Ludlum Steel Co. v. Terry

37 F.2d 153 (1928)

LUDLUM STEEL CO.
v.
TERRY.

District Court, N. D. New York.

October 10, 1928.

Niele F. Towner, of Albany, N. Y. (Richard Eyre and Herbert H. Dyke, both of New York City, of counsel), for plaintiff.

C. Bertrand Race, of Albany, N. Y. (Marshall A. Christy, of Pittsburgh, Pa., and George W. Manierre, of Los Angeles, Cal., of counsel), for defendant.

COOPER, District Judge.

This suit is brought by the plaintiff company for the infringement of three patents issued to Armstrong, plaintiff's assignor. The patents are numbered 1,322,511, 1,456,088, and 1,533,782, respectively, and will be referred to herein as the first, second, and third patents.

The first patent is for a stable surface alloy steel, the essential components of which are silicon, chromium, carbon, and iron. This patent was applied for May 24, 1919, and the patent was issued November 25, 1919. The second patent, application for which was filed December 12, 1919, and the patent issued May 22, 1923, is for a heat-treated, stable surface alloy steel of the same constituents as in the first patent, but in more specific proportions and of a more limited range. The application for the second patent was divided, and a new application filed February 2, 1923, and the third patent granted thereon April 14, 1925. This third patent is for exhaust valves for internal combustion engines, made of the same composition steel and in accordance with the formulæ of the second patent, and other formulæ of varying proportions of the component elements. These formulæ are also within the first patent. This third patent will also be referred to herein as the "valve patent."

The following claims of these three patents are alleged to be infringed:

Of the first patent:

1 (as restricted by disclaimer). "An alloy steel of high surface stability containing chromium 3 per cent. to 20 per cent., carbon .05 per cent. to 3.5 per cent., but not more than one-tenth of chromium and silicon taken together up to about 13 per cent. thereof and not more than one-sixth thereof above about 13 per cent. silicon under 8 per cent. and over .5 per cent. and also over twice the carbon, silicon and chromium together 5 per cent. to 58 per cent., and the principal portion of the remainder iron."

4. "An alloy steel of high surface stability, containing chromium 10 per cent. to 20 per cent. silicon .5 per cent. to 8 per cent., chromium and silicon together 13 per cent. to 28 per cent., carbon, over .05 per cent. under 3.5 per cent. and also under one-sixth of chromium and silicon taken together, and the principal portion of the remainder iron.

5 (as restricted by disclaimer). "An alloy steel of high surface stability, containing chromium over 10 per cent., and under 20 per cent., carbon over .05 per cent. and more than twice as much silicon as carbon, and the principal portion of the remainder iron."

Of the second patent:

3. "An alloy steel of substantially high surface stability, when cooled down from a high temperature, containing silicon over about .5 per cent. and under about 7 per cent., chromium over about 3 per cent. and under about 10 per cent., silicon and chromium taken together over about 5 per cent. and under about 13 per cent., carbon over about .05 per cent., and under one-tenth the chromium and silicon taken together, and also under one-half the silicon, and the principal part of the remainder iron."

4. "An alloy steel of high surface stability when cooled down from a high temperature containing silicon about 3.7 per cent., chromium about 8.9 per cent., carbon about .46 per cent., and the principal part of the remainder iron."

And of the third patent:

6. "Valves for internal combustion engines made from alloy steel containing silicon about 3.7 per cent., chromium about 8.9 per cent., carbon about .46 per cent., and the principal part of the remainder iron."

9. "As a new article of manufacture, a valve comprising a head and a stem integral therewith made from ferrous alloy containing upon analysis silicon between 1 per cent. and 4 per cent., chromium between 3 per cent. and 10 per cent., chromium and silicon taken together between 5 per cent. and 13 per cent., carbon between .30 per cent. and a maximum of one-tenth the sum of silicon and chromium contents, and the principal part of the remainder iron."

11. "As a new article of manufacture, an internal combustion engine valve, comprising a head and stem integral therewith, made from ferrous alloy containing upon analysis silicon between .75 per cent. and 3.7 per cent., chromium between 5 per cent. and 10 per cent., carbon between .30 per cent. and the maximum of one-tenth the sum of the silicon and chromium contents, and the principal part of the remainder iron."

As granted, claim 1 of the first patent was broad and covered a chromium content of from 3 per cent. to 50 per cent. This chromium content was subsequently narrowed to a range of from 3 per cent. to 20 per cent. by a disclaimer filed during the pendency of this action. Claim 5 was also affected by the disclaimer, which limited the chromium to 20 per cent. The defendant challenges the intent, good faith, and effect of this disclaimer, as appears more fully later herein.

The alleged infringement is the sale by the defendant Terry of certain automobile exhaust valves made of an alloy steel claimed by plaintiff to be almost identical with that covered by certain claims of plaintiff's second and third patents, and to be within the first patent. The valves were sold to Terry by the Rich Steel Products Company of Battle Creek, Mich., which is undertaking the defense of this suit and which will hereafter be called the defendant. The steel from which the valves were made was sold to the Rich Steel Products Company by the Crucible Steel Company of America.

The defenses are:

(1) Lack of invention.

(2) Anticipation in the prior art.

(3) Prior public use.

(4) Invalidity of the second and third patents, because applied for 18 days after the issue of the first patent and covering subject-matter disclosed in, but not claimed, by, the first patent.

(5) Noninfringement, except as to the following claims, if found valid: Claim 1 of the first patent, claim 3 of the second patent, and claims 9 and 11 of the third patent.

The particular alloy steel made by plaintiff under the Armstrong patents, which defendant is alleged to have duplicated and thereby infringed the plaintiff's patents, is called by the plaintiff "silcrome No. 1," and will be referred to hereinafter as "silcrome." Silcrome is used chiefly for the manufacture of exhaust valves for internal combustion engines. While plaintiff also manufactures and sells to various other licensees other alloy steels of somewhat different proportions in their composition under the first and second patents, these other alloy steels were used for other purposes, and are not claimed to have been infringed by the defendant.

Silcrome falls within all three patents, if they are valid, and follows more particularly the specific formula set forth in claim 4 of the second patent and claim 6 of the valve or third patent. The formula as given in these two claims is "Silicon about 3.7 per cent., chromium about 8.9 per cent., carbon about .46 per cent., and the principal part of the remainder iron." Defendant's steel, from which the valves in question were made, is called "C. S. 93." It is made by the Crucible Steel Company according to a formula submitted by the defendant. The formula is as follows, as given by the Steel Company's superintendent: "Silicon, 2.75 per cent. to 3.50 per cent.; chromium, 8.75 per cent. to 9.75 per cent.; carbon, .38 per cent. to .45 per cent.; manganese, .40 per cent. to .60 per cent.; phosphorus and sulphur, under .022 per cent." Of course, the principal part of the remainder is iron.

