BUFFINGTON, Circuit Judge.
This case concerns the instantaneous transfer of the speed power of a swiftly revolving shaft to an at-rest machine, without abrasion or wrecking of such at-rest machine. This instantaneous transmission and reception of power may best be appreciated if some future invention — as no doubt will eventually be done — should make it possible, by the application of brakes, to instantaneously stop an automobile traveling at the rate of say seventy miles an hour. Of course,' the two problems involve different facts, forces, and agencies, but in the instantaneous absorption of shock forces the two situations are mechanically analogous. It will therefore be seen that in this patent case we face a major mechanical problem.
In practice a large row of sewing machines got their power from the swift rotating shaft by means of an exposed belt connection controlled by the operator. Such operators, the proofs R 106 show, if the women were not alert, “would have their fingers caught in the transmission, their hair torn or their dresses torn.” As lint from sewing was present, this made the conditions unsanitary. Such method, which is styled the indirect drive, consisted of the main revolving shaft extending along a row of sewing machines. The latter were each provided with a counter shaft with a clutch device which was belted to the driving shaft and also to each sewing machine by a second belt. In addition to its danger incidents, the device was hampered by another incident, namely, the slow stopping of its sewing. This objection is thus correctly summarized in the patentee’s brief: “A sewing machine runs fast, for example, at.the rate of three or four inches per second. [544]*544Many parts of clothing are small and in the case of them and in other cases, it is essential to stop the machine quickly at the end of the work or to turn it. If this is not done the machine will make a chain of looped threads which is apt to break the thread in the shuttle of the machine which requires stopping and re-threading the shuttle with loss of time of the operator and with defective work, which may have to be ripped out and done over.”
That these grave dangers were recognized in the art is evidenced in the statement of Herbert Corrall in his American patent No. 1,147,377, granted July 20, 1915, to his assignee, the defendant, Singer Sewing Machine Company, a corporation of New Jersey, for a sewing machine driving device :
“This invention relates particularly to an improvement in'under drivers for use with power benches carrying a number of sewing machines driven from a rotating shaft beneath, and it has for its object to equip a bench with all the moving parts, such as shafts, pulleys and belts fully inclosed so that the operators are guarded from danger of accident, while also insuring that on stopping any one of the sewing machines the belt and driving pulley for that machine will also be stopped. * * *
“A guard or shield secured to the under side of the bench protects the operator from the pulley when rotating, and a tubular non-rotating sleeve extends over each length of the shaft between such guard or shield and the adj acent under driver.”
In that connection we note that the avowed, and indeed sole stated, purpose of Corrall was to protect the operator from danger by shielding the operating parts.
The Corrall device, although the patent was applied for and was owned by the Singer Company, the defendant, and though that company had ample resources, was never placed on the American marketi In the eighteen years succeeding its application, the Singer Company never used it, but continued to use the old indirect method and means. In that regard we have not overlooked the fact that the Corrall device went into extensive use in England, where it was manufactured by an affiliated British Company of which Corrall, the inventor, was a director. There is no testimony that the machine as there made was capable of instantaneous stoppage, and there is convincing testimony that when one sewing machine got out of order, the main shaft and all the sewing machines operating from that main shaft had to be stopped. Thus the witness Jones, who was a manufacturer of machine parts and for fifteen years familiar with the Corrall machines since he was called on to repair, testified: “At certain times its failed to engage or disengage when the operator by means of the treadle intended that it should do so, and the only way to correct that when it happened was to stop the line shaft to make proper repairs and adjustments. This line shaft drove machines other than the one where the failure might occur, so that the whole number of machines driven by the line shaft would be stopped while a particular one was corrected. In the event that the clutch failed to disengage, the sewing machine would continue to run even though the operator desired to stop it. This would so affect the sewing that the operator was working on that s'he would have to run the material off of the needle and very often the case was that they jammed the bobbin case and loaded it up with stitch cotton.”
