WIDENER, Circuit Judge:
Appellant Prudential-Grace Lines, Inc., presents the court with three issues:
I. Whether the district court, 346 F. Supp. 1116, erred in determining that plaintiff-appellee’s patent is valid as against the defenses of (1) obviousness and (2) failure to recite a necessary structure.
II. Whether the district court erred in determining that defendant-appellant’s ship cargo gear infringes appellee’s patent.
III. Whether the district court erred in determining that plaintiff-appellee is not barred from recovery on the theory that the patent was misused.
This patent case involves the field of ship loading and unloading apparatus, commonly known and referred to as cargo gear. Plaintiff-appellee Blohm & Voss AG (hereinafter Blohm & Voss) is a West German shipbuilder. On February 15, 1966, Blohm & Voss acquired the assets, including patent rights, of H. C. Stulcken Sohn, the original assignee of the patent in suit. The patent in suit, United States Letters Patent No. 3,236,390, was issued to H. F. C. Sprengel, a German inventor. Sprengel '390 was granted February 22, 1966, based upon patent application No. 403,441 filed by Sprengel ■/on October 2, 1964. Such patent application was a continuation-in-part of Sprengel’s then co-pending United States patent application No. 361,422 filed April 24, 1964. No. 361,422 was abandoned when Sprengel ’390 was filed.
[233]*233Sprengel ’390 discloses apparatus which is mounted on the deck of a ship to facilitate loading and unloading of cargo.1 While the gear has many components, it is basically assembled around three main structures, a boom and two kingposts. The kingposts are massive vertical structures affixed to the sides of the ship's deck, one port and one starboard directly across from one another. Located in between the kingposts is a long pole of lesser diameter called the boom. The boom is attached at its bottom end (foot) to a universal joint2 on the ship’s deck so that it may be pointed not only fore and aft but also outboard to port or starboard, and led up and down. Movement of the boom is controlled by cables which lead from winches to the upper part of the king-posts and thence to the upper part [234]*234(head) of the boom. These are topping lifts. The top of the boom in each of the three embodiments disclosed by Sprengel ’390 is constructed so that a cable or cables suspend therefrom. This cable and associated blocks to multiply power is the purchase (purchase gear). Affixed to the lower end of the purchase is a hook to which the cargo to be lifted is attached.3
Sprengel ’390 is devised so that the single boom may tilt forward to accommodate loading and unloading of the forward hatches, and also tilt aft to accommodate the after hatches. At all times, including while the boom is in motion, it is affixed at its foot to the universal joint. Activating the winches which control the cables leading from the kingposts to the top of the boom controls the movement of the boom. To illustrate, assume the boom is dead center and perfectly upright. Activating the winch on the starboard side to take in cable and the winch on the port side to pay out cable would cause the boom to lean toward the starboard kingpost. If both winches were activated at the same time to pay out cable, the boom would swing either forward or aft depending upon which way (forward or aft) it is shoved off dead center. Once it is off dead center, its own weight causes it to gravitate toward the deck and the cables prevent the boom from crashing down onto the deck like a felled tree. By paying out the proper amount of cable, the boom may tilt forward or aft any desired number of degrees. To raise the boom, both winches take in cable.
Once the cargo is attached to the hook suspended from the purchase, and raised from the hold, the boom will be swung either to port or starboard to unload the cargo. Assuming the unloading is to port, the portside winch would take in cable and the starboard winch would pay out cable which would cause the boom to point to port, thus swinging the suspended cargo over the ship’s side and above the unloading pier or barge.4
The cables previously referred to which suspend vertically from the top of the boom are rigged through blocks. A [235]*235block is a more sophisticated and proficient version of what is commonly known as a pulley. Sprengel ’390 discloses the use of blocks which are suspended from the top of the boom. One block, called the upper purchase block, is suspended from and attached to a fitting which fitting is affixed to one side of the top of the boom. The cables which suspend from this block lead down to the other block, which is called the lower purchase block. The cargo is attached to a hook which is affixed to the bottom of the lower block. The fitting from which the upper purchase block suspends is attached to a cross pin which runs at right angles to, the longitudinal axis of the boom and through the boom near its top in a plane parallel to the deck.5 It protrudes slightly out from the boom. The fitting rests on the protruding part of the pin alongside the boom so that the purchase blocks which suspend from the fitting likewise suspend down alongside the boom. Simply stated, configuration is not unlike the human body with the arms considered the purchase gear, the shoulder joint the fitting, and the body the boom.
Sprengel ’390 discloses three embodiments, two of which employ the above described gear. In one of the embodiments, the purchase gear suspends from only one side of the boom — like a man with one arm, if you will. In another embodiment, purchase gear suspends from both sides of the boom and accommodates heavier loads. In both embodiments, the purchase gear suspends like a pendulum from the fitting attached to the cross pin which extends through the top of the boom. In both embodiments, the fitting, from which the purchase gear suspends pivots on the cross pin to allow the purchase gear to swing alongside the boom as the boom moves through its vertical plane fore and aft.
