(dissenting), with whom MARTIN, Judge, joins.
I feel that the majority decision will unlawfully extend Allen’s monopoly on the combination claimed in his patent No. 2,952,060. Normally, in the absence of special circumstances not present here, upon the expiration of a patent, the public is free to practice the claimed invention without fear of infringement. In this case, however, a person using the combination claimed in the Allen patent after it expires will be infringing the later issued patent claiming the waler bracket alone. The doctrine of double patenting was developed to prevent such a result. Yet, the majority has, in effect, written double patenting out of the law *488when the relationship between the inventions in issue is that of combination-sub-combination (a single element here).
At the outset, let me say that I find no fault with the majority’s detailed analysis of the claims because I think it merely shows that the appealed claims are drawn to a novel single element, the waler bracket, while the patent claims a novel combination including the very same waler bracket. The majority calls the element and combination “two separate inventions.” The word “invention” has been subject to many definitions in the patent law recently. The word “invention” as used by this court in the past few years does not connote a patentable invention but merely that which is the subject of a patent application whether or not it is novel, useful, unobvious or within a statutory category of things which may be the subject of a patent. I will, therefore, attempt to use the majority terminology indicating what I think it means. I also feel it necessary to go to some lengths to point out precisely where I feel that logic has become subservient to and distorted by the “magic words” used in double patenting analysis.
To return to the majority opinion, I reiterate that I agree that two separate inventions, i. e., a waler bracket and a combination including the waler bracket, are claimed in the application and patent. Apparently, this finding of fact ipso facto dictates the result — no double patenting. We are not given any analysis or any rule that requires such a result. Thus, the opinion states a rule, a new and dangerous rule, I think, that after obtaining a patent on a combination an inventor can always obtain a patent on an element of the combination claimed in a later, copending application.
As previously indicated, this rule would allow an inventor to control the use of his combination invention long after the seventeen year limitation set by Congress had expired. Such an evasion of the limit on the patent monopoly should not be allowed without legal justification. No such justification is present here. While I am not aware of a precise legally accepted definition of “element” and “combination,” I know that the terms seem applicable to and often are applied to specific chemicals and chemical compositions. Thus, as I interpret the majority opinion, an applicant would be allowed to patent a chemical combination, e. g., a wonder drug and a suitable carrier for the drug, and then later to patent the drug alone based upon an application filed after the combination application. Seventeen years after the combination of drug and carrier had been patented, the public would still not be free to use the drug with the carrier or in any other obvious manner because of the dominance of the later issued patent to the drug alone. That result is an unlawful extension of the patent and is condemned by our decision in In re Hadsel, 173 F.2d 1010, 36 CCPA 1075. Such an extension of the monopoly has also been held unlawful by other courts. See, e. g., Pierce v. Allen B. Du Mont Laboratories, Inc., 297 F.2d 323, (3rd Cir. 1961); Pierce v. Hewlett Packard Co., 220 F.2d 531 (1st Cir. 1955); Palmer Pneumatic Tire Co. v. Lozier, 90 F. 732 (6th Cir. 1898) and cases collected by Stringham, Double Patenting § 2866 (1933).
All of the cases cited here and in the majority opinion recognize that extension of the monopoly is undesirable. In those cases cited by the majority which do allow an extension of monopoly by issuance of the broad patent last, the applicant was the victim of some quirk of Patent Office procedure which the court considered in allowing the applicant to receive broad coverage. In In re Coleman, 189 F.2d 976, 38 CCPA 1156, the court found that there was not necessarily an extension of monopoly because, as the result of a limitation required by the examiner, the claims of the first invention could be practiced without infringing the later issued patent. In In re Sarrett, 327 F.2d 1005, 51 CCPA 1180, the application and patent had been filed on the same day. It was not possible to claim both inventions in the same application because the inventors were different. Thus, the fact *489that one application issued before the other was primarily a result of Patent Office procedure. At least, no evidence of delaying tactics by the applicant was cited. In In re Carlton, 77 F.2d 363, 22 CCPA 1223, the Patent Office had first required division and had then rejected one of the divisional applications on the patent into which the other divisional application had matured. This court reversed the rejection. The same result would most likely be reached today under 35 U.S.C. § 121. In In re Stanley, 214 F.2d 151, 41 CCPA 956, a narrow improvement application, filed for two years after the basic broad application, issued first and was used by the Patent Office to reject the basic patent. I thus think that each of the cases cited by the majority can be distinguished from the issue presented here.