As to the noninstantaneous stoppage of the machine, his testimony was : “When the Corrall apparatus was working properly, the time required to stop the sewing machine, in distance of stitches or revolutions, would be anything from a quarter of an inch to three-quarters of an inch — depending on the length of the stitching. It would be correct then to say that, with the Corrall mechanism, the machine did not stop instantaneously; and this failure to stop instantaneously continued the sewing beyond the point at which the operator desired to stop it.”
He testified there were some fourteen thousand Corrall machines in use in England, but that “all' of the troubles encountered in connection with the Corrall transmitter were in the screw mechanism (the cam) and the braking, due to not stopping at the correct time it should.” We find no testimony in the record controverting the above statement that the machine did not stop at the correct time it should. Later on we comment on this basic feature of instantaneous stopping and how and by whom this all-important objection was overcome. The witness testified that he knew of two factories — naming them — where the Corrall device was thrown out and the companies went back to the old indirect drive method. Lewis, a dealer in .secondhand machines, testified he found them no success what[545]*545ever; that there is no market "for secondhand Corrall machines; that they required “a big staff of maintenance men.” He testified that except in what are called “cooperative factories,” which have numbers of mechanics at call, the Corrall tables have gone out of use for ten years. He referred to another large concern “that used to use Corrall tables, but they threw them on the scrap heap and went back to the indirect drive, using probably about four thousand sewing machines all told.” He testified further: “The other concern that scrapped that table was World Wear, Birmingham, whose transmitters were replaced by those of Wilcox and Gibbs, Wheeler and Wilson, and Jones, and various other makes of two-speed transmitters. In the cut showing a typical scene in the plant, there is shown a scrap heap in a considerable amount, probably over a hundred benches. Lewis said that they looked like old iron; that they weren’t new or relatively new, but that they could have been replaced, -he supposed, with little parts, to have them fitted up all right —but they were of no use. The firm didn’t like that type of machinery and that is the reason why it was on the scrap heap because it wasn’t practical.”
From the above — and we find no testimony to the contrary — we are constrained to hold that the Corrall machine did not satisfy the British market; that it did not effect instantaneous stoppage, and inferentially that for these reasons the Singer Company, the patentee defendant, though amply able, did not attempt to put it on the American market; and that it had no effect on the art, and the Singer Company itself, as we have seen, for eighteen years after its grant, continued to follow the indirect system device.
Continuing the status of the art, it appears that Lewis Frankel had one of the largest men’s clothing factories in the country and was using therein “the old belt style tables, belt-driven Singer tables,” such as had been in use for twenty to twenty-five years.
We now turn to the proofs showing the attempts to improve the art. They show that Frankel started in 1912 a ten years expensive, futile and unsuccessful effort to do so. His superintendent, Axilrod, told him of the Jones direct drive, an English direct drive, tie placed six of them in his factory and in a week they collapsed. He then directed Axilrod to go ahead, irrespective of cost, and devise a direct drive device. This Axilrod did, with the result he obtained patent No. 1,188,385 on June 27, 1916. As showing the desire of the art to improve and its failure to do so, the proof is: “Frankel formed a company around 1916 and called it the Safety Machine Company, and at that time it was heralded all over the country as a progressive step, particularly by' the insurance companies, and particularly the liability insurance companies — they were very much interested in the project because they felt it would reduce a lot of accidents — and in 1917, in January, Frankel received a certificate of approval from the State of Pennsylvania, which was the first certificate of approval for safeguards for sewing machine devices. Of these Axilrod tables, probably thirty or forty thousand were sold principally in the East, but they did not work satisfactorily.”
The General Electric Company took up the problem of the lubrication of the Axilrod device, but “they gave the job up as a bad job, after spending considerable money.” Referring to the Axilrod device, the testimony of Frankel was:
“Q. Then the tables that preceded yours, as asked before, the Axilrod table and the Corrall table, as far as you know, from the testimony and your own knowledge of the Axilrod table and the Corrall table, did not stand up and do its work for an appreciable length of time, is that (right? A. To my knowledge, it did not.
“Q. Just to get some idea of the thing, when you had completed your machine, during the time you were completing it, you had dealt with a great many Axilrod tables ? A. Yes.
