KOHLSAAT, Circuit Judge
(after stating the facts as above). Preliminarily to a consideration of the patent as set out in the claims, specification, and drawings thereof, it may be well to look into a certain assumption of the patentee, the experts, and counsel, wherein it is sought to read into the patent features which are neither claimed nor suggested therein. Of these, the most prominently asserted are those which claim for the patent the invention or first use of what is variously termed an annealed- mongrel-sized pipe, or pipe of a new, or characteristic, or off-size in dealing with the element of telescoping one pipe within another. The merits of such a pipe are effusively reiterated in the brief of appellee, wherein it is asserted that not only'is telescoping thus made easy, but that the pipe, being thin and annealed, can be readily bent to meet situations requiring adjustment. It also appears that, as a matter of fact, appellee uses this so-called annealed, off-size pipe in manufacturing his device. A reference to the claims, specification, drawings, and the proceedings in the Patent Office wherein the claims were several times rejected and as often amended, however, fails to discover even a suggestion of these ideas. Even when rejected on references which involved the standard pipe, no distinction on account of size and annealing was raised. On the contrary, it is stated, commencing at line 33 of page 2 of the specification, that:
“From the foregoing description, as well as from the drawings, it will be seen that the coupling-pipe is essentially a pipe or section of a pipe, being of the same cross-section at both ends and throughout its length, and distinguished from an ordinary piece of pipe by having a fixed eollar at or near one end. This is the feature which makes it practical for coupling purposes as herein described, and its belongings are the two gaskets, e and d, and washer, e', when used.”
Appellee himself testifies that his invention is not limited to any particular size of pipe. The expert McElroy testifies (Ans. to cross-question 10, page 210 of record) that the thinness of the pipe is a vital feature. At page 191 of the record, the same witness, speaking of appellee’s purpose in constructing the device of the patent, says:
“With this object in view, the first and most vital step was to employ a radically different coupling-pipe from those of the prior art, as just discussed. It was a thin annealed pipe, of a characteristic diameter, and small enough so chat it could telescope into the smallest size service-pipe that might be. employed. As he dispensed with all threaded joint connections on the pipe, he was able to make the pipe of thin material, and as the pipe was also of small diameter and annealed it could be readily bent at any point to as much of an angle as might be necessary to accommodate the pipe to the relative locations of the shanks of the bath-cocks and the ends of the service-pipes to be connected.”
It nowhere appears in the record that any one skilled in the art would know from the patent that the using of an annealed off-size pipe or tube is an essential element of the invention; nor is the court [614]*614able to make such a deduction. ■ The specification itself (lines 38 and 39; p. 2) says the coupling-pipe of the patent differs “from an ordinary piece of pipe by having- a fixed collar at or near one end." Appellee’s experts and counsel insist that such a construction as they place upon the patent is vital. In so contending- appellee’s brief is amplified to an extent not justified by the real issue. Yet little attempt is made to disclose the reasoning by which they reach their conclusions. It certainly is a far cry from the language of the patent to that of ‘the experts. It appears in the record that ordinary pipe may vary so in size as to permit of slip-joints. While the pipes must be of different sizes in order to telescope, yet that does not at all suggest that the pipe walls shall be thin and annealed — that is, tempered and toughened- — so thin as to be readily adjustable by bending, etc. The record contains several patents, as, for instance, the Smith patent, the Totham patent, the Moore patent, which involve the slip-joint, or the telescoping of one pipe into the other, yet there is no claim made that the question of off-size, or mongrel, or annealed pipe is contemplated. Nor was it. The situation before us here is in no wise different.