The analysis of 317 different shipments of this C. S. 93 steel, made by the Crucible Steel Company, to the defendant during 1926 and 1927, shows a range of silicon from 2.63 per cent. to 3.66 per cent. and of chromium from 8.60 per cent. to 10.03 per cent. Only three of these were over 10 per cent. chromium. The carbon content is practically identical with that of silcrome. It does not appear that the traces of manganese, phosphorus, and sulphur have any bearing on the questions at issue. Analysis of 11 out of 12 valves purchased by plaintiff from Terry showed a silicon content ranging from 3.08 per cent. to 3.41 per cent., and chromium ranging from 9.10 per cent. to 10.10 per cent.; two of the valves having over 10 per cent. chromium. The carbon content ranged from .37 per cent. to .46 per cent. In order to determine whether or not Armstrong invented something in the art of making ferrous-chromium alloy steels with a substantial silicon content, and may rightfully claim validity for his patents, it is desirable to trace briefly his work and the state of the art, as disclosed in the records of this case.

What is claimed for Armstrong is that he discovered new and important properties which the product had when silicon was used in certain proportions with chromium in an alloy steel, the principal other elements of which were iron and carbon. Chromium had for some time been used in alloy steel making. What place silicon had in the art was an unwelcome and a supposedly unavoidable one. Silicon, when in more than the least possible quantity, was supposed to make an alloy steel brittle, less desirable, and less available for most purposes, including valve manufacture.

Armstrong's experiments and researches into the properties, qualities, and proportions of metals in alloys began while he was engaged in quasi are welding as early as 1914-1915. Meeting difficulties, due to rusting and scaling away under heat of the welds made by him, he began to experiment to produce weld products which were noncorrosive and nonscaling. Certain patents covering composite electrodes resulting from his researches were applied for by him in 1915, granted in 1917, and assigned to his then employers, and known as No. 1,241,899 and No. 1,241,900. Continuing his experiments, he made a copper crucible and melted materials in it by means of the electric arc. Chromium was used by him from the time of his early experiments. But his use of silicon came from his discovery that the reduction of silicon from the silica of the asbestos envelope of his electrode had introduced silicon into his melts, and he observed the effect of silicon on the melts.

He subjects these melts to tests for acid effect, corrosion by rusting, and to scaling under heat. Recent analyses of samples of the melts those days indicate that what he made was substantially a stable surface silicon-chromium alloy steel, within the range of his first patent. American entry into the World War of necessity restricted Armstrong's experiments. Steel makers at this time were too engrossed in the production of weapons and ammunition for the more certain destruction of their country's enemies to find time to discuss the merits or compositions of an alloy of steel alleged to have been produced by a characteristically impoverished inventor, who, confident of the ultimate value of his discovery, consistently refused to disclose its exact nature and formulæ. Thus he was forced to defer any serious attempt at the exploitation of his product until the close of the war. But this period was not entirely without fruit, for Armstrong had the opportunity of making and forming connections which later proved useful to him in his experiments and researches, and in obtaining his patents.

Immediately after the Armistice, Armstrong, who had been in plaintiff's employ since about the middle of 1917, went to work again on his silicon-chromium alloys. He put an assistant at work at Columbia University, and he himself worked with an assistant at the plaintiff's plant. About 2,500 melts were made to test the properties of the silicon-chromium alloy, to which his attention was chiefly directed, and to determine the proportions of the elements entering into the alloy. In May, 1919, he filed the application for his first patent, which was granted November 25th of the same year, 18 days before the application for the second patent was filed. The third patent was a division of the application for the second patent.

All three of these patents cover a stable surface alloy steel as first described, the differences as between the first and second being chiefly one of formula or range, and necessary heat treatment, and, as between the second and third, one of use for a specific purpose, viz. for an exhaust valve for internal combustion engines.

The patentee concedes in the specifications of the first patent that the presence of chromium in a ferrous-chromium alloy has some effect in making the product more stable or resistant to deterioration, such as surface corrosion; but the patentee states that such material has certain serious drawbacks or defects, such as blemishes, cracks, and fissures, scaling badly when heated, and the necessity of having a very low carbon content.

The plaintiff's alloy, consisting of substantial quantities of silicon added to chromium, carbon, and iron, and proportioned as plaintiff's patents outline, has properties, most of which are new and hitherto unknown, viz. resistance to rust, scaling, or corrosion, ready forgeability and workability, freedom from cracks, fissures, or blemishes, high critical points, capability of being heat-treated to any degree of hardness, durability at high temperatures, and ability to use a greater percentage of carbon. No alloy previously known had all these properties, or many of them.

Armstrong's claim to invention can best be envisioned if the practice and knowledge of the art, before the discovery of silcrome in making exhaust valves for internal combustion engines, is considered. The internal combustion engine is essentially an engine for automobiles, aeroplanes, and other gasoline propelled machines. The exhaust valve is a very important part of such an engine. In the earlier automobiles these engines were crude, and perhaps the exhaust valve and material from which it was made were not of so much importance. But with the improvement of the automobiles and their engines, giving higher speed and greater efficiency, the necessity for better valves and valve materials became urgent. It was about 1912 before an alloy steel was used for valves.

From 1912 to Armstrong, manufacturers of internal combustion engines had to content themselves with valves made from various alloys, none of which equaled Armstrong's alloy in serviceability, and endurance, or performance. Metallurgists and steel experts connected with this industry experimented with the new and different combinations, and the use to which a valve was to be put largely determined the material of its construction. For example, valves were made of cast iron, nickel steel, tungsten high-speed steel, and the stainless steel disclosed in the Brearley patent, No. 1,197,256, the last two being alloys most generally used prior to the invention of silcrome.

In 1913, George Rich, the defendant's president and managing head, obtained a patent for high tungsten steel valves and began making valves of such steel. He found the makers of engines using cast iron heads and steel stems. The high tungsten alloy steel valve became very popular in the valve business, and has never been entirely replaced.

Besides the high tungsten steel, which was apparently expensive and heavy, about 1917, the so-called stainless steel came into use for valves. This largely displaced the high tungsten steel. Stainless steel was a steel alloy composed of 11.5 per cent. to 14 per cent. chromium, iron, a small quantity of carbon, and about .30 per cent. silicon. Stainless steel lasted as a valve steel until it was largely displaced by silcrome, commencing about 1925.

Neither of these former alloy steels, or any other on the market used for valves, had any considerable quantity of silicon in it, unless by accident. So far as valve uses were concerned, Rich, the defendant's president, usually rejected steel with more than a fraction of 1 per cent. of silicon in it, sharing in the belief that a larger quantity of silicon tended to make the valves brittle and unsuitable for use.

It is well known that internal combustion engines were called upon for a strenuous use during the World War for aeroplanes, automobiles, trucks, tanks, and other uses, and every effort was made on both sides to get the best and most efficient engines of war. Had any improvement in valve-making or valve materials suggested itself, or been discovered, it would have been found and used, at least in our efforts during the World War, to make the Liberty motor the best in the world. But the same old materials were used as before, except that stainless steel began to be used in 1917.

In November 1919, Mr. Leslie Aitchison, metallurgist of the British Air Board, published an extensive article in the Automobile Engineer on "Valve Failures and Valve Steels." Defendant strenuously objected to the receipt of this article in evidence in this case, but it seems competent, as illuminating the state of the art at that time, even though the author could not be subjected to cross-examination. The Aitchison article came out in November, 1919, just at the time of the issue of Armstrong's first patent, and one month before the filing of the divided application for the second and third patents. This article treats of the valve art from the metallurgical side, rather than from the mechanical; consequently its authoritative statements are helpful, particularly in considering the valve patent in this case.

