This appeal from the United States District Court for the Southern District of Ohio attacks the final judgment of Judge Druffel entered for the defendant, dismissing the complaint. Appellant contends that some of the findings of fact are clearly erroneous. The issue of validity is not before us, because of the decision of the district court on the question of infringement.

Corning is the undisputed owner of two patents, specifically No. 2,177,336, 'Glass Vessel,' and No. 2,329,057, 'Tempered Hollow Glass Article,' hereinafter referred to as the 'Shaver Patent' and the 'Kiehl Patent,' respectively. Both patents relate solely to tempered glass articles having certain characteristics which render them less susceptible to fracture due to mechanical and thermal shock. It is not claimed that the process of tempering glass is novel. The subject matter is solely in 'end product.'

It has long been known that the durability of glass articles can be improved by certain heat-treating processes. Glass products, when originally formed into shape, have random stresses as an inherent quality which make them tend to break spontaneously, or upon slight shock. It was discovered long ago that if the article were evenly heated to the point of softness and slowly cooled, the random stresses were permanently relieved. This process is known as 'annealing.' It was discovered later that an annealed product could be reheated to the point of softness, followed by a rapid chilling effected by immersion of the article in a coolant, or by playing jets of cooling fluid or air upon the portions to be tempered. This procedure has the effect of setting permanently the outer surfaces of a glass article in a condition of compressive stress which is balanced by an interior zone permanently set in tension so that the resistance to breakage from shock is increased.

Appellee manufactures a heat-treated tumbler, or common drinking glass, known as the 'Sturdee' tumbler. Appellant contends that this tumbler infringes both its Shaver and Kiehl patents. Claims 1, 3, 5, and 7 of the Shaver patent are in suit. That patent has to do principally with the degree of tempering or quantitative maximum of permanent tension which should not be exceeded if a durable product is to result. It is well known in the prior art that if a certain degree of tempering is exceeded, the potential energy due to the permanent stresses in the glass would, upon fracture, be released with near explosive force and the glass would be shattered into many very small fragments. This phenomenon is known as 'dicing.' There is then a degree of tempering which, if exceeded, will result in a more fragile and dangerous article. The Shaver patent defines in terms of a formula this degree of tempering, above which dicing will occur. The variables of this formula are the ratio of maximum compression to maximum tension, modulus of elasticity, tensile strength, thickness and other properties of the glass. The formula is quite complicated and could not practicably be used for quality control purposes in manufacture. The trial court found and the evidence shows that the defendants never resorted to measurements of compression and tension in the production of their article, but they have utilized the drop and impact tests of the prior art exclusively.

Drinking glasses are more susceptible to fracture at the rim. The Kiehl patent pertains to strengthening that portion of the tumbler by the above stated tempering process. All its claims are in issue. It contemplates an article having its rim and an adjacent portion of its side wall set in a permanent condition of stress, such that its surfaces are in compression, the body of the glass tumbler being substantially free of permanent stresses (annealed), and an intermediate zone in the wall of the glass between the tempered rim and the annealed body having a maximum stress gradient in terms of kilograms per square millimeter per centimeter. This gradual change of the degree of stress in the intermediate zone of the glass is the crux of the Kiehl patent. The gradient prevents the top of the tumbler from 'ringing off,' as often happens when radical changes in degree of tempering occur in close proximity.

The burden is upon the appellant here to establish infringement of its patents by a preponderance of the evidence. The district court found that this burden had not been carried and, therefore, dismissed the complaint. Appellant takes the position that it introduced uncontroverted evidence to the effect that the appellee's 'Sturdee' tumblers, without exception, had characteristics contemplated by the wording of the patents and that it, therefore, had proved a prima facie case of infringement.

The district court found, with respect to both patents, that the appellee follows the prior art methods and products in the manufacture of its tumblers and therefore does not infringe either patent. No measurements were used by the appellee to determine stress values for purposes of controlling temper, as the polariscopic observation, rim impact, and drop tests of the prior art were sufficient for appellee to produce good commercial ware. It was found and there was evidence tending to show that the measurements submitted by the plaintiff as proof of infringement were unreliable. The testimony is uncontradicted that two factors are required for the computation of stresses: a measurement of birefringence in millimicrons by polarimeter observations and a value of 'stress optical constant.' As to the former there is evidence to support the finding of the trial court that the cutting of a strip of glass from the wall of a tumbler releases to some degree the permanent stresses in that sample, so that a measurement of birefringence is unreliable. There is evidence also that casts some doubt upon the reliability of the alternative method of measuring stress distribution in the walls of a tumbler by observations of an intact tumbler immersed in a liquid having the same index of refraction as the glass.

Not only was the trial court justified in finding that the measurements of birefringence were unreliable, but there is evidence also that the appellant's calculations of the necessary conversion factor 'stress optical constant' may have been inaccurate. District Judge Druffel had before him extensive documentary evidence, models, motion pictures, optical instruments, and other physical exhibits. His findings leave no doubt that he was convinced that the methods of measurement employed by the plaintiff were insufficient and, therefore, had only slight probative value.

On the record, the evidence, the briefs, and the oral argument of the attorneys, we are convinced that the findings of the trial court were supported by substantial evidence and were not clearly erroneous. Its judgment therefore, should be, and is, affirmed.