SANBORN, Circuit Judge.
This appeal involves a decree for the usual injunction and accounting for the infringement of claim 1 of letters patent No. 626,746, issued June 13, 1899, to Charles R. Bird, claim 1, of letters patent No. 668,175, issued February 19, 1901, to .Anton S. Froslid, and claims 1, 2, and 3, of letters patent No. 684,751, issued October 15, 1901, to Anton S. Froslid. These patents illustrate an advance in the art of separating oats from wheat, and disclose improvements in sieves and aprons, and in the arrangement of sieves and aprons in fanning mills, to accomplish such a separation. The fact that oats pass endwise but cannot pass sidewise through perforations no larger than required to permit the passage of wheat was obvious, and rigid and flexible aprons had been used as riders upon the mixed wheat and oats as they passed over vibrating sieves to prevent the oats from taking a vertical position, and thus passing through the perforations with the wheat, long before these patents were issued. The use of these aprons aided, but it failed to effect the separation satisfactorily. A fanning mill provided with a gang of sieves or screens inclining slightly downward from their receiving ends or heads and secured in a vibratory shoe was old in the art. A top sieve provided with a flexible apron resting upon the succotash upon it as a rider to keep the oats in a horizontal position, called a “scalping sieve,” had been frequently used with an underlying gang of sieves, but these had proved ineffectual to accomplish well the result sought. Froslid attained that result in this way: He placed a gang of five sieves inclined downward from their receiving ends or heads in the vibratory shoe of a fanning mill, so that each underlying sieve extended farther to the rear than the sieve next above it. He hung upon a transverse bar over the head of the upper sieve by one end a long flexible apron or flap which extended the entire length of and served as a rider upon this uppermost or scalping sieve. When the succotash was fed to this sieve the kernels of wheat, which were smaller and heavier than the oats, commenced to rattle rapidly through it and through the screens beneath it near their heads, while the kernels of oats, which were lighter, tended to pass on toward the tails of the screens. Many of these oats, however, assumed a vertical position as they fell through the screens, and slipped through the parts of them more remote from their heads and then through the screens beneath, and mingled with the cleaner wheat which had dropped through near the heads of the screens. The problem he sought to solve was to find some way to keep the cleaner wheat that came through the head ends of the screens separate from the other parts of the stock until the latter could be completely cleaned, to deliver these other parts under aprons riding upon the screens which would keep the kernels of oats horizontal iipon the screens below the scalping sieves, so that these parts of the stock might be separately subjected to the process [262]*262until the wheat therein was completely extracted from the oats, and to prevent the stock from falling upon any parts of the screens not covered by imperforate aprons. He conceived such a mode of operation and embodied it in these means. Beneath the scalping sieve and its single long, imperforate flexible apron riding upon the succotash which was fed to this sieve near its head he placed four sieves so that the tail of each one extended a short distance beyond that of the screen next above it, and over each of these sieves he hung four short, dividing flexible imperforate aprons by their heads upon transverse wires so that the four transverse wires were about four inches apart, and the tail of each apron extended a short distance over the head of the corresponding apron beneath it between the screen on which it rode and the next screen below, and so that each of the aprons, except the four tail aprons, extended a short distance under the head of the next apron toward the tail of the sieve upon which it rested. The principle of this combination was (1) that the succotash which fell through the scalping sieve was immediately divided by these short aprons into parcels according to its grade, (2) that the cleaner parcel which fell through near the head of the scalping sieve was practically prevented from again mingling with the dirtier parcels until all the wheat had passed through the four screens and the process of separation was completed, (3) none of the oats or succotash fell upon any part of any of the four screens below the scalping sieve, but all of it fell upon imperforate aprons or imperforate heads of screens, and after being shaken into a horizontal position was slid off the heads of the screens and off of the imperforate aprons upon a screen beneath a riding apron, and (4) that each of the dividing aprons performed the offices of a rider, keeping the oats beneath it in a horizontal position, and of a carrier bearing along and delivering beneath the next rider upon its sieve or over its tail the stock which fell through the screen above it. The fanning mill which Fros-lid constructed in this way satisfactorily separated wheat from oats, went into general use, and became commercially valuable, and then the defendant constructed a mill which accomplished the same result by placing upon each of the three lower sieves beneath the scalping sieve a single flexible imperforate apron with four transverse slots in it about four inches apart and with pieces of wood upon the tail sides of these slots to divide the stock and cause it to pass through the slots, and by so arranging these slots that the delivery end of each of the sections of the apron beneath them overlapped the receiving end of the section of the apron upon the next screen below it.
