Appellants seek to reverse the decree of the District Court enjoining them from infringing any one of claims 13, 16, 19, 24, 28, and 38 of reissue patent No. 13,436 of patent No. 1,020,655, termed the acid, patent, and also of claims 2, 6, 7, and 9 of patent No. 1,020,656, known as the alkali patent, and for other relief; the patent last named and said patent No. 1,020,655 having both been granted March 19, 1912, upon applications filed respectively November 2, 1908, and said reissue having been granted July 2, 1912, upon application filed May 17, 1912, subsequent to the death of the original patentee, Frank G. Perkins, who died in September, 1910, testate, leaving liis widow, Gertrude S. Perkins, as his executrix, who duly qualified as such and assigned said inventions to appellee on March 31, 1911.
All of said claims pertain to starch glue and the manufacture thereof. It will be sufficient for our purposes to set out claims 7 and 9 of patent 1,020,656, which are typical of the subject-matter of this suit. Said claims read as follows:
7. The process of making a wood glue which consists in agitating a starchy carbohydrate or its equivalent with a solution of sodium peroxid and caustic soda 1o decrease the water absorptive properties of the carbohydrate without rendering the carbohydrate materially soluble in water to properly proportion the viscosity, adhesiveness and cohcsivcness resulting when the ea rbohydrate is dissolved to form glue and dissolving the product thus produced with caustic soda or its equivalent and about three parts of water or less to produce a glue l'or application.
9. The process of making glue which consists in treating a suitable amylaceous material in two stages, in the first stage treating it with sodium peroxid *66or its equivalent and caustic soda or its alkali equivalent, and in the second stage treating with caustic soda or its equivalent to form an alkaline wood glue.
The gist of appellee’s contention, as stated by counsel, is that Perkins discovered that there is a point of degeneration of starch at which, when degenerated to that degree, the starch is treated to not to exceed three parts of water and a solvent there is produced veneering glue as good as or better than animal glue, and that appellants have appropriated that, discovery and the application of it.
Some of the claims in suit cover the process and product of the .so-called first step, or glue base of the patent. Others cover the process of the so-called second step of the patent, while others cover the process and product of the complete processes and product of the patent.
The defenses are: (1) That both patents are invalid in view of the prior art; (2) that both patents are invalid by reason of prior use more than two years before the applications were filed; (3) that infringement is not shown.
[1] Owing to the facts that the inventor died’in 1910, before the amendments were made which enter into the claims in suit, and that such amendments were made' by his executrix without their being sworn to as required by statute, appellants insist that the same are invalid, as new matter. For appellee it is insisted that said amendments ai'e not new matter but merely elaborations' of the process and product claims of the original patents in suit.
While the added matters may in a sense be new, we do not regard them of so important a character as to constitute that class of new matter which would be invalid' for want of an oath under the circumstances. They are related to the original purposes and objects set out in the original specification and claims, though some of them approach the border line.
For some time prior to Perkins’ alleged invention, it was old to form adhesives by dissolving starch in different degrees of fluidity and strength, according to the use it was sought to put them to, as substitutes for animal glue. The increasing scarcity of animal substances which had theretofore been adequate and the consequent growing cost thereof stimulated this development in the art, and the result was a number of patents purporting to cover adhesives of every degree of strength and for every purpose, made from starch or starchy substances. It was also an established and well-known fact years prior to the patents in suit that raw, undertreated starch was not suitable for the manufacture of the better grades of adhesives, such as glue; and also that all grades of adhesives could be obtained from starch somewhere between raw starch and overtreated starch, just as they could be obtained from animal matter. It is the theory of the patent that the starch grain and its enveloping pellicle, are rendered less amenable to water absorptiveness by degeneration or treatment under the patent process, so that the fluent glue may result from the addition of a minimum of water, since the water must eventually be eliminated.
