delivered the opinion of the court:
Henry K. Bichardson has appealed here from a decision of the Board of Appeals of the United States Patent Office affirming that of the Primary Examiner in rejecting claims 1 to 11, inclusive, 15, and 16 of his application for a patent relating to “Improvement in Lamp Base.” No claims were allowed.
The instant application, serial No. 132,956, was filed March 25,1937, and is said to be a continuation-in-part of an earlier application of appellant, serial No. 692,587, filed October 7, 1933, now patent No. 2,075,057, granted March 30, 1937. The patent contains only method claims, while the claims of the application on appeal are directed to the article. It may be said here that Bichardson also filed, on May 14, 1936, an application, serial No. 79,645, said to be a division of the parent application of October 7, 1933. The said divisional application has been abandoned in favor of the one on appeal.
In 1938 appellant’s present application became involved in an interference, No. 75,785, with a patent to one Daniel K. Wright, No. 2,098,-080, granted November 2, 1937, on an application filed January 2, 1936. The interference comprised six counts corresponding to claims 14,15,16, 23, 24, and 25 of said Wright patent, copied by Bichardson into his application. In May of 1941 the interference was finally terminated, unf avorably to Bichardson on all the counts, by a decision of the Patent Office Board of Appeals. It was-there held that the Bichardson parent'patent would not support the counts, so as to permit Bichardson to take advantage of its earlier filing date, and it was further held that Bichardson had not proved reduction to practice prior to Wright’s'filing date. No appeal was taken from that decision, and the claims in appellant’s application corresponding to the interference counts were finally rejected. Appellant then made slight modifications in two of the said claims, viz., 15 and 16, and continued with the ex farte prosecution of those and the other claims now on appeal.
Bichardson’s parent application and patent, his said divisional application, the counts of said interference between Bichardson and *1154Wright, and the decisions of the several Patent Office tribunals in the interference proceeding are all of record in this case.
The subject-matter of the instant application relates to the manufacture of electrical devices such as high-wattage incandescent lamps, power tubes, rectifiers, etc., where metal contact elements are used for the passage of heavy currents. Appellant uses a hollow cylindrical contact pin adapted to fit into a socket connected with a source of electricity. The pin is closed at one end and has a shoulder at its other end. From this shoulder a flaring metal skirt portion or thimble extends upwardly toward the cup-shaped glass base of the lamp. The thimble portion tapers to a sharp edge at its upper end and fits around an apertured glass boss depending from the base of the lamp. By the application of heat, the glass is melted sufficiently to “wet” the metal of the thimble and adhere thereto, so as to form a gastight seal. A metal post member or lead-in conductor extends from the hollow interior of the contact pin up through the thimble, the aperture in the glass base, and into the glass envelope of the lamp. These post members serve as the supports for the filament of the lamp.
In the prior art, as disclosed by some of the patents of record and also as pointed out in the instant application, it was customary first to bead the edge of the thimbles with a ring of glass and then to fuse that glass to the glass of the base. Richardson, by his method, dispenses with the beading step.
It is important to an understanding of some of the issues that we describe briefly the differences between the' drawings of the Richardson parent patent, No. 2,075,057, and the draAvings of the instant application. There are four figures in each. In the patent, each figure shows a base construction in which the boss portion surrounded by the thimble of the contact member does not extend directly from the normally horizontal Avail portion of the base but is connected thereto by a bulging or approximately conical portion. The sharp edge of the thimble is not embedded in tlie glass but merely extends to the conical portion. Figure 1 of the patent does not show a cross section of the contact member, but dotted lines representing the lead-in conductor, though unnumbered, extend from above the cup-shaped base portion down to the top of the thimble, so that the exact physical relationship between the lead-in conductor and the interior of the contact member is not there delineated. Figures 2, 3, and 4 illustrate the contact members in section and show the hollow interior portion thereof in association with the aperture of the boss. No lines representing the lead-in conductor, however, appear in these figures. In the instant application, figure 1 is sectioned to show how the lead-in conductor extends down through the base of the lamp and into the hollow portion of the contact member. In each of figures 1, 2, and 3, as dis*1155tinguished from tlie corresponding figures of the patent, the bottom wall of the base is shown as entirely horizontal. The thimbles surround the depending bosses,, as in the patent drawings, but the sharp edges of the thimbles are embedded in the glass wall of the base, there being no conical or bulging portion separating the normally horizontal portion of the base wall from the boss proper. Figure 4 of the application is identical with figure 4 of the patent, except for numerals. Figure 4 of the application is reproduced hereinafter.
