This suit was originally brought upon two patents, both issued to Robert Conrad, No. 822,723, on June 5, 1906, and No. 838,303, on December 11, '1906. The first is for an article of manufacture, an improved ball-bearing, and the second for the method of assembling its parts. The method patent was withdrawn early in the litigation, and at the same time the charge of infringing the' other-patent was limited to the eighth and ninth claims. This is the fourth suit in this court upon No. 822,723, but the second and third suits (which are not reported) need not be referred to. In the first suit claims 8 and 9 (and one other) were involved; the parties being the present plaintiffs and the Standard Roller-Bearing Company. The case was decided in March, 1910, and is reported in 177 Fed. 435. As stated by Judge Holland, the only issue there was lack of invention:
“The defendant makes the sole defense of lack of invention. Hence the only Uuestion for the court is whether or not the three claims of the first and the one claim of the second (method) patents are valid; that is to say, do they cover patentable subject-matter, and did it involve invention to produce the hearing in view of the prior art?”
The prior art then receives some general consideration, and the •patents to Lechner and to Pettee are specifically referred to; but the present plaintiffs concede that the English patent to Gentry, and the Oldfield patents and prior use, were not then before the court. Neither does it appear from Judge Holland’s opinion (and it is most un*868likely) that he considered the file-wrapper of Conrad’s patent, upon which the present defendants lay much stress. The Gentry and the Oldfield references are now relied upon to invalidate the claims altogether, or at least, in connection with the file-wrapper, to limit them to a solid, unbroken structure that is not infringed by the defendants’ bearing. Admittedly these defenses are open in this suit, and they have been fully discussed.
The point of difference between the parties is exceedingly narrow, for the plaintiffs are obliged to admit that if the notch, or opening, that is a vital feature in the defendants’ bearing, extended 44/10000 of an inch farther than it does, there would be no infringement. If, then, the case for the plaintiffs depends upon sd minute a measurement, which can barely (if at all) be observed by the unaided eye, it is necessary, I think, to determine with precision just what is the scope of the patent in suit. Let us turn, therefore, to Conrad’s own description, as he laid it before the office, in language that varied from time to time, but never lost sight of the solid, unbroken rings that lie at the heart of the patent.
. But a few preliminary observations may be useful. The plaintiffs do not assert, and could not successfully assert, that ball-bearings were invented by Conrad. In 1904, when the application was filed at Washington, every element contained in the patent was well known (concentric rings, balls, or rollers, and spacers or spreaders), and these elements had already been frequently combined in one form or another. To make a ball-bearing at all, it was obviously necessary to begin with two concentric rings, and these needed to be grooved, or the balls could not be confined as they rolled. It was just as obvious that these grooves, or curves, must be upon the inner periphery of the outer ring and the outer periphery of the inner ring; in other words, the sides of the peripheries must overhang the balls to prevent them from escaping. The grooves need not be deep, but in some degree they must exist. It was also elementary that the balls must be put into the raceway formed by the grooved peripheries, and it was seen without difficulty that they could readily be inserted through a filling-opening, or notch, cut into the side of one ring or into the sides of both rings. In this manner the raceway could be completely filled with balls, and at first this was believed to be the necessary, or at least the most desirable, method of operating the bearings. But it was soon discovered that the friction between adjacent balls distorted them so that the bearing quickly became ineffective. Thereupon various devices, such as a small coiled spring, or a rigid cage, or some other spreader or distributer, were adopted to keep the balls from touching each other.
