delivered the opinion of the court:
This appeal is from the judgment of the United States Customs Court, Second Division, C.D. 2248, sustaining the protest of the importer to the classification of merchandise invoiced as “Finished hardened steel balls,” which had diametral bore holes in them.
These articles were classified by the collector, in accordance with their intended use, under paragraph 368 of the Tariff Act of 1930, which, as modified by the trade agreement with Switzerland, 90 Treas. Dec. 174, T.D. 53832, provides for “any mechanism, device, or instru*8ment intended or suitable for * * * regulating, indicating, or controlling the speed of arbors, drums, disks, or similar uses * * They were, in fact, imported for use as parts of a device manufactured by the Cleveland Worm & Gear Co., Ltd., the actual importer, and known as a “speed variator.” We shall describe it later.
The importer protested that the steel balls should be classified under paragraph 321, as modified by the Annecy Protocol to GATT, 85 Treas. Dec. 116, T.D. 52462, as “Antifriction balls * * *, whether finished or unfinished, for whatever use intended * *
Finding the merchandise to be “antifriction balls” within the meaning of paragraph 321, the lower court sustained the protest. The only issue before us is whether the imported articles are such “antifriction balls,” no question having been raised as to the characteristics of the balls or their use. Since it is beyond question that they are balls, the single simple issue is whether the adjective “antifriction” aptly describes them, within the common meaning of that term.
There appear to be no controlling precedents. The parties and the court below have referred to only two cases, one in this court and one in the Customs Court. This court’s decision was in United States v. J. A. Freeman & Son, 29 CCPA 103, C.A.D. 177. The relevant part of that decision involved %" steel balls imported with a variety of other parts claimed to be free of duty as parts of agricultural implements. They could not be free of duty under the statute if they were “specified by name in Title I” of the Tariff Act of 1930. The question as to the steel balls was whether they were specified by name in paragraph 321 as “antifriction balls.” The evidence was that they were “for use on Deering mowers; that they would ‘reduce friction’.” The court held that “they are used to ‘reduce friction’,” and hence were specified by name as “antifriction balls” in paragraph 321 and were dutiable thereunder. This is not helpful in deciding the present case because one thing we have to decide is whether the balls involved here are ever used to reduce friction, so as to be antifriction balls.
The Customs Court case is J. F. Goldkamp & Co. v. United States, 39 Cust. Ct. 420, Abstract 61194, wherein 1 millimeter diameter steel balls for ball point pens, invoiced as “Stainless Steel Anti-Friction Bearings,” were held to be antifriction balls dutiable under paragraph 321. Although we do not have the record before us, we think it reasonable to conclude that a ball for a ball point pen is an antifriction ball since, in writing, it substitutes for the sliding friction of an ordinary pen point the rolling friction of a rotating ball. If proof were necessary that a substitution of rolling for sliding friction in the same use results in a reduction of friction over what it otherwise would be, and that the device responsible therefor is an “antifriction” device there is testimony to that effect in the instant case. It by no means follows, however, that the instant balls are antifriction balls.
*9We do not consider that a ball which is never used for the purpose of reducing friction is an antifriction ball. “Ball” is a very broad term indeed, including every kind of ball from a popcorn ball to this earthly ball itself, according to the dictionary. If “antifriction” is to be given any meaning at all, the term “antifriction balls” in paragraph 321 must be deemed to be restricted to such balls as are, or may be, at some time or other in some normal use, employed to reduce friction which would otherwise exist. We do not think it is sufficient, to classify a ball as an “antifriction ball,” to say that the ball rolls, and that rolling produces less friction than sliding, as every ball rolls; but clearly not every ball can be an “antifriction” ball under the statute or the term “antifriction” would be superfluous. What, then, are the balls at bar ?
The importer concedes that the imported steel balls are designed for use in and are used as essential parts of the “speed variator” manufactured by the actual importer. It is clear to us from the record, and it is not contended otherwise, that they have no other known use. What these parts are and what they do requires consideration of what the “speed variator” is, what it does, and how it does it, with particular reference to the part played by the balls.
The “speed variator” is a mechanical device for transmitting power which provides for infinitely variable output speed, relative to input speed, over a range of up to 9:1. It has an input shaft and an output shaft in axial alinement therewith. The input shaft rotates a beveled drive disc and the output shaft is rotated by a similar beveled disc, these discs being parallel ¡and spaced apart with their conical, beveled rims facing each other. In contact with the beveled rims of both discs are four (or more) hardened steel balls, all exactly the same size and evenly spaced. Viewing a disc head on and imagining it to be a clock face, the balls would be at the 3, 6, 9 and 12 o’clock positions. Each ball has an axle on which it rotates and the axles are adjustably carried in the framework of the device so that the balls remain in the same locations, but their axles can be uniformly tilted from positions of parallelism relative to the axes of the input and output shafts. When tilted, the axes of the ball axles, by reason of the design of the “variator,” have a common point of intersection. Surrounding all of the balls and in point pressure contact with each of them is a circular ring or raceway. Thus each ball has three points of contact, one with the raceway and two with the beveled rims of the discs. In operation, rotation of the input shaft and its drive disc causes the balls to rotate on their axles and as they rotate they drive the other disc on the output shaft with which they are in frictional engagement, causing it to rotate. The balls are, therefore, power transmission elements. As they turn, the outer raceway, which presses them into frictional contact with the driving and driven discs, *10turns freely with, them, its function being only to maintain frictional engagement between the balls and the discs. Quoting from descriptive literature in evidence,
If the input shaft is presumed to run at a fixed speed, the speed of the output shaft is determined -by the angular positioning of the axles on which the drive balls rotate. With the ball axles parallel to the drive shafts * * ♦ the speed ratio between the drive shafts is 1:1. However, * * * the ball axles can be tilted uniformly in either direction * * * so that the effective rolling radii of the balls with the engaging drive discs can have a ratio variable up to 3:1.