[2] - The two patents in suit differ mainly in the chemicals used in securing the desired degree of degeneration, the reissue patent calling-*67for sulphuric acid treatment among other chemicals, and patent No. 1,020,656 calling for treatment with an alkali- — leading the District Court to the use of the term “acid patent” with regard to the reissue patent, and “alkali patent” with regard to the other patent, designations which we find it convenient to follow. The same degenerated condition of the starch is obtained in each patent, it being claimed by the patentee that there is less danger of over degeneration in the use of an alkali. Resultants from treated starch and similar carbohydrates supply the whole gamut of adhesives from the strongest glue to dextrine and glucose according to the treatment applied and result desired. This was clearly set out in the prior art, though the record does not show that the.Perkins article was ever exactly produced prior’ to the glue in suit. Perkins experimented in the glue production from starch extensively. His present claims, as above stated, are for a veneer glue as good or better than animal glue. Por a long time prior io filing the applications for his patent he furnished an article for the manufacture of veneered barrel heads, at Poplar Bluff, Mo., and other places. This glue is not fully described in the record, but it seems dear that such as it was it proved fairly satisfactory, supplanting other articles and holding its place to the present time. It can hardly, under such circumstances, be said to have been experimental, having been supplied in vast quantities. It does not, however, yield to less than three or less parts of water and a solvent, requiring four parts of water. It appears from the evidence that it was not suitable for fine work on furniture, and the like.
Counsel for appellants thus sums up the prior art situation at the l mm the alleged invention was discovered, from the appellants’ point of view:
“At the time Perkins entered this field, it had long been old to form adhesives by dissolving starch in a suitable aqueous solution in different proportions of slareh to solution, according to the use to which the adhesive was to be put. The starch solutions so formed were used in various relations as substitutes for animal glue, as in sizings and coatings of various kinds, as well as in the gluing together of bodies of wood, and tlio like. The solvents more commonly employed were hot water and cold dilute aqueous solutions of caustic soda, the former being used where it was desir»bio to avoid staining or other injury by the caustic and the latter to dispense with heat. The caustic solvents were also sometimes heated.
"SIarches as found upon the market, then, as now, had different degrees of solubility, depending on the sources from which they wore derived, the processes by which they bad been separated from the plant and their subsequent’ treatment. Consequently, when dissolved under the same conditions, they would yield adhesives of different degrees of fluidity, dependent upon the solubility of the starch. It was established commercial practice to treat iho less soluble raw starches, winch would not, without excessive water, yield Huiiicieiiily Said solutions for the various purposes for which they were designed, with reagents to Increase their solubility, after which they were dried and placed upon the market. The degree of solubility of the starch depended on its original condition and the extent to which the treatment was carried, and varied from the substantial insolubility of raw slareh to the ready solubility of soluble dextrin or glucose. For some uses of such adhesives one degree of solubility was desirable, and for other uses, another. Thus, where & very thin, penetrating, easily absorbed coating, readily soluble in cold water, was desired, as. for ('sample, Cor envelopes, postage stamps and the like, glucose or soluble dextrin was best. For starches for certain laundry *68purposes and sizings, a material less soluble in cold water was advantageous. Different sizings demanded different degrees of solubility. By stopping the process of degeneration or modification of the starch at the proper point any required degree of fluidity within wide limits conld be attained. The process is a progressive one, as before pointed out, and since some of the starch grains are more readily affected by the reagent than others, at any particular stage of the treatment a batch contains grains in different stages of degeneration or solubility. Thus, at one point of the treatment the batch will contain dextrin along with other less modified starch granules. It had long been known at the time Perkins entered the field that dextrin was a weaker adhesive than less fully treated starch, and that, therefore, ¿o get a maximum strength of product, the treatment should ndt be carried to the point where dextrin, which is ' soluble at ordinary temperature, is formed. Therefore when an adhesive with great strength was desired, care was taken to stop the treatment at a point where little or no dextrin had been formed in the batch. Now, as before stated, it is desirable in ail adhesives to obtain the necessary degree of fluidity with as little water as possible, since the latter must be eliminated by drying in setting the material. Por this reason it was recognized as desirable to carry the process as far as'possible without the formation of the objectionable dextrin, and thus secure a maximum fluidity without detriment to the strength and adhesive qualities of the material.
