Full opinion text
SOPER, District Judge. The bill of complaint in this case was filed on August 31, 1928, by the Fleisehmann Yeast Company, for which Standard Brands, Incorporated, was subsequently susbtituted as complainant. The bill charged the Federal Yeast Corporation, the defendant, with infringement of United States letters patent Nos. 1,449,102, 1,449,105, and 1,449,106, and prayed for an injunction against continuance of the infringement and for an accounting for damages and profits. For convenience* these patents will be hereinafter referred to as patents 102* 105, and 106, respectively. The defendant’s answer sets up the defense of invalidity of each of the patents, and also denies infringement thereof. The trial and oral argument of the ease was had during the period from October 1 to October 14, 1920. The patents relate to processes for the manufacture of bakers’ yeast and the product of such processes. A prior suit between the Fleischmann Yeast Company and the same defendant, involving United States letters patent No. 1,449,103, also relating to the manufacture of yeast, resulted in a decree in which the validity of the patent and infringement by the defendant were established. In the same litigation United States patent No. 1,449,127 was declared invalid. Both parties appealed. The decision of the District Court is reported in 8 F.(2d) 186, and the affirmance thereof on appeal in 13 F.(2d) 570. Patent 103, as well as the three other patents involved in the pending suit, was based on applications filed by Friedrich Hayduck, of Wilmersdorf, Germany, assignor to the Fleischmann Company, and were granted under the provisions of the Nolan Act of March 3, 1921, 41 Stat. 1313, USCA tit. 35, § 80. The four patents were issued on the same date, March 20; 1923. The effective priority dates, however, corresponding to the respective applications in Germany, are as follows: For patent 102, February 24,1915; 103; March 15, 1915; 105, March.16, 1915; 106, March 19, 1915. Patent 103 covered the process of manufacturing bakers’ yeast, which comprised the preparation of a yeast nutrient solution containing sugar material with inorganic ammonium salts as an additive source of nitrogen, and, during the period of propagation, neutralizing, by an antacid substance, the deleterious acidity created by the use of the salts. Some description of the nature of yeast apd of the ait of yeast manufacture may be found in the prior opinion of this court. The discussion therein, infringement being admitted, was confined chiefly to the 'validity of the patent in relation to the prior art. The gist of the patent was the discovery that inorganic ammonium salts may be used to great advantage in the yeast industry, and that the harmful acidity which attends their use may be overcome by the process of neutralization. The trial of this case was held during the week of October 7,1924, and the decree of the court was entered on May 9, 1925. After it had been affirmed on appeal, an order was passed on July 24, 1926; at the request of the parties, referring the ease to a special master to state an account of the profits and damages to which the plaintiff was entitled. The report of the special master was filed on November 16, 1929, and exceptions thereto were filed in the following month. Before the hearing on these exceptions took place, the second case involving the three remaining patents was heard. Subsequently, on December 27, 1939, the exceptions to the master’s report in the first case were argued before the court, and it was then stipulated that the evidence taken before the master in the accounting proceedings and that before the court in the second case might be used by the court interchangeably in either case. It will be necessary to discuss each of the three patents at length in view of the character of the defenses raised. For convenience, however, it is well now to state generally the character of each patent: Patent 103 resembles the process of patent 103, in that it involves the use of a nutrient solution, consisting of essentially sugar material and inorganic ammonium salts, but differs from patent 103 in that it seeks to overcome the harmful effeet of mineral acidity by the high dilution of the nutrient solution instead of by neutralization. Patent 105 has reference to the method in which the nutrient material is handled during the period of propagation, and is applicable generally to worts suitable for the growth of yeast. In the method usually employed in yeast factories prior to the adoption of the invention, all of the nutrient material was placed in the propagating vat at the beginning of the operation, and such a mash is described in this case as a straight set mash. By the method disclosed in patent 105, this process is abandoned and propagation of the yeast is initiated in a dilute portion of the wort, and thereafter the remainder of' the wort of higher concentration is added substantially continuously during the period of propagation. Patent 106 combines the disclosures of patent 103 and patent 105. There is employed the manipulative feature o£ patent 105 in the propagation of yeast in a nutrient solution consisting of sugar material and inorganic ammonium salts, together with the neutralization feature of patent 103. The detailed description of the patents in suit and a discussion of the defenses of invalidity and infringement may be considered most conveniently in the following order: Patent 105, patent 106, and patent 102. Patent 105. The important feature of the process disclosed by this patent relates to the method in which the mash or wort is introduced into- the propagating vat. A feeding process, referred to in the testimony as the Zulauf feed, is prescribed, whereby a comparatively small part of the nutrient solution is diluted, placed in the vat and seeded with yeast, and, after it has begun to grow and propagate, the remainder of the wort is fed in gradually to supply its increasing needs. As we shall see, this change has been accompanied by a great increase in the yield of yeast from raw material in the modern yeast factory. The experts now agree that, in the growth of yeast, it is desirable to have neither too much nor too little of the necessary foods. During the propagating period, the microscopic organism of yeast grows to a certain maturity, and then divides or splits to form additional cells, so that under ideal conditions there is a progressive geometric increase oin the number of cells until all the nutritive material has been consumed. It is obvious that, unless sufficient food is produced for the growing organisms, they cannot thrive to the best advantage. On the other hand, it is equally important that not too- great an amount of food be introduced. There must be sufficient space in the propagating vat to enable the respective cells to reach maturity without undue interference with each other; for it is known that, in the course of growth, the cells excrete alcohol and carbonic acid gas, which have a poisonous effect when present in too great quantities upon other cells in the neighborhood. Furthermore, it is equally well settled that too great a supply of sugar material will bring about the production of more alcohol than the yeast can assimilate and thereby retard the growth-of the yeast. It has been picturesquely said that the Hayduek procedure avoids the extremes of feast and famine, and that it aims to maintain throughout the period of