That the specification may not be read into a claim for the purpose of changing it, or to escape anticipation, or establish infringement, is settled beyond question by the authorities. Westinghouse v. Boyden Power Brake Co., 170 U. S. 537, 18 Sup. Ct. 707, 42 L. Ed. 1136; Gen. Sub-Construction Co. v. Netcher et al. (C. C.) 167 Fed. 549 (affirmed at the October session of this court). Much less will one be permitted to read into.his claims a feature not shown in either the specification or drawings, and therefore the attempt so to do is without merit in the present controversy, and may not be considered in determining the question of validity and infringement herein. Dealing with the patent as granted, it is apparent that the unitary device alone and in combination presents an article of manufacture drid merchandise vthich has very considerable merit. Whether, in view of the prior use and art set up, there is disclosed a true combination- or only an aggregation, involves an investigation of the history of pipe coupling. It-will be observed that claims 1 and 2 make no mention of the kind of pipes to be coupled, while claims 3 and 4, respectively, speak in terms of “water-pipes and the like” and “water and other pipes.” Appellee (Q. 13, p. 22, of the record) says his devices “are generally used for a connection to couple the service-pipe to a faucet. They may be used for connecting other members .that may require connection.” The specification (line 49, p. 2) says:.
“ * * * And ag these pipes are used for open plumbing, they are nickel-plated and finished alike from end to end.”
The expert McElroy (Record, p. 209) says:
“X. do not. know of any other use in open plumbing to which these coupling-joints could be put except for connecting bathtubs and wash basins with service-pipes, and I accordingly understand that that was what they are designed for.”- .'....
[615]*615And again on the same page:
“As a mechanical expert, however, f should say that the invention is. designed solely for use in connection' with water systems where there is a" high pressure to be taken care of, as the specification of the patent emphasizes the fact that there is no possibility of the coupling being Mown out, and this, of course, is a trouble which can only be apprehended in connection with water under high pressure.”
Later on he testifies that it might be applied to a gas machine.
The pertinency of all this is found in the attempt to narrow the field of the prior art to devices for use in connection with faucet-spuds and service-pipes. Five patents were cited in the Patent Office against, the Glauber patent in suit, i. e., Bower patent. No. 337,-120, granted March 2, 1886, “for pipe-connections designed more especially for wash basins, bathtubs, traps, etc.”; Schamp patent, No. 685,691, granted October 29, 1901, for a “simple coupling for connecting various kinds of pipes for the conveyance of water, steam, or other fluids”; Thomas patent, No. 657,712, granted September 11,’ 1900, for a combined union and elbow for coupling steam or water pipes; Sexton patent. No. 546.906, granted September 24, 1895, for a pipe-joint “for uniting pipes without threads cut on the ends thereof, and without solder, calking, or flanges”; and the Goodale patent, No. 130,216, granted August 6, 1872, for a joint or coupling for connecting steam or air-conveying pipes beneath cars, etc. On these references, the examiner rejected the application of Glauber three times. Claim 2 was rejected because the examiner was of the opinion that it would not require invention to apply to the elbow of the Bower patent the pipe-end of the Schamp pipe-coupling.
The Bower patent is for a waste-pipe which may well he deemed within the term “open or exposed plumbing.” It shows slip-joints at each end of the elbow-shaped coupling-pipe and the same cross-section from end to end. No collar is provided. It is stated that;
“The pipe P is arranged to extend more or less into (he respective sockets, as may he required, according to the relative positions of the trap and basin or bathtub.”
Coupling-nuts are used as well as rubber gaskets, differing somewhat from Glauber’s, as the situation required. The Schamp patent covers a straight coupling-pipe, having a fixed collar near one end against which a coupling-nut contacts, the end of the coupling-pipe beyond the collar projecting into another pipe member, which end is drawn by the coupling-nut and washer into water-tight relation with the other pipe-end as required. The end is thus telescoped into the neighboring pipe. The Thomas patent merely shows that. slip-joints were old. The Sexton and Goodale patents show slip-joints and gaskets suited to the coupling provided for therein. Of the patents relied on by defendant, it will he necessary to consider only the following, viz.; Tyden patent, No. 696,383, granted March 25, 1902, which is primarily for a device to prevent stealing of gas, .water, etc., but showing a one-piece coupling-joint of elbow shape of the same cross-section from end to end, and provided with a collar near one end. This end is plainly to be telescoped into its neighbor[616]*616ing pipe or meter connection, and fastened with a coupling-nut. It is provided with a gasket which makes a water-tight joint.
The connection with the pipe corresponding to the water service-pipe of the patent in suit is made by a solder-joint, just as Glauber says his connection may be made, although the Tyden joint is effected without telescoping the coupling into the service-pipe.