“Many different reagents can be and have been employed to effect this degeneration, and among them the very reagents employed by .defendapt. We find in a single prior art patent the preparation of the identical glue base or material by degenerating starch to the same point to which defendants’ material is modified, by the same reagents and by the same method. This material, as stated by the patent, was put upon the mark.et in the same form that defendants’ material is put upon the market and was sold, to be dissolved by the user in a solvent containing the same aqueous solution, of the same reagent, of the same concentration as defendants’ customers employ. The only difference between the process of this particular patent and the practice of defendant is in the use to which the glue material is put and the consequent difference of dilution of the solution thereof. The patentee refers to the use of his adhesive for dressing or finishing, i. e., sizing or stiffening fabrics, while defendant and its customers employ the material for sticking veneer layers together. Therefore the prior art patentee uses and describes a relatively dilute solution, while the defendant employs a relatively heavy solution, just as the more familiar adhesive, animal glue, is used in dilute' solution for sizing, etc., and in heavy solution as a glue for veneers, etc. No other difference exists. It is the defendants’ view that this difference is an immaterial one, especially in view of the fact that similarly modified starch had been- dissolved in aqueous solution in substantially the same proportion^ of starch to solvent employed by defendant, to make a glue for the same purpose as that for which defendants’ glue is employed, to wit, to glue wood together, and otherwise as a substitute, for animal glue.”
This summary the trial court found to be substantially correct, subject to certain equities and presumptions favorable to appellee. In view of the fact that large quantities of starch in a natural state are in a more or less degenerated state, some much below the degree required for the manufacture of the glue of the patents^ it is not deemed possible to establish a fixed standard of treatment. Some starch must be fully treated, some partially, and some restored to a better grade. The former process is accomplished by the use of chemicals and the varying of the degree of application thereof, according to the quality of the starch. The overdegenerated must be restored mainly by the intermixture thereof with more or less untreated starch — a matter of general average. There is no way of knowing what the condition of this starch to be used is, except that when experimented with it responds satisfactorily to tests furnished by the patentee, calculated to deter*69mine whether the proper degree of degeneration is attained.' The most satisfactory of these is to apply the second step of the patent, and, if the result is the veneering glue of the patents, the operator has guessed right. If the glue is not up to standard, further treatment must be applied. In the prior use case of the Poplar Bluff barrel head, workmen claim to have been able to tell by the feeling of the glue in the hand and on the rollers of the distributing machine whether the starch was in the effective condition. At that plant no treatment seems to have been applied, except to thin the starch by the addition of raw- or partly degenerated starch, if it ran too thick, or to strengthen it by admixture of other starch if too thin. Nothing but experiment avails in the successful production of the glue base. If the patent were for the preparation of a proper glue base from entirely raw starch, it may be the processes of the two patents in suit might be valid. As it is, we see no disclosures which entitle appellee to a patent for any of his claims for the manufacture of the glue base. It is a hit or miss formula and not such a disclosure to those skilled in the starch glue or adhesive art as would enable them to practice its manufacture without experimentation. They may not be required to resort to experimentation. Panzl v. Battle Island Paper Co., 138 Fed. 48, 53, 70 C. C. A. 474; General Electric Co. v. Hoskins Mfg. Co., 224 Fed. 464, 140 C. C. A. 150; Chemical Rubber Co. v. Raymond Rubber Co., 71 Fed. 179, 182, 18 C. C. A. 31. The patents irr suit disclose no advance upon the prior art in the creation of a proper glue base. That must be discoVered anew on each occasion. For the same reasons the glue base- process cannot be deemed an invention or an original discovery, under the facts of this case.