propagation an optimum concentration of materials. The specification expressly declares that the patent relates to a low alcohol yeast process. Prior workers in the field, it is admitted, had already pointed out that, to accomplish this purpose, the wort should be diluted to such an extent that the alcohol produced from the sugar could be assimilated by the yeast as nourishment. It had also been disclosed by the Rainer patent, hereinafter 'discussed at length, that one method of-achieving the desired result was to add the peptone (nitrogenous material) continuously, and the sugar at intervals as- required. The particular method disclosed by the Hayduek specification is substantially as follows: A wort having, for example, a gravity of 12° Balling is prepared according to the procedure commonly followed in the industry. A part of the wort is placed in the vat and highly diluted, for example, to about Io Balling. Seed yeast is added and aeration follows. Thereupon the remainder of the concentrated wort is added slowly and substantially continuously at such a rate that the alcohol formed in both portions of the wort is immediately assimilated by the yeast. The quantity of the diluted wort originally placed in the vat is so chosen that the mean content of the entire wort would be from 3° to 5° Balling. The patentee states that by this method the yeast is constantly given new supplies of all nutriment necessary for its propagation, and there is always a wort concentration by means of which the nutriment is absorbed most efficaciously and quickly by the yeast. In other words, there is offered the advantage of great dilution, while the concentration of the wort calculated for the entire quantity remains the same. The advantage of this procedure is contrasted with that customarily used in the manufacture of yeast, for therein propagation begins in a relatively concentrated solution and the concentration is decreased gradually during propagation. The plaintiff relies on all the claims of the patent, of which claims 4, 6, and 10- are considered typical. They are as follows: “4. A process of propagating yeast with a relatively low yield of alcohol which comprises preparing a wort containing all essential yeast nutrients in solution, initiating propagation of yeast in a diluted portion of said wort, aerating the diluted portion, and substantially continuously adding during the period of propagation, a wort of higher concentration, at a rate such that the yeast may propagate and substantially all of the alcohol which may be formed is assimilated by the yeast.” “6. A process of propagating yeast with a relatively low yield of alcohol which comprises, preparing a wort containing all essential yeast nutrients in solution, initiating propagation of yeast in a diluted portion of said wort, aerating the diluted portion and slowly and substantially continuously adding the wort of high concentration, which addition is effected until near the end of the period of propagation.” “10. A process of propagating yeast with a relatively low yield of alcohol which comprises, preparing a wort containing all essential yeast nutrients in solution, initiating propagation of yeast in a highly diluted portion of said wort, aerating the diluted portion, and substantially continuously, adding during the period of propagation, the wort of higher concentration at a rate such that the concentration of the dilute wort remains substantially constant, whereby substantially all of the alcohol which may be formed is assimilated.” Claim 6 is typical of claims 1, 2, 3, 5, and 7, while claims 4 and 10 are illustrative of the remaining claims of the patent. A consideration of these claims shows a distinction between claim 6, on the one hand, and claims 4 and 10 on the other. Claim 6 contains the following three elements: (1) Preparing a wort containing all the essential yeast nutrients in solution; (2) initiating the propagation of yeast in a dilute portion thereof; (3) substantially continuously adding a wort of higher concentration during the period of propagation. Claims 4 and 10, on the other hand, contain another element in addition to those above described, namely, a description of the rate at which the concentrated part of the wort is added to the more dilute portion. In claim 4 the rate is described as such that the yeast may propagate and substantially all the alcohol formed may he assimilated by the yeast; while in claim 10 the rate is such that' the concentration of the dilute wort remains substantially constant, whereby substantially all the alcohol formed is assimilated. It is obvious from the foregoing description that the maintenance of the food supply at the proper level by regulating the rate of addition is a most important feature of the patent; and hence claims 4 and 10 cover the invention much more comprehensively than claim 6. Indeed the parties have directed the greater part of their discussion to claims 4 and 10 in. considering the validity of the patent. This matter will now be discussed in the light of the prior art. The Validity of Patent 105. The Rainer Patent. The defendant’s chief reliance to support the charge of invalidity is the German patent of Rainer of 1879. Hayduek refers to this patent in his specification as one which has the same object in view and employs a somewhat similar process. The defendant goes further and asserts that the Rainer and Hayduek patents are on exactly the same basis, that both contemplate the starting of propagation in a wort of high dilution, so that yeast production is favored at the expense of alcohol, and that both direct that throughout the greater part of the period of propagation nutritive substances should he supplied to the fermenting liquid continuously and at such a rate that the solution in the vat may be kept substantially constant at the chosen concentration. If this is the. case, the Hayduek patent necessarily fails, since it is anticipated in every important point; and hence it will he necessary to examine this reference with care. The details of the Rainer method are disclosed inIhe patent in some confusion; but the preparation and treatment of the mash seem to he substantially as follows: Hirst a process of peptonization is employed in which vegetable albuminous substances from cereals, corn, legumes, or from industrial refuse,' such as brau, malt, sprouts, etc., are treated with acid and a diastaee, for the purpose, as explained in the testimony of the experts,a of breaking down or degrading the complex chemical compounds in these substances into simpler forms so as to render them assimilable by tbe yeast. The preparation of the sugar material necessary for the propagation of yeast, and the addition thereof to the pep-tone material is then described in the following language: “Corresponding to the peptone content of the liquid, there is added an amount of liquified starch which has been transformed into dextrose by cooking in the usual manner with addition of mineral acid; or maltose, molasses or, finally sugar, equal to about half of the dry weight of the peptone, together. with some pitching yeast (beer yeast or pressed yeast), for initiating the growth of the yeast; however, the carbohydrate thus added must always amount to no more than % of 1 per cent, of the total weight of the liquid; it then serves exclusively for the formation of cell tissue.” The patent then describes at length the temperatures at which yeast may be best