The alleged inventiveness of Sprengel ’390 rests in the method by which it serves both forward and after hatches without rerigging.6 Assume that the gear is set up to serve the forward hatch, in which case the boom would be tilted forward and the purchase gear suspended from the top of the boom down to the hatch. To service the after hatch, the boom would be raised until it is upright between the kingposts (dead center) and then lowered toward the after hatch. During this maneuver, the purchase gear (in the single gear and the double gear embodiment) which suspends from the side or sides of the top of the boom would simply swing alongside the boom and when the boom is tilted so that its uppermost part is over the after hatch, the purchase gear would be lowered down to the hatch. In the double gear embodiment, a fitting attached to both lower purchase blocks is uncoupled on one side to allow the purchases to pass on opposite sides of the boom.
As previously mentioned, cables which run from the top of the kingposts to the top of the boom control movement of the boom. Such gear is collectively called span gear, by Sprengel, but is nothing more nor less than topping lifts. The two span gear winches are located on the deck of the ship, one near each king-post. The cables lead from the winches up through the kingposts and out the top and then over to the top of the boom. The span cable from each king-post is attached to an end of the cross pin which extends through the head of the boom. There are pulleys attached to the ends of the cross pin called span blocks through which these cables run. The cables enter the top of the kingposts on circular runners called guide sheaves.7 The taking in and letting out [236]*236of these cables raises and lowers the boom as well as slewing it. The cables, which together with the upper and lower purchase blocks comprise the purchase gear, likewise have their origins at winches on the ship’s deck, one near each kingpost. The hauling part of the purchase tackle is that part which is led from a purchase block to the winch for hauling it in and out to raise and lower the lower block. The cable is led by guide sheaves from the winch into and up through the top of the kingpost and then down through the pendulum fitting and around the upper and lower purchase blocks.
Sprengel disclosed his concept in a third embodiment which is for all practical purposes the same as the two previously discussed except for the configuration at the head of the boom. The boom in the third 8 embodiment resembles a giant needle with an oversized eye. The eye is formed however with straight lines so that it resembles a simple fork or “Y” which almost closes at the top instead of remaining totally open like a “Y”. A pin (again perpendicular to the longitudinal axis of the boom and in a plane parallel to the deck) extends through the top of the fork connecting each side of the top of the “Y”. A fitting rests on this pin, inside the fork of the “Y,” just like the fitting which rests on the cross pin at the head of the boom in the two other embodiments. This fitting rotates freely around the pin. Suspended from this fitting is the purchase gear, an upper purchase block and a lower purchase block. To illustrate the working of the third embodiment of Sprengel ’390, assume that the boom is tilted forward so that the top is over the forward hatch. The purchase gear would be suspended from the top of the boom down to the hatch. To service the after hatch, the boom would swing through the kingposts as in the other embodiments. The method of moving the purchase gear is, however, different. The upper and lower purchase blocks are drawn up closely together and the hauling part is taken in so that the entire suspended length of the purchase gear does not exceed the length of the opening of the “Y” at the top of the boom. The fitting on the cross pin rotates around the axis of the pin to allow the purchase gear to swing through the opening in the top of the boom as the boom swings through the kingposts. After the boom clears dead center, the purchase gear may be lowered; and once the top of the boom is over the hatch, it is lowered down to the hatch. If the crew attempted to swing the boom to serve the opposite hatch without drawing up the purchase gear, such gear would strike the boom as the boom swung through. By drawing up the purchase gear, such gear swings through the opening in the top of the boom.
Blohm and Voss sued the appellant Prudential-Grace Lines (hereinafter Grace), claiming that the cargo gear on Grace ships infringed Sprengel ’390. Grace denied infringement and claimed that the patent was invalid for failure to recite a necessary structure and invalid for obviousness. Grace also claimed that certain license agreements entered into by Blohm & Voss rendered the patent unenforceable on grounds of misuse. Because we conclude that Sprengel ’390 is invalid because of obviousness, we do not reach the other issues.
The conditions of patentability are set forth in the Patent Act in three sections, 35 U.S.C. §§ 101, 102 and 103.9 [237]*237Sections 101 and 102 express the new and useful tests.10 Section 103 expresses the non-obvious test with which we are most concerned here. In Graham v. John Deere Co., 383 U.S. 1, 17, 86 S.Ct. 684, 694, 15 L.Ed.2d 545 (1965), the Supreme Court carefully analyzed the content and history of the standards for patentability and enunciated the following standards for decisions in patent cases which must guide us here:
“While the ultimate question of patent validity is one of law, A & P Tea Co. v. Supermarket Corp. [340 U.S. 147, 71 S.Ct. 127, 95 L.Ed. 162 (1950)], at 155, the § 103 condition, which is but one of three conditions, each of which must be satisfied, lends itself to several basic factual inquiries. Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy.”
