The Pittsburgh Wire Company and Thomas W. Fitch, President, were sued by Henry Roberts, George T. Oliver, and Andrew J. Day for an alleged infringement of the first, third, fourth, and sixth claims of letters patent No. 392,365, dated November 6,1888, granted to Henry Roberts, and which were subsequently acquired, through mesne assignments, by the defendants. On hearing on bill, answer, and proofs, the circuit court held the first and third claims to be invalid, and that the sixth claim had not been infringed. The fourth claim was sustained, and the defendants were enjoined from infringing the same. Both parties have appealed (69 Fed. 624), — the defendants, from so much of the decree as sustains the validity of the fourth claim; and the complainants, from that part of the decree which dismisses the bill as to the first claim. The patent in dispute relates to an improvement in rodmills, by which metallic rods are passed through a train of finishing rolls preparatory to their being drawn into wires. The specification describes, in part, the improvement as follows:

*707"In rolling metal rods in a mill, as now commonly practiced, It is customary to employ a series of trains of rolls, set in line with each other, and to pass the metal back and forth between these rolls. As the metal is reduced in diameter and increased in length, there is always difficulty in controlling it and preventing it from kinking, especially so when it is attempted to roll more than one rod at a time in a single mill. For this purpose it has been usual to employ boys, who stand with hooks opposite to the rolls, and guide the elongating metal loop, with a view of preventing it from kinking or injuring the workmen. The work of these boys is very dangerous, and requires the closest attention, and therefore the wages paid them are quite high. Besides this, the least neglect on their part is apt to cause the kinking of the rod, and when this happens the delay in its passage through the mill chills the metal, and unfits it for use, necessitating the cutting of it into pieces for scrap. The object of my invention is to diminish these evils, and to provide means for guiding the metal loop as the rod is passing from one set of rolls to the next. To this end, it consists mainly in the combination, with two sets of rolls arranged on different lines of feed, of an inclined mill floor common to said rolls, whether the same be used either with or without an overfeed regulator leading from the rolls to the inclined floor of the mill, or with or without a guide arranged on the floor, as shown by me in the dra wings, and the inclination of the floor being in a direction transversely to the axis of the rolls, and being such that, as the metal loop elongates, it shall be forced by the action of the primary roils away from the rolls, down the inclined floor, on which it travels freely, thus preventing kinking, and lessening the labor necessary to be employed in guiding it.”

The claims involved in these appeals read as follows:

“(1) In a wire-rod mill, the combination, with the main mill floor having an inclined surface which extends in a plane transversely to the rolls, substantially as described, of a series of rolls arranged on different lines of feed, whereby the propelling force of the rolls, and the gravity of the loop, are utilized to cause the loop to travel freely over the floor, substantially as and for the purposes specified.” “(4) In a wire-rod mill, the combination, with the main mill-floor having an inclined surface extending in a plane transversely to the rolls, substantially as described, of a series of rolls arranged on different lines of feed, whereby the propelling force of the rolls, and the gravity of the loop, are utilized to cause the loop to travel freely over the floor, and a guide extending along the said inclined main floor transversely to the delivery side of the primary rolls, and adapted to guide the primary branch of the loop, substantially as and for the purposes specified.”

The defenses are anticipation, lack of patentable novelty, noninfringeinent, and fraud on the part of the patentee in obtaining his patent. The application for the patent was filed on May 12, 1888, and to prove anticipation the defendants produced (1) a German treatise entitled “Das Eisenhuttenwessen,” by Prof. Joseph von Ehrenwerth, published in Leipzig, Germany, in the year 1885; (2) the McCallip patent, No. 331,516, of December 1, 1885; and (3) the Lenox patents, No. 351,836, of November 2, 1886, and No. 351,840, of the same date. The German publication contains the drawing and description of a mill at Dumnarfvet, Sweden, from which it appears that (translation):

“The cast-iron floors on both sides of these rolling mills are inclined, and between every two sets of rolls there are provided horizontal steps, S', rounded towards the outside. By this arrangement the hooker boys are entirely done away with, since the wire loop, in consequence of the inclination of the floor itself, draws or bends outwards, and thereby is perfectly conducted or guided by the steps. But it Is easy to understand that this arrangement is practically only applicable where the rolls (calibers) are arranged in one line in the pairs of standers (roUs) following one another. It *708gives, besides tbe advantage of a saving of labor to tbe process of rolling, a finished aspect.”

