Opinion
Patent Appeal No. 7053.
June 11, 1964.
Norman E. Schrader, Rochester, N.Y., W. Houston Kenyon, Jr., New York City (Richard K. Parsell, Kenyon Kenyon, New York City, of counsel), for appellant.
A. Russinoff, Princeton, N.J., for appellee.
Before WORLEY, Chief Judge, and RICH, MARTIN, SMITH and ALMOND, Judges.
Walkup, junior party in Interference No. 89,802, appeals from the decision of the Board of Patent Interferences awarding priority of invention of the eight counts therein to Greig, senior party. Walkup states here that he seeks reversal of the board's decision as to counts 1 to 7, but not count 8. Accordingly the appeal is dismissed as to that count.
Greig is involved on the basis of his patent No. 2,811,465, having a filing date of April 30, 1952, and assigned to Radio Corporation of America, hereafter RCA. The counts correspond to claims of that patent copied in Walkup application Serial No. 401,811, filed January 4, 1954, and owned by the Xerox Corporation, which was formerly the Haloid Company, hereafter Haloid.
The name Haloid applied at the time of the events described in the testimony.
Greig took no testimony to prove conception and reduction to practice prior to his filing date and thus is restricted to that date. Walkup took testimony and relied on certain activities which took place between August 25 and September 5, 1950, as proving conception and actual reduction to practice.
The counts are directed to a device for use in the reproduction of visual images by the process of xerography. Testimony for Walkup shows that xerography involves the use of a base plate of conductive material coated with a thin layer of photoconductive material which is relatively insulating in darkness and relatively conductive when illuminated. The entire exposed area of the photoconductive coating is first given a uniform electrical charge in the darkness, which charge remains as static electricity. Then a light image of the item to be reproduced is made to fall on the surface of the photoconductive layer with the result that the illuminated areas of the layer become relatively conductive and the charges on those areas pass through the layer to the underlying base plate and discharge to ground. The non-illuminated areas continue to be nonconductive, and the charges there remain in place on the surface of the photoconductive layer. There is thus created on the photoconductive layer a pattern of electrostatic charges corresponding to the applied light image, which pattern is referred to in the counts as a "latent electrostatic charge image."
The latent electrostatic charge image is developed by turning it into a material and visible image. To accomplish that, a finely-divided pigmented electroscopic powder of opposite polarity to the charges on the photoconductive surface is transported into the field of attraction of those charges. The particles, called toner, are attracted to and held on only the charged area of the plate. Subsequently, the toner image is transferred to another surface, as to paper. In another form of xerography, the toner is not applied to the plate bearing the latent image, but is applied directly to a paper web placed on the plate after its exposure, the toner being held on the paper by the electrostatic charges on the plate acting through the paper.
The counts relate to a development device for conveying toner particles into the field of attraction of the latent image to develop a material image from the latent image. In general, the device embodies a cylindrical member having a surface which is capable of retaining toner means received from supply means and is rotated into contact with the image bearing surface. The construction is brought out more specifically by counts 1 and 7 which are representative and read:
"1. A device for applying electroscopic developer material to a record receiving medium bearing a latent electrostatic charge image to be developed, said device comprising a cylindrical member presenting a peripheral surface capable of retaining a quantity of electroscopic material thereon, means for uniformly distributing a quantity of electroscopic developer material on said surface, and means for bringing said surface into contact with the surface of the record receiving medium.
"7. A device for applying electroscopic developer powder to a record receiving medium bearing a latent electrostatic charge image to be developed, said device comprising a cylindrical roller member having a surface provided with a multiplicity of small pits substantially uniformly distributed about said surface, means for applying a quantity of electroscopic developer powder to said roller, said powder being received and held by an electrostatic attractive force in said pits, and means for passing the record receiving medium into intimate rolling contact with said roller member whereby to transfer, through electrostatic attraction greater than said first mentioned attractive force, portions of said developer powder from said roller onto selected areas of said record receiving member corresponding to said charge image."
In 1950, Battelle Memorial Institute was conducting research on xerography under the sponsorship of Haloid. The work was carried on in the Graphic Arts Division of Battelle under the direction of one William T. Reid, and Walkup was Assistant Supervisor of that Division in charge of the work done on xerography.
