Claims 1-16 of appellant’s application1 stand rejected, and the threshold question is, on what ground? Neither the examiner nor the Board of Appeals specified the statutory basis for the rejection.
In concluding its opinion, the board stated: “we deem claims 1 through 16 to be unpatentable over Link2 and accordingly, the rejection of the claims will be sustained.” The solicitor takes the position that this holding constitutes reliance upon 35 U.S.C. § 103, and in this connection states:
“Even if the examiner’s rejection were deemed limited to 35 U.S.C. 102, the Board’s reliance on Section 103 was not a new ground of rejection, as *185referred to in Rule 196(b) * * *. The ground is the same, the claims are unpatentable over Link. Appellant has not, in any sense, been deprived of any ‘procedural rights’ * * *»
Even though the statute was not named, scrutiny of the examiner’s answer makes it clear that the examiner’s action was based on section 102:
“Applicant is correct in stating that the only issue is the question of what Link teaches. It is agreed that if Link teaches a final product having two pieced interleaved spiral core strips than [sic] all claims are anticipated. On the other hand if Link only teaches a series of concentric closed loops to form a laminated core then the claims are not anticipated. [Emphasis added.] ******
“The portions of Link’s specification referred to by the examiner clearly disclose a final structure of a plurality of pieced interleaved spirals, and though a diagrammatic showing of the drawing may allegedly disagree therewith; such disclosure constitutes a valid anticipation of the claims.” [Emphasis added.]
Our appellate jurisdiction is restricted by 35 U.S.C. § 144 to hearing and determining the appeal “on the evidence produced before the Patent Office.” It seems basic to the concept of procedural due process that an applicant at least be informed of the broad statutory basis for rejecting his claims, so that he may determine what the issues are on which he can or should produce evidence. In order to insure that an applicant be so informed, 35 U.S.C. § 132 requires that whenever a rejection is made, the Commissioner “shall notify the applicant thereof, stating the reasons for such rejection, or objection or requirement, together with such information and references as may be useful in judging of the propriety of continuing the prosecution of his application * *
We are here confronted with a situation wherein the examiner, although not in so many words, relied on section 102 as the statutory basis for his rejection, while the board, if the solicitor’s interpretation of its opinion is correct, relied upon section 103 as well. We think that consideration by this court of the merits of the board’s alternative position would deprive appellant of important procedural rights.
We do not mean by this to imply that the Board of Appeals may not, if it sees fit, enter new or alternative grounds for rejection of claims in an appealed application. Such practice is not uncommon and is often of great value in clarifying the examiner’s position or in correcting errors and oversights. And if the new ground of rejection is properly identified as such, the applicant is in no way prejudiced, for he has the benefit of the adequate procedural safeguards of Rule 196 (b).
Here, however, the board’s opinion did not provide adequate notice that a new statutory ground of rejection was being relied on. For all that appellant knew, from the vague language used by the board, the only rejection being passed upon was the one made by the examiner under section 102. Thus, if we were to consider, on this appeal, issues which arise under section 103, we would be doing so without affording appellant an opportunity to be heard in the administrative tribunals of the Patent Office.
We cannot agree with the solicitor that there is no material difference between sections 102 and 103 with respect to an applicant’s right to a day in court. A rejection under section 102 involves a comparison between the subject matter disclosed by the reference and the claimed invention, in order to determine whether the claimed invention is “described” in the reference. See, e. g., In re Sheppard, 339 F.2d 238, 52 CCPA 859. Section 103, on the other hand, requires consideration of the differences between the claimed invention and the prior art, for the purpose of determining whether the claimed subject matter as a whole would have been “obvious” to one of ordinary skill in the art. Thus, the *186issues arising under the two sections may be vastly different, and may call for the production and introduction of quite different types of evidence.
Accordingly, to the extent the board’s decision goes beyond the examiner’s rejection, which we have found to be for want of novelty under section 102, it will not be considered, since appellant never received adequate notice of any other ground of rejection which may or may not have been relied upon by the Board of Appeals.
Basically, the issue under section 102 turns upon the interpretation of the Link disclosure. As the examiner pointed out, if Link discloses pieced, interleaved-spiral core strips, the claims are anticipated. If, however, Link discloses a series of concentric closed loops to form the laminated core, the claims are not anticipated.
In electromagnetic induction apparatus of the general type shown in Link, a core is assembled about preformed conductive windings. The appealed claims 3 specify the structure of the core element in such apparatus as consisting of a plurality of wound, pieced, interleaved-spiral magnetic strips. Such a core is made by concurrently winding a plurality of faeeabutting strips of magnetic material upon an arbor to form interleaved spirals. The wound structure is heat-treated to improve its magnetic characteristics. The strips are unwound without destroying their permanent set, and are cut at predetermined intervals. The severed pieces are then assembled on the preformed conductive windings. The issue as to what Link discloses arises at this point. It is clear that appellant assembles the several pieces in their original interleaved-spiral relationship, with the ends of one segment abutting the ends of adjacent segments.
Link, however, after winding two magnetic strips about a trapezoidal mandrel to the form in which the strips are annealed, drills an indexing hole through one end of the core, after which he separates the core into the separate spiral strips and inserts indexing pins through the respective holes in the strips. The two strips are then separately clamped and one is cut on the left side of its indexing pin and the other on the right side of its indexing pin. The cut pieces are reassembled around the preformed windings to provide the completed core.