grown and the aeration which is necessary far efficient propagation. There is next set out the following paragraph, which is the first suggestion in the document that additions may be made during the propagation of the yeast to the mixture of peptone and sugar worts above described: “Depending on the larger or smaller density of the nutritive liquid employed, on the higher or lower temperature of this liquid, and on the ozonized air that is introduced, the growth of the yeast after each larger addition of dextrose, maltose, etc., is finished within 6 to 8 hours. The peptone content of the mass may be 1 to 2 per cent, or more of its weight, whereas of dextrose or other carbohydrates there should be added each time only % to 1 per cent, in order to prevent the occurrence of lactic or alcoholic fermentation and to surely prevent such losses in my process. “When, after 6 to 8 hours, the added dextrose or other kind of sugar has been wholly or for the most part consumed in the growth of the yeast, a further portion thereof, amounting to %o to )4o per cent, is added; and similarly the peptones, after they have been consumed, may be added in portions or run into the liquid slowly and continuously, so that the same nutritive solution, by successive replacement of the substances consumed, may remain in use for weeks and months insofar as it is not contaminated by other substances or by side fermentations and thus rendered unsuitable for further use.” The procedure thus outlined is referred to in this case as Earner’s first illustrative example. There follows provision for the successive removal of the yeast as grown from the fermenting vat, and then a second illustrative example or method is given in the following -language: “In order to obtain a yeast as pure as possible and free from side ferments, the nutritive liquid may, to be sure, be made still thinner, that is, prepared with a peptone content of % to 1 per cent, and the carbohydrates (dextrose, maltose, etc.) also in smaller amounts (the first portion about. per cent., then about V20 per cent, every 3rd hour) ; then, when there is introduced a sufficiency of pure air that has been led through sulphurie acid, or of ozonized air, the pep-tones that are present will for the most part be transformed into yeast within 12 to 15 hours at a temperature of 16 to 17° E. The entire liquid is then cooled by means of a cooler or by addition of cold water or ice to about 6 to 8° if possible; when the depth of the liquid is from 86 to 961 centimeters the yeast settles to the bottom almost completely within 12 to 14 hours, whereupon the exhausted nutritive liquid may be disposed of by way of the sewer.” After this example, directions are given for the purification and pressing of the yeast, and finally the yield of yeast to be expeeted from the processes described is expressed in the following language: “From each 166 kilograms of pure pep-tone (dry weight) there are produced, according to experience thus far, -256 to 306 kilograms of pure and strong pressed yeast for the production of which there are required only about 206 kilograms of dextrose or sugar.” It is not easy to follow the sequence of the steps in this procedure. What is meant by the phrase, “the growth of yeast after each larger addition of dextrose, etc., * * * is finished within six to eight hours,” which is found in the second quoted paragraph of the patent. Only one addition of sugar material is previously mentioned in the patent, to wit, that of the so-called sugar wort to the peptone wort. Perhaps this paragraph was intended to follow rather than precede the succeeding paragraph in which a smaller quantity of sugar (one-twentieth to one-tenth per cent, instead of one-half per cent.), is added at the end of the sixth to eighth hour when it would seem that the growth of yeast from the materials first supplied in the pep-tone wort and sugar wort has been finished. If this view is correct, then the patent means that first the peptone and sugar worts should be combined to form a nutritive solution with a peptone content of one to two per cent, of the weight of the mass, and a sugar content of one-half of this amount, and that the mixture be then seeded, and used as an ordinary straight set mash for the growth of the yeast. Accepting this construction of the first step in the successive replacement process, the next query is, how many additions of one-twentieth to one-tenth per cent, of sugar are to be made during the subsequent growth of the yeast, and what amount of peptone is to be added in portions or slowly and continuously, and, if in portions, at what intervals. The answer may be that the sugar is to be added in the quantities mentioned at intervals as consumed, and the peptones either at intervals or continuously in such quantities that at no time shall the ratio of sugar to pep-tone exceed one to two, and that, at the end of the growth of each crop of yeast, it will be found that the amount of each of the worts actually consumed will be in the ratio of one of peptone to two of sugar. This seeming contradiction in the ratio between the substances actually consumed and that existing during the growth of the yeast is due to the fact, as the experts agree, that all of the peptone in the material is not assimilable by tbe yeast, and that consequently the sugar is consumed faster than the peptone. This reading of the patent would seem to be the one. most favorable to the defendant’s contention. Yet, if it be accepted, it is still not clear that Rainer’s patent contained the elements whieh constitute the substance of the Iiayduek procedure. It does not appear that the ratio between the materials remains constant throughout the propagation of the yeast. Nor does Rainer show that there is any particular virtue in the continuous addition of the nutritive material, as consumed, rather than the replacement of it at intervals, when consumed. It seems to be certain that during the first interval of six or eight hours, the peptone and sugar material are mixed together as a straight set mash. It is not explained why the feeding process is substituted for the straight set mash in the subsequent intervals of the replacement process. If the yeast grows satisfactorily during the first interval of six or eight hours in a straight set mash, so long as the amount of sugar is kept down to a point where all of the alcohol is consumed by the yeast, it might be supposed that the same method could be safely used in the succeeding intervals of the continuous process. Moreover, it clearly appears in the second example that only the sugar is added at intervals, whilst all of the peptone is placed in the vat at the beginning of the period of propagation. The patentee seems to prefer this process, for he advised that it be used when it is desired to obtain a yeast as pure as possible and free from side ferments— that is to say, from alcoholic fermentation. Not only is all of the peptone added at the beginning of the period, but the sugar itself is not added continuously but only in portions at three-hour intervals. There is a total sugar content of about one-half per cent., with pep-tone of three-fourths to one per cent., but it is obvious that, during the twelve to fifteen hours prescribed for the growth of the yeast in the second example, there is a continuous diminution in the quantity of the peptone material as it is consumed by the growing yeast, and a fluctuation in the amount of