B. F. Sturtevant Co. v. Massachusetts Hair & Felt Co.

122 F.2d 900 | 1st Cir. | 1941

122 F.2d 900 (1941)

B. F. STURTEVANT CO.
v.
MASSACHUSETTS HAIR & FELT CO.
MASSACHUSETTS HAIR & FELT CO.
v.
B. F. STURTEVANT CO.

Nos. 3644, 3645.

Circuit Court of Appeals, First Circuit.

September 29, 1941. Rehearing Denied November 21, 1941.

Harry Dexter Peck, of Providence, R. I., and Melvin R. Jenney, of Boston, Mass., for B. F. Sturtevant Co.

Ralph L. Chappell, of Kalamazoo, Mich. (Robert J. Keating and Roberts, Cushman & Woodberry, all of Boston, Mass., and Earl & Chappell, of Kalamazoo, Mich., on brief), for Massachusetts, Hair & Felt Co.

Before MAGRUDER, MAHONEY, and WOODBURY, Circuit Judges.

WOODBURY, Circuit Judge.

These are cross-appeals from a final decree of the District Court confirming the report of a master in a patent suit.

The B. F. Sturtevant Company, which will be referred to hereafter as the plaintiff, is the assignee and present owner of two letters patent, one granted on February 23, 1932, and the other on January 29, 1935, to one Harold F. Hagen as its assignor. These patents are numbered 1,846,863 and 1,989,413 and are respectively for "Improvement in Fans and Methods of Operating the Same" and for "Improvement in Centrifugal Fans." Hereinafter they will be referred to as Hagen's first and Hagen's second patent.

On May 18, 1936, the plaintiff filed a bill of complaint in the District Court charging the Massachusetts Hair & Felt Company with infringing both of these patents by using in its plant at Peabody, Massachusetts, a so called Vortex control fan made by the Clarage Fan Company of Kalamazoo, Michigan. The Clarage Fan Company openly assumed and conducted the defense of this suit and, unless otherwise noted, will be referred to in this opinion as the defendant.

In issue are claims 1 and 2 of the first Hagen patent and all four claims of the second Hagen patent.

The master, after full hearing, found, with respect to the first patent, that claim 1 was invalid both for anticipation and for absence of patentable invention; that claim 2 was valid; that both claims were infringed; and that the plaintiff had not been guilty of laches. With respect to the second Hagen patent he found that although none of its claims were anticipated, all were invalid for absence of patentable invention; and that all had been infringed. The District Court in its final decree adopted the findings of fact contained in the master's report as its own and confirmed that report. The plaintiff appeals from so much of the final decree "as dismisses the bill of complaint herein as to claim 1 of Letters Patent No. 1,846,863 and as to Letters Patent No. 1,989,413." The defendant appeals from the portions of that decree "which hold claim 2 of patent No. 1,846,863 valid and infringed."

In order to understand the questions of mechanics, physics and law raised by these appeals, it will be necessary to consider at the outset the construction, function and uses in industry of centrifugal fans or blowers. These machines, in their simplest form, consist of a wheel, called a rotor or impeller, attached to a shaft for rotation inside a scroll-shaped casing. Rigidly fastened around the circumference of this wheel are a series of identical blades or paddles which are usually curved backward away from the direction of rotation. In each side of the casing are circular openings around the shaft upon which the rotor is mounted. These openings are concentric with the rotor shaft and are called the air inlets, or eyes, because through them enters the air or other gas to be moved by the fan. In order not to obstruct the eyes, the bearings in which the impeller shaft is journaled are generally set some distance out from the casing.

Fans of this sort act upon the air, predominantly by centrifugal force, to increase its pressure and deliver it through the outlet which is the straight portion of the scroll-shaped casing. They are used for various purposes, including ventilation and furnace draft, and are of two general types. When such a fan is located at the entrance end of a system, that is, when it takes air directly from the atmosphere and forces it under pressure into the system, it is called a forced draft fan. When it is located near the delivery end of the system, that is, when it draws air out of the system and discharges it into the atmosphere, it is called an induced draft fan. Fans of either sort as used commercially are frequently of substantial size, the rotor or impeller in some cases being eight or more feet in diameter.

As a centrifugal fan operates and forces air through its outlet, atmospheric pressure forces air into the fan through its inlets. This air, if the inlets are open and unobstructed, enters axially, that is, parallel to the shaft upon which the rotor is mounted, but in order to replace the air being forced out, it, when inside the casing, turns at right angles to its path at entry and moves radially outward to the blades of the impeller. As the master describes it:

"Air enters the eye of the impeller (sic. casing) in a generally axial direction, and if it were to continue to flow entirely axially, it would never reach the fan blades and no work would be done. But as the air passes into the fan it moves radially toward the blades, and the air has both an axial component and a radial component, one perpendicular to the other."

The output of a fan, which the master defines as "the product of the pressure and the volume of air delivered", depends upon the amount of work done on the air in the fan by the blades of the impeller, and the amount of this work depends upon the speed relative to the entering air at which the impeller is revolved. In industry it is frequently desirable if not essential to vary this output, that is, to vary the amount of air being forced or drawn through a system, and "It is with this matter of control of fan operation to meet changing requirements of the same system, that the Hagen patents are concerned."^[1]

Obviously a reduction of the speed of the rotor of a fan will reduce the speed of its blades relative to the entering air and thus "reduce the pressure, with accompanying reduction of volume delivered", but although this, so far as the fan itself is concerned, is the best and most efficient way yet devised of controlling its output, there are distinct disadvantages accompanying the use of variable speed control in modern industrial plants.

"If the fan is driven by a steam turbine, variable speed can be easily and efficiently effected, and prior to 1922 or 1923, steam turbines were commonly so used. About that time plant designers introduced what is known as the regenerating cycle, which involved using the steam for heating the feed water, and the use of turbines for fan drive has, since then, largely gone out.

"With direct current (electric) motors, speed control can be accomplished with efficiency, although the motor control may have to be supplemented by some damper control to give control between the fixed steps of speed provided by the motor.

"But direct current is little used in industry, except in a few places such as Detroit. And with alternating current in use most everywhere, difficulties are encountered in providing for variable speed control. Such control with an alternating current motor involves high initial cost, and considerable space for large banks of resistors. What is known as a slip-ring alternating current motor can be used, but its efficiency falls off rapidly as speed is reduced. As with direct current motors, it is necessary to supplement the motor control with damper control, to provide control between the steps provided by the motor."

