WALSH, District Judge.
Appellants sued appellees in the United States Court, Southern District of California, Central Division, charging by their Second Amended Complaint that appellees had infringed claims 1 and 7 of United States Letters Patent No. 2,332,1821 (hereinafter “the Stearns patent”), and praying for an injunction, an accounting, and damages. Appellees, by answer, raised the defenses that: the claims in suit were invalid for want of invention and because they do not particularly point out and distinctly claim the invention; the claims in suit were not infringed by appellees; and appellants were barred or estopped from obtaining any relief because of their misuse of the Stearns patent by employing it to exercise a monopoly over unpatented materials. Appellees also filed a counterclaim against appellants predicated upon [51]*51the affirmative allegations of their answer and praying for (a) a declaratory judgment that the claims in suit were invalid and that appellees had not infringed said claims; (b) an injunction restraining appellants from acts of unfair competition and from asserting the Stearns patent against appellees and their customers; and (c) a judgment against appellants for damages occasioned appellees by appellants’ misuse of the Stearns patent. Appellants made answer to the counterclaim and, upon the issues formed by the pleadings outlined above, the cause proceeded to trial before the court sitting without a jury.
Following the trial, the court below filed a memorandum opinion2 and, thereafter, made and filed Findings of Fact and Conclusions of Law. Having concluded as a matter of law that claims 1 and 7 of the Stearns patent were invalid for want of invention over the prior art and that, accordingly, appellants’ Second Amended Complaint should be dismissed, the court entered its judgment dismissing the Second Amended Complaint upon the merits. No findings or conclusions were made upon the issues of invalidity of the claims because of failure to particularly point out and distinctly claim the invention, or infringement, or misuse of the patent; nor was any judgment entered on the counterclaim and the answer thereto. The appeal here is by appellants from the judgment dismissing the Second Amended Complaint.
The Stearns patent is for an “Insulation Testing Device” and is in the field of what are styled “holiday detectors”, devices to detect flaws or “holidays” in the coal tar or asphalt enamel coatings which are applied to steel pipelines prior to their burial in the ground. The need for holiday detectors had its origin in the pipeline industry devoted to the transportation of petroleum products. The tremendous investment in the miles of steel pipe utilized in the industry, buried in the ground, and the fact that the pipe, if unprotected, is subject to rapid deterioration by rusting or electrolysis dictated that every feasible step be taken to protect the pipelines and thereby lengthen their life span. Therefore, at a comparatively early date in the development of the industry, there came into use the practice of coating the pipe, before it was interred, to safeguard it against corrosion. Originally, the coating was applied by hand and imperfectly, with the result that there were “pin holes”, “thin spots”, and other flaws or breaks — all called “holidays” — in the coating. Holidays would, of course, permit corrosion and, ultimately, leakage at their points of location. Thus was the industry led into efforts toward and a search for improved means of protectively coating pipelines and means of detecting flaws in the coating so that they might be repaired before the pipeline was placed underground. The efforts and the search continued for years, in both directions. Means of protectively coating the pipelines have been improved ; but holidays still occur and hence there is still need for detectors. Detectors, too, were discovered and improved, and out of the history of these devices this case arises.
In the matter of holiday detection, visual inspection of the coatings was at first relied upon to detect the flaws, but with highly unsatisfactory results. Then, because the organic material used for coating the pipe is, if sound, a poor conductor of electricity, while a flaw or holiday therein is, relatively speaking, a good conductor of electricity, the possibility of making an electrical inspection of the coated pipe was hit upon. In 1931, there was devised one of the first holiday detectors which made an electrical inspection of pipe coatings. It was a cloth saturated in a brine solution to conduct electricity and connected to the terminals of a battery to supply electric current. [52]*52The cloth was slung underneath the pipe 3 and cords attached to its ends were held by a man standing atop the pipe. As the man walked along the pipe, the cloth was moved along the pipe also; and as the cloth passed over a holiday, the passage of electricity from the cloth electrode through the holiday was registered upon a meter carried by the man and connected to the battery and the cloth electrode. This primitive holiday detector did disclose holidays better than was possible by visual inspection, but it was still very inefficient and efforts to improve the device and its several parts thenceforth engaged the attention and the efforts of the industry. Over the years, improvements and advancements were made in holiday detectors and in the elements which make up such detectors; but one of the large questions in this case is the precise point which had been reached by these improvements and advancements at the time of the advent of the Stearns holiday detector in early 1941.
