This is an appeal in a patent infringement suit. The patent is No. 2,742,327 which was applied for by Theodore Marks on April 20, 1951 and was issued to him April 17, 1956. We shall refer to it as the Marks patent, although by mesne assignments it is now owned by American Technical Machinery Corporation. The claims relied upon are 2, 16 and 17. They are printed in the margin.1 Judge Mishler2 held the claims valid but not infringed. The plaintiff has appealed from the holding of non-infringement. The defendant counter-claimed for a declaration of invalidity of the patent, and has appealed from the dismissal of his counterclaim.

Plaintiff and defendant are competitors in making and selling brushes, and plaintiff also manufactures brush-making machinery.

The Marks patent is a combination patent for a fully automatic machine for making brushes of the type commonly used in the household and in women's hair curlers. Judge Mishler's opinion clearly describes the Marks' machine as follows:

"* * * the machine is designed to produce wire-twisted brushes of varying lengths not exceeding approximately 3". An upper and lower wire are continuously fed into the machine from two spools, the upper and lower wires engage pre-cut fibers or bristles fed from a hopper; the two wires holding the fibers or bristles in a more or less firm position, are then drawn into a predetermined length; the wires holding the bristles are then twisted by a rotating chuck; the twisted-wire brush is then cut. The brush is then complete and ready for packaging, except that in some varieties handles are added."

The court found that the Marks machine was an advance over the prior art; that it produced brushes faster than any previous brush-making machine, did not require the degree of skill in its operation that machines then known demanded, substantially reduced the cost of making twisted-wire brushes, was commercially successful, and the need for a low-priced brush was heightened by the widespread use of twisted brushes in hair curlers. Such considerations are consistent with the standard of patentability in this circuit. See Reiner v. I. Leon Co., 2 Cir., 285 F.2d 501, 504, cert. den. 366 U.S. 929, 81 S. Ct. 1649, 6 L. Ed. 2d 388; Lyon v. Bausch & Lomb Optical Co., 2 Cir., 224 F.2d 530, 535-537, cert. den. 350 U.S. 911, 76 S. Ct. 193, 100 L.Ed 799; The Standard of Patentability — Judicial Interpretation of Section 103 of the Patent Act, 63 Columbia L.Rev. 306, 313-16.

These findings of fact are not clearly erroneous, and we must accept them. We also accept the finding that "The means disclosed is not found in the teaching of any of the patents nor were the means obvious to one skilled in the art of brush-machine making." Accordingly, we agree with Judge Mishler's decision as to the validity of the patent.

We do not agree with his holding of no infringement. The proper test of infringement is whether the claims involved read directly on the accused machine, for "the claims measure the invention." Continental Paper Bag Co. v. Eastern Paper Bag Co., 210 U.S. 405, 419, 28 S. Ct. 748, 52 L. Ed. 1122; see Smith v. Snow, 294 U.S. 1, 11, 55 S. Ct. 279, 79 L. Ed. 721; 3 Walker, Patents § 450, at 1681 (Diller ed.). In the case at bar the accused machine was the Briglia.3 Judge Mishler compared plaintiff's machine with defendant's or, which is really the same thing, compared plaintiff's claims with defendant's machine but limited the claims to a range of equivalency substantially defined by plaintiff's "best mode." Both tests are erroneous; the former because it begins with plaintiff's machine and not plaintiff's claims; the latter because it too narrowly construes the protected range of equivalency.

It is true that narrow patents in a crowded field are not accorded as broad a range of equivalence as are pioneer patents, 3 Walker, Patents § 475. But here both defendant and his expert testified that all of Marks' claims in suit could be read on the Briglia machine.

Nor could Briglia's simultaneous pulling and twisting the wires be conclusive evidence of non-infringement. Marks' claims did not limit these operations to separate steps, and there was testimony that the Marks machine could also so operate. The simultaneous operation was regarded as equivalent by defendant's expert, and the twisting operation was not only known to the prior art but was thought by the trial court to be "of no consequence * * *."

Defendant is not saved by the difference in feeding wheels. Not only was defendant's mechanism thought equivalent to plaintiff's, but also both had been taught by the prior art. Besides, the mechanism was not the gist of Marks' novel combination as defined by the court.4

The court placed no weight on any "presumption of non-infringement" attaching to Briglia's patent. Defendant would treat that patent as establishing "more than a presumption" that the accused machine does not infringe. But the statutory presumption of validity of a patent is limited only to the existence of a patentable improvement. Cantrell v. Wallick, 117 U.S. 689, 694, 6 S. Ct. 970, 29 L. Ed. 1017. It need not be controlling once sufficient evidence of infringement of the basic patent has been introduced, as was done in the case at bar.

While we must accept findings of fact which are not clearly erroneous, we are not so limited as to erroneous views of the proper legal tests of infringement. See Up-Right, Inc. v. Safeway Products, Inc., 5 Cir., 315 F.2d 23; Hansen v. Colliver, 9 Cir., 282 F.2d 66. Here the finding of non-infringement was based on an unduly narrow definition of the patent's protection.

The decision as to validity of the patent and dismissal of the counterclaim is affirmed. The decision of non-infringement is reversed and the cause remanded for further proceedings consistent with this opinion.