Prior to August and September of 1950, Walkup had been active in work on a technique of xerographic development known as "cascade." He describes cascade development as employing a mixture of small dust-like particles of toner, which it is desired to deposit in charged image areas, and larger bead or marble-like particles called "carrier," which are predominately of spherical shape. The mixture tumbles or slides across a sloping image-bearing plate under the influence of gravity. The particles and toner have such properties, described as "tribo-electrical" relationship, that the toner particles will have a polarity causing them to detach themselves from the carrier particles in charged portions of the image-bearing articles and be attracted to the photoconductive surface.
For conception of the different type of development of the invention in issue, referred to as brush development, Walkup relies on a June 14, 1950 entry in his laboratory notebook which reads:
"It has been known an [sic] used that a granular material can be used as a vehicle for developer powder both in the development and the cleaning of a xerographic plate. If a lot of developer powder is mixed with the larger granular material then the resulting mixture is rich in powder and if it is cascaded over a plate carrying an electrostatic image some of its powder will be given to the image and the image `developed'. If, on the other hand, there is little or no powder mixed with the granular material it will be poor or "lean" in powder. If the granular material is cascaded over the plate it will take powder off of the plate. This removal of the powder from the plate is used in cleaning the residual image off of the plate.
"The basically new idea suggested here consists in substituting fine fibers, as of a brush or fur, for the granular material. In this capacity, the fibers of the brush should work in a manner similar to action of the granular cleaner. Thus a soft brush should clean a plate, and a brush loaded with powder should `develop' a plate. For either use the brush or fiber element may take a number of physical forms of which a cylinder with the fibers extending out approximately radially is typical and perhaps desirable. Departures from the radial direction may be desirable so that the fiber touches the plate as a line rather than as a point.
* * * * * *
"In using the brush as a developer powder will have to be added to the brush as it is used up. A simple means of placing powder on the brush might be to have the brush pass over and rub against a mass of developer powder."
In August 1950, one Carlton was working on an assignment by Walkup to build and test apparatus for determining the feasibility of using a rotary brush for cleaning xerographic plates which had become dirty in use. Carlton interrupted that work, from August 25 to September 5, 1950, to build and test the brush type development machine relied on here by Walkup as a reduction to practice. The development machine was made by modifying the brush-type cleaning machine, using the same brush comprising fur wrapped around a wooden cylinder.
The most nearly contemporaneous record of the August-September experiments as noted by the board, is that set out in an entry made by Carlton in his notebook under date of October 13, 1950. There Carlton illustrates a rotatably-mounted cylindrical brush underlying an advancing plate and contacting the under side of the plate with movement opposite to the direction of the plate. The portion of the brush diametrically opposite the plate contacts the upper surface of a body of material in a container with an open top, which container is labeled "Powder." Carlton's description of that apparatus and its operation follows:
"* * * a preliminary trial of fur development was made because it may be more practical to use with a continuous xerographic process, and possibly it may be simpler for all xerographic development. Good images were developed with the fur but they were not as dense as those developed by the cascade method.
* * * * * *
"The fur is mounted on a wooden cylinder which rotates slowly while the plate passes over it. The fur speed is variable from 110 RPM to 440 RPM. The brush is 3 in. in diameter. The plate is pushed through by hand. The distance between the plate and the brush is determined by the height of guide bars above the fur. This distance is changed by placing spacers under the guide bars. The wooden cylinder is electrically insulated from ground and therefore it is possible to apply a potential to the fur. The beater at the right of the cylinder rotates eight times as fast as the fur and its tips have a linear speed twice that of the fur. Its purpose is to reverse the lay of the fur so the fur goes to the xerographic plate in the direction of rotation. Rubbing the plate turns the fur over and electrostatic attraction pulls the developer powder out of the fur. The developer is added to the fur by brushing it in a pan of powder under the fur. This is not a satisfactory method of adding powder to the brush but no other simple way of adding powder was available.
"The plate was passed over the fur at several speeds. At high speeds the image was very light; at low speeds the image was darker. Very slow plate travel causes slightly darker images and a faint background. A plate travel of two inches per second is satisfactory.
"The brush speed should be 120 r.p.m. Slower [sic] speed causes powder to be thrown from the fur by centrifugal force. Some of this powder lands on the plate and forms a dark background.
"The selenium surface of the xerographic plate should be mounted about 5/16 inch above the surface of the wood roller. If mounted higher, light images are formed and the letters are thinner than on the original. If mounted lower, the image is light and there is background. However, the entire black area of the original reproduces black.
