This is an equity suit for the infringement of the ninth claim of letters patent granted to J. H. Schulte, for improvements in type-writing machines and relates particularly to that part of the machine known as the tabulating' attachment, which is described in connection with a “caligraph” machine. It may, however, readily be adapted for use in other well known writing machines which employ a sliding carriage in which is mounted a roller for supporting the paper to be printed upon.
The patentee’s main objects were, first, to produce a machine in which the means for disengaging the carriage from the step-by-step spacing mechanism are connected thereto and operated automatically thereby. Second, to provide a stop or abutment for arresting the carriage at a predetermined point after being released from the step-by-step spacing mechanism. The invention consists in combining with the step-by-step spacing mechanism means for releasing- the carriage therefrom, the carriage being provided with a stop, to coact with the spacing mechanism so that it may be detained automatically after disengagement from the latter.
*577The specification contains the following statement of the prior .art:
“This movement of the carriage is generally "éffeeted by a step-by-step action through the., escapement spacing mechanism of, the machine, and is necessarily slow; but in some machines there is pivoted in connection with the racks a device which may be reached (though not'"Without inconvenience) and operated to throw out of engagement the stop-by-step feeding mechanism and, permit the carriage to travel quickly toward the left of the machine.”
This is a distinct recognition of a fact which is established by the5 proof that, prior to Schulte, there were machines provided with a de-vice which enabled the operator to throw out of engagement the step-by-step feed mechanism, thus permitting the carriage to travel quickly towards the left to be stopped by his hand at the desired point.
The Claim Involved.
The ninth claim, which is the only one in controversy, is as follows:
“In a type-writing machine, the combination, with the carriage, of an adjustable column-stop, a dog to engage the same, and a finger-piece or key to actuate said dog, substantially as and for the purpose set forth.”
The claim is for a combination, in a type-writing machine, containing four elements as follows:
First. The carriage.
Second. An adjustable column-stop.
Third. A dog to engage the column-stop.
Fourth. A key to actuate the dog.
The last three—the column-stop, the dog and the key, are the features relied on by Schulte to form the tabulating attachment.
Contention of Parties.
The complainant insists that the claim covers a pioneer invention and is entitled to a liberal construction and a wide range of equivalents. The defendant contends that the claim is anticipated and that, In any. view, it must be limited to the mechanism described and shown and, as so construed, it is not infringed.
Date of 'Schulte’s Invention.
Schulte’s application was verified July 15, 1886, and was filed July 22, 1886, but the complainant contends that the invention -was actually made about the middle of May, 1883. We are of the opinion that the record fails to establish any earlier date for the invention than May, 1886. The so-called proof to the contrary, if entitled to any consideration whatever, wholly fails to establish the date as early as 1883 by the clear, convincing proof required in suca circumstances.
Schulte was sworn in the action brought by the Wagner Company against Wyckoff and others, argued at the same time as the present action and by stipulation the testimony in that case may be considered here with the same force and effect as if taken in this • suit. He was asked:
“During the time that you worked for the American Writing Machine Company, did you or not, devise or make any device or attachment adapted for the •use upon a .typewriter, for column spacing or tabulating work?”
*578■ The answer was':
“íes, sir, in the. middle of May, 1886.”
We are unable to find any intimation of an earlier date except the testimony found in the record of interference before the Commissioner of Patents between Schulte and U. Sherman McCormack, introduced by the defendant without any stipulation that the' testimony found therein shall be regarded as taken in this action or considered as proof of the facts herein issue. In the interference proceeding Schulte swore that he conceived the invention in the middle of May,’ 1883, at Corry, Pa.' Assuming that we are permitted to consider this testimony it is enough to say that it is vague, uncertain, at variance with his later statements arid absolutely insufficient to carry the date of his invention to 1883.
Yost Anticipation.
The defense of anticipation is based principally upon letters patent to G.’W. N. Yost, granted April 23, 1889, upon an application filed June 28, 1880, and the model accompanying the same, filed September 3,1880. The patent does not disclose the tabulating attachment of the Schulte patent, but the drawings show an adjustable stop and the model does, we think, disclose fully the combination of the ninth claim in controversy unless confined to the identical mechanism described and shown. The model, which at the present time is capable of being operated, shows, in a type-writing machine, the combination with a carriage of an adjustable column-stop, (which is an internally threaded block mounted on a screw rod) a dog to engage the same and a key to actuate the dog. The dog and key are almost the exact counterparts of similar elements in the Schulte combination and operate in the same way.