The damper referred to by the master in the above quotation is defined by him as "a restriction placed in the air passage." In the fan art its action is commonly termed "throttling." This method of control is "accompanied by a waste of a part of the power developed by the fan, which is used in overcoming the added resistance imposed by the damper." Because it is wasteful, throttling, except in conjunction with speed control, is seldom used except in small installations when waste is of no great consequence.

The Hagen patents pertain to a third method of regulating the output of a fan called "vane control." This method, as the name implies, "involves controlling fan operation to meet changes in system requirements by the use of adjustable vanes at the inlet or eye of the fan, operating to impart to the entering air a velocity of spin in the direction of blade rotation, by which, at constant speed, the work done on the air by the fan can be varied, with a corresponding variation in power necessary to drive the fan."

In further detail the master describes the essential principle of operation of vane control as follows:

"In an ordinary fan in which the air enters through an open unobstructed eye, the air will approach the rotor blades in a radial direction without any appreciable spin or whirl. And the speed of the blades relative to the air is the actual speed of the blades. The rotor will exert considerable force in changing the direction of flow and in impelling the air in the direction of its rotation, such work being reflected in an increase in pressure in the air as it leaves the rotor.

"A rotor cannot impart as much energy to a stream flowing in the direction of rotation of the rotor as it can to fluid which is not already flowing in that direction. If the air, by some means or other, when it enters the fan, is already whirling or rotating in the direction of fan rotation, and at the same speed, then the rotor blades can apply no force to it, and there is no increase in pressure in the air. No work will be done by the fan, and the power necessary to operate the fan is at a minimum. If the air is spinning in the direction of fan rotation, but at a speed less than that of the fan, then the fan will exert some force on the air to speed it up, and there will be an increase in pressure imparted to the air."

"It results that in a fan running at constant speed, the fan output and the power input are reduced with an increase in spin velocity, and are increased with a decrease in such spin velocity."

The master finds that vane control, as compared with damper control, effects a substantial saving in horsepower, and that "variable speed control also effects a substantial saving over damper control in horsepower input to the fan at reduced speeds", such saving being greater than that effected by vane control.

"Therefore, whenever it is possible to effect variable speed control efficiently, as by direct current motor or by steam turbine, vane control has no advantage. Indeed, vane control is less desirable than such variable speed control. But in comparison with speed control by an alternating current motor, vane control has proved to have no such disadvantage and has gone into extensive use in competition with such alternating current motor control."

"The main reason for this drop in efficiency with vane control lies in the fact that good flow conditions determined for full performance of the fan, with the vanes wide open, are violated when the entering air is given a spin component.

"The fan or impeller is so designed that at maximum performance the air or fluid will enter the blades smoothly and without shock loss, as it is called. Shock free entry is an important design consideration with any impeller such as a fan, pump or turbine and it is the effort of the engineer to so design the blade in relation to the flow of the fluid to the blade that a line tangent the blade surface at its entry edge will coincide or be parallel with the direction of relative velocity of the entering fluid at that edge."

Stated in non-technical language this means that good design of any centrifugal fan requires that the impeller blades be set at such an angle to the radius of the wheel upon which they are mounted that at maximum, or to conform to the expert testimony, at normal or usual load, they will strike the entering air with a slicing rather than with either a slapping or a cutting blow.

Naturally fans are designed for maximum efficiency at their normal load. If their output is reduced by vane control, that is by imparting to the entering air a spin in the direction of rotation of the impeller, air will no longer "enter the fan smoothly along the blade" and "There will be shock at entry and loss of energy by reason of the resulting turbulence." Obviously, as the entering air is given greater spin with further closure of the vanes, the more its direction will deviate from the radial direction in which it must flow to the rotor blades for maximum efficiency at normal load and the greater will be the shock loss with its resulting turbulence and loss of energy.

This is the background against which the Hagen patents are to be projected.

Hagen makes no claim to have discovered the principle of controlling the output of a fan by controlling the spin of the entering air. In fact it is admitted, as the master found, that "Knowledge of the principle of operation of spin control of a rotating impeller is of ancient origin."^[2] Neither does Hagen claim "to have invented broadly the method of or an apparatus for producing and varying the spin component of the entering air of a fan by an arrangement of adjustable vanes at the eye or inlet." What is contended by the plaintiff is that Hagen developed "a particular method and apparatus, which for the first time utilizes with practical success and efficiency the known principle of impeller control by adjustable vanes."

In claim 1 of his first patent Hagen describes his method of vane control as follows:

"1. A method of operating a centrifugal fan which consists in rotating the fan wheel at constant speed and controlling the output of the fan in accordance with the variations in demand thereon by supplying fluid to the fan with a spin component of velocity in the direction of rotation of the wheel, adjusting the spin through a range to preduce variations in output from substantially the full capacity of the fan to a minimum capacity thereof with a substantially tangential admission of fluid, said variation from maximum output to lower outputs being effected by increasing the spin component while reducing the area through which the fluid passes at a rate not proportionately greater than the rate of increase of the spin component."

This is the claim which the master and the District Court found invalid for both anticipation and lack of patentable invention.

Claim 2 of this first patent, which is the only one of those involved in this suit which was found valid, is an apparatus one. In it is described the mechanism which the patentee devised to apply vane control to an induced draft fan. This mechanism consists of two scroll-shaped inlet chambers, one mounted on each side of the fan casing itself, both of which are connected to a common supply conduit. Extending across each inlet chamber and around each eye are a plurality of vanes. The inner ends of each set of vanes are pivoted on studs arranged in a circle around each eye of the fan casing and the outer ends of each set are similarly pivoted on studs arranged in similar circles on the inside of the outer walls of the inlet chambers.

The vanes are curved, or of cylindrical contour, and those in each set are mounted quite close together on their pivots which are located near the trailing or inner edges. The vanes in each set are connected one with another and each set of vanes is connected with the other set on the other side of the fan, by a mechanism of links, studs, springs, rods, shafts and gears, which need not be described, in such a way that they can be rotated on their pivots simultaneously either by hand or mechanically. When fully opened the vanes in each set radiate outwardly around each eye of the fan. In this position they have but little hampering effect and they impart no spin to the entering air but rather take from that air such spin as may have been imparted to it by the scroll-shaped inlet casings. When closed the vanes in each set overlap one another and so form the sides or walls of horizontal cylinders one end of each of which opens into an eye of the fan and the other end of each of which is closed by the center portion of an outside wall of an inlet chamber.

