The question is whether the determination by the District Court that two patents on slide rules are void for want of invention is clearly erroneous.

Plaintiff, a manufacturer and designer of slide rules, is the owner of Patent No. 2,170,144, issued to Lyman M. Kells, Willis F. Kern and James R. Bland on August 22, 1939, upon an application filed April 17, 1937, hereinafter referred to as the 'Kells Patent,' and Patent No. 2,422,649, issued to James R. Bland on June 17, 1947, upon an application filed October 17, 1944, hereinafter referred to as the 'Bland Patent.' Plaintiff brought an action against defendants for infringement of the patents by the manufacture of five separate models of slide rules. The defenses were invalidity and that both patents had been illegally issued. The case was tried by the court without a jury. The court held the patents invalid and void for lack of invention, and dismissed the complaint for want of equity.

Slide rules are instruments, used by engineers and others, to make computations involving multiplication and division, to determine square and cube roots, squares and cubes, to find the values of trigonometric functions, and to make computations involving them. The determination of squares, cubes, square roots, cube roots, and all other computations made with the slide rule of the type in controversy are merely special forms of multiplication and division. Slide rules are quite old, centuries old. The one essential mathematical principle of the slide rule is that two scales are always used together.

The first slide rule was invented in 1630 and consisted of two logarithmic scales made to slide along each other, kept together by hand. The essential physical parts of a slide rule are the scales, a body, a slide and a runner. The body, the slide, and the runner are so arranged that all three may be longitudinally displaced with respect to one another. In its bare rudiments, the slide rule consists of only two identical numbered scales, the basic C, D scales, two relatively sliding bars on which to mount them, the body and slide, and a sliding indicator, the runner. The scales are mounted on separate sliding bars so that a selected distance on one of the scales may be quickly added to, or subtracted from, a selected distance on the other scale. By this process of adding distances, the slide rule accomplishes multiplication and division by addition or subtraction of distances is made possible by the fact that the scales are what are known as 'logarithmic' scales instead of 'arithmetic' scales. Modern general purpose slide rules include more scales than the two basic scales. But all of the scales have been well known on slide rules in one form or another for at least 100 years, and it is well known that any desired number of mathematical scales may be engraved on a single slide rule if the rule is big enough. The scales are laid out on the body and slide, graduated according to logarithmic functions of numbers. These logarithmic values are based on and laid out in accordance with standard logarithmic tables. The trigonometric scales are laid out according to and based on the logarithms of trigonometric functions which are also found in standard mathematical tables.

An examination of the front face of Kells' rule discloses the two basic scales, C and D. The C scale is mounted on the slide, and D scale, on the body at the bottom seam. Additional scales are the so-called 'folded scales' CF and DF, which are mounted at the upper seam, and the 'inverted scales' CI and CIF, which are mounted on the central portion of the slide. Also on the fact are the 'log log scales,' LL1, LL2 and LL3. The face also shows a scale known as 'equal parts scale.'The back face of the rule shows the basic D scale on the body along the bottom seam between the slide and the body, and two scales known as 'square scales' A and B. This face also shows the 'cube scale' K, which is a third of the unit length of the basic C and D scales, and hence is repeated three times in the length of the rule. It also is mounted so as to cooperate with the basic C, D scales and is useful in connection with cubes and cube roots. Also, on this face, on the slide, appear the 'trigonometric' scales T, ST, and S. These are all laid out to cooperate with the basic C and D scales and are useful in connection with problems involving trigonometric functions. The T scale is for the trigonometric quantity known as 'tangents' of angles. The S scale is for 'sines' of angles. The ST scale is for both sines and tangents of small angles. We think it advisable to note here that Kells' alleged invention has to do with the position on the rule of the sine (S) and the tangent (T) scales and the relationship between those scales and the other scales on the rule.

In addition to the slide and two body members there is an 'indicator' or 'runner.' As already observed, the runner is an essential element in the manipulation of the rule. It is mounted on the body and moves thereon independently on the slide. It has transparent faces on both sides of the rule through which the scales can be seen. It contains two hairlines inscribed across each transparent face perpendicular to the slide and body, so positioned that a setting indicated by the hairline on the D scale on one face is simultaneously indicated by the other hairline on the D scale on the back, thus making it possible to use the scales on the front with those on the back.

The mechanical manipulation of the rule itself is simple. The user selects the scales for the functions, values or sums in which he is interested, locates the desired points on the scales, and continues to add or subtract the log values, completing step by step the required acts of multiplication and division until the computation is completed when he reads the result on the last scale he used.

In the specification accompanying application No. 2,170,144, the patentees states that the object of their invention is 'to provide a slide rule in which problems involving numerical and/or trigonometric terms may be solved, irrespective of the number of steps therein, in a continuous manipulation, that is, without the necessity of having to set down a result in order to retain the same while a new setting is made.' And in application No. 2,422,649 the patentee states the invention 'has as its object to provide a rule of enhanced power and in the use of which by a process requiring generally one but no more than one movement either of the hair-line or of the slide for each number in the expression to be computed.'

Defendants in their brief admit that if the patents are valid, they, by the manufacture of their rule, infringe. But here, as in the District Court, they contend that the patents are invalid for want of invention and because they are anticipated by the prior art.