“Q. Can you give any idea how many of them were returned and that you had to dispose of? A. I do not know actually in numbers, I know we took two hundred out here and two hundred out there, but if that gives you any idea, I did scrap three railroad cars full of clutches at one time.
“Q. That gives me an idea of quite a number of them. So that actually these three carloads were dispensed with by you ? A. Yes, by me personally.”
Frankel in 1919 hired one Voigt and set him to work and the evidence as to what he did and the commercial success of his efforts is thus shown in the proofs: “Frankel hired him in 1919 and gave him carte blanche orders — money no object — to perfect this transmitter, which Voigt did in [546]*5461921 or 1922. Plaintiffs spent upwards of three hundred thousand dollars before they ever saw a cent back. Axilrod tables were replaced by these new Voigt tables. Upwards of one hundred and fifty thousand Voigt transmitters were sold to the present date. A partial list of Frankel’s customers shows upwards of a thousand names of the leading manufacturers in all industries using sewing machines. Each transmitter costs about fifty dollars with the motors and accessories.”
In pursuance thereof, patent No. 1,431,-198, issued to Louis Frankel, assignee of Voigt, was granted for a transmission apparatus.
As we have seen, the patent to Corrall was classified as a “Sewing machine driving device,” while that of Voigt was called a “Transmission apparatus.” But in the Voigt patent the Patent Office did not cite it against Voigt. Without at this point describing the means Voigt used in combination, we inquire what progress and improvement in the art did Voigt make, as shown by the proofs. In that regard he testified, and there is no proof to the contrary, as follows:
“Q. * * * At about what rate does the fabric travel through an ordinary machine of the kind with which this patented device is intended for use? About how many yards a minute does it sew? A. Well, I would say that the average speed of a sewing machine is about three thousand stitches per minute.
“Q. In yards how much is ,that per minute? A. I would say at the rate of eighteen stitches per inch. This would amount to travel of approximately six yards per minute, assuming that the sewing machine is in constant operation for a minute.
“Q. Plow much would that be in a second ? A. Per second would be approximately three or four inches per second.
“Q. Now, in practice, with the sewing machines of the kind that this device is used with, some of the pieces that the operators are called upon to sew are quite short? A. They are rather short, depending on the type of garment.
“Q. For instance, my collar would be from the point back? A. It would begin on the short side. It would be turned on the point — the long side and then the short side again.
“Q. Then, as you say, for example with a collar, when you have sewed to the point you desire, and you want to send the machine in another direction, would you stop the machine there? A. Yes, you must stop quickly.
“Q. In the operation of the machine it is required, I understand, that it run for a ways and stop either at the end of the machine or at the point of turning? A. That is correct.
“Q. Suppose the machine does not stop accurately, what happens? A. It overruns the seam. It might spoil the work. It might have to be reopened again and done over. It would certainly impair the efficiency.
“Q. If the fabric runs out from under the needle, the needle will begin to make stitches ? A. That is correct.
“Q. That, I believe, has some disadvantage. Can you say that it does? A. Well, in some cases it might break the thread, and it might have had bad effects on the sewing machine itself.
“Q. I understand you, from what you said here, that sewing machines, although on the same table and driven from the same power, are desired to make different kinds of stitches or run at different speeds; is that correct ? A. That is correct, depending on the type of sewing machine being, used. The table itself might have from ten to twenty machines, less or more, and these machines must be operated from one shaft.
“Q. At different speeds ? A. At different speeds.
“Q. So, as I understand your testimony, there are two considerations of considerable importance — one of them that the machine must stop at the will of the operator— A. That is the most important, it must stop at the will of the operator.
“Q. The other is that your device, whatever it is, should be capable of adjustment or use in connection with machines going at different speeds? A. That is most important.”
Referring to the elevation of the main and constantly revolving shaft from near the floor to near the bench and what was new and important in Voigt’s plan, the testimony is:
“The Witness: The shaft is elevated toward the bench top. The old way of driving was a shaft close to the floor, and then from this shaft was a pulley toward the transmission. From the belt was a pulley toward the sewing machine.