It is clear from Graham and United States v. Adams, 383 U.S. 39, 48, 86 S. Ct. 708, 15 L.Ed.2d 572 (1965), decided the same day as Graham, that not only must the invention meet the tests of novelty and utility, it must also be non-obvious in accordance with the standards there set forth. In Graham, the court admonished that the non-obvious criteria should be honored by “strict observance” (p. 18, 86 S.Ct. 684). Such strict observance was reiterated in Anderson’s-Black Rock v. Pavement Co., 396 U.S. 57, 62, 90 S.Ct. 305, 24 L.Ed.2d 258 (1969). Not only must the criteria be strictly observed, we are told by the Supreme Court that “Courts should scrutinize combination patent claims with a care proportioned to the difficulty and improbability of finding invention in an assembly of old elements.” A & P Tea Co. v. Supermarket Corp., 340 U.S. 147, 152, 71 S.Ct. 127, 130, 95 L.Ed. 162 (1950). Of like effect is this [238]*238court’s statement in Heyl & Patterson, Inc. v. McDowell Co., 317 F.2d 719, at 722 (4th Cir. 1963): “It is usually difficult to find true inventiveness in a combination patent.”
With these basic considerations in mind, we turn to the patent in suit.
Scope and content of the Prior Art
Obviously, the field of ship’s cargo loading and unloading gear is relevant to the inquiry, and it would seem only reasonable to consider the area of ships’ gear generally and how the mechanical principles of force have been utilized in such gear. Cf. Mott Corp. v. Sunflower Industries, Inc., 314 F.2d 872 (10th Cir. 1963); A. J. Deer Co. v. U. S. Slicing Machine Co., 21 F.2d 812 (7th Cir. 1927).
The basic idea of booms, blocks, lines and winches is, of course, quite old. A perusal of manuals of seamanship and texts crusty with age indicate that principles of hoisting utilizing the mechanical advantage of lines and pulleys have been employed on ships at least as early as the 1700’s. See Steel’s Elements of Mast Making, Sailing and Rigging, from the 1794 edition, W & G Foyle, Ltd. London. A leading text, Knight’s Seamanship, which was cited to the district court and this court, also reveals that use, utility, and the mechanical advantage of upright supports, blocks, and lines is quite old. A drawing in a British Standards Institution publication (BS 302: 1938, figure 23) depicts a typical ship’s derrick rig. Such rig discloses a boom mounted on a universal joint which tilts along a vertical plane and may be slewed outward. Upper and lower purchase blocks suspended from the top of the boom are disclosed. Topping lifts from the mast to the boom are disclosed, as are hauling blocks to take up the purchase cable. Bearing in mind that gear to serve one hatch was well developed at least by 1930,11 we turn to the more pertinent area of two-hatch gear and the concept of swinging the boom and purchase gear through to serve fore and aft hatches.
1. William B. Ashe, No. 1,257,664 (1918).
Ashe ’664 relates to an apparatus known as a boat davit, the purpose of which is to facilitate the employment of small boats carried on shipboard. When the boat is not in use, it is suspended by ropes between two vertical supports called davits. As a practical matter, boats are nearly always cradled on board ship when not in use. Among other advantages, this keeps the strain off the purchase blocks and tackle. The davits are located at or near the side of the ship and the boat suspends between them. They are secured at the base by a pivot joint allowing movement in a vertical plane perpendicular to the center line of the ship. Each davit has a pin which protrudes near its top. A fitting is attached to this pin, freely rotating around the pin, so that the fitting may swing back and forth like a man’s arm if the vertical support be considered the body. A block is suspended from this fitting and the lines from this upper block lead down to a lower block. The ends of the lifeboat are attached to these lower blocks so that it is suspended from the davits. When the boat is not in use, the davits lean inboard so that the lifeboat is suspended over the deck. To lower the boat, the davits are pointed out over the water. As the davits swing outboard, the lifeboat swings in between them and out over the water. [239]*239During this maneuver, the upper purchase block (and lower, of course) swings alongside the davit like a pendulum. Not only do the blocks swing inboard and outboard, the upper pendulum block fitting which is attached to the side of the top of the davit also swings inboard and outboard. The district court discussed the Ashe patent but gave little or no weight to it because, among other things, it felt that davits, while they may loosely resemble booms, are primarily for lowering, and a boom must lower, raise and transfer. To the extent that such analysis was based on the assumption that davits do not raise, lower and transfer, it was clearly erroneous. Davits raise, lower, and transfer the boat both inboard and outboard. The district court took its definition of davit, which mentions only the lowering of life craft, from the Patent Office classification definitions. Never mentioned was the obvious fact that a ship’s boats are taken on board each time they are launched at sea. Webster’s New International Dictionary, 2nd Ed., defines davit as a form of crane, of metal or wood, fixed or movable, projecting over the side of a vessel or hatchway, for hoisting boats, anchors, accommodation ladders, cargo, etc. Knight’s Modern Seamanship, 10th Ed., gives a definition of davit as a curved metal