The vanes in each set, when in closed position, may be likened to the staves of a straight sided cask or barrel set horizontally in an inlet chamber in such a way that the head or open end surrounds an eye of the fan and the closed or bottom end is formed by the middle part of the outside wall of an inlet chamber. The vanes, or to continue the simile of the barrel, the staves, are each pivoted on each end so that they may be rotated simultaneously on longitudinal axes. Thus, as they are rotated from closed position, air passages are formed between them through which air is admitted to the area enclosed by the vanes, that is, to the inside of the barrel, with a spin or whirl which persists in the air as it is sucked through the eye of the fan and into contact with the blades of the impeller. In likening Hagen's vanes to the staves of a barrel it must be borne in mind that, when in closed position, they do not meet edge to edge like barrel staves, but overlap one another.

This construction provides vanes which, as to each set, the master finds to be substantially parallel in all positions and which as they close come into closer parallelism until upon full closure they form a complete cylinder around each eye of the fan. Due to the substantial parallelism of the surfaces of adjacent vanes, partially opening them creates air directing passages of substantially uniform cross-sectional area.

The patentee describes his mechanism in claim 2 of his first patent as follows:

"The combination with a constant speed motor, of a centrifugal fan having a rotor, an inlet, a plurality of vanes in the inlet forming fluid directing passages to admit fluid to the rotor with a spin component of velocity in the direction of rotation of the rotor, the vanes being adjustable to vary the output of the fan through a range from maximum capacity with the vanes open to a minimum capacity with a substantially tangential admission of fluid when the vanes are closed, adjacent vane surfaces continuously approaching parallelism and having extended overlapping portions shaped to define passages of substantially uniform cross-section as the vanes are moved toward closed position."

Before taking up these claims in detail some of the terms used therein will have to be explained.

It will be noted that in both of them Hagen uses the phrase "substantially tangential admission" to describe the direction taken by the air as it enters the eye of the fan when his vanes are almost closed. The master finds that by this phrase Hagen "contemplates and means that as the vanes are closed, you approach a condition where the whole body of air, as it leaves the vanes and enters the fan, is spinning, going around in a circle, tangential to the eye or to the rotor", — that is to say, at right angles to the radius of the eye or rotor. It is admitted that it is mechanically impossible to achieve completely tangential admission, and even if it could be achieved no useful purpose would be served. The reason for this is that if the entering air moved in an absolutely tangential direction after leaving the vanes it would never come in contact with the blades of the impeller. Thus no work would be done by the fan, it would have no output, and so it would be better to shut it off.

In his first claim Hagen says that by his method variation in output is effected by "increasing the spin component while reducing the area through which the fluid passes at a rate not proportionately greater than the rate of increase of the spin component." By this phrase the master finds that Hagen is giving "a mere mathematical description of what occurs in any air directing vane system which supplies good guidance, when the volume of flow is maintained constant, except * * * where the vanes are excessively thick." In elaboration he says: "this vague clause of claim 1 is at best a statement that the vanes are of such number and so arranged as to supply air directing and guiding passages. The vanes cannot be excessively thick, for they could not then close sufficiently far to supply tangential admission as called for earlier in the claim. Therefore, if they are of such size (not too thick) as to be capable of closing to supply tangential admission, then the law as to the rate of increase of air spin and decrease of area inevitably follows." In short, he finds that "The most that can be claimed for this clause is that it is an indirect way of stating that the vanes are of usual thickness and operate to supply substantially complete guidance throughout their range of adjustment."

The last phrases of claim 2 will, for better understanding, be explained later on in this opinion when consideration of that claim is reached.

In connection with claim 1 the master makes the finding that "Hagen was the first to develop a constant speed fan wherein the output is controlled by supplying a variable spin to the entering air by means of adjustable vanes at the inlet. His vane controlled fan constitutes a highly useful and valuable contribution to the fan art, has gone into wide use, and has supplied a very real need." But, in spite of this, he found that claim 1 of the first Hagen patent did not set out a patentable invention. He says:

"Turning to Hagen's claim 1, there is no patentable novelty or invention involved in the subject matter of the first portion of that claim, and I do not understand that plaintiff contends otherwise. The principle of spin component and its effect on the output of a rotor were known and understood as far back as Euler. And in view of the prior art, there was no novelty or invention in the method of controlling the output of a constant speed centrifugal fan by supplying fluid to the fan with a spin component of velocity in the direction of rotation of the rotor, or in adjusting the spin to produce variations in output. The patentable novelty of claim 1 must reside in adjusting the spin `through a range to produce variations in output from substantially the full capacity of the fan to a minimum capacity thereof with a substantially tangential admission of the fluid', such variation throughout this range being effected by increasing the spin component while reducing the area, as defined in the concluding clause of this claim.

"I fail to see wherein these features of the method of claim 1 * * * involve more than the efficient utilization of known principles and methods, or at most the carrying forward of a known method."

"The method of spin control over less than the maximum range being known, it does not in my judgment constitute any invention in method to merely extend that range of control by carrying forward a known method."

"To state the matter in another way, if there is any novelty in claim 1, it resides in result rather than in method. The method of control is the old one of supplying a spin component which is adjustable and which is effected by increasing the spin while decreasing the area, as specified in claim 1. To effect control over the wide range specified may be a new result, but as to method, the steps of claim 1 are old. So far as claim 1 is concerned, if Hagen effected a new result (more extended control), he did it by merely extending and carrying foward an old method by old means."

This question of patentable invention is one of fact. Thomson Spot Welder Co. v. Ford Motor Co., 265 U.S. 445, 44 S. Ct. 533, 68 L.Ed. 1098. As such the master's finding, since it was adopted by the District Court, "shall not be set aside unless clearly erroneous." Rule 52 of the Rules of Civil Procedure, 28 U.S.C.A. following section 723c. See National Biscuit Company v. Crown Baking Company, 1 Cir., 105 F.2d 422. The plaintiff takes the position that the finding below of lack of patentable invention as to this claim is clearly erroneous. It contends that as a matter of law it is entitled to a finding in its favor on this question, first, because Hagen's "advance is a new concept and a change of kind over what was previously known," and second, because all reasonable men must agree that Hagen in far outstripping his predecessors by achieving control over a wide range almost wholly by spin and almost wholly without throttling, must have used an inventive faculty rather than a merely mechanical skill.