There seems to be little doubt or dispute about the fact that most valve trouble is due to overheating. The experts are not all agreed, however, as to what temperature is normal, generally speaking, for a motor exhaust valve when the motor is running. At the trial Jardine testified that 1,500 degrees Fahrenheit to 1,600 degrees Fahrenheit was not unusual; Rich, that 1,000 degrees Fahrenheit was a fair average; while Aitchison in his article writes that a temperature of from 700 centigrade (about 1,290 degrees Fahrenheit) to about 760 degrees centigrade (about 1,400 degrees Fahrenheit) is usual in hotter type engines. Thus we have a range of from around 1,000 Fahrenheit to upwards of 1,500 Fahrenheit. In any event, it seems clear that a valve steel must be durable enough to retain its "cool" qualities at an average temperature of about 1,400 degrees Fahrenheit. Overheating apparently comes from a variety of causes, some mechanical, some functional, and some metallurgical. In the first class of causes it appears that the valve or the valve mechanism is not properly designed for its work and other similar defects. In the second class are included those causes which spring from overheating owing to improper seating, and other defects which interfere with proper heat conduction and cooling generally. But in neither of these are we greatly interested for present purposes.

The main concern is with the metallurgical cause of valve failures. A good example of this class is where the steel chosen for the valve purpose has not been properly heat-treated and tempered, so that, when hot, the valve cracks or warps. Another example is where the steel air-hardens, so that, on being heated to a temperature above its hardening point, during use or subsequent cooling in the air it becomes hard or brittle, or gets surface cracks through which the hot surface gases penetrate and heat or burn away the valve. This is because the surface stability of the valve steel is not sufficiently great at high temperatures to resist the oxidation and scaling caused thereby. Thus it is seen that the choice of valve steels requires a high degree of technical ability and skill.

At this point it might be well to give Aitchison's 12 qualifications for ideal valve steel. They are: (1) The greatest possible strength at high temperatures. (2) The highest possible notched bar value. (3) The capacity of being forged easily. (4) The capacity of being manufactured free from cracks, whether these arise in the manufacture of the steel bar, or whether they are produced during the forging of the steel. (5) The capacity of being heat-treated easily, regularly, and reliably. (6) The least possible tendency to scale, and, if scaling does occur, the scale should be as adherent as possible. (7) The ability to retain its original physical properties after frequent heatings to high temperatures, followed by cooling to normal temperatures; also after being heated to an elevated temperature for a considerable length of time. (8) No liability to harden when cooled in air from the temperature which it will attain when used normally as a valve in an engine. (9) The capacity of being heat-treated after forging, so that it is free from strains liable to produce distortion. (10) Sufficient hardness to withstand extensive wear in the stem. (11) The capacity of being hardened at the foot of the stem with considerable ease, if necessary. (12) The capacity of being machined easily and satisfactorily by ordinary methods.

Of course, no one steel then known possessed all these properties, but in 1919, according to Aitchison, the following five types of steel were in use as possessing some of these qualities in varying degrees: (1) High tungsten ranging from 10 per cent. to 18 per cent. tungsten and from .2 per cent. to .7 per cent. carbon; (2) high chromium steels ranging from 7 per cent. to 14 per cent. chromium, from .25 per cent. to 1 per cent. carbon and varying proportions of silicon, nickel and cobalt; (3) 25 per cent. nickel steel; (4) steels with from 3 per cent. to 5 per cent. nickel, together with chromium from 0 per cent. to 1.5 per cent., and carbon from .15 per cent. to .65 per cent.; (5) ordinary carbon steels. In this list the high tungsten and high chromium were accepted as the best. Aitchison recommended high tungsten alloy for temperatures over 760 degrees centigrade (over 1,400 degrees Fahrenheit), and for lower temperatures high chromium alloy steel. It is noteworthy that as to both these alloys he stated that a variation of as little as .2 per cent. carbon had as much, if not more, effect upon tensile strength as a variation of 5 per cent. tungsten, or 7 per cent. chromium. Thus at the time of Aitchison's writing the important element of the composition of high chromium and tungsten steels was thought to be in the carbon content.

There follows a list of chromium alloy steels given by Aitchison containing silicon used for valves or tested for valve purposes in 1919.

Chemical Compositions of Steels Used.
Steel.   Carbon.  Silicon.  Nickel.  Chromium.  Cobalt.
C-1 ...    0.37     0.25     0.15      12.37
C-2 ...    1.04     0.15     ....      10.42
C-3 ...    1.12     0.85     0.50      11.35       3.90
C-4 ...    0.96     0.17     0.45       13.1
C-5 ...    1.08     0.17     0.50       13.1
C-6 ...    1.18     0.10     0.45       13.1
C-7 ...    1.42     0.35     0.44       13.1
C-8 ...    0.36     0.16     0.23       11.2
C-9 ...    0.54     0.14     0.43        6.3
C-10 ..    0.55     0.75     0.50        7.1
C-11 ..    0.56     0.17     0.42        6.6
C-12 ..    1.09     0.19     0.42        6.3
C-13 ..    1.08     0.56     0.51        6.7
C-14 ..    1.11     0.19     2.96        5.8

But it will be noted that in none of these does the silicon content exceed .85 per cent. and all but that designated as C-2 contain nickel in generally higher proportion than silicon. So it does not appear that before the introduction of silcrome that any one knew, other than in a small way, the effect of silicon on high chromium steels, and that knowledge of such effect was the "philosopher's stone" of the valve art.

In the light of the knowledge of that time, for his ideal high chromium valve steel formula, Aitchison gives the following: Carbon .65 per cent., silicon .60 per cent., manganese .50 per cent., and chromium 10 per cent. But he nowhere states the value or function of even .60 per cent. silicon, so we can almost guess that this figure represented a good average minimum, because, prior to the introduction of silcrome, silicon was known only to add brittleness to chromium alloys, and brittleness is to be avoided in valve steels.

There seems little doubt that down to Armstrong the art of alloy steel manufacture and valve making had no material, and knew of none, adapted to exhaust valve used with the greatly desired properties of Armstrong's silcrome alloy.

Despite the fact that silcrome seemed to fill a long-felt want, much difficulty was met in overcoming the deep-seated prejudice against any steel with more than a minimum of silicon content. At first the progress was slow, but, once the value of the material was known, the sales increased rapidly, and especially after the granting of the second patent.

As before stated, plaintiff limited its sales to its two licensee valve-manufacturing concerns, and to certain automobile manufacturers, presumably those who make their own valves. The record shows that, since the silcrome steel came on the market, its sales to these concerns has steadily increased and is as follows:

1920 ...............     10,000 pounds
1921 ...............     78,000   "
1922 ...............    514,000   "
1923 ...............    996,000   "
1924 ...............  1,312,000   "
1925 ...............  3,874,000   "
1926 ...............  4,446,000   "

The defendant Rich Steel Products Company, of Battle Creek, has bought the alleged infringing steel from the Crucible Steel Company since 1926, and its manufacture and sale of valves made from this steel has apparently caused some loss to the business of the plaintiff. Many of the leading automobile manufacturers now use the silcrome steel or its equivalent for valves, though some of the cheaper cars do not use it.