There is in the record no substantial dispute between the parties as to a great deal of the history of holiday detectors during the decade between the birth of the “brine-soaked cloth plus battery” device and the completion of the Stearns invention. As to the electrode, the first step after the brine-soaked cloth was wire brushes or minute wires 4 attached to a metal frame semi-circular in shape. A handle to be held by the operator was attached to one end of the semi-circular frame, and inspection of the pipe was made by the operator’s passing the electrode along the length of the pipe. Since each such passage of the electrode covered but one-half of the circumference of the pipe, it was necessary, in order to cover the entire circumference, to pass the electrode over the pipe twice. Next came the full circle brush type electrode, constructed by putting together two half-circle brush type electrodes, joined by a hinge at the ends opposite from those to which the handles were attached. During the years under consideration, the brush type electrodes were the most widely used but other types also came into use. One was a copper ring set in a circular steel band but insulated from the band by a ring of reinforced resin. O. C. Mudd built a full circle electrode by joining the ends of a coiled spring. Mudd’s electrode was attached by links and chains to a circular steel frame and was moved over the pipe surface by dragging it.
Developments in electrodes were paralleled by developments in other elements of detectors. One aspect of those developments which is pertinent here relates to the manner in which the electrical apparatus supplying current to the electrode was rendered mobile. Originally, such apparatus sat on the ground and was moved forward as the electrode progressed to the limit permitted by its connection to the electrical apparatus. Next, the apparatus was carried by the operator or someone assisting him. Then, the apparatus was placed in a cart or other vehicle and pulled along beside the pipe. Finally, the apparatus was set into a small platform which had wheels attached to it, and the whole arrangement was placed on top of the pipe to be moved by pushing or pulling it along the pipe as a roadway.
In 1934, R. A. Brannon built and used a holiday detector which integrated the platform on wheels to move atop the pipe carrying the electrical apparatus and a full circle brush type electrode. The circular frame of the electrode was rigidly joined by a piece of pipe to the platform containing the electrical apparatus, and as the platform was pushed or pulled the electrode moved also.
Turning again to the electrodes described above, it is not disputed that they had serious shortcomings. The brush type electrodes tended to wear out rapidly. Those attached to frames could not be held steadily at the proper uniform distance from the pipe coating, either because of inability to keep the frame concentrical with the pipe as the frame was in motion or because of variations [53]*53in the diameter of the pipe. The Mudd coiled spring electrode above described, designed to be dragged along the pipe, tended to separate from the pipe as the “pull” at its several points of connection to the rigid frame became unequal.
The claims of the Stearns patent in suit are for a combination of (1) a platform or carriage on wheels intended to roll along the pipe and to carry the electrical apparatus; (2) an electrode in the form of a coiled spring having its ends secured together to embrace the coated pipe to be tested; and (3) a pusher and contactor, carried by and electrically insulated from the platform, to provide mechanical and electrical connections between the spring electrode and the platform such that the movement of the platform along the pipe will cause a rolling movement of the spring electrode along the pipe.5 From what has been said hereinbefore of the history of the holiday detector art, a reading of these claims raises the inquiry whether the Stearns patent is for a combination of old elements; and, indeed, appellants concede that the elements are old. But appellants took the position upon the trial, and they do here, that the elements in the Stearns combination cooperate differently therein than they did in prior detectors and that new and unusual results are produced by the combination. They point out, as the keystone of a different cooperation of the elements, that in the new combination the coiled spring electrode is detaehed from any frame, is supported only by the pipe, and, most importantly, it rolls as it inspects the coated pipe. The results are, they say, that the electrode at all times completely embraces the coated pipe in operation so that no part of the coating will be skipped, the electrode is always free in operation to change its form to conform to changes in the contour of the pipe, and the electrode rolls so readily and easily that by the very nature of its movement it will not pull away from the coating.