The complainant has taken the testimony of several witnesses, having more or less intimate knowledge of Yost’s connection with the typewriter art in the early eighties, who testify that they do not remember the Yost model or any similar machine, but all this testimony may be rejected as negligible in view of the indisputable fact, as shown by the filing card attached, that" the model was'received at the Patent Office September 3, 1880. By letter, dated June 29, 1889,, the examiner withdrew the Schulte application from issue in view of the Yost patent and model. Of the latter the examiner says:
“See model, showing ah adjustable column-stop in the path of the dog, when the latter is thrown out of engagement with the rack bar by double movement of the space key. This stop is shown also in Figs. 1 and 2 of drawing.”
' In a subsequent communication (August 5, 188.9) the examiner describes the model as “a full sized operative machine, complete in every respect, except that all the keys and type bars have not been put in, a feature that is immaterial in respect to the point at issue. The machine is more than ,a. model) it is clearly .a machine manufactured with others for purposes of use and sale—a completed machine which can in no sense be considered, an abandoned experiment.” Claim 9 ■was thereupon rejected but was afterwards allowed upon receipt of an affidavit by Yost; verified December 19, 1890, in which he states, in *579substance, that after an examination of letters patent, No. 401,990, he is unable to state why the screw and nut were put into figures 1 and 2: of the drawings, his impression being that said devices had relation “to an adjustable means for ringing the gong at predetermined points/’ and that he has no recollection of inventing a device for doing column work. At the date of this affidavit the complainant was the owner of the Yost patent and of the Schulte application.
The complainant’s expert testified in the Wagner-Wyckoff Case that he had examined the Yost model in the Patent Office and after describing it continues as follows:
“From the foregoing I am compelled 'to understand that 'the construction find mode of opernlion of this Yost patent, model, as explained hy the examiner, was sneli that it was adapted for ‘tabulating purposes/ ”
It is argued that the Yost model is only the skeleton of a type-writing machine, the type bar and finger-key being frail and flimsy. It is quite true that only so much of the typewriter necessary to exhibit the invention is shown, but this is also true of the drawing of the Schulte patent where the inking mechanism, roller, type levers and finger-keys have all been removed. Indeed, this is true.of most of the patents in evidence and is generally necessary where the improvement relates to one feature only of a complicated and composite machine. One who has invented an engineer’s airbrake valve is not required to describe and show his device attached to a full sized locomotive. It is not to be expected that a model made in 1880, when the art was in its infancy, will exhibit all the perfections which a quarter of a century of development has since added. Grant that it was defective in each of the particulars pointed out by the defendant, grant that each feature of the combination has been simplified, strengthened and improved, and yet the fact remains that it was, and is to this day, an effective, operative tabulating attachment. It is, we think, a much clearer exemplification of the invention than anything found in the Schulte patent.
On the 17th of July, 1888, Schulte, in his preliminary statement in the interference proceedingsbetween. him and McCormack, swore “that no model of the invention in issue has been made by him; ■ and the invention has not by him been embodied in a full sized machine.” Yost’s affidavit, which resulted in securing the issue of the Schulte, is entitled to little consideration as proof of the statements contained therein. .I t was evidently procured to put an end to an inconsequential controversy which would in no way benefit Yost even though it resulted in ■defeating the Schulte patent. If Yost had been sworn in this cause and given, similar negative testimony it would have availed nothing in face of the existing verity in metal showing precisely what hte accomplished.
Franz X. Wagner, who for many years had been an inventor and manufacturer of typewriter models and machines, made the Yost model in question with the adjustable stop for' the purpose of stopping the carriage at any desired point. He testifies that Yost gave him the idea and left the details of construction to him. He also says that about the same time, between 1880 and -1882, he made another operative machine for Yost-with' a full key-board and a'tabulating *580attachment like that of the model. Wagner is a German and testified that it was- difficult for him to explain intricate machinery in English. At the time he testified he was in the employ of the defendant as an inventor, receiving a fixed salary. That the Yost model was in .existence as early as ^September 3, 1880, has, as before stated, been established by proof which is incapable of contradiction, and although Wagner’s testimony must be weighed in the light of his. relations with the defendant company we see no reason to doubt the statement that he received his instructions from Yost That he made the model there can be’ no doubt.