The validity of the patents in question and their infringement by the defendant are challenged, and the patents which are cited as anticipations will now be noticed.
Bodge in 1865, in' letters patent No. 51,687, had disclosed boards fastened by cleats at the proper distance above screens to allow oats and other grains to rise upon their ends freely, but not sufficiently to enable them to pass through the screens.
In 1866, by his patent No. 58,052, upon which the defendant seems chiefly to rely, he had described an improvement in grain separators.' [263]*263made, as he said, to force longer or broader grains to pass over sieves flatwise while the shorter or narrower grains pass through them, which consisted of a single sieve which had three perforate and three imperforate parts alternating along its length, three wooden covers in the form of parallelograms rigidly secured far enough above the perforate parts of the screen to permit oats and wheat to pass between three flexible aprons secured to the undersides of these wooden covers respectively and the perforate parts of.the sieve, and three chutes just over the imperforate parts of the screen, one at the end of the front cover, one between that cover and the middle cover, and one between the middle cover and the tail cover. In operation one-third of the stock was fed down each chute, where it would fall upon an imperforate part of the sieve and pass towards its tail upon the next perforate part and under the next flexible apron. Counsel argue, and experts testify, that this improvement may be made to accomplish the same result as Froslid’s combination by substantially the same means if one places two or three of Bodge’s screens in suitable positions beneath the single sieve he described. If so, that result would flow, not from any improvement disclosed or any mode of (operation portrayed by the drawings, specification, or claims of the patent to Bodge, but from the mode of operation conceived and described and the mechanical means adopted by Froslid. The patent to Bodge contains no suggestion or description of the simple and effective combination of Froslid from which a mechanic skilled in the art could have constructed the latter. The fact that, while such a device as that of Froslid was constantly sought, all the suggestions of the patent to Bodge failed to inspire any mechanic with the happy thought that the multiplication beneath the scalping sieve of screens in the proper positions covered by short dividing aprons hung at their heads would reach the desideratum, is a demonstration that more than mechanical skill was required to accomplish this end, and that the patent of Bodge did not anticipate the invention of Froslid. In truth, it described nothing but a single scalping sieve divided into perforate and imperforate parts and short flexible aprons secured to wooden covers and riding on the perforate parts of the sieve. The raw stock was fed upon each perforate part as it was upon the ordinary scalping sieve with a single apron. Bodge’s sieve accomplished nothing that the common scalping sieve failed to accomplish, and it performed the function of the latter less efficiently than the ordinary sieve, because it exposed the larger portion of the stock to less perforate surface. It failed to suggest the division for treatment of the material which fell from the scalping sieve, the prevention of the mingling of the cleaner with the dirtier grain until the process of separation was completed, the multiplication of sieves or their arrangement beneath the scalping sieve, or the use of dividing aprons simultaneously as riders and carriers.