[3] With regard to the Poplar Bluff prior use, there seem to be some unexplained causes which differentiated the glue processes of the patent from those of the Poplar Bluff practices. The appellee claims that his formulae for the patent glue base have, as above intimated, some advantage arising out of the acid and alkali processes which makes the starch less absorptive and thus more amenable to solution and smooth distribution in a minimum of water and a solvent, while in the Poplar Bluff case more liquid is and was required to secure the proper spread on the rolls. But a large body of the starch, both of appellee and particularly of appellants, was furnished in a degenerated slate, without treatment. We therefore find no merit in this contention. It will be recalled that the Poplar Bluff veneering was rough work; all that was required was that the veneering of barrel heads should hold. The veneer layers were rough and the glue was variable. There was no need of a high grade of glue. It was, however, necessary that the glue be of a consistency to flow through the pipes onto the rollers of the gluing machine. In changing from animal glue into casein and from casein into starch glue, changes of a more or less serious nature were required in the devices used in order that the work should be smoothly performed. The tendency, in view of these things, would naturally be and was to be less careful in the preparation and use of the glue, resulting in a corresponding uneven result. However that may be, the record shows that the glue used was not found *70suitable for fine cabinet work. Whether any of it was found adequate does not appear. We apply the term “glue” above to the finished prepared article, ready for application. The parties all concede that the Poplar Bluff glue base was treated to four parts of water to one of base in addition to a solvent. This, appellee insists, resulted in a too watery and weak adhesive, hardly adequate for the successful operation of the Poplar Bluff gluing machines. The object sought in the latter case was the same vas that sought in the patents in suit, i. e., a glue that was as good as go'od animal glue. The starch was degenerated. As here, the degree of degeneration was subject to all the conditions above set out.. As in the patent, so in the Poplar Bluff use, the proper article could have been obtained, and sometimes was, by experimentation. From the record it is fairly shown that after making certain changes in the glue and machinery, the result was satisfactory in that plant, and, indeed, it seems, as above stated, that the same glue is still in practically exclusive use there. It was used in such quantities and so generally that, whatever its use was, it was not an experimental use. It was an unqualified use of the glue. We are not at liberty, in the state of the record, to assume to find that there is and was not a substantial difference between the glue of the patent and that of Poplar Bluff.
In 1903 the French patent to Verneisel, No. 337,001, was granted, covering a starch adhesive termed “vegetable agglutinant” in the solid state. “It is known,” says the patent, “that by means of oxidizing agents the pellicles containing the starch can be converted into the state of substances soluble in water and in alkali, and it is upon this property that are based divers processes fori the manufacture of soluble 'starch.” After providing for subjecting the starch to various chemicals, the patent proceeds, “The product obtained by the operation just described forms an intermediate between natural starch and so-called ‘soluble’ starch,” which when shaken with water forms a milk, which indicates that the starch is held in suspense and has not gone into solution. After some further treatment, the starch becomes soluble in boiling water and also in a cold sodium solution, but does not go into solution in cold water at any time. The patentee as well as the prior art caution the manufacturers of glue against permitting the starch to be treated to the point in which dextrine is produced. For weak applications, such as backing postage stamps, the use of glucose is desirable. It is thus apparent that Verneisel was aware that different strengths of adhesives would be obtained between raw starch and dextrine. The same is true of the Gerson & Sachse German edition of the Verneisel patent for a solid vegetable mucilage, which denominates the product of the patent as vegetable glue. The trial judge was of the opinion that the Gerson & Sachse glue base was degenerated to the same degree as the Perkins glue base. If so, of course it would respond to less than three parts of water and an alkali.