Edward M. Johnson, whose acquaintance with Yost began in 1878 or 1879, and who, with his brother Robert and Yost formed the American Writing Machine Company, testified that a number of models and machines were made for Yost at about that time. He remembered two between 1880 and 1881; one a machine for visible writing; the other a machine for tabulating. The latter was an operative machine and was used experimentally by the witness to do writing and tabulating and also by the typewriter operator employed by them at the time. Although the witness, who is a chemist, could not describe the exact details of the tabulating mechanism he knew that the machine “had an attachment that would allow the carriage to be released and stop at a given point that was adjustable.”
Robert W. Johnson testified that the typewriter business in the early 80’s was in its infancy and it occurred to him that if a machine could be devised which would make invoices better than those then in existence it would find a ready market in the commercial world. He constantly suggested the importance of such an improvement to Yost who made a machine designed to accomplish that purpose. This machine “could be moved over an intervening space and stopped at a given.point; which could- be adjusted so that the greatest amount to be made in the invoice—that is, whether the line was at the ten dollar line or one hundred dollar line, or whatever line the main figures—so that it could be adjusted to stop at that line of figures.” The witness describes the machine having the tabulating device in part as follows:'
“On the machine that I saw he had made an adjustable point which could be moved laterally along the carriage so that it could be set at different points: and the dog, as the carriage was carried laterally struck this attachment which was carried on a screw.”
This machine was built by Wagner previous to 1882 and about 1880 or 1881. The witness is shown the Yost patent and recognizes the “A Stop” of figure 1 and of the Yost model as the adjustable screw threaded column-stop which he had previously described. The witness was asked:
“How does the construction as Illustrated In figure 1 of the patent just shown you, in this respect compare with the machines that were actually built as you say about 1880?”
He answered:
. “I cannot be absolutely certain, because It is a long time ago, but; so far as 1 can reinember they had identically the same thing.” '' ,
*581Complete working machines having this attachment were publicly used in the office of the writing machine company for making invoices prior to 1882.
The testimony of another witness, Belcher, is also in strong corroboration of the defendant’s contention that the improvement covered by the ninth claim of the Schulte patent was well known and in actual operation years prior to the earliest date which can be assigned to the Schulte invention.
The complainant’s testimony in reply is negative in character. Persons employed by Yost, and having more or less knowledge of what contributions he was making to the art at the time.in question, testify that the)^ do not remember the Yost model and have no recollection of seeing the tabulating attachment, although several of them recall the fact that Wagner was employed by Yost as a model maker. Of course the fact that these men do not remember seeing a structure which actually existed at the time is immateriality reduced to its las,t analysis. Yost, the inventor, was not called by either party and one of the witnesses testifies that he heard that Yost ivas dead although we have been unable to find direct proof of his death in the record. We are therefore of the opinion that six years prior to the date of Schulte’s invention there were in existence at least two typewriting machines with an operative tabulating attachment which were actually used in making out invoices in the office of the writing machine company. This improvement was probably conceived by Robert W. Johnson; it was given tangible shape and form by Yost, the machines being constructed by Wagner. If the proof of prior invention were confined to the Yost patent alone, or to the oral testimony alone, or to the model alone, it would be probably rejected as insufficient, but the three united present one of the strongest cases of anticipation with which we are familiar. The drawings of the patent concededly show an attachment capable of use as a tabulating device and incapable of any other use, at least the other uses suggested in the Yost affidavit, and elsewhere, are of such a nature as to appeal only to one whose imagination has been abnormally developed at the expense of his reason. The model filed in connection with the specification leaves no doubt as to the meaning of the drawings and is a complete embodiment of the combination of the claim in question broadly considered. It shows a device, which, if attached to a typewriter, is capable of doing tabulation work. The oral testimony supplies the only missing link and leaves no reasonable doubt as to the purpose for which the machines were made.