Briefly, then, silcrome is the most durable, the toughest, and the most generally useful material which has yet been devised for exhaust valves and other purposes where such properties are required; and to Armstrong belongs the credit for being the first to realize and to make use of the formerly despised silicon in the production of an alloy which has such valuable properties. This makes invention.

Besides the known, Armstrong discovered many theretofore unknown, properties of such an alloy. What knowledge the art had of the greater resistivity to an electric current was of an alloy of a higher chromium content than in any of the Armstrong patents as limited, and no such alloys were made or used in a commercial way. The previous knowledge was substantially an accidental discovery of a single property in a proportion different and outside of Armstrong's patent.

That there was patentable novelty in plaintiff's discovery of his alloy, containing substantial quantities of silicon added to chromium, carbon, and iron, with its new, unexpected, and highly desirable properties, is sustained by authority. O'Rourke Eng. Co. v. McMullen (C. C. A.) 160 F. 933; Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 43 S. Ct. 322, 67 L. Ed. 523; Richmond Screw Anchor Co. v. U. S., 275 U.S. 331, 48 S. Ct. 194, 72 L. Ed. 303.

Armstrong is therefore entitled to the first and second patents for the silico-chromium alloys claimed in the patents as limited by the disclaimer and in particular to the alloy known as "silcrome steel" under the first and second patents, unless defendant shall prevail as to his special defense as to the second and third patents, which is later determined herein. Was there not also invention in discovering the value of the alloy of the more specific formula of the second and third patents for exhaust valve used, even assuming that the second patent is invalid as claimed by the defendant because it merely covers things disclosed and not claimed in the first patent?

Did not the developing of this formula from an old (because of his first broad patent) material for a new use constitute invention? The tests in such cases seem to be: Has the old material as applied to the new use produced such results as are amazing, startling and unforeseen; or has the old material as applied to new use accomplished a result which has hitherto not been attained? The second test seems to fit the facts under this valve patent.

It has been shown that, prior to the invention of Armstrong's valve steel, no alloy had been developed which possessed, among other properties, its durability at high temperature and its resistance to scaling and warping at high temperatures. These results were valuable and useful, and had been and were being sought for by expert steel makers and metallurgists. But the introduction of the valve patent formula marked the end of this quest.

The acts and conduct of defendant's president are of interest in this connection. It will be remembered that he was the patentee in 1913 of the valve made from high tungsten alloy steel. For some time, and until the use of stainless steel, he seemingly dominated the valve market. After stainless steel began to be used for valves, he could still compete with the latter, despite the higher cost of the tungsten steel, and he continued to make some valves of the latter material.

But, when silcrome valves appeared on the market in greatly increasing quantities, he evidently found himself losing in the competition. In 1926 he placed his first order for C. S. 93, which is substantially a Chinese copy of silcrome, and used large quantities of it in 1926 and 1927.

Rich was almost the superman of the valve business, but neither his mechanical skill nor that of any one else discovered or used this silicon-chromium alloy for valves before Armstrong, nor was there any more selection of materials by anyone. There could hardly be selection of a material like silcrome for exhaust valves before silcrome came on the market, as such material was unknown and substantially nonexistent. Whatever be Rich's specious reasons for using silcrome under the guise of C. S. 93, the real reason, of course, is that silcrome was getting the bulk of the valve business away from him, and he apparently realized that he had to take up this material to survive in the competition.

The acts of Rich, cannot, of course, be stretched to represent the whole art, any more than one swallow can be said to make a summer, but Rich's activities are illustrative, if not conclusive, of the value, efficiency, utility, and economy of this new alloy, silcrome, with its important properties for exhaust valves.

Keeping in mind the state of the art, the conviction that silicon was to be avoided, and that, when the Armstrong valve patent issued, experts were puzzling over this valve problem, there was invention in Armstrong's adaptation of silcrome steel to use for exhaust valves to obtain a result which was not only new and unexpected from a silicon alloy, but which also filled a need which had long been recognized. This is so, regardless of the validity of the second patent and the third patent upon the grounds urged by defendant and later more particularly referred to.

The record here, as in the Stainless Steel Case (American Stainless Steel Company v. Ludlum Steel Co. [C. C. A.] 290 F. 103, 105), is replete with accounts of speculations by experts upon the properties of an alloy of the metals involved in the patents in suit and the necessary qualities of a desired steel alloy. As in that case, so here:

"* * * Where men, doubtless well equipped for a particular sort of work, have hoped and investigated and even prophesied as to what could be done, but never did it, and other men similarly equipped have by intensive study and skilful experiment succeeded, such success commands and should receive a greater meed of intellectual appreciation than is accorded even to the cleverness of picking up and utilizing an unconsidered or discarded trifle. When to the scientific triumph of succeeding where other scientists have failed is added the development of a new branch of industry, the word ``pioneer' may well be accorded to the patent which describes and defines, even though lamely, the essentials of such success."

Defendant's citations on lack of invention in the third patent are not in point. They are chiefly: Hotchkiss v. Greenwood, 11 How. 248, 13 L. Ed. 683; General Electric Company v. Yost Electric Mfg. Co. (C. C. A.) 139 F. 568; Drake C. P. S. L. Co. v. Brownell (C. C. A.) 123 F. 86, 87. The last was a case where a steel lug was substituted for a cast iron lug, and it was held that there was no invention. But it was said in several of the numerous citations in that case that invention exists where "some new and useful result, an increase of efficiency, or a decided saving in the operation, is clearly attained."

All these cases proceed upon the theory that there is involved no element of novelty unexpectedness, economy, or greater efficiency. But, as shown, all these elements exist in Armstrong's discoveries. Defendant, therefore, must fail as to his defenses of noninvention, not only as to the first and second, but also as to the third, or valve, patent.

The defenses of anticipation and prior public use are now to be considered. What has been said on the subject of the measure of Armstrong's discovery or invention to some extent applies to these defenses.

True it is that some prior patents and publications have spoken of the effect of silicon in a chromium steel alloy for some particular purpose, unlike valve uses, and of its probable or possible effects upon alloy steel generally. But no one had even pointed out that such an alloy had any of the properties discovered by Armstrong, other than high resistance to electric current and some nonscaling effect, and no one had suggested the use of such an alloy for exhaust valves.

In 1910 one Krupp obtained a British patent, No. 6,240 of 1910, designated "Improvements in Safe and Chamber Walls." As to this subject Dr. Waterhouse says:

"It will be seen that the patentee here had no conception that the use of silicon could raise the critical point, effect scaling at high temperatures or obtain any other properties for a valve steel. The object of the invention as stated in the patent is ``to make safe or steel chamber walls' of a steel composed of iron and carbon to which steel so large a quantity of chromium is added that it practically cannot be melted through. * * * The patentee further says that the chromium should not exceed 40 per cent. and also that ``under some circumstances it is well to add tungsten or silicon in addition to chromium.' The tungsten content may be as high as may be desired. The upper limit of the silicon content is entirely determined by how high a degree of brittleness is to be permitted."