The court below found that prior to the invention of the Stearns patent, and apart from the coiled spring electrode of Mudd described above, holiday detectors employing an electrode formed by a coiled spring encircling the pipe to be tested with its ends joined together were openly and commercially used and/or sold in the United States by Alonzo L. Smith and by David Harrell, Sr., and his employer Houston Pipeline Co.; that the rolling property of a coiled spring electrode and the advantages of such property in holiday detection was known prior to the invention of the Stearns patent to O. C. Mudd, to his associates, and to his [54]*54employer Shell Pipeline Co.; that the practical advantages of a rolling coiled spring electrode and the fact that such a spring could be made to roll along a cylinder and be used as an electrical connection were described and clearly taught by the Stevens patent,6 which was much earlier than Stearns’ and which was not considered by the Patent Office during the prosecution of Stearns’ application; and that the practical advantages of a rolling coiled spring electrode were put into open and commercial use in the United States prior to the invention of the Stearns patent by Alonzo L. Smith, by purchasers of Smith’s holiday detectors, and by Harrell’s employer. These subsidiary findings might, if they are supported by the evidence, render unassailable the further finding of the trial court (insofar as it may be said to be actually a finding of fact) that the “combination of old elements * * * does not amount to patentable invention because the elements thereof do not cooperate in any new way or contribute any new and unexpected result”. Consequently, we examine the evidence to determine whether or not the subsidiary findings are supported thereby.
Alonzo L. Smith. This witness, whose deposition was placed in evidence by appellees, testified that prior to 1941 he manufactured, tested and sold many holiday detector electrodes which were coiled springs encircling the pipe to be tested with their ends joined together and which rolled as they were pushed along the pipe. He testified that he made up and tested the first of these electrodes in late 1938 or early 1939. He stated that some photographs put in evidence as exhibits depicted electrodes such as those he manufactured and sold prior to 1941, and he identified an electrode which was received in evidence as one of the first made by him. He explained that an exhibit consisting of a photostatic copy of an invoice dated September 16, 1939, and containing the language “2 16" Pipe Electrodes (Special Full Circle) for Holiday Detector” related to a coiled spring electrode intended to encircle the pipe to be tested.
Smith testified, also, that a great deal of the work connected with his first coiled spring electrodes was done at Pielop’s machine shop, in Houston, Texas, and that one George Aschenbeck was employed in this work. But Aschenbeck, whose deposition was also introduced by appellees, testified that he did not enter into Pielop’s employ until after he had left the farm on June 28, 1940,7 a date from one to two years subsequent to the time when Smith said he recalled As-chenbeck’s working on the electrodes. Further, Aschenbeck stated that it was a considerable length of time after he went to work for Pielop — he would guess five or six months but would be afraid to say just how long it was — before he knew of a coiled spring electrode.
The evidence just summarized is not sufficient to support the finding by the trial court that before the date of Stearns’ invention Smith commercially used and sold a coiled spring electrode encircling the pipe to be tested with its ends joined together and that Smith and purchasers of his detectors put into open and commercial use the practical advantages of a rolling coiled spring electrode. There is no evidence to support the finding except Smith’s own oral statement as to events and circumstances long past. Smith’s testimony was not given in the presence of the trial court, hence there is no possibility of its having been accorded added weight by the trial court because of his having personally observed and', heard the witness while testifying. The-exhibits cannot be said to corroborate-Smith in any real sense. The existence-before 1941 of the electrode introduced', as a physical exhibit and the electrodes, depicted in the photographs is proved, only if Smith’s oral testimony is accepted. The invoice does not speak of a. coiled spring electrode, and it can be-. [55]*55found to relate to one only if Smith’s oral testimony is accepted. Finally, Smith is contradicted in the most vital part of his story by Asehenbeck. There simply is not here the degree of proof which will sustain a finding of prior public use. Paraffine Companies, Inc., v. McEverlast, Inc., 9 Cir., 84 F.2d 335, 339; Rown v. Brake Testing Equipment Corp., 9 Cir., 38 F.2d 220, 223.
