Final hearing on bill in equity charging infringement of the Schmertz reissue patent, No. 12,443, January 30, 1906, original patent No. 791,216, May 30, 1905, and also infringement of Schmertz patent, No. 791,217, May 30, 1905. The defenses are that the patents are invalid, or, if valid, are to be so narrowly construed as not to be infringed. These patents were directed to be issued in a proceeding under section 4915 of the Revised Statutes (U. S. Comp. St. 1901, p. 3392) under a decree in the case of Appert v. Brownsville Plate Glass Co. (C. C.) 144 Fed. 115, decided September 30, 1904. Other cases along the same line are the following : Continuous Glass Press Co. v. Schmertz Wire-Glass Co., 153 Fed. 577, 82 C. C. A. 587; Schmertz Wire-Glass Co. v. Pittsburg Plate-Glass Co. (C. C.) 168 Fed. 73. There have_been other decisions agreeing with those cited in the Western district of Pennsylvania and Northern district of Illinois, which are unreported.
The wire glass art is a development of the plate glass art. Wire glass, or “sandwich glass,” as it has been called, must be made by forming a lower layer of glass upon a table such as is used in plate glass art, then placing thereon a sheet of wire and covering this with a second sheet of glass in such a manner that the whole will weld together into one piece, which is then immediately placed in an annealing oven, and if designed to be polished is then placed in the polishing machine. The making of all wire glass involves the three necessary factors of an upper and lower layer of glass and an intermediate layer of wire. Several almost insuperable difficulties have been encountered in the art. The glass, being heated to a very high degree, cools with great rapidity, and in cooling forms a skin or coating which prevents the lower layér from uniting with the upper, and results in splitting of the final plate, making it practically worthless. This can only be overcome by reducing the time of the process to seconds, and making the three different steps practically simultaneous. This rapid cooling also tends to prevent the glass being made in large' sheets. The more glass poured out, the more difficult it is to prevent the cooling. Defendant, however, has been able to make wire glass experimentally, by a modified European process, in large sheets, and with the improved feed used by it in its machine. The bottom sheet of glass was completed in every part before the wire was applied by defendant’s elaborate feeding device, and good glass was made; but when it was attempted to put the wire on by hand, as in the European method, the experiment was a failure.
Another great difficulty has been the introduction of the wire trellis. The expansion of the wire from the intense heat is about 4½ per cent. The result is that the wire kinks and crawls in such a way as to prevent its insertion in a uniform plane. That is to say, by reason of the kinking and crawling it is likely to come to the surface, or so near the surface that it is impossible to polish the plate. It is necessary to *979make the glass quite thin for some purposes, and this point of introducing- the wire in the exact center of the plate has been found very difficult. It was at first thought that the wire could be uniformly placed hv subjecting it to great tension, and, while this process was not entirely unsuccessful, yet it was found quite unsatisfactory. In an article in the Revue Industrielle, published in 1892, it is said that the result obtained from the Becoulet and Bellet process, however simple and practicable it may appear, was devoid of practical results on account of the wire suddenly entering a very high temperature, and thus being so bent and twisted as to render it very difficult to keep it in the mass of the glass, not only the middle line, but away from the upper and lower surfaces. The defendant has perfected a process by which the front end of the wire is introduced without tension, the central portion under tension, and the last part without tension, all equally successful. By its method the wire is spread out evenly across the sheet, and made to enter the advancing wave of hot glass evenly, so the distortion from the heat is uniform, and the bending and twisting so equalized as not to injure the product. This point is considered later.
A third difficulty is that when the wire is placed upon the hot glass, and thus rendered incandescent, it at once “oxidizes and becomes blackened, unless it can be immediately covered by the upper layer of glass. The appearance of the -wire is thus injured, although its durability is not affected.
The “European process,” so-called, consists in first forming a lower «beet of glass by rolling or compression, then laying the wire trellis thereon, pouring another portion of glass upon the wire, and rolling or compressing the two layers together. By reason of the rapid cooling of the lower layer, it lias been found difficult to produce a proper welding and prevent splitting as above indicated. The authorities seem to agree that glass made in this manner can only he produced in small sheets, unless modern methods of wire handling are used, and for a long time could not be produced at all by this method. It is described in the English patent granted to Newton and the English patent to Take for Becoulet and Bellet, and in the English patent to Armstrong and the German patent to Armin Tenner. Although the Newton patent was taken out in 1855, the Lake patent in 1886, and the Tenner patent in 1888. it w-as not until 1892 that the Siemens Works in Dresden were able by this method to produce wire glass successfully.