The patent warrants this interpretation. The product of the patent never went into practical use, and it is not clear that it would be suitable, even for safe walls. That the patentee apparently did not appreciate the effect of silicon is further evidenced by the fact that, when he applied for a United States patent, he left the silicon entirely out of his formula.

Neither during the war nor at any other time did any one see that this patent covered a good alloy steel for valve or any other purposes, and neither did Great Britain or Germany, of which Krupp was a subject, make use of such material for valves or any other purpose during the stress of the war.

The only other prior art patent urged by the defendant is the Marsh United States patent No. 1,057,756, granted April 1, 1913. This patent is entitled as for an "Electrical Resistance Element." Marsh says in his patent that he had found that iron-chromium alloys were not satisfactory as resistance elements, but that he had discovered that, by adding about 5 per cent. of silicon or aluminum to such iron-chromium alloy, its durability will be greatly increased when heated by an electrical current. He claims an alloy of iron and chromium in the proportions of 3 to 1, to which is added 5 per cent. of other materials having the properties of aluminum and silicon. He says aluminum gives better results than silicon, as the resistability with aluminum is 121 microhms per cubic centimeter, as compared with 112 microhms when silicon is used in the alloy.

The Marsh patent never went into commercial use for any purpose, but in June, 1912, Marsh made experimentally a certain alloy within the patent, which was his preferred alloy, and a piece of this alloy, said to have been made in such experiments, was produced on the trial, and an analysis by one of defendant's witnesses showed chromium 22.81 per cent., silicon 3.98 per cent., nickel 1.02 per cent., and carbon .072 per cent. These figures are the average of the analyses of four samples, except as to nickel and carbon, which are the average of two analyses. Nickel was not mentioned in the patent, and so large a quantity would not ordinarily be present in the alloy by accident. It is clear that Marsh had nothing in mind except an electrical resistance material for heating appliances. He does not mention that his alloy has any of the properties necessary for valves other than the character of this scaling. He does not mention malleability, forgeability, etc., necessary for a commercial product for valve uses.

There is no proof that a material made under the patent would, except as to scaling or durability under high temperature, be serviceable for valves or any commercial use, except heating appliances, and indeed it has never been used commercially, even for that purpose. The low carbon content, also, as specific in the patent, and as shown in the alleged sample, makes it doubtful whether, on heating to a high temperature and then rapidly cooling, the material would have the hardness characteristic of alloy steels. There is no proof on that point as to a material made as specified in this patent. Nor is there proof that any one has made or can make under the teachings of this patent an alloy steel which is like silcrome as to composition or properties generally or suitability for valve uses.

Under the circumstances, this patent, though seeming to have similarity of silicon content and like properties in some respects, cannot be fairly said to anticipate. If the Marsh patent would produce a material of the properties of silcrome, that was not known to Marsh or to any one else, and what Marsh disclosed and claimed was so vague as to give no information without elaborate experiment. This patent, like the Krupp patent, antedated the World War, and no one during that time saw any usefulness of the alloy therein described.

Moreover, plaintiff filed a disclaimer as to claim No. 1 of the first patent of all alloy steels or products by which the chromium in that patent was limited to 20 per cent. and the silicon to 8 per cent. This would remove the Marsh patent from the record as an anticipatory patent, because its chromium content of 24 per cent. to 25 per cent. is beyond the 20 per cent. maximum limit of the first patent as thus limited.

The defendant insists that this disclaimer of all alloys above 20 per cent. chromium is an arbitrary and unfair division, without any distinction between alloys above and those below 20 per cent. The defendant intimates that this disclaimer is for the purpose of avoiding anticipation of the first patent by the Marsh patent. Defendant gave proof of tests as to the relative resistance to scaling at high temperatures of plaintiff's silcrome steel No. 1 of varying analyses from above 20 per cent. to below 20 per cent. of chromium, which defendant contends proves that all are very high in scale resistance. For example, one above 20 per cent., viz. No. 173, with 23.06 per cent. chromium, 2.97 per cent. silicon, and 0.19 per cent. carbon, has better scale resistance than another below 20 per cent., viz. No. 192, with 17.9 per cent. chromium, 2.90 per cent. silicon, and 0.19 per cent. carbon.

The plaintiff asserts that the conditions under which this relative scaling test was made were not the proper conditions for a reasonably accurate test, and that therefore defendant's test is worthless. However that may be, it appears that resistance to scaling is about the only property common to the whole range of silicon-chromium alloys. It is in the other qualities of the alloy that the lower range chromium alloys differ from those of the higher range, such as being more easily rolled and forged, both hot and cold, making them hard and soft at will, at the same time retaining sufficient strength and toughness. Below 20 per cent. the alloy is not so suitable for electrical resistance; above 20 per cent. it is more suitable. There seems, therefore, a reasonable basis for a line drawn at about 18 per cent. to 20 per cent. of chromium in the alloy, and the disclaimer of chromium alloy in excess of 20 per cent. is valid, and not injurious to any person, whatever be its effect in avoiding the Marsh patent.

That a disclaimer may be filed after suit brought, and be given effect without prejudice to the patent, is established. In Marconi Wireless Telegraph Co. of America v. De Forest Radio Telephone & Telegraph Co., 243 F. 560, 565 (1917), the court in the Second Circuit said:

"The contention that Fleming's patent, whatever its original merit or lack thereof, was voided by an unlawful disclaimer, is without substance. The mistake (if there was one) was in claiming something not needed, and the disclaimer abandoned what was not wanted, without broadening or enlarging any claim; it also left the claims fully supported by the original specification. No injury to defendant, or any one else, is shown. The procedure is within Carnegie Steel Co. v. Cambria Iron Co., 185 U.S. 403, 22 S. Ct. 698, 46 L. Ed. 968, and our former decisions in Simplex, etc., Co. v. Pressed Steel Co., 189 F. 70, 110 Cow. C. A. 634, and Strause, etc., Co. v. Crane Co., 235 F. at page 129, 148 Cow. C. A. 620."

Defendant asserts that, if not anticipated by the prior art patents, claim 1 of patent 1 is anticipated by the steel No. 1185 F, in the table on page 13 of the Barrett, Brown and Hadfield article of 1902, by steels Nos. 46 and 54, contained in the article on page 209 of the Burgess and Ashley article of 1911, and by the composition of stainless steel manufactured by the Atha works of the Crucible Steel Company in 1915 and thereafter.

The Barrett, Brown and Hadfield article of 1902 is a report on experiments carried on to determine the conductivity or electrical resistance and magnetic properties of certain alloys. There were only three chromium-silicon samples among a total of 110. The silicon ranged from 1 per cent. to 2.25 per cent., and the chromium from 2 per cent. to 3.5 per cent. No other properties were investigated or disclosed. The mere fact that these three samples disclosed a good electrical resistance suggests little as to their availability for any other purpose, especially for noncorrosive steels suitable for valves, or purposes where the steel must have several particular properties not found in steels generally.

The Burgess and Alston article of 1911 tells of laboratory experiments conducted to determine the conductivity, electrical resistance, and magnetic permeability of various alloys, including some with both chromium and silicon. The tables of Burgess and Alston give the resistance with and without heat treatment, but otherwise give little more information than the tables of Barrett, Brown and Hadfield.