A series of bare screens receding successively from upper to lower (Noah, 31,128, February 12, 1861), bare receding sieves whereby the grain which passed through the upper sieve was divided into three divisions (Brooks & Ogden, 35,360, May 21, 1862; Hurlbut, 77,4-90, [264]*264May 5, 1868; Youngs, 39,090, June 30, 1863), a separate plate perforate or imperforate beneath the tail third of a sieve (Youngs, supra), rigid covers or plates over screens to hold oats in a horizontal position (Putney, 30,183, September 25, 1860; Thomas, 41,872, March 8, 1864), and an endless apron moved over stock upon a screen by rollers to force the wheat through it (Watson, 33,617, October 29, 1861), appear among the improvements described in the prior patents cited; but while in some of them descriptions of some, and in others descriptions of other elements of Froslid’s device may be found, none of them suggested, much less portrayed, the complete .and effective combination he formed. And it was in the state of the art which has now been disclosed that the improvements claimed in the patents in suit to which we now turn were conceived and described.
Bird placed a longer sieve beneath a shorter one, so that the tail of the latter extended nearly the length of the shorter sieve beyond it, and attached to the tail of the shorter sieve the head of an imper-forate flexible apron or flap which extended over the tail third of the longer sieve, rode upon it, and carried the stock which it received from the tail of the upper sieve down over the end of the longer sieve, and thus served as a tailings extension of the former and a rider of the latter; and the claim of his patent was:
. “In a grain-separator, the combination with a shaking shoe of a pair of overlapping sieves of unequal length, and a flexible flap mounted to serve as a tailings extension for the upper or shorter sieve, and as a rider for the lower sieve, substantially as and for the purposes set forth.”
The nearest approach in the prior art to this combination seems to be the rigid board cover of Putney, which acted .as a tailings extension for an upper and a cover for a lower sieve. But that inflexible board lacked the automatic action of Bird’s flexible apron, necessarily failed to conform to the volume and flow of the stock beneath it, and could not have performed its function as cheaply and effectively as Bird’s device. The majority of the court, however, are of the opinion, in which the writer does not concur, that the defendant did not infringe this claim, because the improvement it describes is so slight that a patent protects only against those who use the very device claimed, and the defendant has no pair of sieves of unequal length in combination with a flexible flap so that they fall within the terms of the claim.
We come, then, to the patents to Froslid. In the first one he shows a series of three sieves which receive the stock that has passed the scalping sieve upon the head of the upper sieve. The tail of each of the two lower sieves projects beyond that of the sieve next above it. Each of the lower screens has an imperforate deck which lies under the head half of the screen above it, and receives the cleaner wheat which falls through that part of the overlying screen, and delivers it upon the head of its sieve, where it falls through that screen. Under the tail half of the screen above each of these sieves an im-perforate lap deck is suspended which receives the dirtier stock that falls through that half of the screen above it, carries it over the head end of the sieve beneath it through which the cleaner wheat is passing, and delivers it upon an intermediate part of that screen. This lap [265]*265deck has a flexible apron attached to its head, which rides upon the sieve beneath it and acts as carrier and rider. This combination partially, but not completely, separated the stock into two sections and separately treated those sections. It caused all the stock passing through the screens to fall upon imperforate parts, and caused it to be carried forward on to the screens and delivered under flexible riders. The claim of the patent here in suit reads:
“In a grain-separator, tiie combination witli a señes of overlapping sieves receiving the passed stock, one from the other, the lower sieves projecting successively in a given direction, of a corresponding series of main or head decks underlying each higher sieve for receiving from the head portion of the overlying or higher sieve and delivering to the head of the next lower sieve, and a corresponding series of lap decks underlying the lower portion of each higher sieve and overlying the upper portion of each lower sieve, said lap decks inclining in the same direction as said main decks and sieves, for receiving from the lower portion of the overlying sieve and delivering to the central portion of the lower or underlying sieve, substantially as and for the purposes set forth.”