The question whether an improvement displays the expected skill of the maker's calling involving only the exercise of the ordinary faculties of reasoning upon the materials supplied by special knowledge is a question of fact, and a finding by a trial court upon that question is conclusive on appeal unless clearly erroneous. Rule 52(a), Federal Rules of Civil Procedure, 28 U.S.C.A.; Thomson Spot Welder Co. v. Ford Motor Co., 265 U.S. 445, 44 S. Ct. 533, 68 L. Ed. 1098; Goodyear Tire & Rubber Co. v. Ray-O-Vac Co., 321 U.S. 275, 64 S. Ct. 593, 88 L. Ed. 721; Graver Tank & Mfg. Co. v. Linde Air Products Co., 336 U.S. 271, 279, 69 S. Ct. 535, 93 L. Ed. 672; and Falkenberg v. Bernard Edward Co., 7 Cir., 175 F.2d 427.

It is, of course, true that a new combination of old elements may amount to invention, and the fact that, viewed in retrospect, a conception seems simple, does not justify denial of inventive genius where the need for the device long existed and the solution of the problem did not occur either to experts or to those skilled in the art, Florence-Mayo Nuway Co. v. Hardy, 4 Cir., 168 F.2d 778. But novelty and increased utility in an improvement upon previous devices do not necessarily make it an invention. National Pressure Cooker Co. v. Aluminum Goods Mfg. Co., 7 Cir., 162 F.2d 26, It must, under the Constitution and the statute, amount to an invention or discovery, Thompson v. Boisselier, 114 U.S. 1, 11, 5 S. Ct. 1042, 29 L. Ed. 76. A device which displays only the expected skill or the exercise of the ordinary faculties or reasoning upon the materials supplied by special knowledge and the facility of manipulation which results from habitual and intelligent practice, is in no sense the creative work of an inventive faculty. Hollister v. Benedict & Burnham Mfg. Co., 113 U.S. 59, 73, 5 S. Ct. 717, 28 L. Ed. 901.

In contending that the court erred in holding the Kells patent invalid for lack of invention, plaintiff says that 'the improvements reside in creating a new arrangement of slide rule elements with a new mode of operation which makes possible the solution of old problems with greater ease and speed and with less opportunity for error.' In other words, that Kells' rule for the first time became a calculating device for both trigonometric and numerical values. As to the Bland patent, plaintiff says that it is an improved cooperative arrangement of the log log scales with one another and with other scales- the final step which integrated all the scales on the general purpose rule. And it argues that the essence of the invention of the patents is the presence of both sine and tangent scales on the slide reading to the same length as the basic C, D. scales.

Plaintiff based its claim on the manipulative advantage claimed to have arisen from the particular selection and arrangement of the trigonometric and other scales and the teaching of the patents as to how those scales are to be used.

Conflicting evidence was introduced on the question of novelty, utility and invention. At the conclusion of the trial, the trial judge filed an opinion in which he discussed the history of the slide rule, the relevant prior art, and the oral testimony of expert witnesses, and concluded that there was nothing in the way of invention in the two slide rules in question- nothing in the nature of discovery, authorship or novelty, and that their improvement, if any, in utility was very slight. After the filing of the opinion, defendants submitted findings of fact to the effect that the essential feature of novelty asserted for the Kells patent was the placing of the trigonometric, that is, the sine and tangent, scales on the slide, so arranged that they both read against the full unit length D scale on the body of the rule and usable directly with the C-D scales, whereby proportions involving trigonometric functions might be performed by combined operations in progressive manipulations; that the arrangement of the trigonometric scales with respect to the other scales of a slide rule was based upon mathematicians long prior to the filing date of either of the patents in suit, and involved no substantial change in the mechanical principles of operation of the rule; and that the specific arrangement of scales constituting the essential feature of novelty asserted for the Kells patent was fully disclosed in several prior art citations of record, and it made no substantial difference in the results obtained whether the trigonometric scales were on the body or on the slide, or on the front or on the back; the mechanical functions of the slide rule remained the same and the mathematical results were identical.

As to the Bland patent, defendants suggested that the court find that the patent merely added to the disclosure of the Kells patent the idea of a reciprocal relationship between the log log scales, that is, the scales of the logarithms of the logarithms of numbers greater than unity, and the scales of logarithms of the co-logarithms of numbers less than unity, with such scales reading to the full length D scale, together with identifying symbols therefor, and that this reciprocal relationship of the log log scales merely constituted the application of a known mathematical concept to the scales of a slide rule and did not involve any new principle of mechanical operation of the slide rule itself.

Judge Shaw did not sign the suggested findings of facts, but in a supplemental opinion, said:

'After due consideration I have arrived at the conclusion that my written opinion, heretofore filed in this case, is sufficient under the provisions of Rule 52(a) * * * .

'I can find nothing substantially wrong with the proposed findings of fact and think that the evidence sustains them. I would not hesitate to sign these findings except for two reasons (1) they display an erudition in mathematics which I do not possess, and (2) they attempt to obscure and dilute the reasons for my decision as shown by the written opinion on file.

'That opinion is based entirely upon my conclusion, as set forth therein, that it is necessary to distinguish (1) between the slide rule as a machine and (2) the science of mathematics. If I am wrong in this basic conclusion then these findings of fact might become of some importance, and I think they are supported by the evidence, but for reasons stated I am not signing them and am permitting them to go into the record unsigned. I have signed the conclusions of law and the judgment for the defendant.'

We do not deem it necessary to set forth the testimony of the expert witnesses or to recite the prior art patents relied on. Keeping in mind the principles enunciated above, it will suffice to say that we have examined the record and are convinced that there was ample evidence to support a finding that the essential feature of novelty asserted for the patents in issue was not new and hence did not involve invention, and that the conclusion reached by the District Court was a permissible one and in substantial accord with the law and the facts, and should not be disturbed.

Affirmed.