There were other difficulties also about these structures, especially about the openings for inserting the balls. Obviously the bearing was weakened if part of the rings was cut away; and, moreover, the inequality thus produced in the surface of the raceway, slight though it might be, was found to injure the balls and to impair seriously the usefulness of the bearing. Of course the opening could be closed up, and this expedient was tried by several inventors, but the result was not *869satisfactory. At this stage.of the art Conrad entered the field. _He believed that he was the first to conceive the idea of making both rings solid in all their parts, and this is essentially his contribution toward the solution of the problem. But, with solid rings having no opening for putting balls into the raceway, it is clear that only one way_ remained by which these could be inserted, namely, displacing the rings eccentrically as the first step toward assembling the parts. _ This was the. only practicable expedient, as a moment’s experiment will instantly prove, and this had already been shown by Gentry, whose patent is in precisely the same art. But, if the rings be displaced eccentrically, it is impossible to introduce more than a limited number of balls, and it was necessary to recognize this fact also. By the use of force one or two more balls than would normally fill the eccentrically shaped space might be squeezed in; but even by the use of force it was impossible to make the raceway take as many balls as could be inserted through the openings that had previously been employed. It is clear, I think, that Conrad’s method of construction is inseparably connected with an adequate description of his completed bearing as an article of commerce. Merely to describe the bearing as it should appear after completion would not instruct the public how. to make it, for the eccentric displacement of the rings was the indispensable first step in the process. As the file-wrapper abundantly shows, Conrad himself was well aware of this fact, and to that extent his method of assembling must inevitably be taken into account. In my opinion what Conrad had in mind, and sought to protect, was a solid structure, unbroken in every part, and therefore altogether free from openings into the raceway. He knew that this would overcome the weakness of former constructions, and this was one of his principal objects. Another object was to secure a smooth, continuous raceway, free from the least irregularity of surface. He was obliged to content himself with a limited number of balls, and, still further, he was obliged to use spreaders, or spacers, to perform a new function, namely, to distribute the balls and to keep them apart, in , order that the balls and the spreader, acting in unison, might hold the bearing together as a unitary structure. It was not to prevent friction between adjacent balls, but to make a unitary structure, that he used the spreader. For, if he did not distribute the balls and also hold them apart, they would run together, the rings would become eccentric, a crescent-shaped space would develop, and the bearing would disassemble. All this will be clearly seen to have been in Conrad’s mind.
Let us now examine the file-wrapper. The proceedings in the Patent Office began in February, 1904, when the application was presented. In it Conrad declared himself to be the inventor of .new and useful improvements in ball-bearings, and went on as follows (Record, pp. 470, 471):
“Ball-bearings having the balls running between grooved carrying rings, one on the inside and one on the Outside of the race course, answer (as proved by Professor Stribeck in Neubabelsberg) in general the most extensive technical requirements with regard to carrying power, decrease of friction, and wear. There is, however, a material drawback connected with this construction, consisting in that one of the rings or both must be provided with some lateral or *870radial recesses or excavations, in order to allow of the introduction of the balls into the race groove of the bearing. Although the recess can be under circumstances closed up again, after the balls have been brought in, nevertheless the following evils arise hereby:
“(1) Upon the part through which the balls are introduced no pressure is to be exerted.
“(2) The strain of hardness of the damaged balls frequently causes the bursting of the balls, which can also occur when the balls are not well mounted, or when incidentally a pressure is exerted upon the part through which the balls are filled in.
“(3) For the screw bolts adapted to fill out the filling holes, excavations must be provided to allow the screws of being mounted in the casing.
“(4) As the filling-hole or recess must always be arranged in the stationary ring, it is always necessary to have the rings made double with an inner filling-hole and an outer one. The same as the above-mentioned evils, this evil will be removed by the application of the new bearing.
“The subject of the present invention is a ball-bearing with two grooved uniform full rings, having no recesses and being not broken. This construction is thus a material improvement over the well-known bearings.”
Describing the drawings,- he said (Record, pp. 472-474):
“As seen from Fig. 1 the space between the rings is not quite filled out with the balls; moreover the latter are situated in their guides at a certain distance from each other and separated by the well-known cages. By the arrangement of a smaller number of balls than that which would have place between the carrying rings, it is possible to bring in the balls between the carrying rings without the necessity of providing the latter at any part with a passage, or weakening it at any part. The filling in of the balls is, as seen from Fig. 3, effected by placing both carrying rings eccentrically to each other and bringing in the rings (balls?) into the crescent-shaped space d. The number of the balls to be filled in in the above-stated maimer can be even increased from 8 (Fig. 3) to 9, when a free space for this additional ball is created by a slight elastical deformation of one or both of the rings; the deformation or slight tilting allowing of at least one of the balls being pressed between the others.