The majority would distinguish the two Pierce cases and the Hadsel case on the ground that only one invention was present in those cases. If the majority means that exactly the same combination of elements was claimed in the patent and application, it is clearly in error. If the majority meant something else when it said there was only one invention, it would have been well to have set forth the test by which one may tell whether or not a subcombination is the same invention as a combination. It is the failure of the majority opinion to set forth any test other than “different invention,” if indeed that is a test, for determining whether double patenting exists which disturbs me most.
A review of this court’s decisions over the last few years discloses at least four different relationships which have been said to exist between claims in a double patenting analysis.
I. Both claims in question may either be identical or differ from each other by a mere colorable variation. In such a case double patenting exists and nothing apparently will save the claims. In re Siu, 222 F.2d 267, 42 CCPA 864, and In re Robeson, 331 F.2d 610, 51 CCPA 1271.
II. The claims may be drawn to different inventions. Although the difference between the claims may be an obvious one, it must be more than a mere colorable variation. When there are different inventions, a terminal disclaimer will obviate a double patenting rejection. In re Robeson, supra, In re Kaye, 332 F.2d 816, 51 CCPA 1465.
III. If the claims are patentably distinct, there is no double patenting. In re Sarrett, supra, and In re Christiansen, 330 F.2d 652, 51 CCPA 1236. In Sarrett, the claims were allowed without terminal disclaimers. Thus, for claims to be pat-entably distinct, the difference between them must be greater than that present when there are different inventions. Furthermore, it should be noted that although this court in Christiansen used an analysis similar to that used in an obviousness case, an obviousness type of analsyis was not applied in Sarrett. In fact, in In re Sarrett, supra 51 CCPA at 1192, 327 F.2d at 1013, we appear to have expressly rejected the obviousness approach in stating:
“We are not here concerned with what one skilled in the art would be aware from reading the claims but with what inventions the claims define.”
Thus to be patentably distinct, two claims need not be unobvious, one in view of the other, and yet the differences between the claims must apparently be greater than when different inventions are claimed. A more precise definition of “patentably distinct,” I cannot glean from reading the cases.
IV. If the claim of the application is obvious (tested by the unobviousness requirement of 35 U.S.C. § 103) in view of the claims of the patent, double patenting exists; if they are unobvious, no double patenting. Judge Rich concurring in In re Zickendraht, 319 F.2d 225, 50 CCPA 1529.
I think that our varied approaches, enumerated above, have overly confused an already confusing subject. For example, my analysis indicates that there is a region somewhere between different inventions and obviousness where claims *490are patentably distinct and thus avoid double patenting. Such a distinction is too fine to be meaningful. I am not unmindful of the fact that the various approaches might possibly be rationalized on the basis of the type of claims to which they were applied, i. e., combination-subcombination, species-genus, basic-improvement. I do not think any such distinctions are set forth, however.
I think that only one test need be applied in the present case — is the claim of the later filed application unobvious1 in view of the claims of the earlier filed application ? If the answer is yes, then the applicant is entitled to a second patent under the law even if it will extend the monopoly of the first patent. If the answer is no, the applicant is not entitled to a patent which will extend the monopoly because he has given the public nothing, in consideration for the added years of monopoly, which was not obvious to the public. Here the description of the waler bracket in the claims of the patent and the appealed claims is almost identical. There can be no question that the bracket of the appealed claims is obvious in view of the patent claims.
It may well happen that after filing an application for a narrower combination, the applicant realizes that a broader subcombination claim is patentable and necessary for proper protection of the invention. If the broader claim is obvious in view of the earlier filed narrower claim, the applicant is not without recourse. He may be able to add broader claims to his first filed application. Such a solution should have been possible for the appellant here. If the first application has already issued as a patent, it may be possible to add broader claims by reissue under 35 U.S.C. § 251. If neither of these approaches works, he may file a terminal disclaimer under 35 U.S.C. § 253, thereby avoiding an extension of the monopoly. Thus, it seems unlikely that an applicant would be unduly penalized for failing to properly cover his invention (or inventions) in one application.
As indicated above, those cases relied upon by the majority which allowed an extension of the monopoly involved technicalities of Patent Office practice which placed the applicant in a difficult situation. The appellant here was not faced with any of those problems. Thus, this case presents a clear issue — is Allen entitled to claims in his later filed application which will extend the monopoly of his earlier filed patent when these later claims are obvious in view of the patent claims? I do not think that he is, and would therefore affirm the board’s decision.
. Since the patented invention is not prior art, it is not technically correct to say that 35 U.S.C. § 103 is applied. However, the test is the same as the unobviousness requirement of 35 U.S.C. § 103.