Hagen's "new concept" was not that the output of a centrifugal fan could be regulated by imparting to the entering air a variable spin in the direction of the impeller's rotation. It is admitted, as the master found, that this concept was of ancient origin. In so far as method is concerned, it seems to us that the most that can be said for Hagen on the score of novelty is that he found a use in modern industrial plants for an old and well known but relatively inefficient method of regulating the output of a centrifugal fan. That is to say, learning that steam turbines, which could readily and efficiently be regulated as to speed, had become, through changes in plant design, no longer available to industry for the purpose of driving fans, and learning that in consequence industry in general had been forced to use for that purpose electric motors of a type incapable of cheap and efficient speed control, he found that the known method of controlling the output of a fan by placing adjustable vanes in the inlets, although inefficient, had become less inefficient than either speed control or hampering. In our opinion this, although a contribution to the fan art and a benefit to industry, is not an invention or discovery of a nature to entitle Hagen to the protection afforded by a patent.

The question remains as to the wider range of control, primarily by spin, which Hagen's method provided. Admitting that he went further than his predecessors in disregarding shock-loss, with its attendant inefficiency and waste, and admitting that he improved greatly upon the method of control by spin over what was known before, still it does not follow that there is anything patentable about his method. While some improvements may be protected by a patent, it is well established that not all of them can be. Few principles of the law of patents are better established than "that a mere carrying forward of the original thought — a change only in form, proportions, or degree, doing the same thing in the same way, by substantially the same means, with better results — is not such an invention as will sustain a patent." Market Street Cable Railway Company v. Rowley, 155 U.S. 621, 629, 15 S.Ct. 224, 228, 39 L.Ed. 284.

The following language of the Supreme Court in the case of Smith v. Nichols, 21 Wall. 112, 118, 22 L.Ed. 566, the basic case on this subject, is peculiarly appropriate:

"A patentable invention is a mental result. It must be new and shown to be of practical utility. Everything within the domain of the conception belongs to him who conceived it. The machine, process, or product is but its material reflex and embodiment. A new idea may be ingrafted upon an old invention, be distinct from the conception which preceded it, and be an improvement. In such case it is patentable. The prior patentee cannot use it without the consent of the improver, and the latter cannot use the original invention without the consent of the former. But a mere carrying forward or new or more extended application of the original thought, a change only in form, proportions, or degree, the substitution of equivalents, doing substantially the same thing in the same way by substantially the same means with better results, is not such invention as will sustain a patent. These rules apply alike, whether what preceded was covered by a patent or rested only in public knowledge and use. In neither case can there be an invasion of such domain and an appropriation of anything found there. In one case everything belongs to the prior patentee, in the other, to the public at large."

This is still the law. De Forest Radio Co. v. General Electric Co., 283 U.S. 664, 51 S.Ct. 563, 75 L.Ed. 1339. See, also, Walker on Patents, Deller's Edition §§ 17, 31.

The case of Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45, 43 S.Ct. 322, 67 L.Ed. 523, upon which the plaintiff principally relies, is not in point. In that case the plaintiff obtained a patent for an improvement of a paper-making machine which consisted in increasing the downward slope of an endless wire-mesh belt which carried liquid paper-making stock from a reservoir at one end into the machine itself at the other, and, as it carried that stock, drained the water out of it. This improvement made it possible to increase the speed at which the machine could be operated without damage to the finished product. It was held patentable.

The reason for this holding was that the change in the machine, while only one of degree, introduced a new principle into its operation. Before Eibel the liquid stock was deposited on the end of the practically level moving belt with some "head" and it was then carried or dragged along by the belt into the machine. If these machines were operated at high speed ripples or waves formed in the stock as the water drained out of it through the meshes of the belt and the finished product was defective. By giving the belt a substantial downward tilt Eibel added the force of gravity to the forces of "head" and "drag." That is, after Eibel, the stock during the early part of its travel from the reservoir to the machine, and while it was still liquid, was not picked up and carried along by the belt, but flowed by gravity down it. In this way it was possible, by speeding up the machine, to equalize the speed of the liquid stock on the belt with the speed of the belt itself, so that ripples would not be formed, and thus the speed of the machine was increased without doing damage to the finished product. The court clearly intimates (page 63 et seq. of 261 U.S., 43 S.Ct. 322, 67 L.Ed. 523) that it was the addition of the principle of gravity to the principles of "head" and "drag" which resulted from Eibel's change in the degree of pitch of the belt that made his innovation patentable. In the case at bar we can discern, as to method, no change by anything which Hagen did in the principle of control over the output of a fan by vanes, and consequently the Eibel case is not in point.

It follows that the conclusion of the master and of the court below that claim 1 of the first Hagen patent is invalid for lack of patentable invention must be sustained and we pass on to claim 2 of that patent.

This claim, as appears above, is an apparatus one in which is described the patentee's mechanism for applying vane control to an induced draft fan. It is the only one of those under consideration which the master and the court below found valid. Specifically the court and master concluded that it had not been anticipated; that it embodied patentable invention; and that the plaintiff had not been guilty of laches.

On the question of anticipation the defendant submitted to the master evidence of six prior patents, — three domestic and three foreign. The domestic ones are Eickhoff No. 928,034 and Moody Nos. 1,460,428 and 1,578,843. The foreign ones are Brown Boveri (Swiss) No. 99,575, Beaudrey (French) No. 589,469, and Keller (British) No. 17,098. The master made detailed findings with respect to each of these patents as a result of which he concluded that none of them anticipated this claim of Hagen's first patent.

This conclusion the defendant vehemently assails in an argument which is long, involved and highly technical. We do not need to exhaust and confuse the reader by considering it, however, because an examination of the record discloses that the position taken by the master is impregnable.

The question of priority of invention, even when the attempt is made to invalidate a patent by showing an earlier one rather than by merely showing prior general knowledge and use, is, like the question of patentable invention, one of fact. 1 Walker on Patents, Deller's Edition, § 61; United States v. Esnault-Pelterie, 303 U.S. 26, 30, 31, 58 S.Ct. 412, 82 L.Ed. 625. The reason for this is that "It is not the construction of the instrument, but the character of the thing invented, which is sought in questions of identity and diversity of inventions." Bischoff v. Wethered, 9 Wall. 812, 76 U.S. 812, 816, 19 L.Ed. 829. Consequently we are not at liberty to reverse the finding of lack of anticipation made below unless that finding appears to be clearly erroneous. Such is not the case. The record contains the testimony of an admitted expert on the subject of centrifugal fans to the effect that in his opinion the mechanism described in Hagen's patents is not disclosed or suggested in any of the prior art patents listed above. The admissibility of testimony of this sort has long been established (Bischoff v. Wethered, supra), and the credibility of the person who gives it presents in each case a question not for us but for the trier of the issue of fact. Rule 52, supra. After an attentive examination of the testimony and of the master's careful and detailed report on this phase of the case, we cannot say that the conclusion reached below was clearly erroneous.