David Harrell, Sr. This witness, whose deposition was placed in evidence by appellants, testified regarding his work in the year 1939 with two types of coiled spring electrodes for holiday detectors. In the first type, a coiled spring was threaded through a number of contact rings which were attached to a rigid circular frame. The coiled spring encircled the pipe to be tested with its ends joined together. Harrell testified that as the frame and the electrode were pushed or pulled along the pipe, the coiled spring rolled, but the rolling actually had a deleterious effect because the spring was imparted a circular motion around the pipe, which caused the spring to “hang up” in the contact rings. He said, also, that it was found impossible to adjust the tension on the spring so that it would not drag in places and thus bite into the pipe coating. He testified that this type electrode was never used in the field but only experimentally in the plant; and that the experimental use of the coiled spring did not continue for any length of time but was discarded in favor of a wire screen electrode. He characterized his work on the coiled spring electrode as an “unsuccessful attempt to apply a coiled spring to the operation”.
As to the second of his coiled spring electrodes, Harrell described it as a coiled spring threaded over two semicircular pieces of welding rod, ends of the respective pieces of rod being joined by a hinge at one point. The ends of the coiled spring were not joined in this electrode, and a gap of several inches between the ends of the spring occurred at the point where the unhinged ends of the welding rod frame approximated each other. Harrell said that this type of spring electrode was actually used in the field on a mile or two of pipe and that it rolled as it was pushed or pulled along the pipe; but again the circular motion imparted to the spring as it rolled caused it to “hang up” at the hinged portion and at the unjoined ends of the frame, with the result that “skipped spots” occurred in the inspection of the pipe. Harrell mentioned other unsatisfactory features of the electrode but perhaps its story is best summed up in his statements that “it was very unsatisfactory”; “It was discarded” in favor of the wire screen electrode; “That spring * * * didn’t work, it was no good”; and “We didn’t use this thing long, we threw it away”.
The foregoing does not support the finding that Harrell and his employer, before Stearns, openly and commerically used a coiled spring electrode encircling the pipe to be tested with its ends joined together and that Harrell’s employer put into open and commercial use the practical advantages of a rolling coiled spring electrode. The evidence does not show use, commercial or otherwise, but shows only unsuccessful experiments; and such experiments cannot anticipate or negative invention. Coffin v. Ogden, 18 Wall. 120, 124, 21 L.Ed. 821; Smith v. Snow, 294 U.S. 1, 17, 55 S.Ct. 279, 79 L.Ed. 721; Pacific Contact Laboratories, Inc., v. Solex Laboratories, Inc., 9 Cir., 209 F.2d 529, 532; Paraffine Companies, Inc., v. McEverlast, Inc., supra; Picard v. United Aircraft Corp., 2 Cir., 128 F.2d 632, 635; Oswego Falls Corp. v. American Seal-Kap Corp., D.C.E.D.N.Y., 65 F.Supp. 338, 355.
O. C. Mudd. It was not contended by appellees, and the court below did not find, that Mudd or the Shell Pipeline Co. ever used a rolling coiled spring electrode commercially before Stearns. However, the depositions of Mudd and two of his associates with Shell Pipeline Co. were put into evidence by appellees and the testimony was that, in 1940, in working with Mudd’s nonrolling coiled spring electrode which has been described herein-[56]*56before, they found that when the coiled spring with its ends fastened together was placed about a roll of roofing paper it rolled easily with a moderate amount of applied force. No experiments were made with the spring as a rolling electrode at that time, the witness Mudd explained, because Shell’s pipe coating process then in use left protrusions of the coating material adhering to the pipe and a rolling coiled spring could not apply enough force to remove them. In early 1942, the witnesses testified, Shell for the first time wrapped its pipe with paper and Mudd and his associates used a rolling spring electrode as a holiday detector on that job.