True, Burgess and Alston state that the electrical resistance is highest in alloys having silicon present in addition to chromium, and that the resistance is greatest where the silicon content is highest. But that is about the only advance in knowledge disclosed over Barrett, Brown and Hadfield. These men were interested only in a material which had high electrical resistance for use in certain kinds of electrical machinery.

While at this date, looking backward, one might say that any one could see that such steels would have qualities making them serviceable for valves and like uses, the fact is that at that time there was no valve problem, and neither then nor later, before Armstrong, did any one see that by certain proportions of silicon, with chromium, iron, and a limited quantity of carbon, a superior alloy steel for certain purposes might be obtained. Any mention of a silicon-chromium alloy before Armstrong was limited, and was but little more than speculation and dissertations by learned men in metallurgy and steel making.

The defense of prior public use merits but little discussion. The high tungsten alloy steel valve made by Rich under his patent, No. 1,050,845, granted in 1913 and assigned to defendant, is perhaps the first alloy steel used in valve making, and replaced to a large extent the valves theretofore in use, made with a cast iron head and steel stem.

This patent is worthy of note, because of its recitals that valves up to that time were made of iron and nickel, and did not work successfully, and that a valve made of high-grade tungsten steel "retains its extreme hardness, does not warp or pit or collect carbon on its bearing surface when subjected to the high temperatures of gases developed in internal combustion engines, which gases usually reached a temperature of 1200 degrees Fahrenheit, sometimes higher."

It is clear, then, that Mr. Rich then knew nothing about the effect of silicon on an alloy steel for valves, and also that he recognized some of the properties, at least, which exhaust valves must have. The Rich patent makes no reference to silicon, but stresses a tungsten content of at least 6 per cent. as an essential element. It has been previously shown that Mr. Rich, in common with many others, thought the presence of silicon was harmful in an alloy steel, and rejected a steel containing any appreciable quantity of silicon, and that he never used an alloy with substantial silicon content until 1926, when competition apparently drove him to duplicate silcrome.

Stainless steel was used for valve purposes from about 1917. The Carpenter Steel Company sold various quantities of this steel to the Rich Tool Company and to the Thompson Products Company. The silicon in this steel was under .30 per cent., and the history of the art shows, as is apparent from Mr. Jardine's chart, that stainless steel began to replace the more expensive tungsten steel. The record shows that the Carpenter Steel Company sold no more stainless steel with low carbon content for valves after the year 1925.

The evidence of manufacture by the Atha plant of the Crucible Steel Company in 1903, when the witness offered in that behalf was a boy 17 years of age, is not at all convincing. In all these melts tungsten and molybdenum were present in considerable quantities. The defendant's evidence in this respect does not meet the degree of proof required to show prior public use.

Again in 1915 the Crucible Steel Company made an alloy steel containing according to the evidence of the manager 13 per cent. of chromium and .76 per cent. of silicon. This witness testified that this melt was made from a sample obtained from England, which had been independently analyzed at the Atha works. He further testified that in 1916 the Crucible Steel Company at its Atha plant commenced to make silicon steel in an electric furnace with carbon .45 per cent. to .55 per cent., chromium 12.50 per cent. to 13.50 per cent., silicon .60 per cent. to .75 per cent., and that that composition was followed for about seven years. But it appeared that during this seven years not very many heats of this steel were made. It appears, also, by this witness' testimony, that some time after they copied this sample from England they returned to the standard stainless steel with a low silicon content under .5 per cent., which they have since continued to make. It does not appear that this steel, even with silicon content of .60 per cent. to .75 per cent., was ever used for valves, nor does it appear that any one could have learned from the use of this percentage of silicon anything of value regarding this alloy, if such steel had ever been used for valves. It reasonably appears that a percentage in excess of .5 per cent. of silicon was something to be avoided, doubtless because it was considered undesirable to use it. The range of carbon content .45 per cent. to .55 per cent. is not very clearly proven. The witness is very doubtful about how many melts were made and about the analyses. As far as can be determined from the proof, both silicon and carbon were, except by accident, under .5 per cent. In view of the vagueness of this evidence and the degree of proof required, it cannot be said that the defendant has shown prior public use, against any of the claims of any patent.

As was said in Tilghman v. Proctor, 102 U.S. 707, 26 L. Ed. 279: "It is not an anticipation of a patented process, if the products of such process were previously produced accidentally and unwillingly while the operators were in pursuit of other and different results."

It is not claimed that the patents of Brearley and Haynes, which were the subject of litigation in the case of the American Stainless Steel Company v. this plaintiff, 290 F. 103, 108, are anticipations of the Armstrong patent. Such claim could hardly be made in the face of a decision on that very point recorded in American Stainless Steel Company v. Ludlum Steel Co. (D. C.) 16 F.(2d) 823-826.

It is not necessary to discuss the Smith and Clement applications, for the reason that Armstrong's application was the earlier filed, and that their composition contains an additional element, titanium, and that their product lacks malleability and is suitable chiefly for castings. Nor is it necessary at all to discuss the Rolls-Royce specifications or any other claims of prior use or manufacture.

The defendant must fail in both of these defenses. As was said in Pittsburgh Iron & Steel Foundries Co. v. Seaman-Sleeth Co. (C. C. A.) 248 F. 705, 708: "The defendant found in the prior art and produced in evidence formulæ and analyses of many iron and steel products, which were either within the analysis of the patent alloy or as close to it as is the analysis of the product of the defendant upon which the plaintiff bases its charge of infringement. * * * The alleged anticipating alloys nowhere showed or even suggested when made or published that they were Adamite or were metals which possessed in structure or performance the Adamite characteristics of cast iron and steel, or suggested to the art either the existence of such a metal or how to make it. If any one of the alleged anticipating alloys was Adamite, that fact, so far as the record shows, was not known to those who produced it or used it, and not being recognized as a new product with its distinctive characteristics, its production was purely an accident without profit to the art and without value as an anticipation."

If we read "silcrome" for "Adamite" we shall have an analogy. Other references supporting the views here expressed are: General Electric Co. v. Hoskins Mfg. Co. (C. C. A.) 224 F. 464; Haynes Stellite Co. v. Chesterfield (C. C. A.) 22 F.(2d) 635; American Gramaphone Company v. Leeds & Catlin Co. (C. C. A.) 170 F. 327; American Stainless Steel Co. v. Ludlum Steel Co. (C. C. A.) 290 F. 103; Chisholm v. Johnson (C. C.) 106 F. 191.

We come now to what is defendant's most important defense. The defendant asserts that as to the two later patents, the second and third (valve patent), the things claimed by Armstrong in these two later patents were disclosed in the first patent and not claimed; that the applications for these two later patents were not made until the first patent was actually issued, and therefore the two later patents are void.

It is the law that where a patentee in his earlier patent makes disclosures, unaccompanied by any claim covering such disclosures, and without reservation of such claim, or notice of intention to claim them in a later patent, the patentee is presumed to dedicate to the public all such unclaimed disclosures, and cannot later obtain a patent for them. Ball & Roller Bearing Company v. F. C. Sanford Mfg. Co. (C. C. A.) 297 F. 163, McClain v. Ortmayer, 141 U.S. 419, 12 S. Ct. 76, 35 L. Ed. 800.