The American method (including therein the process invented by the Frenchman Appert) consists in the Shuman process under patent of September 20, 1892, the Appert patent July 26, 1898, patented in France January 12, 1894, and the Schmertz patents above referred to, The Shuman process is the so-called “single-pour three-roll process.” His machine consisted of a casting plate or rolling table on which the glass is to be formed. Above this table is arranged a series of three rolls connected together and suitably geared to be moved over the casting plate. The first and third of these rolls are smooth finished, and the middle one corrugated parallel to the axis of the roll. A pour of glass is put in front of the first roll, and the three rolls then moved forward together over the casting plate and over the first pour of glass. *980The layer of glass is thus pressed down and rolled out upon the table and thus formed into the lower sheet or layer. Between the first and second rolls there is loosely fed down upon this lower sheet a wire trellis from a chute. The second roll follows immediately, and the corrugations passing over the wire force it down into- the body of the glass below. The third roll then comes along and smoothes over the ridges and irregularities caused by the corrugated roll and wire, thus completing the sheet. This process was regarded as a great advance in the art, as it enabled wire glass to be made in larger sheets, and, the process being practically simultaneous, the lower layer had no time to ,cool. It was, however, found impossible to make a smooth sheet of i glass in this way. The testimony is practically undisputed that the distortion of the wire trellis and the uneven placing of it in the finished product, the scarring of the surface by the wire and the corrugated roll, the abnormal chilling and hardening of the surface rendering it very difficult to cut, and the roughness and unevenness of the upper-surface being so pronounced, all combined to render the glass dull and 'lusterless, the surface uneven, and to introduce bubbles and cracks, and also the size of the sheet was limited.
Defendant has cast an anchor to the windward by obtaining a patent on a modified Shuman process, and has tried it experimentally, apparently with complete success. This patent is No. 938,385, issued to Clement J. Jungers October 26, 1909. It differs from the Shuman process in having an improved feed, being the same as in defendant’s other machines, in using only one roller for thick and two for thin glass, and particularly in the construction of the roller which forces the wire into the glass. Shuman used wide, blunt corrugations around the circle of the roll, while Jungers uses sharp, long flanges also extending around the circle of the roll, making sharp cuts in the glass parallel to its length, and at the same time forcing the wire down into the middle of the single glass' layer. The operation is performed so quickly that it is claimed that the molten glass will fill up the cuts made by the flanges, and make a surface sufficiently smooth to be successfully polished. When thin sheets, less than three-eighths inches, are made, a second smoothing roll follows the flanged roll. One very great improvement over the Shuman method is that he first rolled his glass plate, let it cool, and then tried to force the wire into it, while defendant, if the last Jungers patent process be followed, makes the sheet and puts the wire into it at the same time. However, in the actual experiment of this modified Shuman method three rolls were used on the single pour.
. The Shuman process was followed by those of Appert and Schmertz. The Appert machine consisted of a casting plate,' as before, and two rolls, instead of three. The first roll was elevated above the table one-half the thickness of the proposed sheet of glass, and the second roll just double the distance. The table could be made movable and the rolls stationary, or vice versa. Before the operation commenced, a sheet of wire was fastened to the end of the table where the process commenced at the level of the top of the first layer of glass, and was carried up under tension in front of the first roller,, then, perpendicularly *981over an idler pulley to a higher pulley, and a weight was attached, to the end of the wire so that it could he kept taut under all conditions. A pour of glass was then put in front o f the first roller which moved forward, flattening the glass down with the wire trellis in its upper surface, immediately followed by the second roller with a second pour of glass in front of it. By reason of the fact that the process was practically a simultaneous one, the glass had no time to cool, and two sheets could thus be pressed together and made one. The great difficulty. however, of the process, was that the wire could not be evenly inclosed in the plate on account of the tension and the kinking and distortion of the wire; no means of overcoming its expansion being provided. The process was never a practical or fully successful one.
The Schmertz process was the first to he a really successful one, and this by reason of the improved feeding of the wire. The two Schmertz patents, by the decision of the Circuit Court for the Western district of Pennsylvania, were made to include the Appert process; Schmcrtz being found the first inventor. This is fully described later. Under the first application of Schmertz, the wire could .he introduced either in front of the first roll, as in Appert, or between the two rolls, as in Shuman, and nothing is said in the specifications or claims as to whether the introduction of the wire is to be under tension, gravity, or by a combination of the two. However, the specifications recommend the insertion of the wire, not at the end of the plate, hut a few inches from it, thus indicating that the wire was to be fed down without tension. In the other application embodied in patent No. 70 J ,217, the claims were so amended as to require the feeding of the wire by gravity without tension. In the file wrapper of the Schmertz gravity patent this point is greatly emphasized; it being stated by counsel for Schmertz that the feeding of the wire under tension was a failure, but that when it was fed down loosely the web of wire could be caught by the advancing wave of glass under the second roll and laid evenly on the top of the bottom layer, and the wire thus made to assume a central position in the finished sheet.
Defendant's wire glass is made under the Jungers patent of October 1, 1907', No. 867,510; hut the process has been modified by lessening the tension on the wire feed, two or three rolls or rods having been taken out, reducing the pull on the wire from 17½ to .12 pounds. The chief claim of novelty by defendant is that the improved tension-feeding apparatus inserts the wire in tile bottom surface of the second pour at the. very instant that this pour reaches the lower layer, and never allows the wire sheet to touch the lower layer, thus keeping it cool and avoiding distortion, until it is firmly embedded in the upper sheet, and that this is done so evenly across the casting plate, each part simultaneously entering the glass, that the wire is laid in the exact center, the second pour being larger than the first, in order to bring 1he wire into the middle plane of the finished sheet. Further explanation of the process appears below, in connection with the discussion on infringement.
