SMITH, Judge.

The invention for which appellant seeks a patent comprises means for safely and effectively controlling a relatively large neutron output by varying a small and easily controlled neutron input source. The application, serial No. 654,837, filed April 24, 1957, is aptly titled "Neutron Amplifier." Claim 27 is illustrative of appealed claims 27-33 and reads:

"27. A subcritical neutron amplifier having a controllable neutron source and, associated with said source in cascade, an input region of neutron moderator material in which neutrons of epithermal energy from the source are moderated to thermal energy levels, a sequent fuel region containing neutron fissionable material in mass concentration and geometric configuration adapted to augment the neutron flow by a subcritical reaction, and an output region comprising a thermal neutron barrier substantially opaque to thermal neutrons but transmissive to epithermal neutrons, whereby an amplified neutron output is subcritically produced."

Claims 30, 31 and 32 are quite similar to claim 27 and need not be separately discussed. Claim 28 is dependent upon claim 27 and calls for means to vary the transparency of the neutron barrier. Claim 29 further provides that the transparency-varying means of claim 28 be such that the neutron flow may be modulated as a function of time. Claim 33 adds to claim 32 the limitation that the fuel be in an hydrogenous environment.

In the embodiment illustrated and described in appellant's application, the amplifier employs a Van de Graff accelerator as a source supplying neutrons to an initial stage composed of three regions or zones. The first of these, the input or moderator region, retards the speed of the neutrons entering from the source so that only thermal (low-energy) neutrons are passed to the second region. The second, or fuel region, consists of fissionable material, such as uranium or a water solution of uranyl nitrate, which is responsive to the entering thermal neutrons to effect fission by which more epithermal (high-energy) neutrons are evolved. The second region is not of crtical mass and thus by design it is incapable of exploding while producing neutrons. The third region is a neutron barrier which functions to prevent transmission of thermal neutrons to the moderator region of the next succeeding stage, and allows only epithermal neutrons to pass.

This third region may be provided with means to vary its transparency by a shutter-like array of cadmium strips adjusted to provide feedback control. In addition, the transparency-varying means may be provided with means responsive to a time-varying function so that the feedback neutron flow may be modulated as a function of time. It is also disclosed that a number of such amplifier stages may be cascaded to any given degree of neutron flux enhancement.

Appellant asserts that the claimed invention affords a revolutionary approach to the safety problem in the nuclear reactor art. As the amplifier is said to be inherently safe from divergent nuclear chain reaction, the intricate systems needed to monitor and control the operation of conventional neutron amplifiers to prevent an explosion are unnecessary.

The single reference relied upon by the Patent Office in rejecting the appealed claims is an Atomic Energy Commission document entitled "KAPL-M-RWS-1, A Stable Fission Pile with High Speed Control." The document is in the form of an unpublished memorandum authored by one Samsel, and will hereinafter be referred to as "Samsel." Samsel is dated February 14, 1947 and was classified as a secret document by the Commission until March 9, 1957, when it was declassified. In essence, Samsel sets forth and discusses the problems present in the control of a nuclear reactor, the concept of use of successive fuel stages to effect such control, and a description of the arrangement, composition and relative proportions of materials required to obtain the sought-for results. Samsel is prefaced by a statement that it was made to record an idea, and it nowhere indicates that the idea had been tested in an operating reactor.

The Patent Office does not invoke Samsel as a publication (which it apparently was not, at any pertinent date). Rather, the contention is that Samsel constitutes evidence of prior knowledge within the meaning of 35 U.S.C. § 102(a).

While there seems to be some disagreement on the part of the solicitor, we think the most reasonable interpretation of the examiner's rejection, and one which is concurred in by the board and by appellant, is that claims 27, 30, 31 and 32 are fully met by Samsel and thus the subject matter defined therein is unpatentable because it was known by another in this country prior to appellant's invention thereof. As to claims 28, 29 and 33, even though not fully met by Samsel, they are said to be obvious within the meaning of 35 U.S.C. § 103 in view of the prior knowledge evidenced by Samsel.