The defendant claims that the application of this doctrine to this case makes the second and third Armstrong patents invalid. In support of this contention, the defendant relies particularly on the case of Swift v. Jenks (C. C.) 29 F. 642, decided by former Judge Coxe; also upon Doig v. Morgan Mach. Co. (C. C. A.) 122 F. 460; Gladding-McBean Corp. v. N. Clark & Sons (C. C. A.) 16 F. (2d) 50; Model Bottling Machinery Company v. Anheuser-Busch Brewing Ass'n (C. C. A.) 190 F. 573; James v. Campbell, 104 U.S. 356, 26 L. Ed. 786; Miller v. Bridgeport Brass Company, 104 U.S. 350, 26 L. Ed. 783; Mahn v. Harwood, 112 U.S. 354, 5 S. Ct. 174, 6 S. Ct. 451, 28 L. Ed. 665.

In May, 1919, Armstrong filed the application for his first patent, which was issued as No. 1,322,511, November 25th of the same year, an unprecedentedly rapid grant. His claim for specific formulæ under two different parts of the range of the general claims, as well as for the general claims, caused the Patent Office to reject the specific formulæ under one part of the range as not patentable in the same patent as the general claim, and the specific formulæ under another part of the range of the general claims.

The application was left with the general claims and specific formulæ or claims under the upper range, or over 10 per cent. chromium. Armstrong thereupon, on October 9, 1919, gave notice that he reserved the right to, and would, file another application for patent covering the rejected claims.

Eighteen days after the unexpected grant of the first patent, and on December 12, 1919, Armstrong filed his application for a patent covering such rejected claim, and also other formulæ, including the specific formula which he had come upon about July 2, 1919, all limiting the chromium to the lower range under 10 per cent. This application covered also a cutlery article; while referring to the use of the alloy for exhaust valves, it did not claim the valves themselves.

Under the Patent Office rules a divisional application had to be made to obtain a patent for exhaust valves made of this alloy steel, because claims for both cutlery and valves could not be included in the same patent. On February 2, 1923, divisional application for such patent for exhaust valves was made, the application retaining the original date of December 12, 1919. The lower range formulæ patent issued on May 1, 1923, and is called herein the second patent. The valve patent after some delay issued April 14, 1925, and is called the third patent.

The reservation thus made on October 8, 1919, did not appear in the first patent itself, nor in the application, because it is claimed that rule 44 of the Patent Office so forbids; but it does appear in the file wrapper of the first patent, and the plaintiff rightfully says that the file wrapper is just as much a matter of public record as either the application or the patent, and is substantially as much of a notice and warning to the public as if it were contained in the patent itself. Rule 44 of the Rules of Practice is as follows: "A reservation for a future application of subject-matter disclosed, but not claimed, in a pending application, will not be permitted in the pending application."

This rule supports plaintiff's contention. Had this patent taken the usual time in the Patent Office, the divisional application would undoubtedly have been made before the first patent was granted. But as a matter of fact the first patent was granted six months after the application was filed, an expeditious handling of an application almost unheard of in these latter days.

It is conceded by the defendant that, if the application had been filed 19 days earlier than it was, that is, before the issue of the first patent, the second patent would not be invalid for the reason claimed by it. Defendant contends that the rule applies if the application for the second patent was filed at any time after the first patent was granted. No cases cited by the defendant held a patent invalid under such circumstances as exist here.

The defendant's cases may be distinguished from the case at bar by reason of the following differences in fact:

In Swift v. Jenks (C. C.) 29 F. 642, no reservation was made in the file wrappers or elsewhere of the right to file another application.

In the three United States Supreme Court cases, a reference to the last one will show that these involved reissue patents, and were not cases involving the validity of later patents for things claimed to be disclosed, but not claimed, in the earlier patent by the same inventor. Moreover, the court said in the last of these three, viz., Mahn v. Harwood, 112 U.S. 354, 5 S. Ct. 174, 178, 28 L. Ed. 665:

"The taking out of a patent which has, as the law requires it to have, a specific claim is notice to all the world, of the most public and solemn kind, that all those parts of the art, machine, or manufacture set out and described in the specification, and not embraced in such specific claim, are not claimed by the patentee; at least, not claimed in and by that patent. If he has a distinct patent for other parts, or has made application therefor, or has reserved the right to make such application, that is another matter not affecting the patent in question."

In the case of Gladding-McBean Corp. v. N. Clark & Sons, in the Ninth Circuit, reported in 16 F.(2d) 50, there was no reservation of the right to file an application for any part of the thing disclosed or any claim in the first patent, nor was there any requirement by the Patent Office that the subject of the second patent be divided and eliminated from the first patent.

Model Bottling Machinery Company v. Anheuser-Busch Brewing Ass'n (C. C. A.) 190 F. 573, does not support the defendant's contention. That case held that a patentee of an earlier machine patent and of a later process patent for the use of the same machine could not claim the date of the application of the earlier machine patent for his later process patent; that each were separate inventions, and that the process patent, controlled by the date of its own application was void, because anticipated by a German patent of another inventor.

The court did not hold the process patent void because disclosed, but not claimed, in the earlier machine patent. The case, so far as any inference may be drawn, supports the plaintiff's contention rather than that of the defendant. Defendant cites no other case having anything approaching close application. Plaintiff's authorities have closer analogy than do the defendants.

Holmes Electric Protective Co. v. Metropolitan Burglar Alarm Co. (C. C.) 33 F. 254, was a case decided by Judge Coxe later than defendant's case, decided by the same judge, of Swift v. Jenks, supra. The later patent in the Holmes Case was held invalid, but so held because the facts were unlike those in the case at bar in important respects. The inference is that, had the facts in that case shown the distinguishing features existing in this case, the later patent would have been declared valid. In that case the court at page 258 used this language:

"It is conceded that every feature of the patent of 1871 is described in the patent of 1870. This description was given to the world in * * * 1870, without a word of reservation. * * * Nor is it the case of a patentee who is compelled to disclose his second invention in order to describe the first intelligibly, the two being distinct and not the proper subject of a single patent; though in the latter case it would seem that he should adopt some means to protect himself and warn the public. The proposition which the court here decides is that where a patent fully describes an invention which could be claimed therein, and makes no reservation, and gives no warning to the public, a second patent, granted upon an application filed months afterwards, which claims simply and solely the invention thus made public, is invalid. It is freely conceded that there is a wide diversity of opinion upon this question, but it is thought that the weight of authority sustains the foregoing proposition."

It will be seen that in the Holmes Case there was no reservation, and consequently no warning to the public that there was no compulsion to disclose the second invention in order to describe the first, and also that the two were not sufficiently distinct to be the proper subject of separate patents. In all of which respects that case differs from the case at bar.