Another process which has never been practiced is found in the specifications of the .British inventor Hyatt in 1874. Hyatt was an inventor *982of great merit, having introduced the “patent lights” used in building construction and building materials generally, including the sidewalk lights so common in large cities. In his long specification is found the following:
“According to another part of my invention for making fireproof glass, I make plate or rolled glass with a netting of wire incorporated in the body of the plate; one mode of doing this being as follows: I use two drums or rollers, as is done sometimes in making ornamental face glass, one behind the other. The former carries a wrap or web of wire netting, which as it rolls down the ‘metal’ is spread out upon the face of the flattened mass; but this ‘metal’ is only half of what is needed to make the sheet. Another dumping of another mass is made immediately upon the first flattened sheet with its overlaying of wire netting, the second mass falling upon the wire netting, where it is met by the second or following roller and flattened out, thus incorporating the wire sheet between two plates of glass welded together.”
This paragraph is relied upon as a complete anticipation of the Schmertz patents, and if a workable process appears' in it the defense would be a good one.
Upon the hearing and in the evidence taken the question how far the European method was successfully practiced, and to what extent it could be considered as an anticipation of Appert and Schmertz, were very fully considered. The evidence on this point is somewhat conflicting, although the solution of the question is not a difficult one. It is claimed by the defendant that wire glass was largely made in 1892 by Siemens in Dresden under the Tenner patent, and that the Schmertz invention can only be given effect as an improvement to a prior established art, and that his claims are therefore reduced to a narrow range of equivalents. This European method was examined in 189'!' by a commission appointed by the German Patent Office, in the Siemens Works in Dresden, who viewed a considerable number of operations. The process is thus described in a German Patent Office decision, affirmed by the Imperial Court of Justice in proceedings brought to annul Appert’s German patent: A layer of glass was first produced by rolling or compression, and completely finished in every part. Then a sheet of wire netting was laid on the glass sheet, and a second quantity of molten glass poured over it, and then rolled and compressed; thus making three operations. But as the first layer must not be allowed to cool too much before the second pour, or the two pieces will not weld together, the glass plate can only be made of moderate length. The German Imperial Court say in their decision that the Tenner process was the first to produce a product fit for use. They do not refer to the Hyatt specification; and they decided that the Tenner process was entirely distinct from Appert’s. In 1892 several publications in England, France, Germany, and the United States speak of wire glass from the Siemens Works in Dresden as extensively used for skylights, translucent floorings, and similar purposes. Window panes are also referred to in an article from Brock-haus’ Encyclopaedia, 1892. The Engineering News of July 28, 1892, states that wire glass is being extensively used in Germany for skylights and similar purposes. A like assertion is made by Fire Chief Stude of Berlin, Germany, June 5, 1893, and the record shows one or two other similar statements about the same time. It appears from *983these articles that the glass was recently put on the market, so these authorities fix the date of its introduction as 1892, the same year that Shuman commenced his operations. On the other hand, the Revue Industrielle in 1893 says that in France there were numerous attempts to make wire glass without success. Referring to the Shuman glass, they say, “It is certain this is a novel material,” and they are skeptical as to the result of the Shuman invention. Further, they state that the Becoulet and Belief process was a failure by reason of the bending and twisting of the wire; and other difficulties are referred to, tending to make the product useless. The Scientific American of November 5, 1892, in an article on the Shuman process, refers to his glass as a new product. The Iron Age of November 21, 1892, says:
“The idea of making a strong combination of wire and glass is said to have liad its inception some years ago in England, but it never seems to have gone beyond the region of experiment, although some small panes, not more than 1 foot square, of an imperfect kind, were made, but the material never obtained recognition as an article of commerce on account of the great cost of production and limitation in the size of the sheets. It remained for an American to carry out the idea to a practical issue.”
An article by Daniel Belief, in Da Nature, in 1893, states that a new material is coming on the market, in the shape of wire glass, from the Shuman plant. And the Encyclopaedia Brittanica says that there was no wire glass until Shuman made it in 1892. The Franklin Institute of Philadelphia awarded to Shuman the John Scott legacy and medal for his discovery, with the statement that his process and ■ machine represented a marked advance in the art, and met for the first time the requirements of a practically operative method. And in a book published in 1900 by Lippincott, Shuman is given the credit of discovering the first method which was commercially practical.
Several judicial decisions have reached the same result. In Streator Cathedral Glass Co. v. Wire Glass Co., 97 Fed. 950, 38 C. C. A. 573, the Court of Appeals of this circuit held that:
“There was no wire glass in use or upon the market, either in Europe or in this country, when Shuman obtained his patent.”
The same conclusion was reached by Judge Buffington in Schmertz Co. v. Pittsburgh Co. (C. C.) 168 Fed. 73, 76, and the question was again examined by the same judge in 1909, and the same result affirmed and elaborated in Schmertz Wire Glass Co. v. Highland Glass Co., 178 Fed. 945.