We are ready now to consider the defendant's contention that this claim is "devoid of any patentable invention." The master posed the question as follows:

"Did Hagen's vane adjustment and arrangement involve invention or was it merely a matter of good design and good workmanship, and within the expected skill of the art?"

He answered it thus:

"If anything akin to genius is essential, Hagen was not an inventor. But Hagen was an engineer who was not content to merely try and improve the known commercial forms of fan control. He sought to find out if and how an old principle could be made practical for complete fan control. He was not a mere routines. Whether it can be said that he possessed intuition or courage, or mere persistence and a capacity for patient experimentation, is not clear. But to him the art is indebted for substantial, practical progress. He had a goal and he attained it, to the great profit and benefit of the art. The particular apparatus which he developed embodied an arrangement which was new and satisfactorily supplied a successful and useful vane control to a centrifugal fan for the first time in the practical art. While this is a close case and the question of patentable invention, as distinguished from mere skill or good designing, is not free from very real doubt, in my judgment the evidence is sufficient to resolve that doubt in favor of the patent and to warrant the conclusion that defendant has not sustained its burden of proof."

There are in the books a multitude of objective tests by which to determine this question of patentable invention. Several of them are urged at length by counsel in the case at bar, but in view of the following language of Judge Learned Hand in the case of Kirsch Manufacturing Co. v. Gould Mersereau Co., 2 Cir., 6 F.2d 793, 794, with which we heartily agree, we see no use in expanding this opinion by considering each such test in detail:

"Objective tests may be of value vaguely to give us a sense of direction, but the final destination can be only loosely indicated. An invention is a new display of ingenuity beyond the compass of the routines, and in the end that is all that can be said about it. Courts cannot avoid the duty of divining as best they can what the day to day capacity of the ordinary artisan will produce. This they attempt by looking at the history of the art, the occasion for the invention, its success, its independent repetition at about the same time, and the state of the underlying art, which was a condition upon its appearance at all. Yet, when all is said, there will remain cases when we can only fall back upon such good sense as we may have, and in these we cannot help exposing the inventor to the hazard inherent in hypostatizing such modifications in the existing arts as are within the limited imagination of the journeyman. There comes a point when the question must be resolved by a subjective opinion as to what seems an easy step and what does not. We must try to correct our standard by such objective references as we can, but in the end the judgment will appear, and no doubt be, to a large extent personal, and in that sense arbitrary."

Since the background or basis for any determination of this question of patentable invention must necessarily be the prior art we must give that matter somewhat detailed consideration.

Before Hagen the only devices known to the art for controlling the output of a centrifugal fan or pump by the use of adjustable vanes in the inlet were those described in the six patents enumerated above in connection with the discussion of the question of anticipation. All of them are of the "paper" variety, that is, the master found upon sufficient evidence that none of the devices described therein had ever been manufactured or used.

But before taking up the master's findings with respect to these patents the language of the last part of Hagen's claim 2 must be explained. This language is "adjacent vane surfaces continuously approaching parallelism and having extended overlapping portions shaped to define passages of substantially uniform cross-section as the vanes are moved toward closed position."

Obviously vanes arranged around the eye of a fan in the way described in claim 2 of Hagen's first patent will, when open, radiate outwardly around the eye. In this position they will not be parallel and so the areas between their adjacent surfaces will not be of uniform cross-section but will taper toward the eye. If the vanes are straight, or of rectangular contour, like those described in the patent to Eickhoff, they will, when viewed end on, that is, in cross-section, appear to radiate around the eye of the fan in the same way that the spokes of a wagon wheel radiate around the hub. If the vanes are of cylindrical contour, like Hagen's, they will, when viewed in the same way, appear to radiate around the eye of the fan as the curved spokes of the ordinary flywheel radiate around its hub. Although straight vanes and cylindrical ones have many characteristics in common, they differ in one very important respect. With straight vanes mounted on pivots similar to Hagen's the angle between adjacent vanes remains constant in all positions. Therefore these vanes are never parallel and never approach parallelism and the areas between their adjacent surfaces can never approach uniformity of cross-section. Vanes of cylindrical contour, on the other hand, as they are turned toward closed position continuously approach parallelism until, upon complete closure, they are almost parallel, and thus the areas between their adjacent surfaces approach uniformity of cross-section as closure is approached. Vanes of either type may be constructed so as to have "extended overlapping portions" as called for by Hagen's patent. This feature, although not essential to the formation of air directing passages, aids in attaining that end.

Now the six patents mentioned above can be considered. The master found that four of them, that is, the Brown-Boveri one, both of Moody's, and Keller's, taught nothing with respect to the type of vanes used by Hagen. Specifically, he found that the language of the Brown-Boveri patent was too vague and inconclusive to teach anything, and that in the Moody patents there was no teaching that, upon closing, his vanes "continuously approach parallelism, defining passages of uniform cross-sectional area, as called for by Hagen's claim 2," and, in addition, "without further and more specific information as to vane form, there is no showing of substantially tangential admission on closing." He found that in the Keller patent "It does not appear that Keller contemplated a plurality of walls in overlapping relation, forming fluid directing passages which would give and adequately control a spin component to the entering fluid, as in Hagen."

With respect to Eickhoff the master found that although he described a vane arrangement very similar to that of Hagen, his vanes were straight or of rectangular contour, and thick, instead of being of cylindrical contour and thin, as Hagen's were. The master found that Eickhoff in his patent described vanes which were "adjustable to increase or decrease the spin component of the fluid from maximum to minimum as means for controlling output"; that he described "the use of such vane adjustment for this purpose"; and that his vanes could, with only a mechanical change, i. e. by being made thinner, be made to provide substantially tangential admission without appreciable throttling. But the master found that Eickhoff's vanes differed from Hagen's in the important respect referred to above. That is, he found that since Eickhoff's vanes were of rectangular contour the angle between their adjacent surfaces remained constant in all positions, so that his vanes could never approach parallelism and so could not define passages of substantially uniform cross-sectional areas when they were moved toward closed position.