sugar from a maximum sugar content at the beginning of each three-hour interval to a minimum sugar content at the end of each such interval. It does not seem possible to conclude either from the first or the second example, or from both together, that the patentee had discovered that either a continuous addition of materials or a relative constancy of concentration was particularly desirable. The main admonition of the patent was that the operator should not at any time add so much sugar that the alcohol produced could not be consumed by the yeast. This conclusion is fortified by the disclosures of Rainer’s Belgium patent, whieh was granted on October 28, 1879-, shortly after the issuance of the German patent on October 12, 1879. The two patents are obviously based on the same subject-matter and have the same design. The Belgium patent, like the German, contemplates that, if desired, the same nutritive liquid may serve for weeks or months by successive renewals of the substances consumed, but the addition of the nutriment, either at intervals or continuously during the growth of the yeast, is not directed. The operator is advised that the proportion of peptone may be 2 to 3 per cent, of the weight of the mass, or even more, where as more than 1 to 1.25 per cent, of sugar material should not be added each time in order to prevent the production of alcohol. In place of the third quoted paragraph from the German patent, the following is set out: “When, after 6 to 8 hours, the dextrose or other added saeehariferous substanee has been entirely or for the most part consumed in the vegetation of the yeast a new portion thereof is added, and in the same way there may be added to the liquid portions of the peptones when these substances have been consumed, in such a way that the same nutritive liquid, by successive renewals of the substances consumed, may serve for weeks or months provided that it is not rendered impure by other substances and unsuitable for renewed use.” Hence it would seem that the additions of nutritive substances were to be made only at the expiration of the period during whieh the yeast had been fully grown, and that the patentee did not consider the addition of materials either at intervals or continuously during growth of sufficient importance to be inserted in his Belgium document. He obviously contemplated a succession of straight set mashes and a progressive diminution of concentration therein during each six to eight hour period. Substantial constancy of concentration during the period of propagation was not achieved. The defendant points out that there is some inconsistency in the description by Hayduek of the Rainer process as it is found in the German counterpart of the Hayduek patent 105, and in the original application for United States patent, on the one hand, and the description q£ the process as it appears in the final amended specification of the United States patent on the other. Whereas in the first-mentioned documents the feed is described as being either continuous or at intervals, with reference to both the peptone and the sugar material, the feed is described in the last-mentioned document as being at intervals with regard to the sugar material and continuously in regard to the peptone. The interpretation last given is, however, the correct one, and the differences serve chiefly to show the difficulty of understanding the Rainer patent. The defendant also urges that the differences which Hayduek points out in his specification of patent 105 between his process and that of Rainer pertain particularly to matters other than continuity of feed and constancy of materials. Hayduek says in susbtanee that (1) the Rainer procedure requires analysis in order to determine when additions of material should be used; (2) the Rainer procedure requires large tanks and apparatus, a conclusion based upon the idea (probably mistaken) that the Rainer process contemplated the addition of the material in a highly diluted form rather than in a concentrated form; and (3) that t£e Rainer process did not provide for a wort of usual constitution because of the peptone sugar ratio of two to one. The. defendant urges that none of these distinctions of Hayduek are based on solid foundation; that materials could be added in a concentrated form rather than in solutions of abnormal size; that analysis would hardly be needed since the action of the materials could be foretold and such supervision or tests as were necessary would be simple and quickly made; and further that the actual constituents of the wort, as consumed by the yeast, would be in a peptone sugar ratio of one to two, which is customary in the art. It may be conceded in this case that Hayduek placed an exaggerated importance upon the items discussed. But it does not follow that his patent may not be differentiated from that of Rainer on the two main particulars which have been already discussed. That discussion has shown that Rain-er had no clear perception of the advantages to be derived from a continuous feed and a substantial constancy of material constituents of the wort, and that, on the other hand, Hayduek disclosed these elements. The defendant’s subsequent references to the prior art are far less explicit than Rainer. There is cited from La Far’s Handbueh der Teehnisehen Mykologie, published in 1907, an excerpt entitled “The Character of the Nutritive Solution.” It relates to the production of alcohol. The author shows that taxation of the alcohol industry is sometimes regulated by the size of the fermenters employed, and hence in such cases it is to the interest of the distillers to make their mashes as concentrated as possible. There is, however, a limit to this expedient, because too great a sugar content will impair the growth of the yeast and interfere with its fermentative powers. To- obviate this difficulty, manufacturers in molasses distilleries resort to a so-called drauflassen process; that is, they establish in the molasses such a dilution as experience shows will still admit of a smooth progress of the fermentation, and then gradually add thick molasses to the fermenter as the sugar present therein is fermented. The object, of course, is not to prevent but to promote the formation of alcohol. The article then states that in the breweries the process is not infrequently resorted to when the amount of seed yeast is insufficient and has to- be quickly multiplied. It is plain that the object in view is the manufacture of yeast, not as an end product, but for use in the fermentation of a beer mash. The article teaches that too much sugar in the wort makes alcohol at the expense of the yeast, and that the difficulty may be overcome by gradually adding to a solution of molasses of a dilution not clearly stated a thick material as the sugar is used up. But there is no statement that the process is practicable in the commercial manufacture of yeast. Nor is it clearly set out that the addition should be made continuously, and no suggestion of Hayduek’s idea is found, that a constancy of concentration of all of the constituents necessary for the gro wth of yeast should be maintained by regulating the rate of the continuous feed. The Woehenschrift für Brauerei, which, appeared January 18, 1903, contains an article by M. Delbriick entitled “Some Fundasmental Principles of Yeast Multiplication.” This publication is evidently devoted to the interest of the brewing business, and the article