[547]*547“By Mr. Stoughton:
“Q.'What is generally the purpose of having a drive shaft that drives all the machines located up underneath and o toward the top of it? A. There are innumerable advantages — cleanliness, sanitary — it is more sanitary because you have a clean floor. Then it eliminates accidents. To my knowledge there are quite a few accidents with the shafting on the floor and the belting reaching up to the transmitter. It was not properly guarded. Operators losing a pocketbook and reaching under, their garments were torn off and things like that. In fact, there are State laws demanding guarding of moving powers underneath the table.”
In Voigt’s device the line shaft, located near up to the bench, obviates danger, as stated in the patent and testified by Voigt: “The line shaft is preferably positioned in proximity to the under side of the bench whereby it will not contact with the person or the clothing of the operator when the latter is seated at the bench.” We here note the suggestive and somewhat unusual situation that while defendant’s answer averred five American and five British prior patents, and while the patents of Axilrod and Corrall had been issued by the American patent office, that no one of these patents, or indeed any others, were cited against Voigt by the office — a rather unusual situation — so that we have not only the prima facies created by the issue of the patent, but the fact that the office regarded his patent as a disclosure in a new field.
Turning to Voigt’s device, we are aided by the multicolored sketch prepared by him. In connection with it we note the contention of the plaintiff as follows:
“It is apparent from the preceding description of the Voigt transmitter that each sewing machine is independently driven from the-same power shaft by a separate belt and pulley. Accordingly, there is required a transmitter by which the pulley can be operated and stopped from rotating at the will of the individual operator of any machine independently of the operation of the other machines. As clearly brought out in the testimony of the inventor, Max T. Voigt (1 R. pp. 28, 29, 42) it is necessary that the rotation of the pulley be stopped instantly at the will of the operator in order to prevent over-running of the stitches and to prevent injury to the sewing machine itself. With a proper appreciation thus had of the undisputed necessity for instantaneously stopping the operation of the machine, the importancé of the Voigt structure, and particularly the importance of the stationary sleeve 22 and of the spring means for automatically disengaging the clutch members and automatically engaging the brake members 34 and 21, can be more readily understood. First, with respect to the stationary sleeve 22, it will be observed that said sleeve provides for the rotation of the body which carries the pulley 29 out of contact with the continuously rotating main shaft, that is, the shaft 3. As the result of this mounting of the body on sleeve 22, there is no frictional engagement between said body and power shaft and consequently when the clutch member carried by the body is disengaged from the clutch member fixed to the power shaft, there is no tendency of the power shaft to rotate the shiftable clutch member and the pulley rotated thereby. On this account, in order to stop the rotation of the pulley after disengagement of its clutch member, it is only necessary to overcome the momentum of the body and the parts rotated thereby and this, as • the testimony shows, is readily accomplished by engaging the brake members 34 and 21 automatically by the spring means which disengages the clutch members from each other. Sleeve 22 has additional value in that by keeping the rotatable body out of engagement with the power shaft and thereby preventing said body from wearing the shaft, it considerably simplifies the lubrication problem (1 R. p. 67).”
After patient study of the proofs, we agree with this statement. That his device was novel in result is another proven fact, namely, that if one sewing machine got out of order, it could be repaired without stopping the main shaft or suspending sewing on other machines. We venture to say that in our opinion the most effective subagency in his device effecting what it did lies in the nonrotating tubular sleeve or bushing 22' on which the pulley carrying body is slidably and rotatively mounted out of contact with the shaft and the provision for sliding such body from clutch engagement to brake engagement and stop position by an axially applied, oppositely disposed means. Assuming all mechanical agencies were in themselves old, Voigt’s combination was new. Indeed, the trial judge found: “The defendant admits that no one of the prior art structures combines all of the elements which the patentee has selected and put together.” In the light of its accomplishment, his combination seems comparatively [548]*548simple, but in the light of going to England, in the laborious work of his achievement it took an experienced engineer such as Voigt several years of experimental work and the expenditure of some three hundred thousand dollars by Frankel, who believed an improvement was possible. Basically, in our view, the embodiment of the sleeve 22, in combination with the other elements of Voigt’s device, was the key to the problem. Excerpts from his testimony tend to that effect. In reply to the pertinent questions asked by the trial Judge, Voigt’s testimony was:
“Q. Mr. Voigt, let me ask you something. You were employed to inyent a machine? A. That’s correct.