"A potential applied to the fur acts the same as biasing the plate during cascade development. A positive potential will prevent development and a negative potential will cause excessive background.
"A new rabbits [sic] fur and a used one were tried. Originally the fur came from the same pelt. The old fur had cleaned 11- by 14-inch plate 1500 times. The new fur developed a better image than the old fur. This may mean that a long haired fur will work better than a short haired fur.
"Development with fur will be practical for a Xerox Copier if it is possible to develop a denser image."
On January 12, 1951, Battelle submitted to its sponsor, Haloid, a monthly report on work conducted for Haloid in Battelle's laboratories during September 1950. The description of the work in that report is similar to Carlton's notebook description. A sketch, reproduced in part below, is included in the report as a representation of the equipment used during the tests:
A small section of copy is included in the report as a sample of the xerographic copy produced by Carlton.
The record does not show any significant activity with respect to brush development by Haloid or Battelle between the time of Carlton's experiments in 1950 and the preparation of the Walkup application, filed in January 1954, even though both organizations continued rather extensive work on other aspects of xerography during that period. The major activity appears to be Haloid's suggesting a form of brush developing equipment to IBM (International Business Machines), a licensee of Haloid, in a letter written in June of 1953. However, the brush developer described in that letter differed from Carlton's experimental model in several respects, the principal one being that the brush is in the form of an endless belt rather than a rotary cylinder. Also, the toner was fed by gravity from a hopper and the brush moved in the same direction as the element carrying the latent image.
The board pointed out that Walkup has made no claim to diligence coupling the filing of his application with conception before Greig's filing date. It concluded that the controversy must be determined on "the single issue" of reduction to practice on the basis of the experiments performed between August 25 and September 5, 1950. Walkup refers to such statement of the board, quoting it in part, without questioning its correctness. We think the board's statement of the issue is correct.
It was the board's conclusion that Walkup failed to prove his case by the necessary preponderance of the evidence. In arriving at that decision, it gave consideration to Carlton's testimony, the notebook description of Carlton's work and the report to Haloid on that work along with the comments of other witnesses, and it referred to the conduct of Walkup, Battelle and Haloid subsequent to that work. Such consideration led the board to conclude that Walkup had not provided satisfactory proof of a reduction to practice on the basis of Carlton's experiments.
Consideration of the record in light of the arguments of the parties leads us to the conclusion that Walkup has not met his burden of proving reduction to practice by a preponderance of the evidence.
In the first place, the entries in Carlton's notebook do not demonstrate that the experiments described therein met with such success as to constitute a reduction to practice. Thus, the operation is described as a "preliminary test" of fur development and the extent of the work indicates it was no more than that. The entire project of adapting the brush cleaning machinery and testing apparently amounted to only a short interruption, involving no more than 43 hours of Carlton's time, to the brush cleaning project. Carlton's testimony indicates he made from 20 to 100 copies on the machine. In contrast, he testified that, in making a plate life check with the fur cleaning machine to determine how the xerographic plate would stand up during brush cleaning, he had conducted 1500 operations and "apparently made quite a few runs prior to that, also." That Carlton was not without reservations whether his experiments demonstrated the practicability of brush development, is indicated by the following statements:
"This is not a satisfactory method of adding powder to the brush but no other simple way of adding powder was available.
* * * * * *
"Development with fur will be practical for a xerox copier if it is possible to develop a denser image." (Emphasis supplied)
The Battelle report to Haloid likewise fails to substantiate Walkup's case for reduction to practice. In the first place, the report concludes with the following comment:
"A number of improvements will be necessary before brush development is practical. However, the results of these preliminary experiments are promising. Additional work should be done to improve the results obtained and to devise better means of feeding powder to the brush."
Further, the board characterized the sample of xerographic copy including in the report as "pale, partial and erratic." Walkup disputes that evaluation, stating that the sample is clear and completely legible and "differs only from the adjacent mimeographed material [the mimeographed textual matter on the sheet to which the sample is pasted] in that the black areas are lighter and in some places are not continuous, though the discontinuities are so slight that legibility is not impaired." He urges that the samples prove that Carlton's device "actually performed the function for which it was designed."
Two copies of the report, and thus two sample sections of copy, are in the record as exhibits.