In Eastern Paper-Bag Company v. Standard Paper-Bag Co. (C. C.) 30 F. 63, 65, Judge Colt held a patent valid, though its subject-matter was described, but not claimed, in a patent to the same inventor issued 13 months before the patent in suit was filed, and claiming a machine whose operation involved the same process. The decision says:

"Under these circumstances, has the invention been abandoned to the public? The language of the supreme court in Miller v. Brass Co., 104 U.S. 350 [26 L. Ed. 783], James v. Campbell, 104 U.S. 356 [26 L. Ed. 786], and Mahn v. Harwood, 112 U.S. 354, 5 S. Ct. 174 [6 S. Ct. 451, 28 L. Ed. 665], is cited as sustaining the proposition that the omission of Appel to claim his process invention in the machine patent was in law a dedication of that invention to the public. The supreme court in these cases was dealing with the subject of reissues under the statute. In this case we are not dealing with the law of reissues. * * * I do not understand that the Supreme Court have held that such prior description is a dedication to the public of the second invention. The invention of a machine, and a process employed in the use of the machine, being different things, it is difficult to see how the application for a patent on one should operate as an abandonment of any claim to a patent on the other: Provided, of course, the application for the second patent is made before the statutory forfeiture of two years' prior use has run. This view is in harmony with the decisions of the circuit courts where the question has arisen. Vermont Farm-machine Co. v. Marble [C. C.] 19 F. 307 [Id. (C. C.)] 20 F. 117; Graham v. McCormick [C. C.] 11 F. 859; Graham v. Geneva Lake C. Manuf'g Co. [C. C.] 11 F. 138; McMillan v. Rees [C. C.] 1 F. 722; Hatch v. Moffitt [C. C.] 15 F. 252; Cahn v. Wong Town On [C. C.] 19 F. 424; Collender v. Griffith, 18 Blatchf. 110, 2 F. 206. The patent in suit having been applied for within two years from the date of the machine patent, there was no abandonment of the second invention, though a description of such invention was found in the prior patent."

In Sugar Apparatus Mfg. Co. v. Yaryan Mfg. Co. (C. C.) 43 F. 140, the patents in suit have to do with evaporating apparatus to be used in connection with the manufacture of commercial sugar. Complainant's assignor obtained two successive patents for similar mechanisms, the latter being an improvement on the former.

Aside from the question of infringement, defendant claimed that the later patent was invalid because it covered matter disclosed in the specification of the prior patent which apparently was not claimed. Plaintiff, however, had withdrawn this matter from his first patent, though it all appeared (by inference) in the file wrapper. In sustaining this feature of the second patent, the court said at page 149, last paragraph:

"I attach no importance to the fact that the specifications of the prior patent originally embraced this subject. Mr. Lillie [the patentee] had a right to withdraw that part, as he did, and present it subsequently. I see nothing to justify the allegation of abandonment."

In the case at bar the withdrawal was compulsory and not voluntary, but the legal effect should be the same. It must be held, therefore, that defendant has failed to overcome the presumption of validity arising from the issue of the second and third patents, and that said patents are not void because they contain matters disclosed, but not claimed, in the first patent.

The matter of infringement is the only remaining question to be considered. It having been decided that all of the claims of the patents sued on are valid, the question arises as to how many of these claims the defendant has infringed. The defendant concedes that, if the claims are valid, claim 1 of the first patent and claim 3 of the second patent are infringed. No contention is made but that, if the patent is valid, claims 9 and 11 of the third patent are also infringed. All these claims may therefore be taken as infringed. Defendant insists that it has not infringed claim 4 of the second patent for the alloy steel of the specific formula of "about 3.7 per cent. silicon" and "about 8.9 per cent. chromium," nor claim 6 of the valve or third patent, which is for valves made of alloy steel containing silicon and chromium in the same percentages.

The defendant's argument is that claim 7 of the valve or third patent sets forth a formula which is slightly broader and includes the specific percentages of claim 6 of the third patent, which latter is the same as claim 4 of the second patent. Claim 7 calls for chromium 8 per cent. to 9 per cent., silicon 3½ per cent. to 4 per cent., carbon .40 per cent. to .60 per cent. The 3.7 per cent. silicon in the sixth claim is within the range of 3½ per cent. to 4 per cent. of the seventh claim; the 8.9 per cent. chromium of the sixth claim is within the 8 per cent. to 9 per cent. chromium of the seventh claim; the .46 per cent. carbon of the sixth claim is within the .40 per cent. to .60 per cent. of the seventh claim. That the sixth claim is included within the seventh cannot be denied. Does it then follow, as contended by defendant, that because there are two claims in the same patent, like the sixth and seventh, the sixth claim being more specific and included within the seventh claim, that the sixth claim must be limited to exact percentages therein recited, and, if defendant's steel does not meet the exact percentage recited in that claim, it does not infringe that claim, but, if it infringes, it infringes the seventh claim, which was not sued on, and therefore plaintiff has failed, because he has selected the wrong claim. The cases cited by defendant do not uphold its contention.

It is true that one might infringe both claims 6 and 7, if one had the exact percentages of claim 6, which would also bring it within the terms of claim 7. It may be, since several of the samples of defendant's valves exceeded 9 per cent. of chromium, that, if claim 7 were sued on, the contention would be made that in view of the latitude represented by the range of claim 7 from 8 per cent. to 9 per cent. of chromium, a fraction slightly over 9 per cent. chromium might with greater force be claimed to infringe the 8.9 per cent. of claim 6, more than it would the about 8 per cent. to 9 per cent. of claim 7.

If it were a matter of mere percentages apart from the properties, defendant's contention might be sustained, but, especially in the valve patent, the claim is for a valve made of alloy steel containing about these percentages. Of course, as elsewhere appears in the patent, the real basis of the discovery is the properties which the valve and the valve material have by reason of the component metals in the alloy, and that these properties are the important parts of the discovery, and the percentages of the component elements is only a method of procuring the properties. This view is not without some authority. Pittsburgh Iron & Steel Foundries Co. v. Seaman-Sleeth Co. (C. C. A.) 248 F. 705. It will be held, therefore, that defendant infringes claim 6 of the third patent, but not claim 4 of the second patent.

As to claims 4 and 5 of the first patent, both of these claims are for an alloy containing more than 10 per cent. of chromium, while the specifications for defendant's steels are for chromium content under 10 per cent., and the analyses of defendant's steel show under 10 per cent. of chromium, except in 3 instances out of 317. These 3 instances are not a sufficient proportion of the total number of samples to indicate that in a substantial proportion of the meltings the percentage would exceed 10 per cent. of chromium.

These claims, therefore, cannot be said to have been infringed. Plaintiff's argument is not without force that defendant has apparently carefully endeavored and nearly succeeded in keeping his chromium under 10 per cent., to avoid the infringement of the two claims of the first patent. Nevertheless that situation must be met when reached. Inasmuch as there are claims which are more specific, and which call for chromium content in the one instance of about 8.9 per cent. and the other from 8 per cent. to 9 per cent., the defendant could not escape infringement by keeping his chromium content above 9 per cent. and below 10 per cent. If there were no claims for specific percentages under and near to 10 per cent., there would be justification for holding that, where the chromium content is so close to 10 per cent. and in three instances exceeds that percentage, infringement of the over 10 per cent. claims might be found, but, under the circumstances existing here, the specifications of defendant's steel being under 10 per cent., the court does not feel justified in finding infringement of the over 10 per cent. claims 4 and 5 of the first patent.

Decree may be entered accordingly.