. “Q. It was not an accidental invention ? A. It was not.
“Q. You were given a particular problem to solve? A. Yes.
“Q. You were told what was wanted? A. Yes.
“Q. I do not suppose you went at it by just taking a lot of these parts and shuffling them around and putting them together in different ways? A. No, I did not.
“Q. I suppose you had some definite obj ect in view; and particularly, you figured out that something was wrong with the machines that existed ? Is that the way you went at it? A. That is correct.
“Q. Did you take the machines that were in existence, the patents that were in existence, and study them ? A. I only took the machines in existence — -mainly the Axilrod table that was talked about.
“Q. You started with the Axilrod table ? A. Yes.
“Q. That was not a successful machine ? A. It was not.
“Q. What did you find wrong with it? A. The mechanical structure was wrong.
“Q. That does not mean anything to me. A. Having the running part mounted on a shaft, it just simply would cut the shaft away.
“Q. The running part— A. The driven clutch member mounted on a shaft.
“Q. It would score the shaft? A. It would not only score it, but it would cut it away.
“Q. To remedy that situation you introduced this member 22? A. That’s correct.
“Q. Then, when you were all through, had you made the invention when you did that? A. Yes, I think so.”
In answer to questions by defendant, Voigt brought out these facts:
“By Mr. Scull:
“Q. In fact, that was one of the difficulties of that sort of construction, that if the lubrication failed inside of that loose pulley it would cut the shaft and be driven inside of the transmission tnechanism? A. The actual results used to be that the shaft used to cut down sometimes to three-eighths of an inch.
“Q. What other difficulties were there in it? A. That was the main difficulty that I know of; it failed for that reason.
“Q. It failed for that reason, and you remedied that by putting in the tube, or bushing, as you call it ? A. Stationary sleeve 22.
“Q. As an extension bearing for the shaft? A. That is correct.
“Q. And let the loose pulley run on that? A. Yes.
“Q. Had you ever seen that done on any transmission before that time? A. I don’t remember — I mean, I have seen constructions, naturally, on other things, not on a sewing machine drive, nor have I never seen it before on sewing machine drives before that time.”
In answer to the plaintiff’s question, Voigt testified:
“By Mr. Stoughton:
“Q. The claims of this patent in suit specify generally,- along with the usual parts of a transmitter, a sleeve fastened to the shaft? A. Yes.
“Q. On that sleeve runs the movable part or the part shown in yellow on your drawing? A. That is correct.
“Q. The clutch on the shaft serves to drive that yellow part? A. That is correct.
“Q. And there is a brake into the range of which this body is moved? A. Yes.
“Q. The thing is started by a treadle— that is the means to throw the drive clutch in. You have a means consisting of diametrically oppositely disposed pins which are axially movable for the purpose of throwing that clutch out and putting on the brake. Now, having done all that, what useful result do we get from that? A. We get a clutch axially operated.
[549]*549"Q. It operates to do what? A. Operates to give us a quick start and stop.
“Q. And it gives you a quick stop, more particularly? I suppose it does not make very much difference about the starting? A. It does not make very much difference about the starting, because that depends on the operator. Sometimes a machine will what we call ride the clutch, to get a slow stop, depending on the operation.
“Q. One result is that you get, as I understand it, almost an instantaneous stop? A. You do.
“Q. A substantially instantaneous stop? A. Yes.
“Q. What other result do you get? A. The operator can control it at will. * * *
“Q. How does it work? Does it continue or stop in a week or two? A. No; I made a permanent device that has been operating ever since, has been sold in large quantities, and is very desirable in the trade.”
Seeing then the device was novel, we next inquire whether it was useful. Of course, if the defendant infringes, it cannot well contend it is not useful, but apart from this, the proofs show its marked utility. In that respect the proof is, as already quoted, that upwards of 150,000 of Voigt’s device were sold by September, 1936, netting $7,500,000, and that a thousand leading manufacturers using sewing machines had bought it.