spar fitting in a socket on deck and projecting over the side or stern for hanging a boat or handling weights. Ch. VII of Knight’s also illustrates various davits, some of which are not inset in sockets, as Ashe is not, and includes subheads on lowering and hoisting lifeboats at sea. Obviously, Ashe ’664 is not precisely the same as Sprengel ’390 but Ashe clearly discloses a load bearing support (the davit) which swings through a vertical plane employing a pendulum fitting (from which the purchase is suspended) which facilitates a pendulum-like movement of the hoisting gear alongside the support. The opinion of the district court gave insufficient credence to what it did teach and too much credence to the differences between Ashe ’664 and Sprengel ’390. Because Ashe did not anticipate Sprengel ’390, the district court seemed to reason that the teachings of Ashe were virtually worthless because of their disclosure in the context of a davit rather than a cargo boom. Ashe was clearly relevant for the mechanical teachings it disclosed and should not have been so restrietively treated. Davis v. Matthews, 361 F.2d 899, at 901 (8th Cir. 1966); Ramirez v. Perez, 457 F.2d 267 (5th Cir. 1972); Beckman Instruments, Inc. v. Chemtron-ics, Inc., 439 F.2d 1369 (5th Cir. 1970).
2. Kohnenkamp, et al, No. 2,914,193 (1959).
Sprengel himself was the co-patentee of '193. It was the first workable solution to the problem to serving forward and after hatches with one boom, and it was a commercial success. Kohnenkamp ’193 is similar to Sprengel '390 with the exception of the boom head configuration and the gear at the head of the boom. The boom at its head resembles a “Y” or a simple fork; in fact, ’193 has been referred to as the “old fork” and Sprengel’s third embodiment in ’390 as the “new fork.” The single set of purchase blocks suspends between the arms of the fork. There is no pin or connecting piece between the arms except at the throat of the “Y” from which is suspended the upper purchase block from a fitting which functions much like the pendulum fitting in Sprengel '390. Each arm has a guide sheave at its end and the cable from each arm runs down to the purchase blocks so that each end of the purchase is a hauling part. When the boom is swung through, the purchase gear is drawn up closely to place the upper and lower purchase blocks together. The boom is then closely swung across the vertical, and to the opposite direction, and the lower purchase block is then lowered down the opposite side of the boom after the boom swings through. Despite the success of ’193, there were shortcomings. Due to the height of the head of the boom at the point at which the purchase gear [240]*240passed through the uprights, it was difficult to view and thus tricky to operate under even good conditions and more so at night or in bad weather. Highly skilled crewmen were required to complete the maneuver. Also, great care had to be exercised to insure that the lower purchase block was drawn up close enough to swing through the fork but not so tight as to foul the rigging of the blocks. It is not necessary to further describe in detail '193. It works exactly like '390 except for the different boom head configuration and the resulting swing through of the lower purchase block above the upper, rather than below it in a pendulum-like motion.
3. Lehmann No. 3,042,222 (1962).
Lehmann ’222 recites in the patent the problems encountered in using one boom to serve both forward and after hatches and proposes to disclose gear which can accomplish the end without the problems. The boom, according to Lehmann '222, has two legs through which the hoisting gear may freely swing when the boom is swung through to serve the hatches forward and aft. A load bearing cross piece, perpendicular to the longitudinal axis of each boom leg in a plane parallel to the deck, is mounted between the legs of the boom at their top. On the cross piece is a fitting which rotates freely around the cross piece and from which fitting is suspended the purchase gear. The boom is mounted on a platform on the ship’s deck which rotates. This is nothing more nor less than a universal joint. As the boom swings through the kingposts, the purchase gear suspended from the cross piece swings through the legs of the boom like a pendulum. A reading of the patent discloses that cables from the kingpost are taken up or let out simultaneously to raise and lower the boom. The finding of the district court that Lehmann ’222 taught a boom which rotates on a 180 degree axis to face one hatch and then turn to face the other hatch was clearly erroneous. Lehmann '222 instead discloses a boom which is raised and lowered, not rotated, to accommodate both hatches, precisely in the same manner as Sprengel ’390. It is also clear that Lehmann ’222 discloses purchase gear which swings between the legs of the boom, alongside each leg, like a pendulum, again precisely like Spren-gel ’390.12 Lehmann ’222 is a paper patent; it has never been installed on any ship. The issuance of the patent, however, creates a presumption that it is workable. Simmons Co. v. A. Brand-wein & Co., 250 F.2d 440, 447 (7th Cir. 1957); Application of Reynaud, 51 C.C. P.A. 1310, 331 F.2d 625, 628 (1964). And the alleged unworkability of a patent does not necessarily preclude its relevance in ascertaining the scope and content of the prior art. Hadfield v. Ryan Equip. Co., 456 F.2d 1218 (8th Cir. 1972); Sutter Products Co. v. Pet-tibone Mulliken Corp., 428 F.2d 639 (7th Cir. 1970).
