This appeal questions only the validity of the patents in suit, the appellants contending that neither patent discloses invention but merely ordinary mechanical skill. The. patents relate to a certain type of metal strip used in terrazzo floors. It is known as “heavy top” terrazzo strip, and consists of a thin, metal base strip of familiar type which is fastened to a thicker, metal top bar; the latter being rabbeted along one side to receive the upper edge of the base strip. The structure disclosed in both patents is the same except for the means used to fasten the top and base strips together. Patent No. 1,914,115, the application for which was filed earlier, suggests spot-welding as the preferred method of securing the top and base together, but only claim 3 requires welding; the other claims are satisfied by any form of permanent fastening. Of the twelve claims in the patent, six specify the rabbeted top strip as an element of the combination, while the other six provide for the equivalent of rabbeting in either the top or base strip. All of them except claim 3 are in suit. Patent No. 1,791,267 requires the fastening to be by hollow or *843tubular rivets. The four claims of this patent are in suit.
Terrazzo strips serve two purposes: They tend to prevent the extension of cracks which develop in the terrazzo flooring, and they form ornamental designs. Strips as thin as 20 gauge (.03 of an inch) will serve the former purpose, but for ornamentation a thicker strip is desired. A strip one-eighth or one-quarter inch thick, if of uniform thickness throughout, is too expensive to sell well. Hence there developed a need for the “heavy top” type. The problem was to fasten a thick top bar to the thin base strip. Fink experimented first with solid rivets, then screws, then welding. For various reasons which need not be here stated, none of the methods were satisfactory. He then tried rabbeting the head piece and found that the two pieces could now be successfully welded, but he “could not get equipment to turn out such type of strip commercially.” So he returned to rivets, first solid rivets and later hollow rivets. The latter produced what he considered a “heavy top” strip “wholly satisfactory commercially.” This -was in 1929. His application for the patent calling for hollow rivets (No. 1,791,-267) was filed June 24, 1929. It was granted February 3, 1931. Application for his other patent (No. 1,914,113) was filed April 12, 1928, and granted June 13, 1933.
Fink was not the first to devise a two-piece • terrazzo strip. The art shows several instances in which the top strip was formed separately from, and attached to, the base strip. For example, patent No. 1,544,733, to Ferrarini (1925), and No. 1,667,313 to Galassi (April 24, 1928), show the top strip set into a trough or channel formed in a fold of the base strip; while No. 1,723,825 to Traitel (August 6, 1929) shows a thick top bar with a channeled bottom which fits over the upper edge of the base strip and is swaged thereto. All of these were issued on applications filed before Fink’s first application, although Fink alleged a date of conception prior to Traitel’s application. Nor was Fink the first to apply rabbeting to the top strip. The British patent, No. 232,728, to iYlarus (1925) discloses a top strip rabbeted on both sides and fitting into a groove in a thick base strip. The Chase Company catalogue also discloses rabbeted nietal strips like Fink’s top piece, though there is nothing to indicate how they were to be fastened to a base strip. Moreover, it is common knowledge that-rabbeting has long been applied in the art of carpentry in joining two strips of wood. It serves no different function in Fink’s terrazzo strips. It is true, as the plaintiff urges, that none of these references except Traitel discloses a structure in which a thick top bar is secured to a thin base strip. But once the need for such a structure arose, the only problem was how to fasten the two pieces together. With the art of the metal worker and the carpenter in the state indicated, when Fink began his experiments in 1926, we cannot believe that it required the genius of invention to secure the rabbeted head to the thin base strip by means of rivets, screws, or spot welding. While an invention should not be invalidated merely because it comes after experimentation, still the experiment which succeeds must require some ingenuity, if it is worthy of being granted a patent monopoly. See Kalamazoo Loose Leaf Binder Co. v. Wilson Jones L. L. Co., 286 F. 715, 718 (D. C. S. D. N. Y.); Hollister v. Benedict Mfg. Co., 113 U. S. 59, 73, 5 S. Ct. 717, 28 L. Ed. 901; Concrete Appliances Co. v. Gomery, 269 U. S. 177, 185, 46 S. Ct. 42, 70 L. Ed. 222. Merely to run down every obvious' alley until the best route to the desired goal is found cannot be deemed invention. We cannot see that Fink did more. All his experiments were within the skill of any competent artisan in metal work. Patent No. 1,914,115 we hold to be invalid.
Little more can be said in favor of the other patent. When equipment for spot welding was not available to him, Fink turned to riveting as an obvious alternative. Solid rivets proved unsatisfactory because they distorted the thin sheet metal. After experimenting with them for about two years, he finally tried hollow rivets and found they obviated the trouble caused by solid rivets. Although hollow rivets do not appear io have been used before in connection with terrazzo strip, they were a well-known substitute for solid rivets. Having exhausted the possibilities of one kind of rivet, Fink turned to the other. Tlieir characteristic absorption of strain, which prevents distortion of the sheet metal into which they are driven, is not so recondite that their use for this purpose can be said to require a degree of perspicuity amount*844ing to invention. We think this patent also invalid.
The decree is reversed, with directions to dismiss the bilk