relates to the use of yeast for its fermentative properties in the production of alcohol rather than the propagation of yeast as a final product. The author shows that yeast cells in a mash mutually exert a distributing effect on one another because of poisonous substances excreted by the cells, consisting chiefly- of alcohol and carbonic acid. It is desirable that there should be a uniform quiet motion in the liquid to insure the uniform distribution of the yeast cells throughout its entire area. If there is.such motion, the individual cells are moved away from the poisonous products in their immediate neighborhood. Motion can be evoked by a stirring mechanism, or by the manner in which the process of fermentation is started. ' If seed yeast is added to a small amount of the liquid to be fermented, it at once brings the liquid into active fermentation and movement because the cells are present in excess. If there is then added to this liquid the main part of the liquid, the strong movement is at once transferred to the latter. “For this reason the strongest multiplication is given by the continuous fermentation process according to which the pitching yeast is first placed in a large fermenting vat along with a little wort and then a stream of fresh nutritive liquid perhaps distributed over a period of twenty-four hours is run in until the vessel is full.” The defendant contends that this is a clear disclosure of a process for the propagation of yeast in which seed yeast is placed in a small part of the wort, which becomes more dilute than the reserve portion by reason of the consumption of nutrition by the yeast; and that thereafter, the more concentrated portion is continuously fed into the mixture for a period of twenty-four hours. The suggested procedure obviously involves a multiplication of yeast in a continuous feeding process; but the article does not clearly direct a high dilution of the initial liquid as compared with the reserve portion to be gradually added, or that there should be a substantial constancy of concentration of food stuffs to make a balanced feed. Even if the article be considered solely with reference to the light which it throws upon the best multiplication of yeast, it is obvious that the author emphasizes the desirability of imparting to the nutritive liquid a certain constancy of movement of the yeast cells so as to separate them from each other and distribute them as evenly as possible throughout the liquid rather than upon a regulation of the quantity or quality of the food supply. Other references of the defendant to the prior art may be briefly mentioned. An excerpt entitled “The Building IJp of Yeast” is taken from the Brennerei-Lexikon, a publication devoted to the distilling industry. The propagation of large amounts of yeast from small quantities of seed yeast for use in beet and molasses distilleries is discussed. Propagation is begun in a small part of the wort, not diluted, and after it has been attentuated to one-half its concentration by consumption of the nutriment, fresh wort is gradually added from time to time. It does not- appear that the author regarded -the process as a suitable one for making yeast as such, for the same publication contains elsewhere directions for the commercial manufacture of yeast which make no reference to the addition process. There is no statement that the addition is made to prevent alcoholic fermentation, or that the writer appreciated the advantages of a continuous over an interrupted feed and the maintenance of substantial constancy of concentration. The British patent to Blake of 1861 covers a process for making beer. The process is started in the larger part of the wort, and, after one-fourth to three-fourths of its saecharif erous matter has been decomposed, the remainder of the wort is gradually added in a continuous stream. The purpose seems to have been to improve the flavor and keeping qualities of the beer. No reference is made to the manufacture of yeast. The British patent to Vignier of 1904 is entitled, “Improvements in the Fermentation of Wines* Wash or Beer and in the Making of Yeast for Distilleries.” It shows that, when a wine of high alcoholic strength is desired, the must may be introduced by automatic fractions into the vat, and thereby rapidly converted into alcohol. The patent involves a certain regulation of food supply, and it may be assumed that a similar procedure might be applied in the manufacture of yeast. But the only dilution of the-initial wort is that caused by the conversion of the material into alcohol; the process is discontinuous, and there is no attempt at constancy of concentration. The United States patent to- Claudon and Yigreux of 1883 discloses a process for the-production of a fertile yeast from the ordinary mixed wort of grain and molasses, for use in the fermentation of other similarworts, or for bread making. A quantity of yeast generally used in distilleries is diluted-in a portion of the wort in the fermenting vat, and the vat is then gradually filled up-with the rest of the wort. The time consumed in the Ailing process would seem to be about three hours, for it is said that the vat is filled in two hours after fermentation has become-fully active; that is to- say, after the expiration of about an hour. The total time consumed in the process is 97 or 98 hours. The-defendant’s expert offered in evidence the report of an experiment purporting to have-been performed in accordance with the patent; but actually thq test differed from the-disclosure in important particulars, and therefore throws no light on the issues involved in this' case. The patent merely shows that it is feasible to begin the propagation of: yeast with a part of the wort and then to feed in the remainder gradually during a very small part of the entire time consumed by the process. There is no dilution of the initial part except such as invariably follows from the consumption of the food materials by the yeast. No attempt is made to maintain a substantial constancy of concentration. Indeed, during by far the greater part of the period of growth, the wort is a straight set mash. The German patent to Kamienski of 1912 covers a process for the preparation of pressed yeast. The gist of the patent is a second seeding of pitching yeast which is added along with concentrated wort while the fermentation is not yet finished. The patent does disclose the addition of a concentrated to a highly diluted portion of the wort, but there is no indication as to the rate of addition; nor does it appear that the feeding process is continuous or the concentration constant. When these scattered citations are considered as a whole, it appears that certain of the features of patent 105, taken separately, were known to the prior art. The division of the wort into two parts, the beginning of propagation in one part, which had been diluted, and the addition of the other part continuously or at intervals, the restriction of alcohol production by regulating the supply of sugar — these several suggestions may be gathered together from the literature. But no publication shows the combination of elements found in claims 4 and 10 of the patent. These claims contain not merely the negative admonition that the sugar material be not fed too fast to the growing yeast. They indicate, when considered in the light of the specification, a process wherein the yeast is constantly given new supplies of all necessary