The master found that Beaudrey in his patent described an arrangement of vanes substantially similar to the arrangements of both Eickhoff and Hagen, but that his vanes were "fish-shaped," or of "streamlined" contour. With respect to this patent the master found: "Beaudrey clearly shows and describes adjustable vanes and describes their use to control output at constant speed" but his vanes "do not provide substantially tangential admission at closing. Nor can it be said that in Beaudrey adjacent vane surfaces, on closing movement, continuously approached parallelism, defining passages of substantially uniform cross-sectional area, although they have extended overlapping portions."

The master stated his conclusions as follows:

"The advance of claim 2, therefore, resides in the design and arrangement of the vanes. As distinguished from Eickhoff, the vanes continuously approach parallelism, and as distinguished from Beaudrey, they can be adjusted to effect substantially tangential admission, a feature present in Eickhoff. And also they continuously approach parallelism, with passages of substantially uniform cross-sectional area.

"By designing his vanes so that on closing movement they continuously approach parallelism to define passages of substantially uniform cross-sectional area, Hagen has provided more definite, predictable and better control than does Eickhoff. In a structure like Eickhoff's, the angle or direction of entry cannot be accurately predicted or controlled. And in Eickhoff there are greater losses by eddy currents between entering jets than in Hagen's arrangement as defined in claim 2. In general, it may be said that Hagen's arrangement of claim 2 gives better flow control and better flow conditions than that of Eickhoff. And there is the same difference and advantage in Hagen over Beaudrey. Also, for the reason that Beaudrey's vanes do not supply the uniform passage of the claim (because of their shape), they fail to effect tangential admission, with the resulting limitation of range of control."

Hagen's contribution, then, was of a mechanism for controlling the output of a centrifugal fan by vanes, not down to a useless capacity of zero, but only to a minimum capacity. This had practical value and it achieved commercial success. It approximated, if it did not equal, the range of control possible by altering the speed of the impeller, as had been done with efficiency by steam turbines but which could not be efficiently done with the usual type of electric motor available for fan drive after steam turbines had become unavailable for the purpose, and it accomplished this range of control almost entirely by power-saving spin and with a minimum of power-wasting throttling.

It does not seem to us that it would have occurred to a mechanic or designer skilled in the art either to disregard increasing shock-loss by radically increasing the spin of the entering air or to construct vanes of cylindrical contour for the purpose of providing passages between them "of substantially uniform cross-section as the vanes are moved to closed position", so that control to minimum capacity by spin rather than by throttling would result. Thus the finding of patentable invention made by the master and confirmed and adopted by the court must stand.

The question of laches can be briefly disposed of. The master found that "Hagen's first patent issued in February, 1932, defendant started putting out its fan in April, 1932, and the complaint in this suit was filed in May, 1936." In view of the lack of any showing that the plaintiff's delay of four years and one month in bringing suit in any way prejudiced the defendant, we certainly cannot say that the master and the court below were clearly in error in finding that the "plaintiff has not been guilty of such laches as to deny relief either as to injunction or accounting."

Hagen's second patent is an improvement one. All four of its claims are for an apparatus.

"The specification refers to the first Hagen patent and states that the object of this second invention is to provide vane control in a fan in which the air is admitted directly from the atmosphere, a forced draft fan."

The particular features of the design of this fan are adjustable vanes arranged on oblique axes in conical inlets. More specifically, this fan consists of an impeller attached to a shaft for rotation inside a scroll-shaped casing in which are circular inlets concentric with the rotor shaft. To avoid obstruction of these inlets the bearings in which the impeller shaft is journaled are set out away from the casing. In these respects, and in the principles of its operation, it is similar to Hagen's induced draft fan. It differs from that fan, however, in that instead of inlet chambers it has conical inlet passages. In each of these passages a set of flat, sector-shaped, overlapping vanes are mounted on pivot pins journaled respectively in the inlet cones and in hubs secured to the impeller shaft bearings. Due to the fact that these bearings are set out from the eye and to the fact that the vanes are journaled in hubs on these bearings "It will be noted that while the axes on which the vanes turn extend in a direction which is generally radial in relation to the shaft, such axes are inclined from the perpendicular, and the vanes when closed form a conical surface."

"By varying the position of the vanes, the spin velocity of the entering air will be varied, and the output of the fan controlled, in accordance with the principles described in connection with Hagen's first patent."

The first of the four claims of this second patent reads as follows:

"A centrifugal fan comprising a rotor, a casing having an eye and a substantially conical inlet passage leading to the eye, and a plurality of vanes in the inlet pivoted on axes which are oblique to the rotor axis and substantially at right angles to the conical inlet, the vanes being adjustable about their axes from substantially wide open position in which they do not appreciably obstruct the flow of gas to the rotor through an angle of nearly 90 degrees to a closed position, throughout which adjustment they direct the entering gas to the rotor with an appreciable spin, the work done by the fan on the gas being variable in accordance with the adjustable spin introduced by the vanes at different positions."

Claim 2 of this patent "adds to claim 1 the provision of a hub on the rotor axis, with the vanes pivoted on the hub," and "Claim 3 adds to claim 1, and claim 4 adds to claim 2, `means for simultaneously and uniformly adjusting the vanes about their axes'."

The master found that "the claims of Hagen's second patent are not directly anticipated," but that this "patent is invalid as to all four claims, for absence of patenable invention."

On the question of anticipation the master considered three prior patents. These are the first Moody one and Keller's, which have already been referred to in this opinion, and United States patent No. 1,490,955 issued to one Bell. With respect to these patents the master, after referring to the findings which he made when he was considering the pertinence of the first two of them to the first Hagen patent, said: "I find nothing pertinent in Keller, for reasons already given," and "for reasons set forth in my discussion of Moody in relation to Hagen's first patent, there is no showing of an arrangement of vanes adjustable from wide open position through an angle of nearly 90 degrees, as called for by the four claims of Hagen. Nor is this feature clearly shown and described in the Bell patent."

Since, as already appears, this question of anticipation is one of fact, and since upon a careful examination of the record we are unable to find that the master's conclusions with respect to all of these prior art patents are clearly erroneous, it follows that they must be affirmed.

On the question of patentable invention the master found as follows:

"Hagen makes no claim to have disclosed any new principle of operation by his second patent. Insofar as he made any invention in the matter of vane arrangement and operation applicable to fans generally, that invention is described and disclosed in his first patent.

"Hagen's second patent is limited to a particular manner of mounting vanes for vane control in the eye of a forced draft fan. The alleged novelty resides in mounting such vanes on oblique axes in a conical eye, with the hub away from the eye."

"Before Hagen got up the vane controlled fan of his second patent, the conical inlet was a common and well known form of inlet. Moreover, it was well understood that for good efficiency the eye inlet should not be obstructed, and for this reason it had been good engineering practice to mount the bearings of the motor shaft away from the eye.