4. Lehmann No. 3,107,790 (1963).
Lehmann ’790 discloses a single boom structure attached at its foot to a universal type mounting between two king-posts. The kingposts may be the massive conical type or they may be constructed like tripods. The boom swings [241]*241fore and aft through the kingposts and may also slew outboard to port or starboard. In at least one embodiment, the hollow boom head of ’790 is fitted with a cap which inserts into the boom head and rotates freely about the longitudinal axis of the boom. The topping lifts are attached to the cap. The hoisting gear is attached to and suspended from a fitting pivotally attached to pins protruding from the head of the boom below the freely rotatable cap. In Lehmann’s own words:
“[A] hoisting gear is preferably piv-otally suspended from the boom head and a hoisting gear lead line runs over a guide block at the heel of the boom to a hoisting winch which is positioned laterally offset from the longitudinal center plane of the rig in the plane of the kingposts. The hoisting gear is swung through a free space between the boom and the adjacent kingpost opposite the position of the hoisting winch or on the same side thereof. In this manner, all running gears remain unchanged in their relative position to the boom, kingposts and winches while the boom is shifted from hatch to hatch.” [Emphasis added] Column 1, lines 43-47, Leh-mann ’790.
Lehmann ’790 was installed on two ships. Lehmann ’790 was cited to the district court but was only referred to in its opinion briefly (“rotating heads” 346 F.Supp. at 1126) and apparently was not heavily relied upon by defendant as rendering Sprengel ’390 obvious. It is, however, a part of the record and is relevant on the concept of suspended purchase gear • which swings through with the boom. It provides for a swing through of the purchase gear, like a pendulum, as does Sprengel ’390. The swing through of the purchase without rerigging in ’790 is accomplished by rotating the boom while in a nearly upright position, the boom being pivoted at both ends.
5. Sparrow No. 3,110,403.
Sparrow ’403 was issued on November 12, 1963, and is commonly referred to as the Newport News gear. The gear was developed in 1962 during construction of five Challenger class cargo ships which Newport News Shipyard built for United States Lines. This gear works; it has been installed on cargo ships, and has enjoyed regular commercial use. Sparrow ’403 discloses a single boom structure mounted on the deck on a universal type joint. The boom is mounted between two kingposts. The boom may swing forward and aft between the kingposts and slew outboard to port or starboard. In Sparrow, as in Lehmann ’222 and ’790, Kohnenkamp, and Spren-gel ’390, movement of the boom is controlled by blocks and cables extending from the kingposts to the head of the boom. Sprengel ’390 calls it span gear; the others call it topping lifts — there is no difference except in semantics. Sparrow utilizes a rotary sleeve just below the boom head. The sleeve, simply stated, has a padeye protruding outward. The purchase gear suspends from the padeye. Sparrow’s language is clear enough to merit quotation:
“The cargo fall
[242]*242Sparrow ’403, while cited to the district court and discussed in the briefs to this court, was apparently not pressed below. Why it was not pressed is not clear, but we may not ignore its relevance as a part of the prior art. Sparrow discloses a successful means for serving forward and after hatches without the highly skilled operators needed and danger which attended the old fork. Sparrow also discloses purchase gear which swings alongside the boom as the boom swings through the vertical, like a pendulum. The sleeve at the boom head may be rotary, but the drawings and the claims clearly disclose purchase gear which swings alongside the boom as the boom swings through the kingposts.
The Differences Between the Prior Art and the Claims at Issue
Blohm & Voss alleges that the gear on Grace’s ships infringed claims 1, 2, 3, 14, 15 and 18 of Sprengel ’390. The district court found that claims 1, 2 and 3 were not infringed but that claims 14, 15 and 18 were infringed.13 Blohm & Voss did not cross appeal the no-infringement findings of claims 1, 2 and 3. Therefore, the validity of such claims is not at issue. The claims at issue are 14, 15 and 18 of Sprengel ’390. The other claims, however, may be relevant because they disclose the various methods by which the inventor’s concept may be employed. If the inventor says that one embodiment of his patent employs the same concept as another embodiment of his patent, then prior art similar to either embodiment would seem to bear on the issue of obviousness.14
At the outset, we note that there are some differences between Sprengel ’390 and each prior art reference, just as there are some differences between the [243]*243prior art references themselves. Ashe’s boat davit is necessarily different in some respects. Yet, in both Ashe and Sprengel ’390 a load bearing support (call it a boom, a davit or whatever) is affixed to the ship’s deck so that it may tilt back and forth (through dead center) . In both, the head of the support is always directly over the matter to be raised, lowered or transferred, be it cargo or a lifeboat. In both, a fitting is pivotally mounted from a pin which protrudes from the support and is mounted at right angles to the longitudinal axis of the support in a plane parallel to the deck. In both, the purchase gear suspends from such fitting and swings alongside the support in pendulum fashion.