I think it is the fair result of this evidence that wire glass was made in some form under the Tenner patent in 1892, but it was in small sheets. Some polished plate in small pieces was undoubtedly made. The manufacture, however, left little or no impression upon the art, so that when Shuman discovered his process it was quite generally regarded as a great advance. The evidence is silent as to the difficulties attending the manufacture by the European method. How much glass was spoiled, what proportion of transparent, as distinguished from translucent, glass was made, and how far transparent window glass could be produced, do not appear. It is quite clear that the *984process was never rendered successful to any great degree until the inventions of Appert and Schmertz appeared.
A very important question in its bearing on the scope and range of equivalents of the Schmertz patents, and hence on infringement, is the effect of Judge Buffington's decree in the Appert-Schmertz interference; although, by reason of the conclusion here reached as to collateral attack, it is not necessary to state the facts in full detail, especially as they are quite fully recited in two of the Pennsylvania cases. 144 Fed. 115, and 168 Fed. 73, 87. Appert’s patent having issued before those to Schmertz, and not being sued on here, would narrow the Schmertz patents very materially were it not for the decree establishing Schmertz’s priority over Appert; hence the importance of that decree. On April 87, 1894, Appert filed his application in this country, and the two applications of Schmertz were filed March 19, and April 1, 1895. An interference between Appert’s claims and all the Schmertz claims in both applications was declared May 28, 1896, .on which conflicting decisions were made, resulting in Appert’s success in the Court of Appeals of the District of Columbia June 7, 1898. 84 O. G. 508. On the hearing I thought that there was no interference in fact, especially as to Schmertz’s gravity feed claim, Appert’s claim being for a feed under tension; but, on further reflection, I am satisfied that the interference was properly declared in view of all the claims on both sides. However this may be, the action of the Patent Office in declaring an interference is in no way reviewable, and the only question decided is priority of invention. Westinghouse v. Hien, 159 Fed. 936, 87 C. C. A. 142. The interference having resulted in Appert’s favor, his patent issued July 2G, 189¿. Schmertz was using his own wire glass process, and Appert promptly sued him for infringement (August 8, 1898). On October 3, 1898, Schmertz answered, and also filed a cross-bill' to establish his claims under section 4915 of the Revised Statutes. Later on the Appert Glass Company took over the Appert patent, October 3, 1899, and on April 1, 1902, the Mississippi Wire Glass Company, one of the complainants, took an assignment of the Appert patent, which it still owns. Seventeen days after filing his cross-bill, Schmertz died, leaving as his heirs at law his widow and seven small children. His estate was utterly insolvent. The infringement suit thereupon became abated. In view of the insolvency, Appert was justified in not reviving it, and Schmertz’s estate was unable to do so. By the statute (section 4896) the right to the Schmertz applications on his death passed to his administrator, in trust for his heirs; that is, the legal title vested in the administrator, and the equity in all his heirs. Newell v. West, 13 Blatchf. 114, Fed. Cas. No. 10,150. Mrs. Schmertz had means, but her children, of course, had none. The eldest being only 14 years of age, they cannot be charged with laches or delay in not prosecuting the cross-bill. She was reluctant to carry on the litigation, and was advised not to do so. Finally two former employés of Schmertz were able to enlist capital to build a glass plant, and go on with the suit to obtain the patents. They procured Mrs. Schmertz’s consent to take out administration on her husband’s estate, which was *985done October 15, 1901. On November 7, 1903, revival proceedings were commenced, and a motion argued, but not decided. On February 11, 1903, a petition for leave to file a cross-bill in the nature of a bill of revivor and supplement was filed, and later argued.
At this point occurred the transaction which made a single party dominus litis, and rendered the suit no longer a judicial contest. As has been stated, the Mississippi Wire Glass Company acquired the Appert patent April 1, 1902, and thus became the adverse party in the cross-bill proceedings; the Schmertz interests , having been acquired, through the administrator so appointed for Schmertz, by the Brownsville Glass Company, which was thus in effect the petitioner in the bill of revivor and supplement, although made nominally by the-Schmertz Afire Glass Company patentee. The Mississippi Company had a large investment in the Appert process, which it would lose if the Schmertz invention was held prior to Appert’s. It was thus confronted both with a lawsuit and .ruinous competition. The Brownsville Company was in a position to push the case. Under these circumstances, the Mississippi Company purchased the Brownsville plant for $500,000, and the Schmertz applications, fi'his was done March 27, 1903. The contract was read to the court June 29, 1904, upon the final hearing, and the relations of the parties fully explained to it, with the statement that every fact known to counsel for the Mississippi Company which in any way tended to disprove the Schmertz claims had been produced, and that the company found itself in the position that, if it should stand on the Appert patent, it was forever open to the claim that Schmertz was the prior inventor. A stipulation briefly stating the facts was made, and ordered filed September 28, 1904. In the transfer it was provided that the Brownsville Company transferred all the capital stock of the Schmertz Wire Glass Company, then owner of the Schmertz applications. This would carry the real title to those applications to the Mississippi Company; but the Brownsville Company reserved the right to prosecute the applications, at its own cost and expense, and if patents should issue it was to have an additional $5,000. Of course, the ownership of the Schmertz stock would carry the title to the patents, when issued, to the Mississippi Company; but the Brownsville Company would be entitled to the whole of the additional $5,000. This sum was paid, however, not to the Brownsville Company, but to the Schmertz Wire Glass Company, all whose stock had passed to the Mississippi Company; and it was used to pay expenses of printing on both sides, and attorney’s fees and expenses of the Brownsville Company, all amounting to $2,982.-40, and the balance of $2,017.60 being retained by the Schmertz Company nominally, but in reality by the Mississippi Company as the owner of all the stock of the Schmertz Company; but later was paid to T. Ramsey Speer, for distribution among his associates in the Brownsville Company. These payments were made more than a year after the decision of Judge Buffington, and six months after the patents were issued. These facts are pertinent only as possibly showing an intention by the parties to keep something back from the court. But as this was done long after the decision, and could not be disclosed *986before, it is difficult to infer any such fraudulent purpose, in view of the full and frank disclosure made at the hearing.