Mr. Kennedy, the complainants’ expert, in speaking of the operation of this mill, testifies, on cross-examination,-as follows:

“X-Q. 172. Please refer to tbe German publication which is in evidence in this case, and state whether that describes a rolling mill in which loops are to be formed in the process of making the rod. A. It does. X-Q. 173. State whether it shows a series of rolls arranged on different lines of feed. A. It does. X-Q. 174. Does it also show an inclined surface which extends in a plane transversely to the rolls? A. Yes. X-Q. 175. Is the construction such that the propelling force of the rolls, and the gravity of the loop, are combined to cause the loop to travel down tlie incline? A. Yes. X-Q. 176. What do you find set forth in that German publication as the object of the construction there described? A. It says that the hooker boys are done away with, and it also gives to the process of rolling a finished aspect. X-Q. 177. Don’t you think that the construction of mill which is shown in this German publication, taken in connection with the description of the same contained in said work, would naturally have suggested to a rod-mill man, on reading the same, at the date of the said publication, the combination of rolls arranged on different lines of feed, and an inclined floor common to said rolls, for the purpose of facilitating the movements of the loops, and also for dispensing with or lightening the labor of the hooker boys? A. I think It would.”

TMs testimony from the complainants’ own expert would seem to be conclusive of the question of anticipation, so far as it relates to claim 1, but we are also referred to other patents which are alleged to be anticipatory of this claim.

The McOallip patent, No. 381,516, of December, 1885, is for an “overfeed regulator of rolling mills,” and consists of an inclined floor in front of two adjacent pairs of rolls, which are placed end to ■ end. In front of the rolls is located a semicircular guide, C, called a “repeater.” The inclined floor is placed in front of this repeater, so as to receive the loop formed by the overfeed of the rod. The “overfeed” is only another name for the loop which is formed by the second pair of rolls running at a slower rate of speed than the first pair, and which causes the loop to flow over the repeater, and to elongate. The object of the regulator is to take care of the overfeed. The language of the specification is:

“The object of my invention is to provide certain improved means for automatically keeping from tangling the loop of working metal which forms between the two stands of rolls, and is technically known as ‘overfeed.’ ”

The inclined floor has a vertical flange at each of its sides, andis provided on its upper face with a series of steps or caps, marked “b,” which are of such dimensions as to leave spaces between their two ends and the lateral flanges of the floor. The patent provides for the spacing of these caps so as to correspond with the length of the loop which is to be formed thus:

“At a suitable distance from the first cap, as may be determined in any given case, the loop passes over the second cap with like effect as in the first instance, and so on throughout the series of caps, according to the length of the loop.”

The McOallip construction is similar to that of the Roberts patent, in that in each the propelling force of the rolls, the gravity of the *709loop, and the inclined floor, are made use of to prevent the loop from kinking-.

Next in order of time follow the Lenox patents, Nos. 351,836 and 351,840, of November 6, 1886. In these the rolls are arranged in a continuous line, instead of end to end, as in the Roberts patent; and the successive rolls are run at an increased rate of speed, in order to complete the process of rolling in one heating. But, to avoid an excess of speed at the last pass, it was found necessary to reduce it at some point; and this is done by separating the primary rolls by a considerable space from the secondary ones, and by speeding the first secondary lower than the last primary. The rods are made to pass from the primary rolls through a double-guided channel, and enter the bite of the secondary, slower-running rolls. The effect of this is that the rods, being checked in their passage from one set of rolls to the other, are forced from the channels by the propulsive force of the primary rolls, and form lateral loops or overfeeds which are provided for by an inclined table on each side of the double-guided channel, and over which the loops spread. In his second patent., Lenox uses a different form of guide channel, with an inclined table on one side of it, instead of on both sides, but the overfeed is provided for by similar means to those employed in the first patent. The proofs place it beyond all doubt that Roberts was not the original inventor of the combination described in claim 1.

The fourth claim is also for a combination, but it is of a different character from the one we have been considering, in the respect that it contains an additional element to those of claim 1, namely, “a guide extending along the said inclined main floor transversely to the delivery side of the primary rolls, and adapted to guide the primary branch of the loop.” This additional feature, however, was old, as is plainly to be seen in the Lenox patents, and is suggested in the McCallip patent; and the question is whether, by its union with the other two, it makes claim 4 a patentable combination. The new combination makes use of the propulsive force of the primary rolls, of the gravity of the loop, of the inclined floor, and of the guide. Prior to the adoption of this method of rolling rods, the operation had been attended with much danger to the hooker boys, and only' one or two rods could be passed through the rolls at the same time. Formerly, it was necessary for the hooker boys to seize the loop as soon as it was formed in front of the rolls, and run with it down the floor, to prevent its tangling or kinking; making the employment a very hazardous one, often resulting in serious personal injuries, and occasionally in a fatal accident. Now, by the use of the Roberts combination only a few hooker boys are required; their work is no longer extradangerous, and by the improved process four or five, and even six, rods can pass through the rolls at once, — the several loops traveling down the guide at the same time without tangling, with the general result that the expenses of rolling are greatly reduced, and the output of the mill is largely increased. The objection to the validity of claim 4, and which is chiefly relied on, is that it contains nothing more than an aggregation or a double use of old devices. This is the common defense in cases like the present one, but the *710valuable results secured by the Roberts combination are not the sum of the independent action of its separate elements, which, for the first time, have been brought together and united in such a manner that by their joint operation they produce a new and useful result, which is a great advance in the art and manufacture of rod rolling. The inclined floor had been used alone with moderate success in preventing kinking of the loop as it passed from the delivery rolls, and thus “lessening the labor necessary to be employed in guiding it.” The guide had also been applied to a level floor without obtaining the desired effect; and it was reserved for Roberts to unite these two devices with the propulsive force of the rolls and the gravity of the loops, and by their joint action and co-operation, to achieve a decided improvement. Experiments had been previously made by Roberts, and by others who were engaged in the same business, to provide for the disposition of the loops, to reduce the danger attending their formation, and to save the waste of material consequent upon their tangling, and he was the first to make the combination described in claim 4. We have not overlooked the contention that the guide channel shown in the McCallip patent, No. 331,516, is substantially the same as that shown in the patent of Roberts. The McCallip guide is not a continuous one. It is used on the square side of the mill, and is adapted for the spreading of thé loop over the side of the caps. It is also designed for a mill in which only one rod is rolled at a time, and would be not only useless where several rods are passed through the rolls at once, on the delivery side, but also dangerous to the workmen. The doctrine relating to patentable combinations is well stated in Loom Co. v. Higgins, 105 U. S. 591, where the supreme court, by Mr. Justice Bradley, say:

“At this point we are constrained to say that we cannot yield our assent to the argument that the combination of the different parts or elements for attaining the object in view was so obvious as to merit no title to invention. Now that it has’succeeded, it may seem very plain to any one that he could have done as well. This is often the case with inventions of the greatest merit. It may be laid down as a general rule, though perhaps not an invariable one, that if a new combination and arrangement of known elements produce a new and beneficial result, never attained before, it is evidence of ■ invention. It was certainly a new and useful result to make a loom produce fifty yards a day, when it never before had produced more than forty; and we think that the combination of elements by which this was effected, even if those elements were separately known before, was invention sufficient to form the basis of a patent.”

We think that the combination of claim 4 comes clearly within the rule above stated.

Under the allegation of fraud in the procurement of his patent, the specific charge against Roberts is that, at the time of making his application, he did not believe himself to be the original and the first inventor of the combination described in claim 1. To prove this charge, evidence was produced to show that, in 1885, Roberts visited the Bewsey Mill, at Warrington, England, and there saw in actual operation the combination of rolls and inclined floor which constitute the principal features of the claim. Without entering on a full discussion of the testimony, it is sufficient to say that it falls far short of satisfactory proof of the charge. It appears that Mr. *711Roberts, while stopping in Manchester, in the month of September, 1885. accompanied a friend to Warrington, and was there taken through the Bewsey Mill by Mr. Morris, the superintendent, who swears that Roberts had every opportunity of seeing, and that he must have seen, the inclined floor placed in front of the foils, both of which were then in operation. Mr. Bleckley, the managing director of the company owning the mill, did not see Mr. Roberts on the occasion of the latter’s visit, but corroborates Mr. Morris to the extent of saying that no one entering the mill could have failed to see the inclined floor, unless he was blind. Roberts, shortly after his first visit: to the mill, made a second one, and was again shown over it by Mr. Morris. On neither occasion was any reference made to the slanting floor, and Roberts positively denies that he noticed it; his attention having been directed, as he says, to the fine adjustment of the rolls, and to the vertical repeaters, which were new to him, and attracted his interest. He also states that he was not on the oval side of the rolls, from which it is said the slanting floor extended, and Mr. Morris admits that a very imperfect view of that iloor could be had from the square side. The latter, however, asserts that Roberts was on4he balcony or bridge, which covers about one-(bird of the slanting or subfloor, and could have seen it from that point. Mr. Roberts further states that he was in Bewsey Rod Mill for a very few minutes at each of his visits. Mr. Garrett saw the Bewsey Mill in 1885, and again, in 1890, without observing the presence of an inclined floor; and Mr. Poole, a practical workman in rod mills, says that the inclined floor in that mill could not well be seen without standing in front of the delivery rolls, or going under the bridge. There are some discrepancies in the testimony of Morris and Bleckley, which may he accounted for from the fact that their depositions were taken nearly eight years after the occurrence of the events to which they testify, and when the accuracy of their recollection might be to some degree impaired. There are other circumstances which make the truth of this serious charge still less probable. Mr. Roberts, after his return from England, in the autumn of 1885, continued his experiments for making improvements in rod mills for more than two years before applying for the patent in suit, and at about the same time, to wit, May 4, 1888, he also obtained British letters patent for the same invention which is described in claim 1. It would be altogether unreasonable and dangerous, on such evidence, to find Roberts guilty of this serious charge.

The refusal of the circuit court of leave to the defendants to take further proof is assigned for error, but this assignment has not been seriously pressed by counsel. The rebuttal proofs in the case had been closed, and leave was denied on the ground of lack of due diligence. and because it was not clear that the new evidence was material. The disposition of the motion was, however, in the discretion of the court, and in general such motions are not appealable. Steines v. Franklin Co., 14 Wall. 22. The decree of the circuit court is affirmed.