• Cornelius patent, No. 776,398, granted November 39, 1904 (after the filing of Glauber’s application, but prior to the last rejection and amendment thereof), is for a one-piece faucet-coupling “having a proper degree of adjustability, and, furthermore, to provide such a coupling having a proper degree of flexibility.” The faucet end of the coupling-pipe is provided with a. collar, a tapered gasket to correspond with the internally tapered seat of the faucet-spud and a water-tight connection effected by a coupling-nut substantially as in the Glauber device. The collar is adjustable, being screwed into the screw-threaded end of the coupling-pipe which extends beyond it to receive the gasket and fit into the tapered seat of the faucet-spud. This collar would seem to be to all intents the equivalent of a fixed collar. Appellee contends that the two last-named patents show only what he terms “tail-pieces,” and do not describe his commercial article.
Smith patent, No. 435,553, granted April 15, 1890, is for a unitary waste-pipe connection between washtubs and soil-pipes. It shows an elbow-shaped coupling-pipe having one arm longer than the other. It is of the same cross-section from end to end. Its long arm telescopes into another pipe to any desired distance, and is held by the usual coupling-nut and washer. It has a fixed collar, which also is held in end to end engagement with another pipe by means of a fiat washer and a coupling-nut. The patentee says his object was to provide a series of expansible and contractable pipes and connections for washtubs. whereby washtubs of varying dimensions may be readily connected to the soil-pipes. His specification closes with the following statement, viz.:
“With the above construction the plugs, a, may be of varying distances apart arid be readily connected by the pipes, D D, and couplings, B, by passing the unflanged end of a pipe farther in or out of a coupling until the required position is found. The parts are then suitably fastened. With this arrangement a system of piping for washtubs may be sold that will fit any tub, and an. ordinary workman can readily set up the piping. This is a great saving on the practice heretofore employed of measuring and cutting each pipe and then soldering the pipes together. This system also overcomes the danger of becoming loosened and water escaping.”
The Murdock patent, No. 193,654, granted July 3, 1877, calls for a combined cock and coupling intended to do away with the existing complicated and expensive system of attaching pipes and faucets to hot-water boilers. It shows a comparatively short and straight piece of pipe, having a fixed collar near one end, which end is telescoped into another pipe, and the two drawn together by a coupling-nut. The other end is smooth and capable of making a slip-joint, but not. so used.
[617]*617Downing patent, No. 167,850, granted September 21, 1875, shows double coupling-pipes, each composed of one piece, and a collar near their ends, which are drawn and held in connection with other pieces of pipe by means of washers and a coupling-nut. It also discloses the .smooth ends which telescope into service-pipes.
Shields patent, No. 433,750 for a pipe-joint, shows a short coupling-pipe having a fixed collar at its end with a tapering gasket fitting into a tapering seat in its adjoining pipe, a coupling-nut engaging the collar, and a smooth surface.
Dockery patent, No. 582,137, granted May 4, 1897, and several other patents introduced in evidence, furnish cumulative evidence of the fact that every element of the patent in suit, separately, or in various combinations, approximately that of the present device, were old and in common use.
In the field of prior use defendant cites the so-called Nelson pipe, claimed to have been manufactured by the N. O. Nelson Manufacturing Company of Edwardsville, Ill., from the year 1899 or 1900 to the present time. The exhibit is not one of the pipes put upon the market by the manufacturer, but is introduced as a fac simile thereof. The witnesses are one C. H. Harkins, formerly in the employ of the Nelson Company, L. D. Dawnin, secretary and manager of the Nelson Company and connected with the company for 23 years, John I'\ Staab, who had been with the Nelson Company since 1902, and Thomas R. Walton, who had been with the Nelson Company for about 15 years.