All that can be made of this testimony is that, in 1940, Mudd and his associates found out that a coiled spring encircling a cylindrical body with its ends joined together would roll easily if it were pushed with a moderate amount of force and they believed that a rolling coiled spring could be utilized as an electrode in a holiday detector if it were tried. Though the witnesses gave reasons for not doing so, the fact is that they never used, or even tried to use, a rolling spring as an electrode, until after Stearns had made application for a patent. Mere belief, or even confidence, that a device will work if it is constructed and tried is not enough to bring into operation the statute8 which renders unpatentable an invention “known * * * by others in this country” before the invention thereof by the applicant for a patent. An invention is “known” as that word is used in the statute, when it is “reduced to practice”. It follows here, on the same principle, that Mudd and his associates and his employer cannot be found to have “known” the practical advantages of a rolling spring as an electrode before they ever rolled a spring for that purpose. Coffin v. Ogden, supra; Clark Thread Co. v. Willimantic Linen Co., 140 U.S. 481, 489, 11 S.Ct. 846, 35 L.Ed. 521; Block v. Nathan Anklet Support Co., 2 Cir., 9 F.2d 311, 312; Harper v. Zimmerman, D.C.Del, 41 F.2d 261, 267.
The Stevens Patent. This patent, a copy of which was placed in evidence by appellees, is for an electric snap-switch, and proposes the utilization of a rolling coiled spring in its operation.9 The intended use of the spring: may be described roughly as follows: The spring encircles a stationary spindle-shaped member which has electrical circuit terminal contacts on opposite sides-at one of its ends. A carrier (means to-push the spring) is fastened to a push, rod and when the rod is pushed the carrier forces the spring over the bilge (greatest diameter) of the spindle. This, expands the spring and puts it under-greater tension, so that the instant the-bilge is passed the contractile energy of' the spring is added to the thrust of the-rod and snaps the spring onward to the-end of the spindle, where it closes the.circuit between the two contacts.
It is our view that the Stevens patent is in an art so remote and non-analogous from that of the Stearns patent that it should not have been considered by the court below upon the question of invention and, consequently, that, the court below should not have made the-finding which it did upon the basis of the evidence regarding the Stevens patent. To us, it seems not reasonable to-hold that an inventor working in the-field of holiday detectors would naturally have looked to the art of snap-switches, for help. General Metals Powder Co. v. S. K. Wellman Co., 6 Cir., 157 F.2d 505, 510. The rule laid down by many of the authorities is that whether arts or uses are analogous depends upon the similarity of their elements and purposes. It is said that if the elements and purposes in one art are related and similar to those in another art, and because and by reason of that relation and similarity make an appeal to the mind of a person having mechanical skill and knowledge of the purposes of the other art, then such [57]*57arts must be said to be analogous; and, if tbe converse is true, they are non-analogous arts. A. J. Deer Co., Inc. v. U. S. Slicing Machine Co., 7 Cir., 21 F. 2d 812, 813; Copeman Laboratories Co. v. General Plastics Corp., 7 Cir., 149 F.2d 962; Allied Wheel Products v. Rude, 6 Cir., 206 F.2d 752, 755; 69 C.J.S., Patents, § 53, page 266. Even if we assume here a relation and similarity of elements in holiday detectors and snap-switches, the purpose of the snap-switch art, i.e., to open and close an electric circuit, has no relation or similarity to the purpose of the art of holiday detection, i.e., to determine the condition of pipe coating by subjecting it to electrical inspection.