The difference between adhesives, such as mucilage, sizing, etc., and glue suitable for veneering, seems to consist largely in the degree of dilution. Higgins patent, No. 579,827, granted March 30, 1897, discloses processes for the production of an improved paste or adhesive *71compound directly from common flour or starch, used in the proportion of one gallon of wafer to four pounds of starch, or two parts of water to one part of starch, which should possess greater adhesive power and durability than pastes made directly from dextrine. “It will,” says the patent, “unite wood to wood nearly as firmly as gelatin, that is, the joint cannot be separated without breaking the wood at some point.” This adhesive, further says the patent, will spread with great ease when hot. The patent further says that it would be hard “to convince the ordinary observer that it is not a remarkably white and fine glue instead of being purely a starch cement.” At line 92, p. 3, it is said:
“By arresting the digesting action earlier or later, that is, anywhere between the point where the mixture becomes a fluent jelly to the point where It becomes a thin liquid, dilt'eront degrees of body or consistency can be imparted to tile finished product.”
The patentee further says that, when from five,to seven pounds oí cornstarch to the gallon of water are used, he produces a glue like paste which is almost equal to the best animal glue, with extraordinary cementing power for fibrous substances. The paste is prepared for use. in one process. The patentee terms his products an improved paste, an adhesive compound, etc. The patentee divides the starch into five stages — -the gelatinous, (he glutinous, in which it becomes glue-like and fluent and has extraordinary adhesiveness, like glue — the soluble stage, the dextrine stage, and the glucose stage. -
Kantorowicz and Neustadt German patent, No. 88,468, issued August 27, 1896, for a process for the preparation of opened-up (degenerated) starch in the dry condition, calls for a cementitious product the power of which is claimed to be almost as great as that of animal glue.
The Brueder & Co. patent, No. 114,978, issued November 21, 1900, sets up its one claim for a process for a starch adhesive material “comparable with animal glue,” in dry form. This patent is termed “process for the preparation of a substitute for glue from starch.” “By the following process,” says the patent, “it is possible to prepare from starch a stuff that can be used as a substitute for animal glue. The process rests on an interaction between starch on the one hand and the solution of a hypochlorite of an alkali on the other. The solution of the material occurs In a way similar to the solution of glue. To one hundred parts by weight of water are added twenty to fifty parts by weight of the adhesive.”
The claim reads:
“Process for the preparation from starch of an adhesive material comparable with animal glue consisting in the following: The starch in the cold condition is mixed with a solution of potassium or sodium hypochlorite. After the reaction is ended the mass is allowed to settle, is decanted and dried.”
The patent provides that:
“If the product is to be used for dressing colored textiles, the material may advantageously he neutralized by a slight addition of ascetic acid.”
*72It is thus apparent that Perkins was advised by the prior art that to obtain veneering glue from starch he need only secure glue with a minimum of water content. It does not appear, however, that any one had obtained a good article of starch or vegetable glue as compared with animal glue. The patents cited were not in search of veneering glue but of other and weaker adhfesives. In course of this quest, several of them hit upon a starch glue good enough for veneering glue. Higgins describes the appearance of it, so he must have had it. A fair article of veneering glue was made at the time of the hearing below from the glue base disclosed in the prior art patents. There may have been some confusion of names-in the terms cement paste, etc. Starch glue, however, as compared with animal glue, does not appear to have been treated as an article of commerce prior to Perkins. Its location somewhere on the line of starch characteristics was pointed out by the condition of the -art. It was known what it was. The steps to it were practically graduated. They led unerringly to it if followed ■ — only requiring the adjustment of liquid to the requisite quality and quantity of degenerated, starch.
The case differs from the cases cited in the opinion of the District Court. Kuehmsted v. Farbenfabriken, etc., Co., 179 Fed. 701, 103 C. C. A. 243. In that case the medicinal properties of the*purified article, as well as the article itself, were unknown to Hoffman. His experiments were so full of new results as to eventuate in not only a valuable article but one having new characteristics and properties. More-ever, the new formula was made plain and certain, requiring no experimentation. In General Electric Co. v. Hoskins Mfg. Co., supra, the patent covered a new device made from old materials. The device in that case was new, and as such was sustained, and the process was plainly disclosed. Here we have a glue base which the coürt finds to be old. To this appellee adds a second step for the purpose of throwing the base into solution — three or preferably less parts of water and a solvent to one part of glue base, which results in a deliverance as good as animal glue, from starch base, a process which appellants employed surreptitious means to discover. If the glue and its production from the base were just at hand, why this underhand invasion? There was nothing new in the second step.
