The here litigated patents particularly concern flowing oil wells in Oklahoma City, Old. That field is exceptional in character. Underlying it are several oil and gas zones, one of which, the Wheaton, which was developed in 1930, presents the unusual problems to which these patents are particularly addressed. The wells were drilled to and through the Wheaton sand, were some six thousand feet in depth, and a tremendous gas pressure was tapped. When struck, the gas, sometimes in volume of a hundred million cubic feet daily, vented itself through the well tubing and casing at the velocity of from three to five hundred feet per second. The vent was seven inches in diameter, and through it the gas carried with it *340in an hour as much as 4,000 barrels of oil and 1,000 barrels of sand. The proof is that “the particles of sand were absolutely round and frosted grains of pure silica sand and were almost microscopic in size, that is they were somewhere around twenty to thirty thousandths, but they were uniform in size.” When such wells were struck and these great elemental forces 'came into play and when the wells were located in populated places, the control of such “wild wells,” as they were called, was necessary because, with their inflammable oil and gas venting themselves in torrential volume, they were exceedingly dangerous. Indeed, these factors created a situation that called for quick control. A former assistant oil inspector of Oklahoma City testified to having inspected more than 400 wells: “One of them was directly across the street from a large school, also we had one on the four hundred block East Main Street, Oklahoma City, that blew the expansion chamber off and cut the gates off. This well was wild; we had no control over it.”
Describing the field generally, this witness testified that the trouble came from the sand cutting the metal controlling agencies, a cutting which can be readily appreciated when we consider the tremendous abrasive power of a sand blast, where in an hour a thousand barrels of fine silica sand passed through a seven-inch orifice impelled by a volume of gas traveling at from three to five hundred feet a second. Some wells gave, no trouble, the inspector testifying: “We had quite a number of wells that gave no trouble at all.. Part of the field there was in the lime but part in the' sand area; in other words, the east side of the field had very little sand. Q. So it wasn’t all the wells in Oklahoma you had this terrible trouble with? A. We had an awful lot of trouble with 411 wells that I know of. 'There is better than a thousand wells there. As I stated above, I had charge of inspection inside the city limits of Oklahoma City, and in this territory I think there are at the present time 528 wells, and we had trouble with four hundred and some.”
The field of Oklahoma presented another difficulty which had to be met. That state allowed only a certain oil pro rata to be produced by these flowing wells. This state pro rata was ascertained by allowing a shut-in well to flow wild for certain test times, so that at these test times control was released and the well ran wild. In that regard the proof is (see record, p. 59): “It is necessary to control or choke the wells' in the Oklahoma City field, and those wells are controlled while the production — the allowable production — is being produced. However, it is necessary also to take as large a flow as possible from the wells at certain given intervals, which are called potential periods. In the Oklahoma City field we take a four-hour open flow test to determine our potential. This potential then measures the amount of oil that we are allowed over a six months period.”
Prior to the patents in suit, the best methods known to the oil industry in wild well control were the Christmas tree devices and the expansion chambers. Both these are based on allowing the escaping gas to expand; the idea being that, the more gas is expanded, the less sand cutting there will be. The Christmas tree device is shown in the accompanying draw-
ing, concerning which the testimony is (page 63): “This drawing illustrates a Christmas tree, of which there were examples similar to this in Oklahoma City field, showing the well pipe coming up through the derrick floor, and the upper master gate; and from this well pipe are several branches which have gate valves, at least two gate valves on each branch, and an adjustable flow bean at the end of each branch, and from the *341flow bean the oil passes through that into the flow line to the tanks. This multiplicity of gates and valves and crosses —also there arc two large crosses there, from which the branches come from the flow line — are necessary in certain fields where a severe condition exists, due mainly jxist as auxiliaries, because of the cutting effect of the sand. Usually just one branch — or possibly two — was used at a time, and when those would cut out the gates would be closed, if possible, and the flow diverted through another branch. Obviously this type of equipment is very expensive, costing from $4,000 to $5,000; and even with all that cost, it was found early m the field that it was almost impossible to even use this type of equip- , ment to handle the severe conditions that existed tnere.