Our own independent consideration of Samsel has convinced us that it contains adequate enabling disclosure of the invention of claims 27 and 30-32, see In re Sheppard, 339 F.2d 238, 52 CCPA 859, and In re LeGrice, 301 F.2d 929, 49 CCPA 1124, and appellant does not appear to contend otherwise. Rather, appellant contends that Samsel is not available as evidence of prior knowledge under sections 102(a) and 103. Appellant also argues that, even if Samsel is available, the subject matter of claims 28, 29 and 33 is not obvious in view thereof. We agree with this characterization of the essential issues presented on this appeal, and will treat them in the order stated above.

In the case of In re Schlittler, 234 F.2d 882, 43 CCPA 986, this court was presented with the following situation: A manuscript containing an anticipatory disclosure of the appellants' claimed invention had been submitted to The Journal of the American Chemical Society and was later published. The date to which the appellants' application was entitled for purposes of constructive reduction to practice was earlier than the publication date of the Journal article, and therefore the Patent Office did not contend that the "printed publication" portion of section 102(a) was applicable. However, the manuscript bore a notation that it had been received by the publisher on a date prior to the effective filing date of the appellants' application. On the basis of this notation the Patent Office argued that the article constituted sufficient evidence of prior knowledge under section 102(a).

After an exhaustive review of the authorities, and of the legislative history of the Patent Act of 1952, this court rejected the contention of the Patent Office, and concluded that such a document was not proper evidence of prior knowledge. In reversing, the court stated ( 234 F.2d at 886, 433 CCPA at 992):

"In our opinion, one of the essential elements of the word `known' as used in 35 U.S.C. § 102(a) is knowledge of an invention which has been completed by reduction to practice, actual or constructive, and is not satisfied by disclosure of a conception only."

And therefore, since the Journal article, "at best, could be evidence of nothing more than conception and disclosure of the invention," the

"* * * placing of the Nystrom article in the hands of the publishers did not constitute either prima facie or conclusive evidence of knowledge or use by others in this country of the invention disclosed by the article, within the meaning of Title 35, § 102(a) of the United States Code, since the knowledge was of a conception only and not of a reduction to practice."

Another aspect of the court's discussion in Schlittler involved the well-established principle that "prior knowledge of a patented invention would not invalidate a claim of the patent unless such knowledge was available to the public." After reaffirming that principle, the court went on to state:

"Obviously, in view of the above authorities, the mere placing of a manuscript in the hands of a publisher does not necessarily make it available to the public within the meaning of said authorities."

However, the court did not go on to determine whether the Journal article was in fact available to the public, since such determination was deemed unnecessary for disposition of the case, under the court's theory.

We shall consider first the public availability aspect of the Schlittler case. Although that portion of the Schlittler opinion is clearly dictum, we think it just as clearly represents the settled law. The knowledge contemplated by section 102(a) must be accessible to the public. In addition to Schlittler and cases cited therein, see, e.g., Minneapolis-Honeywell Regulator Co. v. Midwestern Instruments, Inc., 7 Cir., 298 F.2d 36 (7th Cir. 1961); Rem-Cru Titanium, Inc. v. Watson, 152 F. Supp. 282 (D.D.C. 1957).

In the instant case, Samsel was clearly not publicly available during the period it was under secrecy classification by the Atomic Energy Commission. We note that the date of declassification, however, was prior to appellant's filing date, and it is perhaps arguable that Samsel became accessible to the public upon declassification. But we do not find it necessary to decide that difficult question, for there is a statutory provision which is, we think, dispositive of the question of publicity. Section 155 of the Atomic Energy Act of 1954 ( 42 U.S.C. § 2185) provides:

"In connection with applications for patents covered by this sub-chapter, the fact that the invention or discovery was known or used before shall be a bar to the patenting of such invention or discovery even though such prior knowledge or use was under secrecy within the atomic energy program of the United States."

We think the meaning and intent of this provision is so clear as to admit of no dispute: With respect to subject matter covered by the patent provisions of the Atomic Energy Act, prior knowledge or use under section 102(a) need not be accessible to the public. Therefore, Samsel is available as evidence of prior knowledge insofar as the requirement for publicity is concerned.