It will be remembered that Glauber filed his application May 21, 1904. These four witnesses are relied upon by appellant to establish the fact that the Nelson coupling-pipe antedated the patent in suit. The witness Harkins says he invented the Nelson pipe and recognizes it as easily as he would his own children. His evidence as to the identity is unequivocal. Two or three hundred pair were made at a time prior to the year 1902, when he left the employ of the Nelson Manufacturing Company. The witness does not tell to whom, if any one, the pipes were sold; nor does he swear that they were ever used in the trade. The witness Lawnin positively identifies the coupling as like one made by Nelson Company more than eight years prior to date of deposition, i. e., January, 1908. They were made up in lots of 300 or 400 pairs, both smooth and threaded at the end. He confirms Harkins’ statement that he (Harkins) invented the pipe. To the best of his recollection, Harkins left in May, 1902. He further testifies that pipe was made and sold by Nelson Company for several years before Harkins left; that Nelson Manufacturing Company have made 500 pairs or more of these coupling-pipes per year for the last 8 years; that about 100 pairs were sold to appellant during the last 3 years. Staab entered the employ of the Nelson Company in September, 1902. He saw pipes like the exhibit at the Nelson Company’s place in about a month or six weeks after he began work there. He says he had one order for 100 or 200 pairs, and . that the company has been making them right along; that they were made with or without threaded ends as desired; that he began to [618]*618make them without threads in about 1904 or 1905; that those sold were ordered through the St. Louis store. Walton identifies the exhibit as a coupling-pipe which the Nelson Manufacturing Company had been making for seven years prior to January, 1908. He claims to have manufactured the first one made, under the witness Harkins’ direction; that they have been made up by the Nelson Company continuously since 1900 or before, generally in sets of about 100. He also is unable to give names of purchasers, as the orders came through the St. Louis house; and that the Nelson Company first began to manufacture,the device of the exhibit with a second coupling-nut for use on the longer arm somewhere about 1903 or 1904.
' There is no question but that these four witnesses identify the exhibit as one made by the Nelson Company, though two of them, at least, are somewhat mixed as to time. Their testimony so far as it goes is emphatic. But it suggests a lack of endeavor to procure the best evidence. It is a well-established rule of law that prior use, in order to show anticipation of a patent, must be “proven by evidence so cogent as to leave no reasonable doubt in the mind of the court that the transaction occurred substantially as stated.” Deering v. Winona, 155 U. S. 286, 15 Sup. Ct. 118, 39 L. Ed. 153.
In Barbed Wire Patent, 143 U. S. 275, 12 Sup. Ct. 443, 450, 36 L. Ed. 154, the court, speaking of oral testimony with reference to anticipations, says:
.“In view of the unsatisfactory character of such testimony, arising from the forgetfulness of witnesses. * * * courts have not only imposed upon defendants the burden of proving such devices, but have required that the proof shall be clear, satisfactory, and beyond a reasonable doubt.”
Authorities to the same effect might be multiplied indefinitely. A reading of the testimony of these witnesses raises the inquiry: What has become of the great number of these pipes which they say were manufactured? Were they used at all, and, if so, where? It is not at all probable in the ordinary course of things that they have all disappeared. No attempt is made on the part of appellant to show •that they were in actual use, or had been placed in the hands of the public. No original coupling-pipe made by the Nelson Manufacturing Company at the time of the alleged prior use is produced. The witnesses are shown an alleged duplicate of the original pipe and depose from memory alone that it is the same device. However clearly and emphatically the testimony of such witnesses may be given, yet if it may reasonably be deduced from all the record that other and conclusive evidence might have been obtained, the existence of which is not negatived, nor its absence accounted for, the court is forced to the conclusion that it is purposely withheld. There is no excuse for the failure to follow up the coupling-pipes alleged to have been made by the Nelson Manufacturing Company.
It is conceded that prior use must be established beyond a reasonable doubt. There must always be doubt in the mind of the court, when no effort is made to produce in evidence facts, the existence of which may reasonably be inferred from the record in the absence of any excuse for the omission. In such case the prior use cannot [619]*619be said to be. established beyond a reasonable doubt. We therefore conclude that the prior use contended for, as shown in this Nelson pipe, is not made out with the certainty required in such case.
In view of the prior art, it is evident that appellee’s device is very close to the line which separates invention from mere mechanical skill.