The defendant insists that there was no novelty in this combination, and cites Putney, Nash, Brooks & Ogden, Hurlbut and Bodge. Putney shows a conducting board beneath a sieve, but no flexible apron acting as rider and carrier, and no series of overlapping and successively projecting sieves; Nash, a gang of screens without the arrangement of Froslid and without lap decks; Brooks & Ogden, a series of three overlapping and successively projecting plates with perforate and imperforate sections, but without lap decks with flexible flaps which serve as riders and carriers; Hurlbut, a complicated arrangement of screens without lap decks with flexible flaps; Bodge, wooden decks with flexible flaps over the parts of a scalping sieve, but no overlapping and successively projecting screens with lap decks between them to treat the stock after it has passed the scalping sieve. None of them discloses the combination of Froslid, and it is no defense to a claim of an infringement that one or more elements of a patented combination, or one or more parts of a patented improvement, may be found in one old patent or publication, and others in another, and still others in a third. It is indispensable that all of them, or their mechanical equivalents, be found in the same description or machine, where they do the same work by substantially the same means. Imhaeuser v. Buerk, 101 U. S. 647, 660, 25 L. Ed. 945; Bates v. Coe, 98 U. S. 31, 48, 25 L. Ed. 68; Latta v. Shawk, 1 Bond, 259, Fed. Cas. No. 8,116; Machine Co. v. Pearce, 10 Blatchf. 403, Fed. Cas. No. 4,312; Manufacturing Co. v. Steiger (C. C.) 17 Fed. 250, 252; National Cash Register Co. v. American Cash Register Co., 3 C. C. A. 559, 53 Fed. 367; Rhodes v. Lincoln Press-Drill Co. (C. C.) 64 Fed. 218, 219; Packard v. Lacing-Stud Co., 16 C. C. A. 639, 641, 70 Fed. 66, 68. The combination of the first claim of this patent was novel; it docs not appear in any prior patent or publication; it marked a distinct advance in the art, a step toward the perfected combination of Fros-lid’s second patent which no mechanic with all the references now cited before him had ever taken; and this record is far from making it clear that the findings of the Commissioner of Patents and of the [266]*266court below that it evidenced invention, which are presumptively correct, were erroneous in law or mistaken in fact; and the validity of the first claim of this patent must be sustained.
While conceding that the defendant uses every other element of this combination except the lap deck, counsel insists that its machine does not contain that element, and hence does not infringe this claim. The function of the lap deck is to extend over the imperforate deck and the head of the sieve below it, to catch the stock falling from the part of the sieve just above it, to carry it forward to an intermediate part of the screen below it, and at the same time to act as a flexible rider to the. part of the screen just beneath it. Now the defendant has between its screens long flexible aprons, each of which covers and rides upon the sieve beneath it, catches the stock falling through the sieve above it, and has three transverse slots in it, four inches apart, through which the stock thus caught is delivered to the sieve beneath '\t at intervals of four inches. In operative effect these slots divide the long apron into four short aprons. The head apron upon each sieve, the four inches of apron between the head of the apron and the first or head slot, laps over the imperforate deck at the head of the screen beneath it, catches and carries to the head slot, and there delivers to an intermediate part of the sieve beneath it the stock which it has received from the screen above it, and at the same time rides upon the sieve beneath it, and thus in eveiy way performs the very function of the lap deck of Froslid by means of a flexible apron in the exact relation to the sieves that Froslid’s bears.
But counsel contends that the defendant escapes infringement because in the specification of his patent and in his drawings Froslid shows a lap deck composed of a rigid element, such as a zinc plate, about half as long as the deck, and a flexible apron or flap, the head end of which is fastened to the head end of this plate while the remainder of it rides upon the sieve below and holds the oats in a horizontal position. But the description in a specification or drawing of a form, or a composition, or construction, of a mechanical element, when that form, composition, or construction is not, and is not claimed to be, essential to the combination or improvement claimed, is the mere pointing out of the best mode in which the patentee contemplated applying the principle of his invention under section 4888, Rev. St. (U. S. Comp. St. 1901, p. 3383), and does not deprive him of protection for mechanical equivalents or indicate that he intended to give up all other modes of application. Continental Paper Bag Co. v. Eastern Paper Bag Co, 210 U. S. 405, 418, 28 Sup. Ct. 748, 52 L. Ed. 1122; National Hollow Brake-Beam Co. v. Interchangeable Brake-Beam Co, 45 C. C. A. 544, 566, 106 Fed. 693, 715; City of Boston v. Allen, 91 Fed. 248, 249, 33 C. C. A. 485, 486.