Kohnenkamp ’193 very successfully accommodated forward and after hatches and disclosed all of ’390 except for the boom head configuration and fittings used on ’390’s side mounted pendulum fittings, which constitutes the major difference between the two. The result of the side mounted pendulum fitting was that the danger and requisite skill problems of swinging through in Koh-nenkamp ’193 were solved.
' There are differences between Leh-mann ’222 and Sprengel ’390 side pendulum mounting. Sprengel ’390 discloses a single boom structure, whereas, Leh-mann ’222 discloses a boom with two legs, or, as plaintiff calls it, two booms joined at the top. Either way, Lehmann ’222 discloses a boom or booms with an opening in between, through which the purchase gear swings as the boom swings fore and aft. One could say that it swings through the opening in the single boom, like the new fork of Sprengel ’390, or one could say it swings alongside the separate booms in pendulum fashion like Sprengel’s side pendulum embodiments. Plaintiff, in its zeal to show that Leh-mann ’222 does not anticipate Sprengel ’390, fails to take account of what Leh-mann ’222 does disclose. Its position seems to be that Lehmann ’222 must anticipate Sprengel ’390 or it is irrelevant. Such is not the law. Graham v. John Deere, supra, sets forth the applicable standards, and specific anticipation by a single reference is not the rule.
Although, it may be said that Leh-mann ’790 and Sparrow ’403 are enough different from Sprengel ’390 to the extent that the claims of neither could be said to anticipate Sprengel (at least defendant did not so argue), both disclose quite relevant concepts.
While the methods of obtaining the pendulum swing through of the purchase gear without rerigging in Lehmann ’790 and in Sprengel ’390 are different, in both, when the boom is swung from forward to aft, the purchase gear swings through alongside the boom like a pendulum. When the boom is positioned at a place other than at or nearly dead center, the fittings freely rotatable around the pins protruding from the top of the boom and which support the purchase gear, have precisely the same function as does the pendulum fitting in Sprengel ’390.
Both Sparrow ’403 and Sprengel ’390 disclose purchase gear which swings alongside the boom like a pendulum as the boom swings through the kingposts. In Sprengel ’390, of course, this is accomplished by the pendulum fitting attached to the pin. In Sparrow ’403, the swing through without rerigging is accomplished by having a rotary sleeve near the head of the boom which freely rotates around the longitudinal axis of the boom and from which the purchase gear is suspended by means of a fitting which enables-the purchase gear to always remain vertical.
The ultimate conclusion to be drawn from these comparisons will be considered along with the conclusion of obviousness.
■The Level of Ordinary Skill in the Pertinent Art
The lower court did not specifically resolve the level of ordinary skill in the pertinent art. Nor did the Supreme Court in Graham, Adams or Anderson’s-[244]*244Black Rock, supra.15 Nor have many decisions decided since Graham. Yet, we are of opinion the foregoing discussion has, perhaps by comparison, made the level clear. The matter seems to be one which does not otherwise easily lend itself to specific articulation. The matter also seems to be necessarily related to the inquiry of the scope and content of the prior art. The inventor who ventures into the art is presumed to know the prior art references. Judge Rich stated it picturesquely when he wrote that the inventor should be imagined as seated in his workshop “with the prior art references — which he is presumed to know — hanging on the walls around him . ” Application of Winslow, 53 C.C.P.A. 1574, 365 F.2d 1017, 1020 (1966). Regardless of what the hypothetical man skilled in the art is presumed to know, the evidence before the lower court revealed that the level of ordinary skill is high. Naval architecture is a recognized specialty in the field of design and engineering. The experts who gave evidence at the trial, by testimony or by deposition, were conversant with the elements of the various cargo gear structures. The patentees in the prior art who patented two-hatch cargo gear exhibited working familiarity with the type of components disclosed in the first workable gear, Kohnenkamp '193, and the capabilities and limitations thereof. It is clear that the field of naval architecture, including design of ship’s cargo gear, is a specific area of engineering specialization in which a considerable degree of expertise is not uncommon.