Without the support of the subsidiary findings which we hold are clearly erroneous, the finding of the court below that the combination of old elements in the Stearns patent does not amount to patentable invention because the elements thereof do not cooperate in any new way or contribute any new and unexpected result must also fall. The elements of the Stearns combination do functionally operate differently in the combination than they did in their old surroundings. As we have determined, the spring electrode for the first time in its use in holiday detectors is rolled, instead of being dragged. In the Stearns detector, the pusher rotably engages and forms a movable electrical contact with the spring electrode so as to roll it and connect it electrically to the high voltage test circuit; and movement of the carriage longitudinally upon the pipe imparts a rolling movement to the spring electrode. And this different coaction of the elements produces a new and useful result, viz.: The detection of holidays in a more facile and efficient way.10 Willard v. Union Tool Co., 9 Cir., 253 F. 48, 54; Long v. Dick, D.C.Cal., 38 F.Supp. 214, 220; Application of Ostermann, 179 F.2d 1010, 1014, 37 C.C.P.A., Patents, 891, Imperial Brass Mfg. Co. v. Bonney Forge & Tool Works, Inc., D.C.Pa., 38 F.Supp. 829, 830; 69 C.J.S., Patents, § 68, page 303; 40 Am.Jur., Sec. 19, p. 543.
One additional finding made by the trial court remains to be considered. It reads: “The act of combining the said old elements to produce the combination of said Letters Patent, does not rise to the dignity of invention because it does not, in view of the state of the art as shown by the exhibits thereof and the explanation and analysis thereof by plaintiffs’ and defendants’ witnesses, represent anything more than the skill of the art or calling”. It is evident that this finding (insofar as it may be said to be actually a finding of fact) is also bottomed upon the trial court’s erroneous inclusion in the prior art of the Stevens patent, the Smith and Harrell rolling spring electrodes, and Mudd’s claimed knowledge of the practical advantages of a rolling spring electrode. It is, therefore, also dearly erroneous.
As we pointed out earlier, before Stearns there was a decade long search in the industry for an electrode which would be free of the shortcomings and deficiencies of those then in use. In the search, Mudd and Harrell, the latter particularly, were close to doing what Steams ultimately did. But though they were outstanding workers in the holiday detector field, they failed to solve the problem because their detectors, like all others theretofore devised, employed a frame which supported the contact elements in such manner that the frame could and did pull part of the contact elements away from the pipe coating. Steams solved the problem, basically, by abandoning the frame and using the pipe alone to support the electrode. Stearns was an innovator, not a follower. Sinclair & Carroll Co., Inc., v. Interchemical Corp., 325 U.S. 327, 330, 65 S.Ct. 1143, 89 L.Ed. 1644; Pointer [58]*58v. Six Wheel Corp., 9 Cir., 177 F.2d 153, 161. In addition, the record establishes beyond question ready and widespread acceptance of ' Stearns’ detector on the market and its attendant commercial success. We have here, then, a patent for an improvement which fills a long felt need, which those schooled in the art had not been able to devise before the patentee, and which meets with acceptance in the market. When these in-dicia of invention are taken into account together with the true state of the prior art and what Steams actually did to improve the art, it must be concluded that the Stearns patent is not invalid for want of invention. Expanded Metal Co. v. Bradford, 214 U.S. 366, 381, 29 S.Ct. 652, 53 L.Ed. 1034; Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 56, 43 S.Ct. 322, 67 L.Ed. 523; Johnson Company, Inc., v. Philad Co., 9 Cir., 96 F.2d 442, 444; Research Products Co. v. Tretolite Co., 9 Cir., 106 F.2d 530, 532; Tyra v. Adler, 8 Cir., 85 F.2d 548, 552; Ideal Roller & Mfg. Co. v. Sutherland Paper Co., 6 Cir., 96 F.2d 675, 677.
The judgment of the court below is reversed and the cause is remanded for further proceedings not inconsistent with this opinion.
PER CURIAM.
On Petition for Modification
It is ordered that appellants’ petition for modification of the opinion herein be and the same hereby is denied. The uncertainty which appellants seem to fear will arise in further proceedings in the case seems unlikely in view of the fact that Claims 1 and 7 of. the patent in suit are fully set out in footnote 5 of the opinion. In the light of the footnote, it is obvious that Claim 1 is directed to the electrode and the pusher and con-tactor, while Claim 7 is directed to the elements of Claim 1 plus a platform or carriage on wheels to which the pusher and contactor is attached.