"I fail to see wherein there can be any invention in arranging vanes in the eye of a forced draft fan, and, in view of Hagen's first patent, making them adjustable from a wide open position through an angle of substantially 90 degrees, as and for the purpose set forth in the claims. Nor can I see wherein there can be any patentable invention in the old conical inlet and oblique axes, and setting the hub out from the eye so that it will not obstruct the inlet, precisely as the bearing of the rotor shaft has always been mounted for the same reason."

It is not altogether clear from the above whether the master intended to find that this Hagen patent was invalid for lack of patentable invention or whether he intended to find it invalid because it was anticipated by Hagen's first one. Counsel for both parties interpret the report as meaning the latter and we take it that their interpretation is correct. At any rate, a finding of invalidity for lack of patentable invention can, under the findings quoted above, be based only on the ground that the mechanism described in this patent is merely an aggregation of old parts, and on that ground the finding cannot be sustained. The reason for this is that the Hagen mechanism for applying vane control to a forced draft fan, although it combined old parts, combined them in such a way as to produce a new and useful result, i. e. more predictable and better control over a wide range, such control being almost wholly by spin and almost wholly without throttling, and this new and useful result was the "product of the combination, and not a mere aggregate of several results each the complete product of one of the combined elements." Hailes v. Van Wormer, 20 Wall. 353, 368, 22 L.Ed. 241. See, also, Grinnell Washing Machine Co. v. E. E. Johnson Co., 247 U.S. 426, 38 S.Ct. 547, 62 L.Ed. 1196, and cases cited. Thus the combination is patentable. Seymour v. Osborne, 11 Wall. 516, 20 L.Ed. 33; Webster Loom Co. v. Higgins, 105 U.S. 580, 26 L.Ed. 1177; Barbed Wire Patent, 143 U.S. 275, 12 S.Ct. 443, 36 L.Ed. 154; Krementz v. S. Cottle Co., 148 U.S. 556, 13 S.Ct. 719, 37 L.Ed. 558; Expanded Metal Co. v. Bradford, 214 U.S. 366, 29 S.Ct. 652, 53 L.Ed. 1034; Smith v. Snow, 294 U.S. 1, 6, 55 S.Ct. 279, 79 L.Ed. 721.

On the question of anticipation by Hagen's first patent the defendant, citing Mahn v. Harwood, 112 U.S. 354, 6 S.Ct. 451, 28 L.Ed. 665; Toledo Scale Co. v. Computing Scale Co., 6 Cir., 9 F.2d 823; and others of similar import, argues that since the issuance of a patent operates as a dedication to the public of all that might have been claimed therein but was not, Hagen's first patent, in which he failed to claim his apparatus for a forced draft fan, operates to invalidate his second one. The argument is beside the point because Hagen applied for his second patent before his first patent was issued. Both of his applications were co-pending in the patent office from October 31, 1929, to February 23, 1932. Under these circumstances the rule is that when the claims are for separate inventions, as they are here, one being for an induced draft fan and the other being for a forced draft one, "The issuance of the first patent does not abandon the unclaimed matter in its disclosure, the pendency of the second application rebutting any such inference," and this is so even though the claims of the second patent do not "embody a patentable advance over the earlier disclosure." Traitel Marble Co. v. U. T. Hungerford Brass & Copper Co., 2 Cir., 22 F.2d 259, 260, 261. See, also, Barbed Wire Patent, supra, page 280 of 143 U.S., 12 S.Ct. 443, 36 L.Ed. 154.

But, the defendant urges, the claim of Hagen's second patent calling for a vane adjustment of nearly 90 degrees was not made in the original application but was inserted by amendment on July 19, 1934, almost two and a half years after his first patent was issued. This, in our view, makes no difference because "the specification as filed furnished an adequate foundation for the allowed claims, and there was no new matter requiring an additional oath," and "Hagen's application had from the beginning contained claims directed to the same general subject matter as that of the allowed claims, only broader." Under these circumstances the amendment relates back to the date of the original application and both are to be regarded as constituting one continuous application. Godfrey v. Eames, 1 Wall. 317, 326, 17 L. Ed. 684; Smith v. Goodyear Dental Vulcanite Co., 93 U.S. 486, 500, 23 L.Ed. 952; Wagenhorst v. Hydraulic Steel Co., 6 Cir., 27 F.2d 27.

From this it follows that the master and the court below erred in concluding that Hagen's second patent was invalid in view of his first one. Since the apparatus described in Hagen's second patent produces by comparably similar means the same novel result as the apparatus described in claim 2 of his first one, it follows that his second patent also embodies patentable invention.

We come now to the question of infringement.

The defendant's device, which it calls a Vortex control, is for a forced draft fan. Its version of this type of fan "has the usual casing and rotor. From one side of the casing projects the rotor drive shaft, driven by a constant speed motor. At the other side is the inlet passage or eye, which is shaped like a section of a torus, — that is, it is like the hole in a doughnut.

"The sector-shaped vanes are trunnioned on rods in the inlet passage, these rods at their inner ends being threaded into a spider hub which they support, and at their outer ends secured in holes in the inlet passage. Gears, which need not be described, are provided for turning the vanes about the rods to open and close the vanes, which can be turned through an angle of nearly 90 degrees. When closed the vanes form a continuous surface. It will be noted that, as in Hagen's second patent, the supporting hub is located outwardly from the center of the inlet passage, so that the rods or axes of the vanes are inclined from the vertical, and when closed the vanes form a conical surface."

"As in the fans of the Hagen patents, when the vanes are wide opened, they do not appreciably affect the air flow. As they are closed they direct the air to the rotor with a spin component. The general resemblance of defendant's fan to plaintiff's commercial forced draft fan, like that of Hagen's second patent, is apparent."

There is, however, one very important difference between the defendant's vane control and that of Hagen's second patent. In the defendant's apparatus "there is not so much overlap of adjacent vanes as in the Hagen patent", and "as closing position is approached, there is only a very slight overlap, which, at closing is an overlap of about 1/16", near the outer edge, which decreases toward the hub and finally disappears, leaving a slight crack between adjacent vanes near the hub."