The testimony of Knowlton, one of the patentees, in regard to the Christmas tree device-and in the absence of contradictory proof we accept it — is as follows / 54 55) • ^ J
... , . As I mentioned before, the Christmas tree was not practical, because of the fact it would cut out sometimes m ten minutes; and it was, of course, very expensive to replace these valves. The expansion heads had many disadvantages. In the nrst place, they are very dangerous and very hazardous m expanding thxs volume of gas and oil and enlarging the cross section of the head the breaking strength was very much m-creased, and if anything happened by cloggmg up or sand, or anything m the outlet, the full pressure ot the wel was imposed on that head; and several of them blew up ano. caused a lot of damage and actual loss of life. The principle of these heads was to expand the gas and the fluid so that the velocities would be decreased; but in expansion, of course, they caused a turbulence which started cutting out anything in front of it; and this was experienced in many cases, and caused the well to go wild. Also, there is no way by means of these heads to shut in the flow and keep the pressure and the velocities from the master gates below, and whenever the head cut out, or the flow line cut out between tbe head and the separator, you would have to depend entirely on the master gates to close the well; and in closing the well against this pressure, and with these velocities, the master gates very often cut out, and of course, when all three mas-ter gates cut out you had a wild well on your hands.
“These master gates cost almost a thousand dollars apiece, depending upon the size; and there was a requirement jn the field to put three of these on each Well, just for safety’s sake. Ordinarily, ¡n m0st fields maybe one master gate is sufficient, but where this extreme sand con-dition existed sometimes the three weren’t enough.
«As j inted Qut aboye; ^ christ. mag tree wag very expensive; ^ cut Qut frequently within ten or flfleen min_ utes_ It wag algQ difficuit tQ t a wide opeQ flow th b the Christmas and where ,any sand existed at all ^ christmas tree just didn>t stand up under the gand blagt»
. , , , In addltl?n the above’ tbe °f the/w?tne”“ (p,agc 69: When we entered the field m the spring of 1930. the development of the Wilcox zone was in its infancy, and the stand-ard e ¡ t that was being used at thaf. time wag the so.called expansion headSi In the beginning the operators tried the cust0 Christmas tree, that had been standard for in the oil busi and th fotmd it entirel in_ ad ate t0 controi tbe severe conditions tl had th s0 th started on tbis Mea of sion heads or expansion chambe which are enlarged casings or j tanks laced dircctiy over the well . or cagi in order tQ cut down ye_ lociti g() ^ tb could allow tbe oil to gg thr h the flow line int0 tbe s tor without damage.»
, TTT, , , When the Wheaton sands were devel°Ped ln tbe Oklahoma field m 1930, the patentee Knowlton, who had broad ex-Perience m tnany fields, was sent to Okla-homa to ^ the difficulties due to sand ?uit^nS- lhe results ,01 hls were *n direcL variance to the practice then fol-^owe^- m the Oklahoma field, m that it regarded gas expansion as a means of lessemn£ sand cuttmg, while Knowlton found &as expansion increased sand cut-tmg and therefore was to be avoided, Testifying as to the course pursued, the Pintee says (pages 66, 67).
“It usually took a day and a half to two days to make such replacements. Very early in my experience in Oklahoma City, it occurred to me that the principle of
*342this expansion head was entirely wrong. I felt that by expanding and decreasing the velocities by expansion was wrong, because it set up a turbulence in this head which caused sand cutting rather than eliminating it. In other words, operators were trying to fight back at these forces that were cutting out their equipment, rather than design equipment which would eliminate sand cutting. They were making them thick enough to withstand sand abrasion, rather than thinking of the proper type of equipment that would entirely eliminate the cutting. Therefore, in order to remedy these conditions, I felt that if I could keep the stream line idea of the well casing, which never did cut out — the casing itself, because the sand was going parallel to the' flow — if I kept that stream line condition until after the turn was made into the separators, and confine the flow rather than expand the flow, I felt that that turn into the separators could be made with the least possible cutting. Also, it occurred to me that if a pocket- at the top of the fitting was used, that the sand would pack -more or less, or at least concentrate at that point, and there would be a sand-against-sand action rather than a sand-against-metal action. And it occurred to me that if this were imposed perpendicularly to the line of flow up from the well, rather than at an angle, the sand cutting would be diminished or eliminated as far as possible.