Conclusion as to Obviousness
With the foregoing in mind, we turn to the ultimate question of obviousness. The patent is before the court with its statutory presumption of validity. 35 U. S.C. § 282. The strength of this presumption has been a matter of considerable comment by the courts. We note first that the patent was issued during a time when the Supreme Court “observed a notorious difference between the standards applied by the Patent Office and by the courts.” Graham, supra, 383 U.S., at 18, 86 S.Ct. at 694. In light of the fact that the court, in Graham, stated that the general strictness of the test of patentability has remained invariable in the Supreme Court for 20 or 30 years, perhaps the suggestion may be that the Patent Office was too permissive in its concept of invention.
Grace contends that the patent slipped through the Patent Office, while Blohm & Voss contends that the presumption is strengthened because the Patent Office considered Lehmann ’222, Lehmann ’790 and Kohnenkamp ’193. We note that the file wrapper for Sprengel ’390 shows that Ashe ’664 and Sparrow ’403 were not considered. This court has held, along with others, that the statutory presumption of validity is weakened where pertinent prior art is not considered by the patent office. Blumcraft v. Citizens and Southern National Bank, 407 F.2d 557, at 561 (4th Cir. 1969); Johns-Manville Corp. v. Cement Asbestos Products Co., 428 F.2d 1381 (5th Cir. 1970); see Skirow v. Roberts Colonial House, Inc., 361 F.2d 388 (7th Cir. 1966); Sound Scribner v. United States, 360 F.2d 954, 960, 175 Ct.Cl. 644 (1966). We think the net effect of all these contentions and authorities is that the plaintiff came into court with a good patent and the burden was on the defendant to prove it bad. We are of opinion the defendant met the burden.
The Supreme Court has listed as secondary considerations which may be rele[245]*245vant on the issue of obviousness the following: (a) commercial success; (b) long felt needs; (c) failure of others, etc. Sprengel ’390 was a commercial success, as was Kohnenkamp ’193. Sparrow ’403 was a commercial success, although perhaps to a lesser extent. Lehmann ’790 was installed in ships. Prior to Kohnenkamp ’193, there had been a long felt need in the industry for a gear utilizing one boom to serve both forward and after hatches without re-rigging. Kohnenkamp ’193 successfully met that need. While it is doubtful that the need to. improve the old fork was as great as the need for the old fork itself, there was a need for gear which could serve forward and after hatches with greater safety and lesser skill. Sparrow ’403 accomplished that result and so did Sprengel '390 to a more successful extent.
These secondary considerations do not affect the analysis in this case to an extent requiring us to find nonobviousness. Commercial success without invention does not amount to patentability. A & P Tea Co., supra. Thus, we come to the ultimate legal question of validity, and it is clear that it is, in the final analysis, a question of law. Graham v. John Deere, supra; Heyl & Patterson, Inc. v. McDowell Co., 317 F.2d 719 (4th Cir. 1963); Lemelson v. Topper Corp., 450 F.2d 845, 848 (2nd Cir. 1971); Cf. Stamicarbon, N. V. v. Escambia Chemical Corp., 430 F.2d 920, 924 (5th Cir. 1970).
While the specific claims of Sprengel ’390 at issue are 14, 15 and 18 because they were the infringed claims, all the embodiments are relevant on the issue of obviousness. This is so in this case because the inventor has disclosed a single inventive concept in three embodiments. (Single purchase gear alongside the boom; double purchase gear, one on either side; and single gear suspended from a structure on top of the boom like a closed wishbone). The fact that the same concept is disclosed in the single and double purchase gear is expressed in the following colloquy between the district court and counsel for Blohm & Voss:
“Now, the double pendulum block, as I believe I said at the pre-trial conference, while discussing what is invention and what is not, is nothing more than putting two of those up there. “The addition of the second block was simply conceived of by Mr. Sprengel because he wanted to increase the load that he was required to lift with the gear that was being ordered. He will testify that when required to lift the heavier load, he told his engineers to take the single pendulum block design and make a mirror image of it. That’s all there is to it.
“It has a single inventive concept, as I believe I tried to express to your Honor at the pre-trial conference. The concept is the utilization of a pendulum block mounted by one end of the boom for free rotation as the boom is swung through the vertical from fore to aft or aft to fore.
“Whether you have one of them or whether you have two of them, the invention is the same.” (Appendix (90a-91a).)