Due to our affirmance of the finding below that claim 1 of the first Hagen patent was invalid we do not need to give detailed consideration to the findings below that that claim was infringed. Swift & Sons v. W. H. Coe Mfg. Co., 1 Cir., 102 F.2d 391, 396. But we must mention those findings briefly because the master referred to them in making his findings on the question of the defendant's infringement of claim 2 of that patent. With respect to infringement of claim 1 he says: "It may be that in defendant's fan the guidance supplied by the passages between adjacent vanes is not as complete and perfect as it might be, or as is supplied by the vanes of plaintiff's fans. But I believe that the guidance is sufficient throughout the full range for all practical purposes, and sufficiently effective to control output, even in the lower ranges, predominantly by varying the spin, as distinguished from throttling. If defendant's fan as made does not supply as perfect and complete guidance as might be, infringement is not avoided by impairing its efficiency."^[3]

Turning now to claim 2 of Hagen's first patent we find in the record the following statement of the master: "The first portion of claim 2 is met by defendant's fan, if my finding as to claim 1 is sound. The issue as to claim 2, therefore, is whether or not, in defendant's fan, adjacent vane surfaces continuously approach parallelism and have extended overlapping portions shaped to define passages of substantially uniform cross-section as the vanes are moved toward closed position, as called for by the concluding portion of claim 2."

The master resolves this issue by saying: "In my judgment, in defendant's fan adjacent vane surfaces continuously approach parallelism, at least in substance, within the fair meaning of the claim, and define passages of substantially uniform cross section as the vanes are moved towards closed position. Also, while the overlap of adjacent vanes in defendant's fan is less than in Hagen's patent, and substantially disappears at closing, in my judgment there is a sufficient overlap to fulfill the aim and purpose of an overlap in the functioning of the fan. While, as I have said, in defendant's fan the amount of overlap is less than in Hagen's, defendant's blades still define passages of substantially uniform cross section as they close and with results commensurate with those effected by vanes having a greater degree of overlap, as in Hagen.

"I therefore find that defendant's fan, * * * infringes claim 2 of Hagen's first patent."

This question of infringement is also one of fact. Stilz v. United States, 269 U.S. 144, 147, 46 S.Ct. 37, 70 L.Ed. 202 and cases cited. See also United States v. Esnault-Pelterie, 303 U.S. 26, 29, 58 S.Ct. 412, 82 L.Ed. 625. As such the conclusion of the master and the court below is not to be disturbed by us unless clearly erroneous. Galion Iron Works & Mfg. Co. v. Beckwith Machinery Co., 3 Cir., 105 F.2d 941, 942. One of the experts who appeared before the master categorically testified that adjacent surfaces of the defendant's vanes continuously approached parallelism as the vanes were moved toward closed position and that although the defendant's vanes do not overlap as much as Hagen's, still as closure was approached, they overlapped sufficiently to define air passages of substantially uniform cross-section. A visual examination of the exhibits in the case serves to confirm rather than to refute this testimony. In view of this we consider it unnecessary to discuss the defendant's elaborate argument based upon its analysis of measurements of distances between lines drawn on various planes through its vanes.

On the question of the defendant's infringement of Hagen's second patent the master made the following findings:

"First as to claim 1, the broadest claim.

"Defendant contends, firstly, that its fan does not infringe this claim, because the vanes do not direct the entering air with a spin throughout the entire range of adjustment, as called for by the claim, because as closing is approached, there is throttling instead of spin control. In my judgment this contention should be rejected for the reasons which I gave as to this contention in relation to Hagen's first patent.

"Secondly, defendant contends that the inlet passage of its fan is not conical, as called for by claim 1, but is shaped like a section of the surface of a torus. While a torus inlet is not conical, it is in this connection in every fair sense the equivalent. There is no file-wrapper estoppel which forbids the application of the doctrine of equivalents to this requirement of claim 1.

"I find that claim 1 is infringed.

"Claim 3 merely adds to claim 1 the provision of `means for simultaneously and uniformly adjusting the vanes about their axes'. This element is present in defendant's fan and I accordingly find that claim 3 is infringed.

"Claim 2, as distinguished from claim 1, calls for a hub and a plurality of vanes `pivoted on the hub and the conical inlet'. In defendant's fan, the vanes are mounted pivotally on spider rods, which are screwed into the hub and bolted to the torus inlet, and defendant contends that its vanes, therefore, are not pivoted on the hub and the inlet. In my judgment, this contention is without merit. Defendant's spider rods are part and parcel of the hub and inlet, and therefore the vanes are pivoted on the hub and the inlet within any fair interpretation of the language of claim 2. There is obvious equivalency.

"I therefore find that claim 2 is infringed.

"Claim 4 adds to claim 2 the provision of `means for simultaneously and uniformly adjusting the vanes about their axes.' Defendant's fan provides such means, and accordingly I find that claim 4 is infringed."

We see no reason to add any discussion of ours to the foregoing. What is there said by the master is self explanatory and clearly correct. We might add, however, that a visual examination of the exhibits amply confirms all of his conclusions.

Testimony was closed on April 28, 1938. On January 4, 1939, the master submitted to counsel a draft of his report, and on February 11, 1939, the defendant filed a motion to take additional proof which motion was accompanied by the affidavit of its principal engineer. Hearing was had on this motion and it was denied by the master on the ground that even if the facts alleged should be proved he would not alter his conclusions; that the motion was not seasonably filed; that the defendant did not contend that it was in any way surprised at the hearing, nor did it claim that the evidence was newly discovered; and that it offered no excuse for not producing its evidence at the hearing and made no motion at the hearing for a continuance in order to produce it.

It seems to us self-evident that the grounds given for the denial of this motion are amply sufficient to support the discretionary ruling made by the master and confirmed by the District Court. Eggen v. United States, 8 Cir., 58 F.2d 616, 620, and cases cited.

The decree of the District Court is affirmed as to Hagen's first patent and reversed as to his second one. Neither party is to recover costs in this court or in the court below. The cases are remanded to the District Court for further proceedings not inconsistent with this opinion.

NOTES

[1] Hereafter in this opinion, unless otherwise noted, quotations will be from the master's report, which, since it was adopted and confirmed by the District Court, embodies the latter's findings of fact and conclusions of law as well.

[2] It is agreed that this principle was discovered in the Eighteenth Century by a Swiss mathematician and physicist named Euler.

[3] This finding cannot be successfully attacked because it is supported by evidence (a) of an expert, (b) of a chart showing a horsepower input curve for the defendant's fan similar to that for a Hagen controlled fan, and (c) of the defendant's advertising matter. See Underfeed Stoker Co. v. American Engineering Co., D.C., 35 F.2d 32, affirmed, 3 Cir., 41 F.2d 985.