Returning now to the question of delay or abandonment of the Schmertz applications: It is provided by section 4894 that the failure to prosecute an application within two years after any action by the Patent Office shall be an abandonment, unless it shall be shown to the satisfaction of the commissioner that the delay was unavoidable. This provision applies to the cross-bill proceedings under section 4915, where it must be, shown to the satisfaction of the court that the delay was unavoidable. Gandy v. Marble, 122 U. S. 482, 7 Sup. Ct. 1290, 30 L. Ed. 1223. Now, from the above statement of the situation after Schmertz’s death, it is clear that the delay was unavoidable. Judge Buffington was “satisfied,” and so decided. Certainly his decision cannot be disregarded when it is offered in evidence in another case.
The question of collusion, and its effect upon the cross-bill decree, were ■ considered on the motion for injunction, but are so important that they have been carefully studied anew, and the following conclusions reached:
A judgment may be rendered which is wholly or partly induced by: (1) Collusion between the parties, such as one of them buying out the other, and thus becoming dominus litis, so that there is no further dispute or controversy; or litigating a question but not a case, there being no real contest. (2) Fraud by one party against the other, not by false documents or perjured testimony contested in the case, but by extrinsic fraud. (3) Fraud on a third person practiced by one party to a suit without the other’s knowledge or participation; that is, fraud without collusion. (4) Fraud on a third person by both parties to the suit, or by one with the connivance of the other; that is, fraud with collusion. The interference suit falls under the first class. When the suit on the cross-bill started, it was a real controversy of great importance. Afterwards the Appert interest acquired the Schmertz interest under stress of grave business complications, and thus became dominus litis. No fraud was thought of, no third person injured. It was a case of pure collusion unmixed with fraud. Even the court was informed, of the true situation, or substantially all of it. Now, in the ordinary case the court in which the suit was pending would dismiss the suit as soon as its attention was called to the union of interests. The Supreme Court has uniformly refused to review cases of the kind. Lord v. Veazie, 8 How. 251, 12 L. Ed. 1067; Cleveland v. Chamberlain, 1 Black, 425, 17 L. Ed. 93; American Wood Taper Co. v. Heft, 8 Wall. 333, 19 L. Ed. 379; East Tennessee, etc., Co. v. Southern Tel. Co., 125 U. S. 695, 8 Sup. Ct. 1391, 31 L. Ed. 853; Gardner v. Goodyear D. V. Co., 131 U. S. ciii, 21 L. Ed. 141. However, such a situation in no way affects .the jurisdiction. Should the Supreme Court review and decide the case notwithstanding the collusion, no one could reasonably contend, when its decree was offered in evidence in another suit in this country, that such decree was a nullity. The court of a foreign country might refuse to give effect to it (Hilton v. Guyot, 159 U. S. 113, 16 Sup. Ct. 139, 40 L. Ed. 95), *987but no domestic tribunal could properly decline to do so (Ellis’ Appeal, infra). )
The moment Judge Buffington learned of the collusion, he might properly have dismissed the whole suit, although it would have entailed a great hardship to have done so, or retained the question whether Schmertz was entitled to any claim not in interference, or considered the whole proceeding an administrative one and proceeded to a decree. His jurisdiction to take any one of these courses seems absolutely tin-questionable. He chose the last, with the utmost propriety, as it seems to me, and held that Schmertz was prior -in time to Appert, and his successors or assigns were entitled to patents. How can this decree be impeached by one whose interests arose five years later, and in a proceeding purely collateral? If the collusion had descended to the level of fraud intended against a stranger, such as a then existing alleged infringer, he might probably avoid the decree, at least so far as it held Appert’s patent void by establishing the Schmertz counts as prior inventions. Michaels v. Post, 21 Wall. 398, 22 L. Ed. 520; Cochrane v. Deener, 95 U. S. 355, 24 L. Ed. 514. But a party cannot do so (Michaels v. Post, Board v. Platt, 79 Fed. 567, 25 C. C. A. 87). nor a stranger not injured or defrauded, because collusion, or fraud with collusion, eveti if unknown to the court pronouncing the decree or judgment, makes it voidable only, not void, impeachable on direct proceeding, but beyond the reach of collateral inquiry. (Board v. Platt, supra; Ellis’ Estate, 55 Minn. 401, 56 N. W. 1056, 23 L. R. A. 287, 48 Am. St. Rep. 514; Van Fleet, Collateral Altack, § 533). Many expressions are