“The bearing with the balls being brought in in the manner specified is not yet ready for use, since the balls will come out of alinement as soon as the rings will get any eccentric position to each other. In order to prevent this, the distance between the different balls is fixed by the insertion of special parts o, so-called cages, which, though old in themselves, have been used neither in this combination nor for this purpose.
“The .special form of the cages has no influence whatever upon the construe- * tion and operation of the bearing, and thus any form can be used inasmuch as the same permanently fixes the distance of the balls.”
The claims were as follows (Record, pp. 475, 476):
“(I) In ball-bearing the combination with the race-surface, of inner and outer grooved rings, said rings being unrecessed and unbroken, and adapted to bear a smaller number of balls than that which would actually have place between them, substantially as specified.
“(2) In ball-bearings the combination with the race-surface, or inner and outer grooved rings, said rings being unrecessed and unbroken, and adapted to bear a smaller number .of balls than that which would actually have place between them, the rings being adapted to be placed eccentrically to each other, so as to -allow pf the introduction of the balls from the side of them, substantially as specified.
“(3) In ball-bearings the combination with the race surface, of inner and outer grooved unrecessed and unbroken rings, said rings being adapted to be placed eccentrically to each other, so as to allow of the introduction of balls from the side of them, of some partitions of any form, material, and arrangement, and forming a cage to be placed between the balls, and adapted to keep the balls at the proper distance from each other, substantially as specified.”
*871In March, 1904, the examiner rejected all the claims, and in January, 1905, the applicant made certain amendments and substituted the following claims (Record, pp. 482, 483):
“(1) Roller-bearing, comprising two parallel concentric but eccentrically movable, solid, unbroken rings, a limited number of balls between said rings, and distance pieces alternating with and between and in line with said balls and keeping them at a practically inalterable distance.
“(2) Roller-bearing, comprising two parallel concentric but eccentrically displacable, solid, uninterrupted rings with substantially axially parallel walls, ball-keeping races in the opposite surfaces of said rings, a limited number of balls between said rings, and guides in said races, non-rolling distance pieces alternating with and between and in line with said balls.
“(3) Roller-bearing, comprising two parallel concentric but eccentrically displacable, solid, unbroken rings with substantially parallel walls, ball-keeping races in the opposite surfaces of said rings, a limited number of balls in said rings and guided in said races, springs loosely inserted between and alternating with said balls and in line with the same.
“(4) The method of inserting balls between the races of eccentric roller-bearings, which consists in placing said bearings eccentrically to each other, inserting a limited number of balls into the open space'between the eccentrically-placed bearings, placing non-rolling distance pieces between said balls, and restoring the roller-bearings into concentric position.”
In February, 1905, these claims also were rejected, and after slight amendment were rejected again in July of the same year. In April, 1906, the applicant presented eight new claims (the last 8 of the patent in suit), and afterwards added what is now claim 1. The following argument was offered with the eight new claims (Record, pp. 491, 492):
“The foregoing amendment is made in view of the examiner’s letter of July 3, 1905. In the successive amendments in this case the claims have departed from the real invention, as expressed perhaps informally in the claims originally filed.
“A new set of claims is filed herewith based on the fundamental invention, namely, a bearing of the type known commercially as a ‘closed’ bearing; this being a unitary article of sale consisting of the two concentric rings and the balls and spacers, all held together merely by their relative positions. With this article of sale the assembling is done in the shop of the manufacturers.. The user has only to apply the inner ring to his shaft and the outer ring to the surrounding frame, as for example the axle and wheel of an automobile-, respectively. To make such a closed bearing with rings, each of which is unbroken either at the edge to form a filling-recess or at any other point, is the problem toward which applicant directed his attention and which has resulted in the present invention.”
