are comprehensive, consisting of fifteen paragraphs. They are supplemented by an eight-page memorandum decision. Although no attempt was made therein to deal with each subissue or byway explored during the trial, we find them to be adequate.

We also find them to be supported by substantial evidence. This is demonstrated by the details we have added in the above summary of the facts. These details are based upon substantial evidence. Where the evidence has been found to be in dispute, we have adopted that version which is consonant with the findings of the trial court. This is the only proper course since it is not our function to make findings or to test the trial court's findings through a weighing of the evidence.

The findings of fact are not clearly erroneous.

The trial court concluded from these facts that in the conception and perfection of his process Reading exercised invention and exceeded the skill of the art. Appellant attacks this conclusion. It is argued that, as compared to the prior paint-spray art taught by the Shelburne patent, Reading cannot be regarded as the invention or discovery of a new and useful process. The same comparison is drawn and assertion made concerning the prior liquid cement spray art practiced by Cahill and Hartman.

It has already been noted that the Reading process can be utilized by means of the Shelburne, Gradolph, or McLean device, providing certain adjustments or changes are made. This fact alone, however, is not sufficient to rule out patentability. The term "process" is defined in the patent law as including "a new use of a known process, machine, manufacture, composition of matter, or material." 35 U.S.C.A. § 100(b). A different use of a known substance, machine, or process is not "new" within the meaning of this statute if it is merely analogous or cognate to the uses theretofore made.

The trial court concluded that the use which Reading made of the known paint spraying devices was not analogous to the uses for which they were originally designed. The court based this conclusion on its findings of fact that Reading was the first to use excess air pressure in the independent air line and to obtain emulsion within the tank of liquid, both of these techniques being undesirable in the case of spray painting. In our view these two variances are sufficient to warrant the conclusion that Reading teaches a nonanalogous art and is therefore a "new" use of a known machine within the meaning of § 100(b).

It is true, as appellant argues, that in numerous decisions the rule has been announced that the mere function of a machine is not patentable. But this rule does not apply in the case of a process patent involving the use of a known machine where such use is found to be "new" within the meaning of § 100(b), provided the other conditions of patentability are satisfied.

Appellant is quite right, however, in contending that patentability is not established by showing that the process is new and useful. It is made clear in 35 U.S.C.A. § 101, that to be patentable the new and useful process must be the result of invention or discovery. The same view has been expressed with regard to the patent law as it existed prior to enactment of the new patent act on July 19, 1952.

Invention or discovery is not present where the new use of a known apparatus is the product of the exercise of ordinary professional skill. Pierce v. Muehleisen, 9 Cir., 226 F.2d 200, 204. There must be ingenuity over and above mechanical skill. Schick Service, Inc. v. Jones, 9 Cir., 173 F.2d 969, 974.

Bearing these principles in mind, we think that the trial court was warranted in concluding that Reading's process represents invention as compared to the prior paint-spray and liquid-cement-spray art. True, the Reading process makes use of the essential features of Shelburne and the other devices which have been named. It is a new use, however, not only in the sense that the device is used to spray a nonpaint material, but also because essential elements of the device are employed in a manner different from that originally intended.

The compressed air inlet tube within the receptacle is used not alone to agitate the liquid (as in the case of paint) but also to emulsify it. Such emulsification is peculiarly desirable in coating tires with liquid cement and distinctly undesirable in applying ordinary paint. The independent air line under Reading is used at pressures far above those contemplated by Shelburne and the other devices, though within the physical capability of those devices. Again, the use of this excessive pressure is to be desired in applying liquid cement and to be avoided in applying paint.

Unlike the prior patents such as Shelburne, Cahill's apparatus and method (also utilized by Hartman) is intended for use in spraying liquid cement. But Cahill teaches neither emulsification nor atomization achieved at excessive pressures. The result was, as the trial court found, that Cahill did not overcome the difficulties which had long beset the tire retreading industry.

After a long period of experiments and tests, varying his technique sometimes by plan and sometimes by accident, Reading finally hit upon a way to overcome almost all of these problems. The need for such a development was urgent. Yet, despite the efforts of trained mechanics in this industry, no one had found the answer until Reading came along.

It is true, as appellant points out, that any mechanic can install an air inlet tube or regulate air pressure. But these steps were not taken in the manner and for the purpose contemplated by the Reading method until he discovered an advantage in doing so. When they were taken, new, unexpected, and extremely useful results were achieved.

Chance was a factor in Reading's success, as it often is in the unfoldment of an invention. But the prime credit goes to careful thought, painstaking research, experimentation and testing, and what we consider more than average patience, determination, and ingenuity. In our opinion the trial court correctly concluded that Reading's conception evidences the exercise of the inventive faculty.

Appellant challenges the legal sufficiency of the specific claims set out in Reading's patent. In our view, however, the stated claims of the patent read in the light of the patent specifications are legally sufficient within the meaning of 35 U.S.C.A. § 112.

The facts summarized earlier in this opinion indicate to us that Reading was not anticipated by the prior art. The trial court so concluded, and also held that the patent is not invalid for prior use or sale or on account of any estoppel. We agree.

The facts concerning Elrick's asserted infringement have been stated herein at some length. The trial court's ultimate finding that Elrick practiced a method equivalent to that disclosed by Reading has already been sustained. The conclusion of law drawn therefrom by the trial court, to the effect that Elrick infringes each of the claims of Reading, necessarily follows and is here upheld.

The trial court further found and concluded that Elrick's infringement was "willful, deliberate and intentional." Based on this finding, the trial court awarded attorney's fees to appellees in an amount later fixed at $7,500. Appellant argues that the evidence does not support this finding and the resulting award of attorney's fees.

There is substantial evidence to support the finding that Elrick's adoption of a process substantially equivalent to that of Reading's was willful, deliberate, and intentional. But the purport of the questioned finding goes beyond this, for it refers to "infringement." There could not be infringement unless Reading's process met all the conditions of patentability. The finding of fact, then, is to the effect that, having reasonable grounds for believing that Reading's process was patentable, Elrick willfully, deliberately and intentionally set out to infringe that patent.

We do not believe that there is substantial evidence to support such a finding. Elrick certainly did not concede any such motivation. On the contrary, he testified that before entering into competition he made an examination of the ultimate results of the Reading device and method, and concluded that Reading was not entitled to patent protection.

Considered in the hindsight of this opinion, that was an erroneous conclusion. But it was not a wholly unreasonable or unfounded conclusion, since the question of patent validity, in our view, is close on both the facts and the law. Having arrived at that conclusion, Elrick proceeded in the only way that he could to test the matter — he engaged in competition, and when resisted by Reading instituted this action for a declaratory judgment.

The finding of fact as to intentional infringement is therefore clearly erroneous.

Under 35 U.S.C.A. § 285, an award of reasonable attorney's fees may be made to the prevailing party "in exceptional cases." Having concluded that the finding as to intentional infringement is erroneous, it is our view that there is nothing else to categorize this as an "exceptional" case warranting the imposition of attorney's fees. See Park-In Theatres v. Perkins, 9 Cir., 190 F.2d 137, 142.

The judgment is modified by eliminating therefrom the award of attorney's fees to appellees. In all other respects the judgment is affirmed. The parties shall bear their respective costs on this appeal.

"To Operate

"General