nutriment by a continuous addition, so that there is always a wort concentration for the efficacious growth. The regulation of the rate of addition so that the yeast may propagate and substantially all the alcohol formed may be assimilated, and the maintenance of substantial constancy of concentration of nutriment, through regulation of the rate of addition, are not found in any reference in combination with the other elements expressed in these claims. It does not appear to the court that the information afforded the inventor by the prior art was so clear and complete as to anticipate the patent. Inventions cannot be invalidated by the mere introduction of prior publications .unless the description is so full, clear, and exact as to enable any person skilled in the art to make and practice the invention, and the rule is enforced with particular strictness when the disclosures are found in foreign patents or publications. Permutit Co. v. Harvey Laundry Co. (C. C. A.) 279 F. 713, 718, 720; Loew Filter Co. v. German American Miter Co. (C. C. A.) 164 F. 855, 860. Selectasine Patents Co. v. Prest-O-Graph Co. (D. C.) 267 F. 840, 842; Bone v. Marion County, 251 U. S. 134, 40 S. Ct. 96, 64 L. Ed. 188; and cased cited in 8 F.(2d) 186, 197. The commercial success of the process disclosed in the Hayduck patent 105 tends to show its validity. Where there is doubt as to the clearness and completeness of prior publications, the commercial success of a patent, especially in a crowded field, may be taken as strong evidence that the matter disclosed was not obvious to the ordinary person skilled in the art, for otherwise it would have been reduced to practice. 8 F.(2d) 196, 198, and authorities there cited. It was shown in the earlier ease that, by the use of the process of patent 103, the yields in the plaintiff's factory were increased from 35 to 65 per cent, on the raw material. The present method of the plaintiff is to combine the process of patent 103 with that of patent 105, and the result has been a substantial increase in yield, so that there is now obtained an average of 80 per cent, of yeast from the raw material. The defendant has also achieved a corresponding increase in the yield in its factory by the adoption of the Zulauf process. It is admitted that these yields are unprecedented in the. yeast industry, but it is suggested that the increase is not entirely due to the additive method of food supply. Some part of the increase may have been due to the growing efficiency of the plaintiff’s manufacturing operations; .but the greater part of the increase cannot be accounted for except by the change of manufacturing method which the adoption of the present invention has brought about. It is not unreasonable to contrast with these results what may be expected from the employment of the Rainer process. This was calculated by the plaintiff’s expert on the basis of an experiment performed by the defendant and introduced in evidence in which the raw material for the peptone mash consisted. of sprouts. Using the method prescribed in Rainer’s second example, from which it appears that only one-fourth of the peptone in the material was assimilable by the yeast, and three-fourths thereof was lost in the process, it was estimated that a yield of only 5 per cent, on the raw materials could be expected of the Rainer process and that the cost was therefore prohibitive. This testimony was disputed, but, without discussing the matter in detail, it is sufficient to say that the evidence that the yield would he unsatisfactory commercially is convincing. Moreover, there is the statement in the Handbuch der Presshefen Fabrikation by Kiby, one of defendant’s references, that the Rainer process was not successfully exploited on a large scale, because the attempt was thoroughly unsuccessful. The admissibility of this statement as proof of the faet stated is open to question; but it may be considered in estimating the anticipatory value of the Rainer reference. Certainly a skilled worker, with the Rainer patent before him, together with the statement of its impracticability, would not be justified in concluding that the disclosure of the patent was of much practical value. On the contrary, he would he led to search for the defects in the process so as to take the last step in order to transform failure into success. As the evidence indicates that Hayduek has rendered this service, the patentability of his disclosure is established. It has been pointed out that claim 6, which is illustrative also of claims 1, 2, 3, 5, and 7, is broader than the claims whieh have just been discussed, in that it omits any reference to the rate of feed. The defendant contends that these claims are invalid because, when literally construed, they cannot be distinguished from the information furnished by the prior art already discussed. But this conclusion does not necessarily follow. Meritorious patents are frequently saved by interpreting the language of the claims so as to include restrictions on limitations appearing in the specification. Bisight Co. v. Onepiece Bifocal Lens Co. (C. C. A.) 259 F. 275, 276; Van Ness v. Layne (C. C. A.) 213 F. 804, 807; Fowler & Wolfe Mfg. Co. v. McCrum-Howell Co. (C. C. A.) 215 F. 905, 909; Permutit Co. v. Wadham (C. C. A.) 13 F.(2d) 454, 458; Gibbs v. Triumph Trap Co. (C. C. A.) 26 F.(2d) 312; Smokador Mfg. Co. Inc. v. Tubular Products Co. (C. C. A.) 31 F.(2d) 255, 257. Whether it is necessary or proper to apply this rule of interpretation to claims 1, 2, 3, 5, 6, and 7 in order to save them, need not he decided. Indeed it is not necessary to pass upon either the question of validity or upon the question of infringement of these claims. It appears from what has already been said that the other claims of the patent, limited as they are by reference to the rate at which the concentrated wort is added, are valid, and it will hereafter appear that they have been infringed by the defendant. Hence the question raised as to the broader claims is purely academic. Infringement of Patent 105. In order to prove infringement of this patent, the plaintiff put in evidence seven mash sheets of the defendant company whieh show actual factory operations from 1923, when defendant was. infringing patent 103, to the present day. These mash sheets are identified by the following numbers, to wit:' 245, 420, 668, 1094, 1368, 1901, and 2086. Mash 2086 is offered in evidence only in connection with patent 102. The plaintiff claims that all of the other numbered mashes infringe patent 105. Of these mashes, five may he classified and treated together, to wit, 245, 420, 668, 1094, and 1368. The operations involved therein are distinguished by the fact that the reserve portion of the wort is fed into the initial portion in a single stream, whereas in mash 1901, the reserve wort is divided into two portions in similitude to the Rainer process, as will hereafter appear. A description of mash 245 will serve for the five mashes classified with it. A mash or wort, consisting of grain and molasses, together with sprouts and sour stock, is made up in the usual way. The wort is divided into an initial fraction highly diluted to a balling of 1.9° and a reserve part of high concentration, whieh is not specified in figures in mash 245, hut ranges in other mashes from 15° to 19.8°. Ammonium phosphate, whieh is an inorganie ammonium salt, is added to the initial portion of the wort. Beginning with the second hour and