The present specification was also substituted, and in this shape the patent issued. The specification follows:
“This invention provides a ball-bearing having concentric-grooved rings; the sides of the grooves being uninterrupted throughout their circumference, and the parts being so proportioned and designed that the balls may be admitted to the grooved space by displacing the rings relatively to each other. The term ‘ball-bearings’ is to be understood as including various other known equivalent devices rolling between the rings.
“The principal advantage of the new bearing lies in the continuity of the sides of the groove, which insures the regular running of the balls, and consequently great durability of the bearing, and which also enables the bearing to support a greater pressure than bearings having an interruption or recess for inserting balls through the side of the ring.
“Other features of improvement are referred to in detail hereinafter.
“The accompanying drawings illustrate embodiments of the invention.
*872“Figure 1 is a face view of complete bearing. Fig. 2, a diametrical section of Fig. 1. Fig. 3, a face view showing the manner of introducing the balls. Figs. 4 and 5 are views similar to views 1 and 2, showing the modified construction. Fig. 6 is a face view, and Fig. 7 an edge view, of the cage used in Figs. 4 and 5. Fig. 8 is a face view of another modification.
“Referring to the embodiment of the invention illustrated, the two concentric rings a and B have between them a number of balls e or equivalent rolling devices. Each ring has a groove, the sides of which overhang the balls to a slight extent. The sides of' the grooves, and, in fact, all the parts of each ring, are continuous and practically integral throughout the entire length of the ring. In the normal position of the parts the balls cannot escape; the space between the sides of the grooves being slightly less than the diameter of the balls. Similarly the balls hold the two rings together against axial displacement, so that all the parts are held together and form a unitary device. The edges of the rings, however, are spaced so far apart from each other that they may be displaced eccentrically relatively to each other in the manner shown in Fig. 3, leaving a crescent-shaped space of sufficient width to permit the introduction of a limited number of balls. The crescent-shaped space is marked d. The rings may be then restored to their concentric position, and spreaders or distributing devices introduced into the spaces between the bails, so as to distribute them entirely around the raceway and to prevent their return to a position such as Fig. 3, which would permit the escape of the balls.
“The number of balls which can be introduced may be increased by effecting a slight elastic deformation or tilting, and at the same time pressing an additional ball between the others.
“The exact shape of the groove is not material. It will depend upon the shape of the ball or rolling device or on various other conditions. The spacers also may be of various designs, many of which are known in connection with ball-bearings of other types, where, however, they do not serve the same function of retaining all the parts together in a unitary whole.
“In Figs. 1 and 2 the spacers f are connected to each other by a ring e, so as to.form a cage. One or two of the spacers f are made of extra length and bent over at the end in order to prevent the cage from being removed in an axial direction. It will be seen that the spacers f hold the balls in the position of Fig. 1 (that is to say, in the distributed position), preventing the balls from running together, and thus allowing one of the rings to fall down against 'the other and release the balls through the crescent-shaped space. In Figs. 4 to 7 are shown spacers of another" type. These spacers are in the form of arms projecting from a ring; but in this case a split ring g is employed, and the arms i at their juncture at the ring are reduced to form recesses h, with overhanging edges. To insert these spacers, the ring is compressed and inserted near the innermost points of the balls, so that the overhanging arms i may pass between the balls. The ring being then released expands, so that the balls are caught in the recesses Ji and hold the cage in place.
“Instead of the solid spacers shown, yielding spacers may be employed, and the spacers may be connected to or disconnected from each other. For example, in Fig. 8 there is shown a separate and yielding spacer in the form of the spiral spring 7c, which upon being compressed may be introduced into the groove or withdrawn therefrom between the overhanging edges of the groove.
“I do not claim in the present application the described method of assembling ’ the parts of my improved ball-bearing; this method being claimed in a divisional application filed May 18,1906.”