found in the decisions of the federal and state courts, to the effect that a judgment cannot be collaterally attacked except for want of jurisdiction or fraud; but an examination of the cases shows that the distinctions referred to above are fully preserved. An existing creditor, for instance, or stockholder under liability for corporate debts, or a surety liable over, is not bound by a fraudulent or collusive judgment injuriously affecting his rights, even in a collateral proceeding, because the fraud is aimed at him, and he had no opportunity to contest, nor was the judgment the result of a real or substantial controversy. So, when it is sought in ancillary proceedings to enforce a judgment or decree, the court may look into its equity and fairness. Riverdale Cotton Mills v. Alabama & G. Mfg. Co., 198 U. S. 188, 25 Sup. Ct. 629, 49 L. Ed. 1008. And a like rule governs the effect of foreign decrees whose enforcement is sought in this country. Hilton v. Guyot. Nor does the rule against collateral attack require a court of equity to do an inequitable thing. In an extreme case it might undoubtedly refuse effect to a decree or judgment induced by such fraud or collusion as to make enforcement unconscionable. It might appear that the former decision was not really the act of the court, but that of the parties alone, and that to give it effect would be most inequitable and injurious. Nothing but conscience and good faith can coerce action by a court of equity. McKnight v. Taylor, 1 How. 161, 11 L. Ed. 86; Perkins v. Fourniquet, 14 Flow. 313, 14 L. Ed. 435.
No lack of conscience or good faith appears in the cross-bill decree. Coerced by the exigency of a practical situation of a critical nature *988and large importance, realizing that an adverse decision would mean the utmost disaster, the Appert interest bought up the Schmertz interest, and thus came into control of both sides of the litigation. But no fraud was exercised nor intended, no injury to third persons; simply protection to existing rights of immense value. There was a substantially full disclosure to the presiding judge; no deceit or bad faith of any kind. The only thipg which could be called unconscionable was the act of putting up to the court a sham dispute, and this was entirely overcome by telling the court all about it, and submitting every scrap of the evidence on which the Court of Appeals of the District of Columbia had held Appert to be prior in time. Surely there is nothing here which should make a court of equity hesitate to give effect to a decree made with’full and complete jurisdiction, with full knowledge of the whole situation, and upon all and singular the evidence, and the arguments which had induced the Patent Office to make its various findings. Anything short of this might induce a different result; but, with all these conditions present, no good reason appears for refusing full effect to the Appert-Schmertz decree. The conclusions here reached are further supported by Tuthill Spring Co. v. Smith, 90 Iowa, 331, 57 N. W. 853; Black on Judgments, § 291; Freeman on Judgments, §§ 334-336; Page v. Holmes Burglar Alarm Co. (C. C.) 2 Fed. 330, 18 Blatchf. 118; and American Nat. Bank v. Supplee, 115 Fed. 657, 52 C. C. A. 293.
The complainant Mississippi Wire Glass Company is the owner of the Appert patent, and might have claimed infringement of it in this suit, had it seen fit to do so. This fact makes the attack of defendant on the cross-bill decree merely a technical defense.
Before coming to the question of infringement, it is necessary to consider the important question of the Hyatt specification as an anticipation. This was most thoroughly presented on- the argument, and is made much of in the briefs and 'testimony, but does not seem to have been given much attention by the courts which have heretofore upheld the Schmertz inventions. As already stated, • this specification was never put into actual practice. Before attempting to construe it, to see whether it was capable of operation, the legal rule governing the case may profitably be examined, to determine whether the Hyatt specification shows definite means for making wire glass. It may not be difficult, in the present state of the art, to read the Schmertz invention into the Hyatt disclosure; hut could it have' been done in 1874? No one ever succeeded in doing it, and this is some evidence, at least, that the description was defective.. ’ Was the original conception that of Hyatt or of Schmertz? By using twentieth century magnifying glasses, a nineteenth century method has been found efficient, which never was so before, and the immensely important point of view of an advanced art is thus unfairly used to discover an original conception never acted on or made anything of, and which never had any practical or beneficial existence.