“It also occurred to me that if we could control by means of a head the flow of oil in taking a potential, in a severe case, we could shut the well in by means of this device, and then be able to work our master gates and equalize the, pressure across the master gates, so that they could be closed without danger of having a wild well. It also occurred to me that if the expansion were made after the right-angle turn to the separator,- the velocities could be diminished before entering the separator horizontally rather than vertically. It also occurred to me that if you had a device that you could change from taking these wide open flows to a restricted flow cheaply and safely, that it would be a big advantage, because these other types of devices cost us four or five hundred dollars.”
His discovery/ the means by which sand cutting was reduced, and the desirability of his device were briefly set forth in his patent application, which application remained unchanged in its passage through the Patent Office. His application says (patent No. 1,831,713, first 33 lines):
“This invention relates to improvements in flow heads for oil wells, and more especially to a novel flow head for controlling the discharge of oil with entrained sand from wells operating under high rock pressure. Recently, in order, to combat the series sand cutting effect of the flowing streams of wells in the Oklahoma City field, many different flow heads were designed and tested, most of which depended on the use of an expansion chamber at the top of the casing to retard the velocities and receive the shock of the slugs of mud and rock when the well is flowing. Considerable difficulty was experienced from these expansion heads, cutting out when the direction of flow was diverted at a right angle into the separator. I discovered that the turbulence resulting from the expansion is the basic cause - of the cutting, and I have invented -a flow head having no expansion chamber but a stream line effect with no offsets liable to start a swirling motion. This improved flow head eliminates the disadvantages mentioned.
“This flow head will now be described with reference to the accompanying drawings, in which
“Fig. 1 is a perspective view of the 'upper portion of a well with my improved flow head attached to the upper end of the casing.
*343“Fig. 2 is a vertical sectional view of a portion of the improved equipment”
As to the devices working, he says (lines 107-118 of patent) : “I have found through actual experience that the only cutting that takes place in this device is on the needle valve head 21, and the replaceable choke 20. In two wells, each of which produced over 50,000 barrels through such chokes, very little cutting occurred. On one, the well was flowed several times through a two inch opening. The point of the needle was cut, and the replaceable seat 20 only slightly. This same well, when wide open, cut and materially damaged the cross 7 in four hours.”
The patentee in his testimony conceded that sand cutting was avoided in other deep well fields, yet the proof is that the peculiar difficulties encountered in the Oklahoma field were not overcome by prior devices. In that regard the testimony of Knowlton is (page 73) : “They had bad sand conditions in the California field, in the reservoir itself, but they controlled the sand at the bottom of the hole with slotted and screened pipe and did not allow the wells to produce large volumes of sand, the way we were compelled to do because of the proration laws in the State of Oklahoma.”
That the device went into instant and wide use is proven by the witness Lowrey, viz.: “While I was in Oklahoma City, I think I had a record of having inspected 411 wells, and I would say at least ninety per cent of them used the Knowlton flow head. While I was inspecting wells, I approved of the Knowlton flow head, and so did Mr. Ingolls of the Fire Insurance Underwriters — I believe that is the way they called the office — after we had great difficulty with fire hazards there. I am not in any way connected with the Phillips Petroleum Company, with Mr. Knowlton or with Mr. Defenbaugh.”
Of its financial features as contrasted with previous practice, the proofs as to labor savings are (page 80): “And when this Knowlton control came out it was very easy to handle, we found that it shut with the pressure, and that they could determine the amount of flow they wanted before they would take a potential, by taking our test. Then they would set that in position of about the right size choke that they figured they could handle it by and not produce too much sand, and at any time one or two men could close that choke, where a master gate would take at least sixteen to eighteen men during the high pressures.”
As to lessened cost, the proof is (page 72):
“The Knowlton flow head in Oklahoma City field complete costs less than $400, whereas the expansion heads range from $1,000 to $2,500. A Christmas tree, as before stated, that would be necessary at all in Oklahoma City would range from $2,500 to $5,000. And, of course, that is only a small percentage of the advantage of the Knowlton device, in that each time these other devices were used for taking open flow tests they had to be removed and some other device put on the well, whereas the Knowlton flow head could be used ‘continuously on the well by a very small change, which took only a short time. This change on the expansion heads often amounted to $4,-*344000 or $5,000, which was actually more— the labor alone cost more than the Knowlton flow head. * * *
“As I stated heretofore, at the time when I came into the field, all the wells were either using the expansion heads or Christmas trees, and most of these wells were replaced later by the Knowlton flow head, and after that time the Knowlton flow head became the -regular standard equipment for the wells in the field.”