; Claims 8 and 9 are the only claims in suit. They are practically identical, and are in the following words:
“(8) A bearing comprising two concentric rings, balls between said rings, each ring having a groove both sides of which overhang said balls and are continuous and practically integral, throughout their circumference, the number of balls being such that they can be inserted in the space between the rings when the latter are displaced from their normal position, and means for distributing the balls throughout the length of the groove, whereby the two rings are held together against axial displacement by the engagement of the balls with the *873overhanging walls of the grooves and the parts are held together so as to form a unitary device.
“(9) A bearing comprising two concentric rings, a and 6, balls c, between the said rings, each ring having a groove both sides of which overhang said balls and are continuous and practically integral throughout their circumference, the edges of said sides being separated so far from each other that by displacing the rings eccentrically a limited number of balls may be inserted between them, and distributing devices adapted to be introduced between said edges and into the spaces between said balls when the rings are restored to concentric position, whereby the two rings are held together against.axial displacement by the engagement of the balls with the overhanging walls of the grooves, and the parts are held together so as to form a unitary device.”
I do not think it necessary to discuss the patentability of the device. Gentry’s structure undoubtedly shows the eccentric displacement of the rings and is a dangerous reference, while Oldfield (although he may not have realized the full scope of his invention) comes so near the Conrad bearing that careful discrimination would be required. But I think it clear that, if the validity of the patent be assumed, I cannot treat it as a pioneer; it must be confined to the rings, solid and unbroken throughout, upon which Conrad laid repeated and emphatic stress. 'This is the central thought of his invention, and I am not prepared to follow the plaintiff’s experts and to give the patent a construction now that makes it cover any ball-bearing whatever, even if only the minutest fraction of unbroken surface is shown at the bottom of the raceway. How such a minute fraction as 44/10000 of an inch could “overhang” is not perceptible; and, moreover, any filling-opening is a departure from the patent in an important particular, because it inevitably weakens the bearing, and to that extent is disadvantageous. This weakening was one of the defects in the prior art that Conrad sought to remedy by making his rings solid in all their parts. He nowhere alluded to the “sides” of his grooves in any way that would justify the elaborate theory that has been offered on behalf of the plaintiffs. The “sides” overhang the balls; the “sides” are continuous and are practically integral throughout their circumference; the “sides” are practically integral throughout the entire length of the rings; the space between the “sides” of the 'grooves is slightly less than the diameter of the balls. I do not understand how the “sides” of the raceway can .be continuous and practically .integral if they are cut in half, or so nearly in half that 44/10000 of an inch further would finish the job; and this is the conceded situation in the defendants’ bearing. Still further, Oldfield (No. 674,213) had already shown a notch, “which cuts nearly into the center of said raceways,” and his device was made with metal as well as with rawhide; the latter being the material he preferred. It seems to me that the defendants copy Oldfield much more closely than they copy Conrad; but of course that is Oldfield’s affair and not Conrad’s.
I think the defendants’ supplemental brief summarizes accurately the material differences between the patent and the defendants’ bearings :
“There are three features which the Conrad patent has pointed out and emphasizes as characteristic of his alleged invention. They constitute his invention. They are:
*874“(1) Continuous, umiotched, and unrecessed rings;
“(2) The limited number of balls (only as many balls as can be inserted by eecentricaiiy displacing tbe rings, plus perhaps one); and
“(3) Distributing means, necessary to prevent the hearing from falling apart, and having that function.
“The defendants’ bearings lack each one of these characteristics, and, on the contrary, contain its exact opposite:
"(1) Notched, recessed rings;
“(2) A iarge number of balls (from one to several more than could be inserted by displacing the rings eccentrically); and
"(3) A ball-spacing device, which is not necessary for preventing the hearing from falling apart, and has no such function.”
Without further discussion I state my opinion to be that the defendants do not infringe; and, if this conclusion be correct, the plaintiffs must fail. The defense of laches, which would otherwise need consideration, will not be dealt with.
A decree may be entered dismissing the bill at the costs of the plaintiffs.