continuing until the twelfth hour, the reserve portion is fed into the vat continuously. There is no direct evidence as to the rate at which the feeding process occurs, or as to whether the rate increases during the growth of the yeast. It is stated in defendant’s brief that the rate is constant. The balling of the mixture falls from 1.9° to 1.2° between the first and the fifth hour. It then increases from 1.1° to 2.2° between the fifth and the twelfth hour. During this last-mentioned period, there are eight additions of ammonium sulphate in equal portions. From the twelfth to the fifteenth hour, the balling increases from 2.2° to 2.5°. A portion of this increase is due to the formation of the yeast, and the evidence shows that, were it not for the yeast, the balling at the end of the operation would be 1° less than is indicated. Whether or not there is an increasing rate of feed, it seems fair to conclude that there is a substantial constancy of concentration. Infringement of the patent is denied by the defendant. In the first place, it is urged that there is no proof in the record to show that, as the result of this process, substantially all of the alcohol is assimilated by the yeast. On the contrary, the evidence before the master showed that at times the defendant deliberately regulated its process so as to produce alcohol for its vinegar plant. But this intentional variation of the invention does not relieve the defendant. The plaintiff has demonstrated in its factory that yeast may be produced without alcohol under the patent, and it can hardly be claimed that the defendant is released from the charge of infringement by inefficient use of the plaintiff’s invention. Gibbs v. Triumph Trap Co. (C. C. A.) 26 F.(2d) 312. Secondly, the defendant contends that it did not add the reserve to the initial part of the wort, but used a divided feed; that is to say, the nitrogenous nutriment furnished in the form of inorganic salts was added partly at the outset to the dilute portion of the wort, while the balance was withheld and fed in at hourly intervals. The table above set out of - the materials which entered into the composition of the several mashes of the defendant shows that, in mash 245, there was used, as one source of the nitrogenous nutriment necessary for the yeast, 195 pounds of ammonium phosphate and 195 pounds of ammonimm sulphate. The ammonium phosphate was added at the beginning of the process and the ammonium sulphate in eight portions, as indicated. Hence it is said that there was no infringement of that element of the claims which provides for the substantially continuous addition of the more concentrated wort containing all essential ingredients to the more dilute wort during the period of propagation. Defendant’s expert testified that it was helpful in the propagation of yeast to have a high nitrogen sugar ratio at the beginning of the period of propagation. It was calculated that in mash 1901, which is similar in composition to the mash under discussion, the nitrogen sugar ratio was 100 parts of sugar to 6.36 parts of nitrogen, while the corresponding ratio in the Hayduek patent was 100 parts of sugar to 2.5-3 parts of nitrogen. It may be assumed that these figures are correct, and that some advantage is derived from giving a certain fillip to the growth of the yeast at the beginning by supplying it with a relatively large quantity of nitrogen. It may also be admitted that the hourly additions during an eight-hour period of one of the ammonium salts indicate, as far as they go, an interrupted rather than a continuous addition of the nutriment. Nevertheless it does not follow that infringement of the patent by this variation of the process has been avoided. A substantial part of the nitrogenous nutriment was. contained in the cereals, which not only entered into the dilute, but also into the concentrated portion of the wort, and these were admittedly added substantially continuously throughout the period of propagation, together with other essential yeast food. The Hayduek process was therefore followed by the defendant. The variations indicated, even though useful, cannot alter the fact that the feeding process, considered as a whole, was substantially continuous. Infringement of the patent by the mashes in question is made out. It is admitted that mash 1094 was an infringement of the patent. The ingredients and the procedure were similar to those employed in mash 245, with the exception that no ammonium salts were used as sources of nitrogen, and consequently there was no separate addition or divided feed in any respect. The evidence shows, however, that this was an exceptional mash and was used only on two occasions for the manufacture of seed yeast. Mash sheet 1901 sets out the process of the defendant which has been continuously in use since 1926, displacing the procedure above described. The nutritive material is divided into two separate worts, and a feeding process, admittedly continuous in some respects, is adopted. The plaintiff contends that mash 1901 adopts and follows all of the essential features of the manipulative process disclosed by patent 105 and consequently infringes it. The defendant, on the other hand, contends that the general principle of a divided wort and the addition of one part to the other did not originate with Hayduek, but was shown many years before by Rainer, and that infringement of the patent in suit can be made out only if it appears that the defendant is following those steps of the invention which distinguish it from the Rainer disclosure. The defense is that mash 1901 follows the procedure of Rainer and not that of Hayduek. It is desirable in this connection to refer again to the references to Rainer in the Hayduek patent and to the details in which the patentee showed that his disclosure was to be differentiated from the art of record. In the first place, Hayduek directed the preparation of a dilute nutrient solution containing all essential yeast nutrients. He said in .his specification in so many words that a wort should be prepared according to the procedure commonly followed in the compressed yeast industry. He contrasted the simplicity of this process with the preparation by Rainer of separate worts, called the “peptone wort” and “sugar wort,” respectively, which were not mingled-together at the outset of the process, but kept separate for subsequent addition. Next Hayduek differentiated his procedure from Rainer’s by showing that, after the dilution of a small portion of his wort, the addition of the more concentrated part was made substantially continuously during the period of propagation of the yeast. He again contrasted the .simplicity of this plan with that of Rainer, .saying that Rainer added the peptones continuously and the sugar at intervals as required. In other words, Hayduek pointed nut, not only that in Rainer the materials were separately added, as contrasted with a unitary stream, but also that a part of the materials, to wit, the sugar, was added at intervals and not with substantial. continuity. Bearing these distinctions in mind, it is necessary to examine the steps of the procedure of mash. 1901 in order to ascertain whether it follows Rainer, as claimed by the defendant, or Hayduek, as contended by the plaintiff. The raw materials in this mash were molasses, cereals, and inorganic ammonium salts. The cereal materials on the one hand and the molasses on the other were treated