Thus it will be seen that if one starts with an input shaft speed of 3 r.p.m., at one extreme adjustment, the output shaft could be made to rotate at % that speed or 1 r.p.m., and at the other extreme adjustment at 3 times that speed or 9 r.p.m., and the device therefore has a speed adjustment range of 9:1.
It will be observed that the importer of these balls refers to them in its literature as “drive balls.” This is an apt description. To serve as such they are rotatably mounted on axles. So that this may be done, each ball is provided with an accurately bored diametral hole, of a size to fit the axle on which it is mounted in the “variator.” It performs its driving function by frictional engagement with the two beveled discs and it cannot, in our opinion, be said with reason that in this use the ball serves as an antifriction, ball. The record is devoid of any evidence that these bored steel balls, which are specially made parts of the “variator,” have any other use. If they are “antifriction balls,” as held below, it must be on the basis of their use in the “variator.”
The Customs Court said:
* * * we are of the opinion, based upon the testimonial record and the exhibits, that the balls in controversy are, in fact, antifriction balls.
The exhibits consist of a partially sectioned demonstration model of a “speed variator,” and two advertising pamphlets of the Cleveland Worm & Gear Co., Ltd., describing the device. The testimony is solely that of the president of that company, the actual importer. He was an interested witness. The court below appears to have been influenced primarily by what he said. We have considered this testimony carefully and it leads us to the opposite conclusion. In reading the crucial parts thereof, we find it difficult to decide whether the ideas expressed are those of the witness or of his counsel, whose purpose, of course, was to establish that the imported balls are “antifriction balls” within paragraph 321. Counsel’s brief sums up the testimony on the alleged antifriction aspects of the balls in the following passages, all emphasis being ours:
The balls in question are made to specification with accurate sphericity to reduce the frictional element (R. 11, 17) and with a high degree of hardness to avoid friction when used with heavy loads (R. 11, 18). They are supplied *11in sets and matched very closely to avoid friction and to permit them to operate with a uniform angular volocity (R. 11, 18). They are aged to achieve size stability (R. 12), and they have an accurately finished hole or bore through their diameter (R. 12).
* * # * * * *
We refer particularly to the following testimony:
“Q. Do these balls actually have the characteristics of reducing friction in that apparatus? A. That is why they are employed.”
(Record page 17)
“Q. So you must have a type of ball which will lessen or reduce friction In order for this variator to operate for its designed function? A. We must have a ball that will avoid slippage, sliding, and that will rotate continuously and so minimize the frictional loss.”
(Record page 19)
From the record as a whole, it is clear that the design of the “variator” is such that the elements which transmit power from the input, shaft disc to the output shaft disc must have matched spherical surfaces and must be mounted for rotation on axles, for which reason they must have bores. The sphericity is not dictated by a desire to reduce friction, in spite of what the witness said, because nothing other than a set of elements with spherical surfaces can be used to transmit power in this device. The true import of the testimony on the subject of friction can be summed up as follows: The balls transmit power in the “variator” by reason of their frictional contact with the discs; undesirable and deleterious friction would be introduced into the “variator” if the balls (1) did not have accurately spherical surfaces, (2) were not sufficiently hardened, and (3) were not accurately matched in size.
We do not consider that drive balls, functioning by frictional engagement, become *m£i-friction balls merely because they are made with such accurate adherence to specifications as to size, hardness and sphericity that they do not introduce unwanted friction into a mechanism of which they form a part. For this reason we cannot agree with this statement by the Customs Court:
The fact that the balls have true sphericity, a high degree of hardness, and were evenly matched contributes much to the effectiveness of the balls for anti-frictional purposes. [Emphasis ours.]
The balls per se do not have “antifrictional purposes.” They are frictional driving elements, a function peculiar to the mechanism of which they are a part. The situation is one which is somewhat difficult of expression. We think the government brief puts it rather well in saying:
* * * because of the hardness, roundness and matching of the four balls used, the friction created is held at the amount utilized.
*12Thus, the friction which is created by the balls themselves is not reduced by them but rather is used by them to turn the otherwise stationary out-going shaft.
* * * * * * *
Thus, they are not antifriction balls but are [drive] balls, the sole purpose of which is to create friction and utilize the friction created.
Since we agree with this analysis of the function of the balls in the “variator” of which they are parts, and since there is no other evidence in the record that they have any other use, actual or potential, we conclude that the imported balls are not “antifrictional balls” within the meaning of paragraph 321.
The judgment of the Customs Court is reversed.
Worley, C. J., did not sit or participate in decision.