So also the witness Clark testified (pages 87, 88):
“I am familiar with the Knowlton flow control head, and I know it has effected savings in money for the Phillips Petroleum Company in, I would say, at least two ways. The one which I would consider the most important would be in the saving of gas energy in the formation, which was brought about by the use of a proper control; the second would be by increased gravities, which brought higher prices per barrel for the oil produced.
“The Court: Q. I don’t understand what you mean about the gas. Does that mean you would produce more gas which would go into the line, or just what does it mean? That is what I don’t understand. A. I meant by that that you produce less gas at some certain opening than you would at any other opening, per barrel of oil. That opening was found on a number of wells which I personally observed — that the energy was conserved within the sand in the gas-oil ratio, or the .amount of gas per barrel of oil was reduced by the proper choking, this energy was saved in the sand—
“Q. The gas didn’t come up then? A. No, the gas remained ’in the formation —that is, part of the gas. In other words, the ratio was reduced enough in one year that our calculation showed that we had reserved the energy within the formation to lift, at the present gas-oil ratio, at some later date approximately a million barrels of oil.
“Q. That oil produced under those later conditions then will produce a greater quantity of gasoline and other gaseous products than it would have, had the gas come to the surface by the old method. Is that right? A. I don’t believe I understand that, your Honor, If I may explain — the gas required to lift one barrel of oil was reduced, this amount that it was reduced remained within the formation, and later was to be used in lifting additional barrels of oil.
“Q. Made available. I see; I understand you now. You keep your gas down in the oil sand for the purpose of recovering the oil that would not have been otherwise recoverable ? A. That is correct.
“Mr. Barry: The character of the oil produced by the Knowlton flow bean was improved, in that it was raised in gravity, which brought a higher price per barrel. The saving in price on that oil ran from 2 cents per barrel to in some cases as much as 6 cents per barrel.”
We then turn to the question of invention. That every single implement —valves, master and needle, nipples — were appliances well known may be conceded. But with all these in common use, with the well-known cleverness and resourcefulness which characterized the oil-producing art, no solution of the problem of controlling this Oklahoma City field was reached. In view of the novel and original course which Knowlton took in departing from the gas expansion practice then prevalent, and in view of the means he combined to form his device by using well-known elements in new relations, we are constrained to hold his device involved invention, and that of no mean order.
It is contended his device did not, in view of the prior art, involve invention, and that he was forced to restrict his claims, and that he is now Seeking to so construe his claims as to cover devices embodied in his rejected claims. We cannot accede to this contention. In our view, no prior patent disclosed either his plan of nonexpansion or the means to avoid sand cutting embodied in the claims here in issue. The patent of Watts, which was granted in 1929, was for a particular kind of nipple and was not intended to or capable of avoiding the sand cutting of the Oklahoma, Wheaton sand problem, and no such use was made of it. A study of his patent — which was granted three years before the control difficulties of the Wheaton sand developed at Oklahoma- — shows his object was not to control wild wells, but to regulate the oil of a flowing well by the use of a particular type of flow nipple, the functional purpose of which he thus describes (Record, p. 323, lines 1 to 33):
*345“Although my present invention is referred to simply as a flow nipple, it should be understood that the invention herein described is intended primarily for use in replacement of so-called ‘beans,’ as used in the outlet pipes from flowing oil wells.
“As Is well understood by workers in this art, if a gushing oil well, or well containing much gas, is allowed to flow without restriction, the gas fails to bring up a maximum total of oil, and the oil will moreover often bring up large quantities of sand. Unless the flow is artificially restricted, the resultant caving commonly results in the choking of a well after a comparatively short time, a general result being, frequently a serious damage to or a complete loss of the well — so that a new hole must be drilled, at much expense, if production is to be continued. Tn order to get, from each well and without danger of standing-up, as large an oil output as possible, so-called ‘beans’ or throttling devices have heretofore been employed ; but these are comparatively crude; and it is an object of my invention to provide simple and improved means, permitting small and precise variations in adjustment in accordance with changes in conditions, so to control the outlets from oil wells by varying an exit passage there-through, as to assure, under various or constantly changing conditions, a maximum output or recovery of oil therefrom.”