How is it possible for the court to reject this proof? On what theory can we decline to give it the full weight to which it is entitled? The complainant has failed to discredit or answer the evidence except in a few subordinate and immaterial particulars. It does not dispute the prior existence of the drawings or of the model. It cannot do so. It admits that the model was probably made by. Wagner, but asserts that it was a crude and flimsy structure, incapable of successful operation. No attack is made upon the character of two, at least, of the defendant’s witnesses who testify to the succes.st, *582fill operation of the anticipating machines, for they evidently are gentlemen whose lives are above reproach, but it is asserted that the machines did not work successfully and were subsequently abandoned. That they did not work as successfully as the perfected machines of the- present is, of course, conceded, but the proof does not justify the conclusion that they were abandoned experiments. No one can read the testimony in the light of the Yost model without being convinced that the idea of a tabulating attachment is embodied there as clearly as in anything Schulte has contributed to the art. Neither produced d perfect device capable of competing commercially with the machines of the present, but both had given form to the idea so that the work of the subsequent improver was comparatively easy. ' Both had the idea but Yost had it years before Schulte. The theory that what Yost, or Yost and Johnson, did was experimental and abandoned as a- failure cannot be maintained. If the invention were made it is immaterial how it was made or for what purpose. 'It is only where experiments fall short of the desired result and are abandoned as failures that they are rejected as proof of want of novelty. Waterman v. Thomson, 2 Fish. Pat. Cas. 461, Fed. Cas. No. 17,260; Aiken v. Dolan, 3 Fish. Pat. Cas. 197,203, Fed. Cas. No. 110.
With this evidence before us of the existence of the Yost drawings, model and use in 1880, we are unable tó reach the conclusion that the tabulating attachment was invented by Schulte six years afterwards.
Law Applicable to Model.
It is urged by the complainant that novelty of the combination of the ninth claim cannot be negatived by the Yost model and in support of this proposition we are referred to Walker on Patents, § 61, and tlipiv ritpd.
Cahoon v. Ring, 1 Cliff. 592, Fed. Cas. No. 2,292, was a trial of certain feigned issues of'fact by a jury to determine whether complainant’s patent for an improved seeding machine was valid and infringed. At page 611 Judge Clifford charged the jury:
“You are accordingly instructed to inquire and determine, from the evidence, whether Luce made his alleged invention of the vertical machine public, and if he did not, but had used it for no purpose, except simply for his own private experiments, and if it had been broken up prior to the 14th of May, 1857 (the date of Cahoon’s application) and its materials used for other purposes, and its essential parts lost, and the invention forgotten or abandoned, that such an invention and use would be no obstacle to the taking out of a patent by Cahoon, or those claiming under him, and that the model, or machine, now in evidence, on that state of facts, would not invalidate the Cahoon patent, if he was an original inventor of his improvements, without any knowledge of said machine, .and did not derive any of them from Luce.”
In Stainthorp v. Humiston, 4 Fish. Pat. Cas. 107, Fed. Cas. No. 13,281, the court says:
“The machine of Whitfield never went into practical use. Although, a working model was made in or about the year 1849, and was soon after used for two or three hours in making candles, by way of experiment, the machine may well be considered as but an abandoned experiment.”
" In Johnson v. McCullough, 4 Fish. Pat. Cas. 170, Fed. Cas. No. 7,401, the court said:
*583“I have also examined all the testimony relating to the Shaw model, and find that it does not show more than the making of a model, and not of any practical working machine, which is necessary to overthrow a patent.”
In Stilwell Co. v. Cincinnati Coke Co., 1 Ban. & A. 610, Fed. Cas. No. 13,453, the court said of a model made prior to the invention in question:
“That (the making of a mere model) does not constitute invention, and it can only be used as an item of testimony, reflecting upon the making and using of the wooden heater in 1845 or 1840. Taking it in connection with the testimony on that point, and weighing all the testimony together, I cannot say, that it is of that clear and satisfactory character, as requires me to find that James Armstrong invented the device, as claimed, in 1845 or 1846.”
In Bowers v. Von Schmidt (C. C.) 63 Fed. 572, the court, at page 577, says:
“That models or drawings will not constitute invention, so as to amount to anticipation, may be true, hut models or drawings may constitute invention to avoid anticipation.”
Non constat models and drawings may be used to establish an invention prior to that of the patentee.