If the patent for the product and that for the process are valid, then it must be that tire glue of the patent is new. The District Court, having heard the evidence at first hand, found it was new. It also sustained the claims for a glue base which we do not find to be valid. The question is thus narrowed to' the following propositions: (1) Was the glue of the patents new? (2) Does the process constitute a valid invention ?
It is true: (1) That the adhesive of the prior art patents, although adverted to in those patents as glue, was never placed on the market as*a substitute for animal glue or otherwise; (2) that prior to the patents in suit, while its location on the line of starch adhesive products was pointed out, nobody actually rescued the glue of the patents from its undeveloped state.
*73From the evidence it appears that there was a demand on the market for starch glue as good as animal glue and suitable for veneering, and that appellee had built up a large business in that article; that the claims of the prior art patents, except the Brueder French patent No. 114,978, aforesaid, all were in fact limited to pastes, sizings, mucilage, and other weaker forms of adhesives; that appellee’s process was not so evident as to be discovered by appellants without subsidizing appellee’s employes: that there seems to have been great difficulty in getting at the correct formula; that Perkins labored years to get the latter just right; that, as stated by his expert, Carmichael:
•‘lie proportioned the various steps to one another according to the selection of raw material, in the conversion of the raw material and in the solution of the material, being particular about the various sub-steps by which, in the main step effecting tho solution, he obtained a uniform and homogeneous product.”
All these, considered in connection with the presumption arising from the grant of the patent, lead us to hold, though with some hesitation, that the claims covering the finished glue product as well as the process claims for that article are valid and should be sustained.
[4] With reference to the claim of contributory infringement, there is some difficulty. Appellants manufactured and sold only a glue base. In tlieir effort to get it on the marker they represented to the trade that their glue base was the same as that of the patents in suit, and advised purchasers that it could be used with appellee’s second step to make glue, as witness the letters from appellants to Gorham Bros, of January 5 and April 12, 1912, which respectively read:
“For some time we have been trying to interest you in our Solva glue, and duo to the fact that we are today making glue identically as that which you are getting from Perkins Glue Go., in other words, wé have in our employ the glue maker and chemist who were with the Perkins Go. for eleven years past, wo can furnish the same glue as what you are using.”
“This glue you can use at the rate of 2% or 2%' lbs. of water to one lb. of glue, caustic to be used 7 per cent., to the cold mixture and 3 per cent, with heat. ¡Some of tho trade are using the; 3 por cent, mixture which avoids the staining of wood.”
Can a manufacturer, by producing a glue base under the conditions of this case, which is not an invention but which may be used, and some of which, the court finds from the evidence, is intended to be used in a process which is not in itself, but only in combination entitled to patent protection, be decreed guilty of contributory infringement, notwithstanding no attempt is made by the alleged contributory infringer to so use the final step himself, and which glue base can be and is sometimes used to manufacture other commercial adhesives than glue, though appellee uses it for making glue only? We think so, and hold that appellant, the Solva Glue Manufacturing Company, was a contributory infringer of the final product and of the final process. The rule, of law in such case is that one who makes and sells one element of_ a patented combination with the intention and for the purpose of bringing about its use in such a combination is guilty of contributory *74infringement, and is equally liable with him who organizes the complete combination. Thompson-Houston Electric Co. v. Ohio Brass Co. et al., 80 Fed. 712, 26 C. C. A. 107, citing many cases.
The decree of the District Court sustaining the claims for a glue base process and product and for the so-called second step as such is reversed, and that part of it which upholds the claims of the patents for the final process and the resultant product respectively is affirmed, with direction to the District.Court to proceed further in accordance herewith.