From our own inspection we agree that the samples are legible, but it seems to us that their over all quality was correctly evaluated by the board. Walkup's own argument concedes the existence of lightness and discontinuities in the black areas, as is plainly the case. Additionally, the last sentence of Carlton's description of the experiments that produced the samples, made nearly contemporaneously with those experiments, shows that he himself considered a "denser image" necessary for a "practical device." We see no clear error in the factual evaluation of the exhibits below.
Also Walkup's witness Greaves stated that the samples were undoubtedly sections cut from larger samples. So far as the record is concerned, the two samples before us may have been cut from a single larger copy of the "perhaps anywhere from 20 to 100" copies which Carlton testified he made, thus leaving it open to question whether the other copies included portions of even equal quality. Additionally, it is not clear whether any of the copies produced by the experiments were uniformly of even equal quality to the small samples even though it seems apparent that the function for which the device in issue was designed was to produce satisfactory reproductions of the entire sheet or copy rather than of just a subsequently selected portion thereof.
Walkup urges that the officers of Haloid who witnessed the operation of Carlton's machine "recognized a significant new achievement" and points to testimony of Dessauer, Executive Vice President in charge of Research and Development, and Clark, Manager of Haloid's Physics Research Department. The most favorable comment of Clark referred to seems to be the following:
"* * * I remember it quite well because it was a pleasant surprise, I think, to them and they were eager to show it to us that this worked so well. It was definitely an interesting new development. I had known that in the past they had been able to develop images with hand brushes, but the fact that this worked was very much an achievement and was kind of an event, and it also opened our eyes to some mechanisms of development."
Clark also said the demonstration of the machine was "kind of an occasion because this was a pet idea of Lewis Walkup's and he was delighted to perform." The board regarded that comment as "very strange when it is recalled that Walkup, himself, was not present at any of the tests."
Dessauer commented as follows:
"Most of the results looked promising, and it was clear that this was a very feasible method of developing xerographic plates."
Those statements, made in testimony taken during interference proceedings in January 1961, do not appear to be entitled to the same weight as the statements made more nearly contemporaneously with the experiments in 1950 which indicate there was serious doubt at that time that the Carlton experiments demonstrated the brush developer was practical.
Walkup urges that Carlton's tests "established the utility of a brush developer at least for purposes of manual xerographic copy-making in a dark room." He takes the position that statements indicating hope or doubt concerning some additional areas of utility, as in continuous or automatic xerographic machinery are not evidence of a lack of "conviction of success" as to manual copying. That argument is not supported by the record since the statements indicating doubt of practical ulity referred to hereinabove are not limited to automatic machines. Rather, the statements indicate doubt or uncertainty of the practicality of the machine tested, a manual machine. Likewise, any doubt as to the quality of the copies obviously is chargeable directly to the manual machine that produced them.
Another indication that the device tested was not satisfactory lies in Carlton's testimony. Thus, he described the "fur brush method" as "much dirtier than the cascade operation." stating also:
"Well, when the brush was dipping some toner was also being put into the air. So, I got pretty black during those operations."
Walkup asserts that the forty months delay after Carlton's experiments before filing the Walkup application is without significance and cites Schnick v. Fenn, 277 F.2d 935, 47 CCPA 1174, for the proposition that once reduction to practice has been found, the question of abandoned experiment no longer exists. While the latter proposition is obviously correct it is not applicable here because the evidence in the record concerning Carlton's experiments at best leave it extremely doubtful that those experiments constituted a reduction to practice and compel a finding that Walkup has not sustained his burden of proof. Moreover, it is well established that subsequent conduct with respect to the invention is properly to be considered where there is doubt that the activities relied upon constitute a reduction to practice. Conner v. Joris, 241 F.2d 944, 44 CCPA 772.
When Haloid did disclose brush development to its licensee in June 1953, the apparatus, as shown in Walkup Exhibit 14, embodied an endless belt rather than the rotating cylinder of the Carlton 1950 experiments, and the fur was moved in the same direction as the plate being developed instead of in the opposite direction as in the experimental apparatus. That apparatus is also disclosed in the Walkup application in which the disclosed structure most like the experimental device differs from that device by providing rotation of the cylindrical brush in the opposite direction. We agree with the board that those circumstances indicate that it may well have been experimental work done after 1950, and not disclosed by Walkup's evidence, that convinced Haloid that brush development was worth promoting or filing a patent application on. In any event, the activities subsequent to Carlton's experiments do nothing to relieve the serious doubt inherent in the record that such experiments satisfied Walkup or the others involved of the practicality of the subject matter here in issue.