The rejection is predicated upon Link’s mode of reassembly and the nature of his final core structure. The examiner viewed Link as teaching but one embodiment, i. e., an anticipatory “interleaved-spiral” structure in which the ends of each piece abut the ends of other pieces. The examiner’s position is based on language in the Link specification to the effect that the severed pieces are reassembled in the “identical position” or “relationship” which they assumed in the initial wound core.
However, we find Link’s description of the reassembly operation to be vague and uncertain. In fact, the board recognized this defect in the use of Link as an anticipating disclosure in stating:
“While it may be that with such description and illustration Link would seem to be somewhat ambiguous in respect to the nature of the final core form, nevertheless the implications of those portions of the patent relating to reassembly of the components involved in identical position and relationship to their initial character, coupled with the generic mode of desciption of the reassembly, cannot be overlooked. In our opinion, the very imprecision as to the particulars of Link’s presentation *187would readily apprise the person shilled in the art of the two possibilities of reassembly and resulting final core form, that is, the possibilities of reassembly for either a spiral or ring mode, with the factors noted immediately above clearly constituting a suggestion of reassembly in spiral mode. In this sense, we think that Link would, in essence, represent a sufficient disclosure of a core having pieced, interleaved spiral form. * * * ” [Emphasis added.]
We do not agree that such an ambiguous disclosure “describes” appellant’s claimed invention within the meaning of section 102. Our view is further supported by Link’s description of the reassembly process as follows:
“The reassembly of the core is begun by taking the smallest loop and removing it from its innermost position in the series of loops which make up one of the component strips of the initially wound core. The severed ends of this loop are spread to permit its slidable positioning over the end of the mandrel. The innermost or smallest loop of the adjacent series of loops which make up the adjacent component strip is then applied to the initially positioned loop on the mandrel by spreading its severed ends a sufficient distance to permit its being slid into proper position on the previously applied loop. The inherent resilience of each loop will cause it to return to a substantially closed position after its free ends have traversed the length of the previously applied loop. This process is continued by the application of successive loops in the same sequential order as in their initial formation. From time to time during the reassembling of the core it may be desired to utilize the blows of a soft hammer to maintain the laminated surface of the core substantially smooth and jar the assembled loops sufficiently to bring their severed ends into substantially abutting relationship.”
Thus in Link it seems clear that each successive loop is slid into position on the previously applied loop and each loop returns to a substantially closed position after its free ends have traversed the length of the previously applied loop. However, in the reassembly of appellant’s interleaved, pieced spiral core, each successive loop must underlie a portion of the previously applied loop. We do not find such a core described in Link’s description of the assembly of his final core structure, in which he requires that the core be comprised of a series of concentric closed loops.
The solicitor in his brief seeks to avoid this conclusion by interpreting the reassembly process described in the Link reference as follows:
“ * * * In conformance with the requirement that the loops have the identical relationship that they had in the initially wound core, an end of a loop being applied will be inserted under the end of a previously applied loop where the resiliency of the loop will move it into a substantially closed position adjacent the end of the next innermost loop from which it had been severed.” [Emphasis added.]
The fault in this reasoning is that nowhere does Link disclose inserting the ends of successive loops under previously applied loops. The method of reassembly disclosed in the Link specification cannot result in a final core structure comprised of a plurality of wound, pieced, interleaved-spiral strips. Link’s phraseology, “identical position,” when construed to cover appellant’s pieced, interleaved core structure, is therefore not consistent with the method of reassembly disclosed by Link according to the solicitor’s own interpretation of that language.
Appellant suggests that Link, in using this phraseology, disclosed a method whereby a coaxial, closed-ring staggered-joint core may be constructed without reversing alternate turns. A careful *188examination of the Link reference supports this view. Link describes at length the shortcomings of the prior art, that is, reversing alternate turns, and states that the primary object of his invention is to overcome this inherent shortcoming.
The Link reference depicts his final core structure in figures 8 and 9. These figures show a core structure comprised of a series of concentric dosed loops. We do not find, as urged by the solicitor, that the apparent purpose of figures 8 and 9 is to show the conformation of the core as a whole, and not the details of the interrelationship of the individual loops.
We do not think that Link disclosed structure which anticipates the appealed claims. Words in a reference are to be construed in light of the relevant surrounding circumstances in each case, In re Folkenroth, 275 F.2d 732, 47 CCPA 812, and a reference in any event is good only for that which it clearly and definitely discloses, In re Moreton, 288 F.2d 708, 48 CCPA 875. Moreover, if we accept arguendo the board’s position that Link can be alternatively interpreted to show either of two final core structures, then Link becomes an ambiguous reference which will not support an anticipation rejection. In re Turlay, 304 F.2d 893, 49 CCPA 1288.
The board thus erred, after finding that the Link reference was ambiguous regarding the nature of the final core, form, when it held there was sufficient disclosure of a core having pieced, interleaved-spiral form. A consideration of all phrases in the Link reference and especially of those relied upon by the Patent Office to sustain the rejection leads to the conclusion that the Link reference does not describe appellant’s invention within the meaning of 35 U.S.C. §102.
The appealed decision is therefore reversed.
Reversed.
. Serial No. 739,232, filed June 2, 1958 for “Electromagnetic Induction Apparatus and Method of Manufacture Thereof.”
. U. S. Patent 2,927,366, March 8, 1960.
. Claim 10 is illustrative and reads:
“Electromagnetic induction apparatus comprising an electrical winding structure and a magnetic core structure in linking interrelationship, said core structure comprising a plurality of wound, pieced, interleaved-spiral magnetic strips, each of said strips being formed of a plurality of discrete individual portions in successive end-abutting relationship, each of said end abutments being bridged on each side by portions of two other ones of said strips.”