The opinion
SMITH, Judge
(dissenting).
The majority opinion, like the decision of the board, is predicated upon a combination of the Murray et al. reference and the Minion reference. It is clear from the majority opinion that it is this *1004 combination which the majority relies upon in finding appellant’s invention to be unpatentable under 35 U.S.C. § 103 . My disagreement with the majority stems from what I understand to have been the purpose of the words “at the time the invention was made” in section 103.
While on the present record Murray et al. and Minion are prior art, the point at which I disagree with the majority is that there is nothing in the record except appellant’s invention which suggests the combination of the two references upon which the majority relies. I question whether this record properly supports a rejection for obviousness under section 103 when it was appellant’s invention and only appellant’s invention which taught this combination to the art. There is no question but that under 35 U.S.C. § 102 (e), a patent which “describes” the invention for which a patent is sought speaks as of its filing date. In re Harry, 333 F.2d 920 , 51 CCPA 1541. In other words, to the extent Murray et al. described appellant’s invention it is a proper basis for the rejection. I do not take this to mean, however, that the partial description of appellant’s invention as found in Murray et al. may be “combined” by hindsight with another partial description found in Minion to support a rejection for obviousness under section 103.
The clear intent expressed in section 103 is that in such a situation as the present the obviousness of the claimed invention must be determined solely from the prior art teachings. This excludes the use of appellant’s invention in making selections of the particular portions of the references and combining them to provide a combination which by hindsight can be said to be “obvious.” In other words, we must make the determination required by section 103 without the benefit of appellant’s invention.
As pointed out in appellant’s'brief:
“Appellant’s claim defines a new compound which when administered parenterally is characterized by an extremely rapid onset of hydrocor-tisone activity, making it a drug of choice in case of medical emergency. Neither the compound nor its utility are taught by the disclosures of Murray et al. in view of Minion. Murray et al. discloses in shot gun fashion thousands of esters which give no indication of any differentiation in properties among them and which does not disclose appellant’s specific ester. Murray et al. and Minion taken together do not teach the compound of appellant’s claim or its utility. * * * ”
I will agree that after appellant had disclosed his invention, and with this disclosure to guide, us, the selection of bits and pieces of information from Murray et al. and from Minion can be made and then readily combined. Yet, this does not resolve the central issue which must be resolved under section 103, i. e., was it obvious to one of ordinary skill to so combine these references at the time appellant’s invention was made ?
I think the present record discloses that appellant’s invention was not obvious when so tested. Appellant asserts that Murray et al. embraces “hundreds of thousands of possible esters.” The majority narrows this to “relatively few” compounds. In either event, it is necessary before using the Murray et al. disclosure in such a rejection to make an intelligent selection from among these many compounds. I find nothing in the record except appellant’s disclosures which suggests making such a selection. Does this make the selection “obvious”? I think not. I am reinforced in this analysis by the way in which both the board and the majority draw on Minion in order to arrive at the combination of references on which the conclusion of obviousness is predicated. In the absence of appellant’s teachings from the prior art, there simply is no basis upon which to predicate what seems to be the most unlikely combination of Murray et al. and Minion.
Both Murray et al. and Minion fail to give any indication of either the utility *1005 or the properties of the compound of appellant’s claim except as these properties are attributable to them by hindsight. In evaluating the asserted obviousness, the compound and all of its properties must be evaluated. “[Cjlaims to chemical compounds are drawn to more than structural formulae.” In re Ward, 329 F.2d 1021, 1023 , 51 CCPA 1132, 1134. “[Tjhere is no basis in law for ignoring any property in making a comparison of the claimed and prior art compounds.” In re Lundsford, 327 F.2d 526, 528 , 51 CCPA 1000, 1003. “The problem of ‘obviousness- under section 103 in determining the patentability of new and useful chemical compounds * * * is not really a problem in chemistry or pharmacology * * *. It is a problem of patent law.” In re Papesch, 315 F.2d 381, 386 , 50 CCPA 1084,1091.
The majority opinion, it seems to me, has approached the problem of obviousness in this case as a problem in chemistry or pharmacology in which it has found both the problem stated and its solution in appellant’s disclosure rather than in the prior art. Under these circumstances, it seems to me that a conclusion of obviousness is not warranted, for it was appellant who first disclosed the new compound and who first taught the art its novel utility. There simply is no teaching of appellant’s invention as a whole to be found in the combination of references on which the majority bases its opinion. There is nothing in the record upon which to find that the expected skill of the art could without appellant’s disclosure extract such teaching from Murray et al. and Minion.
In retrospect, almost any invention can be seen to be but a naturally occurring thought leading to an obvious combination of bids of previously acquired knowledge. It is a most difficult task to remove from the prior art the invention for which a patent is sought but this is what we must do in applying section 103 before we attempt to pass judgment as to obviousness. As has been aptly observed recently, “It is not easy for the human mind to put what is now obvious back into the box labelled ‘unknown.’ ” 1 Yet this is precisely what must be done if section 103 is to mean anything. The failure of the majority here to put what appellant taught into the “box labelled unknown” makes its conclusion plausible but wholly without statutory support.
I would, therefore, reverse the appealed rejection.
. Ferguson, “The Origins of the Steam Engine,” Scientific American, Vol .210, p. 103 (1964).