Thus, the inventiveness of Sprengel ’390 does not lie in one set of purchase blocks or two sets. It would also seem that the inventiveness does not lie in having the purchase gear swinging alongside the boom because one of the embodiments has the purchase gear swing through the top of the boom. Nor does the inventiveness lie in the particular type of fitting used at the head of the boom from which the. purchase is suspended. In plaintiff’s own words: “Sprengel’s invention does not reside in employing only pendulum-block fittings having the specific configuration, including hollow passages through which the cargo hoisting rope extends, depicted in the patent in suit. If this were not so, the patent in suit would surely have included at least some claims to such a specific fitting. But it does not.” And later: “There is no evidence that Sprengel ever [246]*246asserted a patent claim to such pendulum block fittings.” (Appellee’s brief, p. 35). It is acknowledged that all of the elements are old in the art, and that there is nothing new in the concept of swinging through. Particularly in view of this, it becomes increasingly difficult to perceive what it is that Blohm & Voss claims as invention. On oral argument, counsel stated that the point of invention is the specific combination of elements. Of course, the fact that the elements are all old in the art does not necessarily mean the patent is obvious, nor is the use of hindsight permitted, Graham, 383 U.S., at p. 36, 86 S.Ct. 684, but as cautioned by the Supreme Court in A & P Tea Co., supra, such a claim must be closely scrutinized, bearing in mind the “difficulty and improbability of finding invention in an assembly' of old elements.” A & P, 340 U.S., at p. 152, 71 S.Ct., at p. 130. This would seem to be particularly true where, as here, the concept employed in the patent is also admitted to be old in the art.
Nevertheless, Sprengel was the first to combine elements in the precise fashion disclosed in ’390, and we must determine whether such combination was obvious to a person of ordinary skill in the pertinent art. We think it was. The combination filled a commercial need and has enjoyed commercial success, but those matters without invention will not make patentability. Anderson’ s-Black Rock, supra, 396 U.S. at 61, 90 S.Ct. 305. The combination resulted in greater convenience and ease of operation but did not produce a “new or different function.” Anderson’s-Black Rock, at 60, 90 S.Ct. 305. Nor are there present here the elements considered in United States v. Adams, supra, and Anderson’s-Black Rock, supra, of the inventor surmounting skepticism in the art to achieve unexpected results16 or providing the key requirement or missing link to achieve a desired end. Adams, 383 U.S., at 51, 86 S.Ct. 708, and Anderson’s-Black Rock, 396 U.S., at 60, 90 S.Ct. 305.
In addition, and by way of summary the kingposts, boom, universal joint, topping lifts, and purchase gear are common to Sprengel ’390, Lehmann ’222, Lehmann ’790, Sparrow ’403, and Koh-nenkamp ’193. Use of a fitting attached to the boom or davit, at or near its head, by a round pin or support whose longitudinal axis is perpendicular to the longitudinal axis of the boom or davit in a plane parallel to the plane of the deck, which fitting rotates freely around such axis of the pin or support, and to which fitting is attached the upper purchase block, is common to Sprengel ’390, Leh-mann ’222, Lehmann ’790, and Ashe ’664; and the fitting at the throat of the fork to which the upper purchase block is attached in Kohnenkamp ’193 is remarkably similar, although without precisely the same function. Swinging the purchase gear, like a pendulum, alongside the boom or davit as it traverses dead center is common to Sprengel ’390, Lehmann ’222, Lehmann ’790, Sparrow ’403, and Ashe ’664. Use of the freely rotating fitting attached to the pin or support at or near the head of the boom or davit to effect the pendulum swing through is common to Spren-gel ’390, Lehmann ’222, and Ashe ’664. In all events, invention is not claimed for the kingposts, boom, universal joint, topping lifts or purchase gear.
Comparing the claims at issue (see esp. n. 14), Graham, 383 U.S., at p. 17, 86 S.Ct. 684, with the prior art, and mindful of the directions in Graham, we conclude that any achievement here was that of mechanical improvement utilizing known elements and known concepts and that such improvement was obvious to one reasonably skilled in the art. Improvement over the prior art does not necessarily rise to the dignity of [247]*247invention.17 Gordon Cartons, Inc. v. Alford Cartons, 218 F.2d 246, at 250 (4th Cir. 1954). We must never forget that it is a constitutional standard we apply and a strict one at that. Anderson’s-Black Rock, 396 U.S., at 61, 62, 90 S.Ct. 305. Attempt here is not made to degrade Sprengel’s achievements, but what he accomplished in ’390 does not rise to the requisite level of inventiveness to entitle him the monopoly jealously guarded by the constitution. We think the following principle from A & P Tea Co., which this court has noted with approval (Gordon Cartons, supra, 218 F.2d at 250), should apply here:
“A patent for a combination which only unites old elements with no change in their respective functions, such as is presented here, obviously withdraws what already is known into the field of its monopoly and diminishes the resources available to skillful men. This patentee has added nothing to the total stock of knowledge, but has merely brought together segments of prior art and claims them in congregation as a monopoly.” A & P Tea Co., supra, 340 U.S. at 152-153, 71 S.Ct., at 130.
The judgment of the district court is Reversed.
. Fall, as used here and for the purpose of this opinion, is synonymous with purchase.