Again, in the claim in suit the patentee does not specify or claim the two elements of the lap deck, but claims the lap deck simply as one element of his combination; while in his second claim, that is not here in suit, he specifies and claims the rigid element and the flexible element of a lap deck as elements of the combination there described; and where a patent contains a general claim for a com[267]*267bination of certain mechanical elements and a specific claim for a combination of a specified form, composition, or construction of one of those elements with the other elements of the general claim, the legal presumption is that the two claims secure different combinations, and the general claim is not limited to the specific form, composition, or construction claimed in the specific claim, hut protects the element and its mechanical equivalents, though in form, composition, or construction differing from that of the specific claim.
Mere changes of the form or composition of a device or of some of the mechanical elements of a combination will not avoid infringement, where the principle or mode of operation of the patented improvement or combination is adopted, unless the form or composition is the distinguishing characteristic of the invention. Columbus Watch Co. v. Robbins, 64 Fed. 384, 396, 12 C. C. A. 174, 187; New Departure Bell Co. v. Bevin Bros. Mfg. Co. (C. C.) 64 Fed. 859; Machine Co. v. Murphy, 97 U. S. 125, 24 L. Ed. 935; Winans v. Denmead, 15 How. 342, 14 L. Ed. 717; Robinson on Patents, § 141, p. 201; Blandy v. Griffith, 3 Fed. Cas. p. 678, No. 1,529; Bonnette Arc Fawn Sprinkler Co. v. Koehler, 82 Fed. 431, 27 C. C. A. 200; National Hollow Brake-Beam Co. v. Interchangeable Brake-Beam Co., 45 C. C. A. 544, 562, 106 Fed. 693, 711.
“If two devices do the same work in substantially the same way and accomplish substantially the same result, they are the same, even though they differ in name, form, or shape.” Machine Co. v. Murphy, 97 U. S. 125, 24 L. Ed. 935. A lap deck composed of the flexible flap alone properly fastened in place by its head performs the same function in the combination of the first claim of this patent in substantially the same way and by substantially the same means as one composed of a rigid element and a flexible element, and hence the one is the plain mechanical equivalent of the other.
The next contention is that the file wrapper and its contents show that the patentee inserted in his specification the two elements of the lap deck in order to obtain his patent, and that he is thereby estopped from claiming one composed of a flexible flap alone as an equivalent of the lap deck he claitned. The file wrapper and contents have been examined, but they fail to support this contention. The original specification describes these lap decks as imperforate deck sections with flexible flaps secured to them. All the claims, 14 in number, attached to this specification were rejected on March 15, 1899. Thereupon the specification and the claims were practically rewritten, and the lap decks were described as composed of a rigid deck section and a flexible flap; but this was no modification or change of the original specification, because the description in that specification clearly showed that the deck sections to which the flexible flaps were described as secured must have been rigid. Three claims were submitted with this specification. The first and second of these claims, so far as they relate to the composition of the lap decks, were identical in meaning and effect with those allowed under those numbers in the subsequent patent. The first claim sought to secure the lap decks in combination without any specification of the elements of which they were com[268]*268posed, and the second claim specifically claimed the rigid element and the flexible element of the lap deck in combination with the other mechanical elements of the first claim. On April 21, 1899, the first claim was rejected. Froslid answered this rejection that the feature of construction in claim 1 was new and patentable, that this fact was practically conceded by the examiner’s statement that he would allow claim 2, that claim 2 differed from claim 1 only in the fact that the lap decks there named called for rigid sections and flexible sections, and that it must be obvious that the results pointed out in the specification would be accomplished if the lap decks were in single sections of suitable material. After this argument, and on July 15, 1899, claim 1 was again rejected. Froslid then amended it, inserting in their places therein these words, “the lower sieves projecting successively in a given direction,” and “said lap decks inclining in the same direction as said main decks and sieves,” but made no change in the portion of the claim relating to the lap decks, and wrote to the examiner concerning it:
“The claim sets a novel combination in which the so-called lap decks operate in a manner which is entirely novel, the remaining claims have been indicated as allowable, and it is therefore respectfully submitted that this case is now in condition for allowance.”