Then showing the difficulties the art encountered, Watts recites (lines 43 to 49) : “Thus, whenever the sand has begun to move unduly, a bushing having a hole (say) l/64th inch smaller than the bushing previously employed could be substituted therefor — successive bushings being tried, if the movement of sand continued, until sand practically ceased to come.”
He then describes his remedy by the use of his flow nipples as follows: “In flow nipples of my design, the restriction of flow can be varied more accurately and by more minute percentages than with a ‘bean’ of the character referred to; and the regulation can moreover be done without interfering with the operation of the well. This makes it easy to obtain that rate which will give the desired or maximum safe outlet; and it is also an important merit of my novel flow nipple that those parts which are liable to wear by attrition are not unduly expensive. — said parts being, moreover, replaceable without disconnection of the main cruciform body containing the same from the pipe line in which the same may be interposed.”
From this it is clear that Watts’ patent of 1929 in no way anticipated what Knowlton did to control the Oklahoma development of 1930.
Patent No. 1,698,826, to Shaffer, also granted in 1929, was for an adjustable flow nipple suitable for use on a Christmas tree which, as we have seen, was a gas expansion device. It bore no resemblance to the Knowlton, either in form, substance, or function. Shaffer’s flow nipple is located at the end of a branch of the Christmas tree and not over the well outlet as in Knowlton’s, for Shaffer’s patent, after describing all of the Christmas tree structure over the well outlet, says: “All of which forms no part of the present invention.” Shaffer’s device is located at a point after expansion takes place, while Knowlton’s acts before expansion. Knowlton’s device is located at the angle point where the oil and gas flow changes from a vertical to a horizontal flow, and at that turning place his specification states that “sand may pack into the chamber 30, so that I obtain a sand against sand action.” If the Shaffer patent was later than Knowlton, the latter would not have anticipated Shaffer and would not be infringed by Shaffer. Indeed, the two devices are wholly different and were addressed to wholly different problems. There being nothing in functional aim either in Watts or Shaffer, Knowlton was free to secure claims for his device, and the Patent Office was justified in granting, as it did, Knowlton’s claims for his device as not embodying either Watts’ or Shaffer’s. So holding, we find Knowlton’s patent, No. 1,831,713, valid. We make the same finding as to Knowlton’s and Defenbaugh’s later patent, No. 1,911,905, for a flow head with an adjustable flow bean. As stated in that patent, it is an alleged improvement of the Knowlton patent we have been considering. In the earlier patent Knowlton had shown a flow bean which could not be reduced or so receded as to allow wholly uninterrupted flow of the oil and gas stream, while in the Knowlton-Defenbaugh device the flow bean can be totally withdrawn from interference with the flow stream. On the *346application for the latter patent, it was rejected chiefly on the Knowlton patent, but, when the attention of the office was drawn to the difference in the applied for device, viz.: “Applicants appear to be the first to mount a flow head with' adjustable flow bean at the top of a vertical well pipe at the junction between said well pipe and the flow line, and to use a metering screw for adjusting a valve head which may be withdrawn when necessary to a position at one side of the vertical passageway of the fitting, to permit working through the flow head without detaching the same from the well pipe and flow line,” the office acquiesced by the issuance of the patent.
The advantages of the device are set forth in the application, and the record contains no contradiction there: “I have found through actual experience that the only cutting that takes place in this device is on the needle valve head 21, the replaceable choke 20. In two wells, each of which produced over 50,000 barrels through such chokes, very little cutting occurred. On one, the well was flowed several times through a two inch opening. The point of the needle was cut, and the replaceable seat 20 only slightly. This same well, when wide open, cut and materially damaged the cross 7 in four hours.”
After the foregoing study and due consideration having been given to the several other questions raised in the case, it remains to briefly refer to the question of infringement. The defendant is a manufacturer of devices for effecting the same functional objects as do the patents in suit. While it does not itself install its devices on wells, it makes and sells them with the sole purpose that they will be so used. Clearly, their purpose is that they shall be used in the way and for the purpose jLhe plaintiff’s device would be used, and, under the authorities, if their device infringes, they are contributory infringers, for the proofs show they have continuously been selling their devices to the oil well trade since October, 1931. Therefore, finding, as we do, that the defendant’s device substantially embodies the elements shown in the claims of the patents in suit, the decrees below must be vacated and the record remanded, with directions to adjudge the patents valid and infringed with consequent accounting.