There are many decisions on this point, and much pointed discussion. Courts have been very reluctant to find anticipation in earlier descriptions, unless of a most complete and precise nature. In the *989following cases an alleged prior invention or description was cited as an anticipation, but was held not to be so: It must be an account of a complete and operative invention, “capable of being put into practical operation.” Seymour v. Osborne, 13 Wall. 516, 20 L. Ed. 35. “Tn the law of patents it is the last step that wins.” The Barbed Wire Patent, 143 U. S. 275, 12 Sup. Ct. 443, 36 L. Ed. 154. If a combination or mode of operation seems an obvious one, but was not adopted or used, the inclination of the court is to sustain a later patent. “It may have been under their very eyes, they may almost be said to have stumbled over it; but they certainly failed to see it, to estimate its value, and to bring it into notice:” Webster Loom Co. v. Higgins, 105 U. S. 580, 26 L. Ed. 1177. “The fact that the known valves were not used, and the speedy and extensive adoption of Richardson’s valve, are facts in harmony with the evidence that his valve contains just what the prior valves lack.” Consolidated Safety Valve Co. v. Crosby Co., 113 U. S. 157, 5 Sup. Ct. 513, 28 L. Ed. 939. The fact that a device has displaced others, and gone into general use. while not alone sufficient to show novelty, is enough to turn the scale in doubtful cases. Smith v. Goodyear Dental Vulcanite Co., 93 U. S. 486, 23 L. Ed. 952. “It is not sufficient to constitute anticipation that the device relied on might, by modification, be made to accomplish the function performed by the patent in question, if it were not designed by its maker, nor adapted, nor actually used, for the performance of such functions.” Topliff v. Topliff, 145 U. S. 156, 12 Sup. Ct. 825, 36 L. Ed. 658. “The invention or discovery relied on as a defense must have been complete and capable of producing the result.” Coffin v. Ogden, 18 Wall. 120, 21 L. Ed. 821. “If the mere fact that a prior device might be made effective for the. carrying on of a particular process were sufficient to anticipate such process, the absurd result would follow that, if the process consisted merely of manipulation, it would be anticipated by the mere possession of a pair of hands.” Carnegie Steel Co. v. Cambria Iron Co., 185 U. S. 403, 22 Sup. Ct. 698, 46 L. Ed. 968, cited in Diamond Meter Co. v. Westinghouse Electric & Mfg. Co., 152 Eed. 704, 716, 81 C. C. A. 630. “'['he prophetical suggestions in English patents of what can be done, when no one lias ever tested by actual and hard experience and under the stress of competition the truth of these suggestions, or the practical difficulties in the way of their accomplishment, or even whether the suggestions are feasible, do not carry conviction of the truth of these frequent and vague statements.” Westinghouse Air-Brake Co. v. Great Northern R. Co., 88 Fed. 258, 263, 31 C. C. A. 525.
The facts or elements contained in the Hyatt specification quoted above are: (1) Two rolls, one behind the other, as in making ornamental face glass. (2) The first roll carries a wrap or web of wire netting. (3) The netting is spread out by the motion of the roll on the face of the flattened “metal” or mass of glass. (4) The “metal” is only half of what is needed to make the sheet. (5) A second dumping is made immediately upon the first: flattened sheet with its overlaying wire netting, where it is met by the second roller and flattened out, thus incorporating the wire sheet between two plates of glass welded *990together. The following distinctions between this and the Schmertz process are referred to in the testimony: The second roll was the “trailer” used in making ornamental glass, which ran on the surface of a formed 'sheet; no relative spaced relation between the two rolls; no raising of either roll a definite distance above the casting table; wrapping the web around the first roll, which would render the process inoperative; no means for centering the wire. Rejecting the point of wrapping the wire as obviously changeable by mere mechanical skill, there are two others which seem to show that Hyatt did not fully conceive a workable process, viz., no spaced roll-relation, and no definite relation between rolls and casting plate. And when it is remembered that Hyatt did not take out a patent on the process so outlined, and that no one ever made wire glass by this process, the case is brought within the rule, above stated, that the earlier process must be complete and operative in order to anticipate; and that, in case of doubt, complete success in the patent process, as against complete failure or nonuser of the other, should turn the scale in favor of pat-entability and nonanticipation. As to -the effect of Hyatt’s omission to patent the process, see Stoner v. Todd, L. R. 4 C. D. 60, 46 R. J. Ch. 33.
The question of infringement depends upon the construction to be given the Schmertz inventions; whether they are to be regarded as broad enough to cover a fair range of equivalents, or so narrowly construed as to be mere improvements upon an established art, and thus covering only the specific apparatus and process. The claims, drawn to include all that both Appert and Schmertz conceived as novel, are quite broad:
“The process of making glass sheets with wire inclosed therein, consisting in simultaneously forming a layer of glass and introducing wire thereto, and completing the sheet by forming another layer upon the first layer of glass; the process being carried on progressively.
“An apparatus for making sheets of glass with wire inclosed therein, consisting of a table, a leading roll to- roll a' layer of glass, means to support and introduce wire to the said layer, a second roll, behind the leading roll, to form a layer of glass on the first or underneath layer; the periphery of the second roll being higher above the table than that of the leading roll, and the two rolls being far enough apart to allow the glass for the second or upper layer to be poured between them.
“In the manufacture of wire glass, the combination of a table, a leading roll with recessed ends, a finishing roil whose body is higher from -the bed of the table than the body of the leading roll, and means for introducing the wire by gravity between said leading and finishing rolls.
“An improvement in the process of manufacturing wire glass which consists in rolling a sheet of glass of less thickness than the ultimate product required; simultaneously feeding by gravity wire upon the top of said sheet at the rear of the leading roll and rolling a second sheet of glass upon said original sheet and the wire, simultaneously embedding the wire and finish--ing the sheet.”