It would not he invention to remove a device from the washtub art (if it be an art) and adapt it to the faucet and service-pipe art (if such an art there he). It will he observed that Glauber’s device is not by any of his claims limited to use with faucet-spuds and service-pipes. This meaning is by the witness and counsel deduced, first, from the statement that the patent applies to open plumbing, and, second, by the expert witness McIntyre (page 209, Kecord), from the fact that he did not know of any other use' to which the nickel-plating, etc., of the invention could be put. That open plumbing is a matter of whim or taste and may be and is applied to any indoor work, as, for instance, tank connections in water-closets, drain-pipes from basins and tubs, etc., would seem to he a matter of common observation, which may be considered by the court and given as much evidentiary force as the expert’s declaration based on his want of knowledge.
The patent to Smith contains in its claim and specification, as above quoted, language which is strangely like that of Glauber. It is elbow-shaped, and of the same cross-section from end to end. It calls for flanged and unflanged ends, coupling-nuts, telescoping, or slip-joints, universal application, to be made fry any workman, the doing- away with cutting and measuring and soldering, and winds up with Glauber’s statement, in substance, that “this system also overcomes the danger of becoming loosened and water escaping.” it lacks the tapering gasket because it could not be used in that connection, but it can hardly be assumed that, were it used in connection with a tapering seat, any one. skilled in the art would not know enough to use a tapering gasket.
The devices of the Schamp patent, the Tyden patent, the Cornelius patent, the Murdock patent, and others, shown in evidence, are termed by appellees as tail-pieces. Many of them, however, lack only the idea of spanning the distance from a faucet to a service-pipe.
The device of the Tyden patent, though called a tail-piece by ap-pellee, is termed by the patentee “the customary form of joint.” It shows, as far as it goes, a section of Glauber’s unitary pipe. Thus, the idea of a one-piece or unitary coupling is not new, but is abun-dantly accomplished in the prior art to the extent the devices involved required it. Was it invention on Glauber’s paid to substitute the tapering gasket of the prior art for Tyden’s washer, and then extend the lower limit of his coupling into slip-joint or solder connection with the service-pipe?
Undoubtedly, the device of the patent as described therein is a most convenient contribution to the householder; it practically claims to make every man his own plumber. Whether as a separate unitary article of commerce, or in combination with the two pipes to be [620]*620coupled thereby, it places in the hands of the user all the parts required to make the connection. That was never done before, so far as the record shows. It was new and seems to have accomplished all that is claimed for it when made in accordance with the terms read into the patent by the appellee and his experts. But, as above stated, this may not be done. So far as appears from the patent, we have to do with a coupling-joint which cannot be assumed to be made with off-size, thin-walled, or annealed pipe — one which cannot readily be bent in order to secure adjustment to spuds and service-pipes. _ For the purposes of this proceeding, it is just a plain piece of ordinary pipe, one arm of which will telescope into the service-pipe, while the other arm is constructed so as to extend to and fit into the tapering seat of the faucet or other spud. It is evident that the size of the service-pipe and the location and angle of the spud-arm must be first determined before the coupling-pipe can be deemed a complete unitary article of trade adapted to all cases. This can only be arrived at by actual measurement, or by the fact, if it be such, that in ordinary plumbing work these conditions are uniform and well known. Thus construed, the question of patentable novelty is a very close one, as above stated. In view, however, of the presumption arising from the grant, the further fact that the prior art fails to disclose any device showing the same combination, the further fact that its utility is süch that it has come into demand, together with the rule of law which requires that doubt be solved in favor of the validity of the patent, it is our opinion that the patent discloses patentable novelty, though of a low order, 'and it is therefore sustained.
There seems to be no doubt of the fact that defendant’s device infringes claims 3 and 4. Its coupling-pipe is not of the same cross-section from end to end, and therefore does not come within the language of claims 1 and 2. The fact that defendant’s coupling is constructed of off-sized pipe having a thin wall, and being readily adjustable by bending, especially when annealed, does not in our judgment relieve it from the charge of infringement. The idea is, in substance, identical, and all these features can, at most, be held to be nothing more than improvements.
The decree of the' Circuit Court is therefore affirmed.