We are familiar with Tansel v. Higonnet et al., 215 F.2d 457, 42 CCPA 732; Leichsenring, Jr. v. Freeman, 103 F.2d 378, 26 CCPA 1153; Mergenthaler Linotype Co. v. Press Pub. Co., C.C., 57 F. 502; In re Hartop, 311 F.2d 249, 50 CCPA 780; Hildreth v. Mastoras, 257 U.S. 27, 42 S.Ct. 20, 66 L.Ed. 112, in addition to Schnick v. Fenn, relied on by appellant, some of which have nothing whatsoever to do with the issue of reduction to practice. We repeat here what has been said so many, many times: each case must be decided on its own facts.
The decision is affirmed.
Affirmed.
The majority opinion seems to depart from the established legal principles regarding what is necessary for an inventor to prove in order to establish a reduction to practice. In the majority opinion this departure is somewhat obscured by and confused with the question of the burden of proof which must be met by a junior party. Here certain facts are well established by a clear preponderance of the evidence. The record establishes that during the period between August 25 and September 5, 1950, certain tests were run by Carlton under Walkup's direction. There is no question of proof as to what these tests demonstrated. Thus, as I see it, the only issue here is whether as a matter of law these tests established a reduction to practice.
This court recently reviewed the legal requirements as to what is required to establish a reduction to practice and in Schnick v. Fenn, 277 F.2d 935, 939, 47 CCPA 1174, 1180-1181, said:
"* * * we do not believe that an actual reduction to practice necessarily requires the making of a commercially acceptable embodiment * * * Tansel v. Higonnet [et al.], 215 F.2d 457, 42 CCPA 732 * *, if the testing of it shows that it will perform its intended function. The amount of testing necessary depends upon the facts of the particular case, Lustig v. Legat, 154 F.2d 680, 33 CCPA 991 * * *.
* * * * * *
"* * * we know of no precedent which requires that proven methods of mass producing a device must exist before a reduction to practice is established. On the contrary, we believe the law to be that reduction to practice is proved when there is corroborated testimony that the device actually performed the function for which it was designed, * * *."
Here the invention in issue relates to an improved device for applying a known electroscopic developer powder to a known record receiving medium which carries a latent electrostatic charge image. The proofs are abundantly clear that such a device was constructed and tested under appellant's direction. The majority finds, however, that Carlton's notebook entries "do not demonstrate that the experiments described therein met with such success as to constitute a reduction to practice."
The majority opinion also states:
"* * * That Carlton was not without reservations whether his experiments demonstrated the practicability of brush development, is indicated by the following statements:
"This is not a satisfactory method of adding powder to the brush but no other simple way of adding powder was available.
* * * * * *
"Development with fur will be practical for a xerox copier if it is possible to develop a denser image." [Emphasis added by majority.]
In discussing the Battelle report to Haloid, the majority finds that it "fails to substantiate Walkup's case for reduction to practice," apparently because of the concluding comment in the report that:
"A number of improvements will be necessary before brush development is practical. However, the results of these preliminary experiments are promising. Additional work should be done to improve the results obtained and to devise better means of feeding powder to the brush."
The majority also adopts the view of the board as to the quality of the samples which it characterized as being "pale, partial and erratic." There is no dispute concerning the fact as to the quality of the samples. The issue, however, is whether the admittedly imperfect samples are evidence of a reduction to practice in which a fur covered cylindrical roller actually transferred the developer powder to the record receiving medium and whether as a result of this transfer the latent electrostatic charge image was in fact developed.
Here the invention at issue involves the discovery of what appears to be a previously unknown electrostatic phenomenon which was embodied in the device tested. It is my opinion as a matter of law that it is without significance that the machine "operated with something less than commercial printing quality." Cf. Tansel v. Higonnet et al., 215 F.2d 457, 42 CCPA 732.