These are all the authorities cited by Mr. Walker and we are convinced that they do not sustain the broad contention of the complainant. The law, section 4886 of the Revised Statutes [U. S. Comp. St. 1901, p. 3382], provides that any person may obtain a patent, inter alia, for a machine invented by him “not known or used by others in this country before his invention or discovery thereof.” It is clear, as pointed out by Mr. Walker, that knowledge of a model of a machine is not knowledge of the machine itself any more than knowledge of a model of Brooklyn Bridge is knowledge of that structure. But we think the rule should be restricted to a model pure and simple as the word is understood in common parlance, viz.: a pattern, a cop}', a representation usually upon a reduced scale. The word “model” should not be construed to mean the identical device which is covered by the patent. If this were otherwise a defendant who produces the exact structure of the claims and proves that it was known prior' to the date of the alleged invention is completely answered if the complainant can show that the anticipating structure was filed as a model. In other words, the question is not one of nomenclature but of fact. In the case of a complicated machine a small model incapable of actual use may be filed for the purpose of explaining and illustrating the drawing; that such a model alone would not anticipate is, of course, perfectly clear. .On the other hand, it frequently happens that the applicant files as his model not a pattern or representation of the thing invented by him but the thing itself. Take, for illustration, an application for a patent for a horseshoe nail where one of the nails made by the inventor is filed as -a model, can it be that a subsequent applicant can hold a' patent for that nail or any feature thereof after proof of its prior existence and the knowledge thereof by the public? During the pendency of. an application a model filed in the Patent Office is supposed to be inaccessible to the public and therefore proof of its filing date is not alone proof of public knowledge at that time, but, on the other hand, such knowledge hav*584mg' been shown by extrinsic evidence, the model is not open to the suspicion that it has been altered and, until proof to the contrary is adduced, must be presumed to be in the condition it was in at the daté of filing. As before stated we think the Yost model was something inore than a mere model, for the reason that it is a full operative embodiment of the tabulating mechanism, and for the further reason that Its' existence and purpose is established by evidence independent of its connection with the Patent Office.
■ In view of what has been said we deem it unnecessary to discuss the other prior patents and evidence of anticipation. The idea of stopping the carriage automatically at the tabulating point instead of by the hand of the operator is one that would naturally occur to any intelligent workman skilled in the art and it is not surprising, as the proof shows, that several inventors at nearly the same time undertook to remedy the difficulty. It would, however, extend this opinion beyond all reasonable limits, with no corresponding advantage, were we to attempt a discussion of all of the questions mooted in this enormous record. If the Yost model, patent and machines and the testimony regarding the same do not completely anticipate the claim in question they so limit it to the identical construction described and shown by Schulte that there can be no pretense that the defendant infringes.
Infringement.
. The defendant’s machine is a modern typewriter containing many improvements which were unheard of at the date of the Schulte application. It is described at length by the experts in all its minute details—those' which are involved in the present controversy as well as those which are in no way involved. It suffices to say that in our judgment" the defendant does not infringe the ninth claim even if the evidencé of anticipation were removed from the case. The dog of the claim is marked in the drawings and referred to in the specification as “the dog e.” This dog engages with the racks to form the stép-by-step mechanism used in ordinary writing and when released from the racks by the downward pressure of the key the same dog e stops the carriage at the tabulating point by coming into contact with the column-stop, adjustable or fixed, on the third rack. The defendant’s machine does not include such a dog. The feed dogs of the defendant are feed dogs only and have no part in stopping the carriage by contacting with the column-stop. For this purpose the defendant, provides separate and distinct stops, strongly and firmly mounted not on the carriage but on a rod journaled to the frame work of the machine, the stops being moved into the path of a lug on the carriage, jFus stopping the latter at the desired point. The work of arresting the' carriage when it moves forward under the influence of the spring is thus placed upon heavy, strong stops and not upon the comparatively small and delicate feed dog. The defendant does not use the dog e to stop the carriage' by engagement with a column-stop and therefore does not have the third element of the claim. If the claim be construed tp cdver a dog, which simply engages the column-stop and performs no other .fufiction, the combination is inoperative, for'such a dog could'not'c'drhe in contact'with'the' stop until the feed dog'is th'ro'wn *585out of engagement witli the racks, and no releasing mechanism is provided in such a combination. The court is not permitted to reconstruct the claims of a patent, and the patentee is bound by the claims as he has written them. In the present case the patentee in drafting the ninth claim has chosen to confine himself to á single movable dog," viz.: the feed dog of the step-lyy-step movement. Only one dog is mentioned in the claim. It is a dog to engage the adjustable column-stop and which is itself actuated by the finger piece or key. If we take this single dog to be other than the feed dog, we may actuate it by the key as much as we please and thus move it into position to engage the stop, but nothing will happen because the claim contains no provision for simultaneously releasing the two engaging parts of the step-by-step mechanism. Upon any interpretation of the claim which makes the single dog a dog other than the feed dog, the combination is inoperative. If, however, the dog of this claim be held to be the feed dog as undoubtedly the patentee understood it to be, the very movement of the key piece which actuates it to engage the adjustable column-stop at the same moment of time releases it from the racks, thus disJ engaging the step-by-step mechanism and making the claim operative.
The decree is affirmed.