separately to make two worts, the first of which the defendant calls “peptone wort” and the other “sugar wort,” obviously in similitude with the nomenclature of the Rainer patent. The peptone wort was prepared by mashing the cereal materials and the sugar wort by diluting, acidifying, heating, and filtering the molasses. A portion of the wash waters from the cereal mashing operation, which consisted in effect of a diluted portion of the peptone wort, was placed in the vat, and there was added seven liters of aqua ammonia, nine gallons of the sugar wort and all of the ammonium phosphate. At the beginning the liquid had.a high dilution indicated by a balling of .8°. It contained about six per cent, of the total sugars and had a volume of 5,500 gallons. Propagation of yeast was commenced in this dilute portion of the peptone wort and aeration was employed. Prom the end of the first hour to the end of the thirteenth hour there was a slow and continuous addition of the peptone wort at an increasing rate. During the greater part of the same period, namely, from the end of the seeond hour to the end of the fourteenth hour, additions-of the sugar wort were made at intervals so frequent that the addition may be considered practically continuous. On the mash sheet, the additions of peptone wort are shown on the left, and the additions of sugar wort on the right-hand side. It there appears that the feeding of the peptone wort was divided into four periods of three hours each. During the first period the peptone wort was added at the rate of one gallon in each 100 seconds; during the second at the rate of one gallon in 50 seconds; and during the third and fourth periods, at the rate of one gallon in 28 seconds. The amounts added during these respective periods were 108, 217, 380, and 381 gallons, respectively. Corresponding with this addition of the peptone wort, there was an addition of the sugar wort at rapid intervals of 15 minutes. The period of addition was likewise divided into four intervals of three hours each. During the first interval 3.3 gallons of sugar material were added every 15 minutes, making a total of 39 gallons; during the seeond interval 6.6 gallons every 15 minutes making a total of 78 gallons; during the third period 10.7 gallons every 15 minutes, making a total of 129 gallons, and during the fourth or last period, 12 gallons were added every 15 minutes, making a total of 144 gallons. Prom these figures it is obvious that the ratio of sugar material to peptone material was substantially constant throughout the periods of addition of the respective worts. The defendant also used in mash 1901, as in its earlier mashes, a different method of adding the ammonium salts. This material is included with all other essential ingredients in a single wort in the Hayduek process, but in the defendant’s mashes a part of the ammonium salts, to wit, all of the ammonium phosphate, was placed in the vat at the outset, together with the initial dilute solution of the peptone wort. The ammonium sulphate was added in three gallon lots at twelve hourly intervals beginning with the third and ending with the fourteenth hour. The defendant attempts to distinguish this procedure from that of the patent in suit in several particulars. It is urged that Hayduek bad reference to the usual grain wort employed in 1914 and 1915, whereas mash 1901 is not composed of grain exclusively, but contains about one-third grain and two-thirds molasses, with ammonium compounds. The patent, however, relates to procedure rather than the composition of the mash. The claims are not limited to grain materials, and it is obvious that infringement cannot be avoided by substituting for tbe usual wort mentioned by Hayduek in his illustrative example a wort of a composition which under the present practice has been found more efficient. In tbe next place, the defendant contrasts the dilution of Hayduek, which has a balling of about Io, with the concentration of the dilute portion of the peptone wort of mash 1901, which had a balling of .8°. Here again the point is insignificant because the claims of the patent 'are not limited to a particular balling but provide for the initiating of propagation in a dilute nutrient solution, a phrase which is descriptive of the concentration of the mash under discussion. In the third place, the defendant points out that Hayduek employed the usual ratio of sugar to nitrogen, to wit, 100 parts of sugar to parts of nitrogen, whereas the corresponding ratio of mash 1901 was 100 parts of sugar to 6.36 parts of nitrogen, and also that the defendant did not add the ammoniaeal nitrogen -with susbtantial constancy. This contention has already been discussed above in considering tbe other mashes of the defendant. Finally the defendant points out that, whereas the Hayduek reserve wort is added in a single stream and with substantial constancy to tbe diluted wort, the defendant in mash 1901 divides the materials to be added into two separate portions of different composition and adds one continuously and the other at intervals. Certain additional minor differences, as indicated, are as follows: The addition of the peptone wort begins at the first and ends at the thirteenth hour, whereas the addition of the sugar wort begins at the second and ends at the fourteenth hour. The peptone wort is added in a continuous stream, whereas the sugar wort is added at rapid intervals. The evidence is not clear as to the length of time consumed by the introduction at each fifteen-minute interval of a portion of the sugar wort, hut it would seem that the time of flow does not exceed two or three minutes. A further minor difference is the fact that, although the proportion of peptone to sugar material, when the additions are considered as a whole, is relatively constant, the rate of addition of the peptone during the third and fourth intervals is the same, whereas the rate of addition of the sugar wort increases from the third to the fourth interval. The query is whether all of the differences that are pointed out amount to such a substantial departure from the Hayduek procedure as to avoid infringement of the patent. It is .of course obvious that inf ringetnent may not he avoided by separating tbe essential materials into two or more parts and tben running tbem into tbe vat simultaneously in separate streams. Tbe defendant has not adopted this precise procedure, but it has approximated it. It is not seriously contended that the frequent addition of sugar material at short intervals is materially different from the continuous addition thereof. It has not been shown that the interrupted as distinguished from the continuous addition served any good purpose in the defendant’s scheme. A certain flexibility is of course possible, when the sugar is separated from the peptone material, whereby the rate of addition of the respective constituents may be varied during the period of growth if desired. As a matter of fact, no variation during propagation was made by the defendant and none would seem to he necessary, since the action of known materials 'after experiment could be definitely foreseen. It is difficult to resist the impression that the real purpose of the interrupted additions of the sugar material in mash 1901 was the creation of a superficial distinction from the Hayduek process and a superficial similarity to the process of Rainer. The fact is, however, that the defendant has carefully avoided t