In the Tansel case this court characterized the invention as "a fundamental advance in the printing art whereby text or copy * * * is composed direct on film." ( 215 F.2d at 458, 42 CCPA at 734.) Higonnet rested upon his record date, a French filing on June 2, 1945. Tansel's rather crude machine was built beginning in 1943 and final successful testing thereof culminated in January 1946. This court held that Tansel had reduced the invention to practice in and by his machine, Exhibit 59, despite the fact that the initial photoelectric controls were changed midway of the test to slower and less efficient mechanical controls to avoid "contact bounce." This court held
"The capacity to perform in the manner intended for the machine in issue, though only in a crude way, is sufficient to discharge the inventor's burden of proof for actual reduction to practice. * * *" ( 215 F.2d p. 463, 42 CCPA p. 740)
The positions previously taken by this court in Schnick v. Fenn, supra, and in the Tansel case have recognized the principle stated in Section 129 of Robinson, Treatise on the Law of Patents, that:
"Reduction to Practice does not Require Mechanical Perfection or Incapability of Improvement.
"* * * it is not necessary that the art or article should be mechanically perfect. Mechanical perfection is the achievement of the artisan rather than the inventor, and does nothing to develop or to illustrate the idea of the invention. Possibilities of greater excellence in shape, location, arrangement, material, or adjustment do not affect the fact that the inventor has produced a practically operative means, and all such possibilities are legally embraced in what the inventor already has accomplished. Nor it is necessary that the invention, as a means, should be incapable of further improvement by the exercise of additional inventive skill * * *."
In considering whether the tests run on Carlton's laboratory machine amounted to a reduction to practice of the invention at issue, the evidence clearly shows that only a single scientific fact needed to be established by these tests in order to convince men skilled in the art that the machine "actually performed the function for which it was designed." This fact was whether development would occur with reasonable clarity when the device was used in the manual making of prints. All that remained to be determined, beginning on August 25, 1950, was whether the same brush which had been used previously for cleaning the plates when supplied with powder, and rotated in contact with the same or similar xerographic plates bearing a latent electrostatic charge image, would transfer the powder to the plate so as to develop its latent image with reasonable clarity. This is the fact which I would find was clearly demonstrated in the Carlton tests of August 25-September 5, 1950.
The board's characterization of the copy matter actually produced in Carlton's tests as "pale, partial and erratic," overlooks the fact that the copy matter actually is clean, completely legible, affords good contrast between white and black areas, and is almost completely free from background smudging. I would therefore accept these samples as proof that Carlton's device "actually performed the function for which it was designed" within the principle stated in Schnick v. Fenn, supra.
The indication of a standard of excellence by which to evaluate the results of a proven reduction to practice which runs through the majority opinion will, it seems to me, but further confuse the legal issue as to what constitutes sufficient proof of a reduction to practice. The premise of the majority seems to me to ignore what has long been the rule as stated by this court in Leichsenring, Jr. v. Freeman, 103 F.2d 378, 382, 26 CCPA 1153, 1158, that it was sufficient if:
"It was clearly demonstrated * * that appellant's device involved the generic principle, and that it worked for a sufficient period of time to show that the application of the principle solved the problem. * *"
Certainly, we should not here ignore the sound principle stated in Mergenthaler Linotype Co. v. Press Pub. Co., 57 F. 502, 506 (C.C.S.D.N.Y. 1893), that
"* * * It would certainly be a novel doctrine to deny to an inventor the fruits of a broad invention because the machine which first embodied it was rudimentary in character and failed to do as good work as improved machines made subsequently. None of the great inventions could survive such a test. Ten years after the invention of Howe, the machine first made by him would hardly have satisfied the least exacting sewing woman. The Dodds and Stephenson locomotive would, only a short time after its construction, have been discarded as behind the age even by the savages of Tasmania. The telephone of Bell is not the perfected telephone of commerce, the Morse telegraph is looked upon to-day as an interesting antique. And yet, it would be an unheard-of proposition to withhold from these illustrious men the credit they deserve because their machines were crude at first and were improved afterwards. * * *"
I would, therefore, apply to the proofs here the legal principle stated by the Supreme Court in Hildreth v. Mastoras, 257 U.S. 27, 34, 42 S.Ct. 20, 23, 66 L. Ed. 112 (1921):
"* * * It is not necessary, in order to sustain a generic patent, to show that the device is a commercial success. The machine patented may be imperfect in its operation; but if it embodies the generic principle and works, that is, if it actually and mechanically performs, though only in a crude way, the important function by which it makes the substantial change claimed for it in the art, it is enough. * * *"
The foregoing observations would lead me to a reversal of the appealed decision.