Its high usefulness is clear and the primary question arises: did it involve invention? In view of the long practice of the defendant in clinging to the old art, bearing in mind the protracted study and investigation of Voigt before he solved it and the large expenditures made in research, and the grant of the patent, which constitutes prima facie proof of invention, we are strongly of opinion the device involved invention, and we so hold.
As applied to the idea that Voigt was a pioneer improver of indirect power transmission, the term “pioneer” would be somewhat inappropriate. He was not an improver of indirect transmission, but so far as the sewing machine art is concerned, he was really the first creator of a practical 'and satisfactory direct transmitter in that for the first time he gave the sewing machine art a device so perfect in action, so simple in construction, so durable in use, so effective in instantaneous or “split second” start and stop, that nothing was left for other designers to improve upon save to make it ball bearing. He was not a pioneer in the old field of indirect transmission in the sewing machine art, but was the creator in efficiency of direct transmission. In that particular he is a creator of something new, rather than an improver of something old. His success in that field is evidenced by the fact that the defendant, with all its means, engineers, and experience, produced nothing new until in 1930 it entered the direct field in which the plaintiff was having universal success. Did it infringe in so doing ? This question the court below did not pass upon. In view of the length of this opinion, we refrain from discussing in detail the plaintiff’s and defendant’s structures. It suffices to say that we have been deeply impressed by the testimony of Voigt. His ability, his fairness, his frankness, and his ultimate success, all combined, lead us to place reliance on his view of the purely mechanical question on which infringement depends and we concur therein. After discussing the two devices, he testified:
“Q. Now, tell us whether in your opinion the Singer diagram embodies substantially the construction described in the patent in suit and shown in the drawings and referred to in the claims. A. In my opinion, it does embody all thé claims made in the patent.
“Q. There are differences that you have pointed out. For example, this sleeve 7 which moves the yellow body back and forth, and the fact that the brake is on the lever instead of on the fixed side of the bracket — of course, the lever is attached to the bracket. Those things, as I understand, in your opinion, are matters of form? A. They are only matters of form. In all important details it is substantially the same as our device”
—a conclusion on which, be it observed, the defendant did not question him on cross-examination. Indeed, had the defendant’s device preceded Voigt’s, it would have so anticipated him as to defeat his patent, and what would have anticipated if earlier, is an infringement if later. Without further lengthening of this opinion, we limit ourselves to holding the patent valid and infringed.
In coming to the general conclusions we have, we again emphasize the views of this court on the subject of alleged anticipations as forcibly and admirably stated by [550]*550Judge Woolley in Skelly Oil Co. v. Universal Oil Products Co., 3 Cir., 31 F.2d 427, 431: “A patent relied upon as an anticipation must itself speak. Its specification must give in substance the same knowledge and the same directions as the specification of the patent in suit. Otto v. Linford, 46 L.T.(N.S.) 35, 44. It is not enough to prove that a method or apparatus described in an earlier specification can be made to produce this or that result. Flour Oxidizing Col v. Carr & Co., 35 R.P.C. 457. A singularly sensible test of the rule of anticipation is given in British Thomson-Houston Co. v. Metropolitan Vickers Electrical Co., 45 R.P.C. 22, by asking the question — ‘Would a man who was grappling with the problem solved by the patent attacked, and having no knowledge of that patent, if he had had the alleged anticipation in his hand, have said: “That gives me what I wish ?” ’ The Pope Alliance Corporation v. The Spanish River Pulp & Paper Mills, Ltd. (Privy Council Appeals No. 33 of 1928,” reiterated in Worthington Mower Co. v. Gustin, 3 Cir., 80 F.2d 594. Here, as there, we can say that had Voigt, grappling with the problem before him and having, as he had, no knowledge of his ultimate device, but having Corrall’s and Axilrod’s patents before him, would he have said, “They give me what I wish”? The negative answer to that question is furnished by the defendant’s eighteen years of inaction, when with ownership of Corrall and knowledge of Axilrod, the defendant continued its use and indorsement of the indirect system it continued to follow. As always, factory practice speaks louder than nonfactory experts.
Reversed with directions.