The remaining consideration regarding the status of Samsel as evidence of prior knowledge directly calls into question the correctness of the unequivocal holding in Schlittler that the knowledge must be of a reduction to practice, either actual or constructive. After much deliberation, we have concluded that such a requirement is illogical and anomalous, and to the extent Schlittler is inconsistent with the decision in this case, it is hereby expressly overruled.

The mere fact that a disclosure is contained in a patent or application and thus "constructively" reduced to practice, or that it is found in a printed publication, does not make the disclosure itself any more meaningful to those skilled in the art (and thus, ultimately, to the public). Rather, the criterion should be whether the disclosure is sufficient to enable one skilled in the art to reduce the disclosed invention to practice. In other words, the disclosure must be such as will give possession of the invention to the person of ordinary skill. Even the act of publication or the fiction of constructive reduction to practice will not suffice if the disclosure does not meet this standard. See In re Sheppard and In re LeGrice, supra.

Where, as is true of Samsel, the disclosure constituting evidence of prior knowledge contains, in the words of the Board of Appeals, "a description of the invention fully commensurate with the present patent application," we hold that the disclosure need not be of an invention reduced to practice, either actually or constructively. We therefore affirm the rejection of claims 27, 30, 31 and 32.

It remains for us to consider the correctness of the rejection of claims 28, 29 and 33 as obvious in view of the prior knowledge evidenced by Samsel. As previously indicated, claim 28 adds to basic claim 27 the limitation that means be provided to vary the transparency of the neutron barrier. Appellant's specification discusses several such means, including the shutter-like array of cadmium strips referred to earlier. Such a transparency-varying arrangement permits controlled feedback so that once the system has reached a desired neutron generation level, the neutron source may be removed and the system will be maintained in self-sustaining but convergent operation.

The examiner's position, affirmed by the board, was that the provision of such means would be obvious to one of ordinary skill in the art. The argument seems to be that since Samsel discloses a neutron barrier, it can be reasoned therefrom that the thicker the barrier the greater the obstruction to the passage of neutrons, and therefore it would be obvious to vary the thickness of the barrier if it were desired to vary the neutron transparency of the barrier. We note, however, that there is absolutely no hint in Samsel that a variable barrier would even be desirable, much less possible to construct. From the evidence of record, it appears that the concepts of a variable barrier and means for accomplishing it are wholly appellant's and were manifestly unobvious at the time appellant made his invention.

Claim 29, as previously pointed out, simply adds a further limitation to claim 28. We think the rejection of claim 29 was erroneous for the same reasons as the rejection of claim 28.

Claim 33 specifies that the fissionable fuel be in an "hydrogenous environment." The examiner felt that it would be obvious to "substitute the equivalent water moderator for the carbon moderator disclosed by Samsel because to do so would involve mere skill in the art." We think this reasoning is erroneous on two grounds. First, there is nothing in the record which shows the equivalency of water to carbon as a moderator. Second, and much more importantly, appellant's "hydrogenous environment" limitation is directed to the fuel zone rather than the moderator zone.

The solicitor, apparently recognizing the inadequacy of the examiner's treatment of claim 33, asks us to take judicial notice of a textbook which allegedly discloses the use of an aqueous solution of uranyl sulfate in the core of a homogeneous reactor. We decline to take notice of this reference, for it is in a highly technical area of subject matter and we have no way of evaluating its accuracy or repute in the art. As we said when confronted with a similar request in In re Petering, 301 F.2d 676, 681, 49 CCPA 993, 999, "Such a reference should have been the subject of a discussion between appellants and the experts in the Patent Office and we do not consider it appropriate for us to attempt to evaluate this reference without the benefit of such discussion."

For the foregoing reasons, the rejection of claims 27, 30, 31 and 32 is affirmed, and the rejection of claims 28, 29 and 33 is reversed.

Modified.

WORLEY, C.J., took no part in the consideration or decision of this case.