The claim was then allowed. Thus it may be seen that, instead of acquiescing in any limitation of his first claim to a lap deck composed of a rigid and a flexible element, the file wrapper and its contents prove that Froslid always insisted that his second claim secured to him that specific composition of a lap deck, and that his first claim was for the combination of the other elements with a lap deck, whether composed of one or two elements, that he pointed out that a lap deck of a single element might be the mechanical equivalent of one with the two elements named in the specification, and that the Commissioner of Patents after that argument and declaration sustained and granted the first claim.
If a patentee acquiesces in the rejection of his claim on references, he may be estopped to maintain that an amended claim covers the combinations shown in those references, or that it has the breadth of the x-ejected claim, but he is not estopped from claiming and securing by the amended claim every improvement and combination which he has invented and which was not disclosed by those references. National Hollow Brake-Beam Co. v. Interchangeable Brake-Beam Co., 45 C. C. A. 544, 565, 106 Fed. 693, 714. The references which induced the amendment of the first claim did not disclose a lap deck composed of a flexible flap in the combination of Froslid, and hence they did not limit his first claim to a combination in which the lap deck is composed of a rigid element and a flexible element.
The result is that the first section of the long flexible apron of the defendant performed the same work in substantially the same way and accomplished the same result as' the lap deck of Froslid’s first claim, and was therefore its mechanical equivalent, and, as the defendant used every other element of this combination, it did not escape infringement.
[269]*269In May, 190.1, 28 months alter he filed his application for the patent just considered, Froslid made an application for a patent on the perfected combination which has proved successful and valuable, and that patent was issued in October of that year. The differences between the three claims of this patent are not material in this case, and claim 2 has already been set forth to illustrate them. The principle of the combination claimed, the means by which the patentee put that principle in operation, the beneficial results he secured, the prior state of the art, the fact that no prior patent or publication disclosed his combination or its equivalent, the reasons why the Bodge patents failed to anticipate it, and the means used by ihe defendant to take the mode of operation of Froslid’s combination, have been set forth in the earlier part of this opinion, and it is unnecessary to repeat them. Certain specific objections to the validity of this second patent remain to be noticed.
It is said that it is demonstrably untrue that the parcels of the stock divided by Froslid’s aprons can never again commingle. This may be literally true, but the proof convinces that while it is possible for a few grains to pass from one parcel to another, and a few may do so, the practical effect of Froslid’s mode of operation and of the machine of the defendant which adopts it is to maintain the separation of the parcels until the separation of the oats from the wheat is completed. and it is not material if thereafter these parcels commingle in the chute which carries them elsewhere.
It is said that the alleged advantage from, the overlapping aprons is purely imaginary, and that they operate as well when they do not overlap. But the evidence convinces that the lapping of the tail ends of the overlying dividing aprons over the head ends of the underlying aprons is not without beneficial effect.