The Schmertz process was a great advance over Shuman, as his was over the European process. Schmertz discovered the spaced relation of the rolls, both to each other and to the casting table, the improved gravity feed, and the important notion of substantial simultanity of process. The Shuman product was rough, in small sheets, not transparent, and could not be polished nor easily cut. All these things *991Schmertz remedied. Shuman was completely displaced. Schmcrtz’s method was eminently successful, about 84,000,000 feet of wire glass having been made by it since 1901, while the total output of Shuman’s process was about 1,000,000 feet. He was able to practically control, absolutely cover and exactly center the light, flimsy netting of wire in large sheets of glass, which were readily polished. The wire feed operates with precision and is capable of nice adjustment. In respect to unpolished glass, he made considerable improvement in size of sheets, quality of glass, ability to cut, centering the wire, finish, and general appearance. All prior processes were displaced, and the great bulk of wire glass was for a long time made almost exclusively by the Schmertz process.
Results like these are not accomplished without the exercise of the inventive faculty. It is not accurate nor fair to call his achievement a minor improvement or a mere variation or slight forward step in a developed art. It was more than this, and while not a generic invention in the broadest sense — rather it was an important improvement on Shuman — it was a good deal more than the mere putting together of old things. Such an invention is entitled to a fairly broad and liberal range of equivalents.
In order to understand fully defendant’s process, it is necessary to explain in more detail how the wire is introduced to the glass. Under both the Schmertz and Jungers (defendant’s) method, the wire generally 3 feet wide, is cut into 11 or 12 foot lengths, and smoothed by rolling. In each process also there is an inclined chute to hold the wire, fixed above the casting table, from which the wire is drawn into the glass, with the assistance of a workman. In the Schmertz machine this chute comprises the whole of the feeding apparatus, since the wire goes in by the pull of the moving roller or table, without tension. If the rollers move over a stationary table, the wire-chute moves with them; if they are stationary, it is also. In defendant’s machine, on the other hand, the wire is made to pass between two tension rolls, under a weight of 34 pounds, so that the machine must exert a pulling force on the wire of about 12 pounds in order to get it into the glass. These tension rolls are placed about a foot above the table, and about 27 to 30 inches away from the point where the wire is going into the glass. In defendant’s first machine other rolls were put in, so that it took a pull of 17½ pounds; but for some time the apparatus has been as above indicated. The wire is thus laid under tension, except when it first enters the sheet, and until the movement of the roller or table takes up the slack, and also except the last 27 to 30 inches which is laid after the end of the wire runs through the tension rolls.
Another important feature of defendant’s wire introduction is the spreading roll, located 2 inches above the table and 7 inches from the point where the wire enters the glass. This roll is threaded to the left in one half and to the right in the other, so that as the wire passes over it is stretched laterally to a slight degree, and held firmly in a suitable position to enter the glass evenly and squarely. Still another improvement is a bar placed across the framework of the machine nearly over the edge of the second pour of glass,' where the wire *992enters it, upon which the workman rests the ladle, and thus makes a straight-edged pour and keeps the advancing line of the glass nearly straight.
The advantage of tension on the wire is said to be that, when the wire is pulled through the tension rolls, it is thereby kept from falling down upon the hot lower sheet of glass, until the very instant when the wire enters the glass, .thus preventing kinking and twisting, and embedding the wire in the undersurface of the top layer.
Apart from the feeding device, defendant’s machine is the same as Schmertz’s, and as to that it is plain that it is interchangeable with the gravity f'eed. By lifting oif the loose upper tension roller, the two devices are practically the same, except for the spreading roll used by defendant. It is also clear that defendant employs both feeding methods; one-quarter of the length of the sheet being fed loosely and three-quarters under 12 pounds tension.-
Under these circumstances, and eVen conceding some- improvement in defendant’s method, it is difficult to escape the conclusion that the two forms are mere equivalents, doing the same work in the same way, dne equally as well as the other. Both are efficient and secure good results. They are practically interchangeable, each for each. While they can be differentiated, yet infringement is not averted. Paper Bag Patent Case, 210. U. S. 405, 28 Sup. Ct. 748, 52 L. Ed. 1122. Though there may be improvement, yet Schmertz’s invention is clearly employed. Morley Machine Co. v. Lancaster, 129 U. S. 263, 9 Sup. Ct. 299, 32 L. Ed. 715. I do not think Schmertz was confined to the means described in his specifications. His claims cover any means for introducing the wire. It was more than improvement, as he invented a new plan. His invention, therefore, covers more than the precise means described, of which defendant’s feed device, is an equivalent, especially as it uses Schmertz’,s plan through part of the process. Westinghouse v. Boyden Power Brake Co., 170 U. S. 578, 18 Sup. Ct. 707, 42 L. Ed. 1151; Royer v. Coupe, 146 U. S. 524, 13 Sup. Ct. 166, 36 L. Ed. 1073.
A decree should be entered for an- injunction and account, as prayed in the bills.