It is said that the second Froslid patent is anticipated by the first, and there is testimony to show how readily the combination of the latter could have been modified to constitute that of the former, and to the effect that the necessary change was one of degree. But it was a far cry from the three sieves of the first patent with the single lap decks between the upper and middle sieves and between the middle and lower screens to the gang of four sieves with their dividing aprons, each about four inches in length, acting simultaneously as riders and carriers covering the entire surface of the three lower sieves with the tail ends of the overlying aprons lapping over the head ends of the underlying aprons, and the conception and embodiment in mechanical means of the mode of operation of the latter were not the product of mechanical skill, but they clearly evinced the intuitive genius of the inventor. They constituted the final step to an efficient and valuable machine to separate oats from wheat, which immediately went, and has since continued, in successful operation, the commercial value of which the defendant did not fail to perceive. The combinations of the three claims of Froslid’s second patent were novel, useful, ail'd patentable, and they must be sustained.
The last question is, does the machine of the defendant infringe the claims of this patent? Its counsel maintains that it does not, because [270]*270it has no dividing aprons, because it has no sieves extending beyond the ends of such dividing aprons, because it has -no delivery ends of overlying dividing aprons lapping over the receiving ends of like underlying aprons, and because it has no dividing aprons which are overlapped by those upon the same sieve. All these objections, except the last, rest upon the proposition that the long slotted aprons of the defendant are not the mechanical equivalents of the dividing aprons of the complainant. But the sections of these long aprons between the transverse slots completely divide the stock into separate parcels according to its grades, and substantially, keep these parcels separate until the process of separation of the wheat from the oats is completed; they are of substantially the same length as the dividing aprons of Froslid; they serve as riders for the screens beneath them, and as carriers of the stock which falls upon them from the screens above; their delivery ends overlap the receiving ends of the sections which lie beneath them, and they perform the same function by the same im-perforate flexible aprons as do the dividing aprons of Froslid. It is true that none of the sections lap under the sections upon the .same sieVe, but they nevertheless deliver the stock through the slots upon sieves inclining downward toward .their tails beneath the next sections, and thus accomplish the function of the overlapping of the corresponding aprons of Froslid by the same mode of operation and with an immaterial change of form. They are, therefore, the mechanical equivalents of the dividing aprons of the Froslid second patent. The truth is that the defendant has taken the principle, the whole substance of the invention claimed by Froslid, embodied it in mechanical means which are the obvious equivalents of those which the patentee described, in means which differ from them only in matters of slight and immaterial form, and the finding of the court below that it was guilty of infringement upon the three claims of this patent may not be lawfully reversed.
The result is that the defendant below succeeds in its defense to claim 1 of the patent to Bird, on the ground that it is not guilty of infringement thereof, and the complainant below succeeds upon claim 1 of the first patent and upon the three claims of the second patent to Froslid. Where a suit is brought upon several claims of one or more patents, and the complainant succeeds in obtaining relief upon some oJj the claims but fails to recover upon others, an equitable division of the costs proportioned to the expense of litigating the respective claims should be made, because the defendant is not justly liable for the costs of litigating those claims upon which the complainant was entitled to no relief. Ide v. Trorlicht, Duncker & Renard Carpet Co., 53 C. C. A. 341, 354, 115 Fed. 137, 150; Willcox & Gibbs Sewing Machine Co. v. Merrow Machine Co., 35 C. C. A. 269, 93 Fed. 206; Thomson-Houston Electric Co. v. Elmira & H. Ry. Co. (C. C.) 71 Fed. 886; Johnson v. Foos Mfg. Co., 141 Fed. 73, 90, 72 C. C. A. 105. The claim of the Bird patent was of minor importance, and the substantial issue in this suit involved the other claims.
The decree below must be reversed, the appellant may recover one-fifth of its costs in this court, and the case nmst be remanded to the [271]*271Circuit Court with directions to dismiss the bill as to the first claim of patent No. 6.20,746 to Bird, and to enter the usual decree for an injunction and an accounting upon claim 1 of patent No. 668,175 to Froslid, and upon claims 1, 2, and 3 of patent No. 684,751 to Fros-lid, and to award to the complainant four-fifths of its costs in the Circuit Court up to the time of the entry of the decree.
It is so ordered.
HOOK, Circuit Judge, concurs in the result.