The references cited by the examiner are:
Kruh et al., 3,564,690, December 8, 1925.
Houskeeper, 1,583,463, May 4, 1926.
Marline, 1,885,529, November 1, 1932.
Wriglit, 1,967,852, July 24, 1934.
Wright, 2,069,638, February 2,1937.
Richardson, 2,075,057, March 30, 1937.
Wright, 2,098,080, November 2,1937.
The last-cited patent to Wright is the one with which appellant’s instant application was in interference and is the principal reference. Wright shows a device somewhat similar to appellant’s, differing only in minor details. He has a glass base for a lamp bulb, having apertures and bosses. One of his types of contact terminals is cylindrical in shape and is provided with a flaring skirt and a prong member. The prong member extends up through the aperture in the glass base and is hollow through a portion of its length for the reception of & lead-in conductor and support member for the lamp filament. The edge of the flaring skirt is sunk in and fused to the glass base around the aperture and boss, and Wright states in his specification that the edge may be tapered.
In view of our conclusion, it is unnecessary for us to consider the disclosures of the other references, except Richardson’s prior patent.
We shall, for purposes of decision, consider the claims in separate groups and take up first, claims 15 and 16, which read as follows:
15. A metal contact terminal device for a b'ulbwall base comprising a rigid post member adapted to extend through the base and a thin flared skirt around said member with its free edge adapted- for sealing to the base by a fused glass seal to effect a gaslight union, said skirt having elastic flexure permitting slight relative rocking between said rigid member and said base. [Italics oiirs.]
16. An electric energy translation device comprising a vitreous envelope having openings therein, a iflurality of metal contact terminals each comprising a rigid prong member extending through one of said openings and a thin flared skirt around said member with its free edge sealed directly to the vitreous material around said opening to effect á gastight seal, said skirt having elastic flexure permitting slight relative rocking between said rigid prong member and said envelope, said terminals having sufficient rigidity to support the device. [Italics ours.]
These claims are very closely modeled on counts 5 and 6 of the interference between Richardson and Wright, No. 15,185. The *1156italicized portions represent the only differences between the claims and the counts. In claim 15 the words “its free” have been substituted for the word “outer” of count 5 of the interference, and the functional limitation “to effect a gasti'ght union” was not a part of the count. The same differences exist between claim 16' and count 6; and the word “terminals,” where it first appears in claim 16, was, in the count, “terminal devices” and, where it last appears in the claim, was, in the count, “prong member.” Additionally, count 6 concluded with the limitation, “and an electric energy translation element in said envelope connected to said rigid members,” which has been omitted from claim 16.
One of the examiner’s grounds of rejection of these claims, approved by the board, was that the slight changes were insufficient to distinguish the claims from the counts and that accordingly the claims were unpatentable over the issues of the interference.
It is obvious from the above comparison of these claims with the interference counts that the only difference which could possibly lend patentable distinction is that between the terms “its free edge” and “outer edge.” According to the construction placed upon the counts in the interference proceeding, the term “outer edge” was held to include a marginal portion of the outer surface of the thimble near its edge. No appeal was taken from the decision of the board in the interference proceeding, and that construction is no longer open to question. The issue here, therefore, is whether or not the words “its free” or “free” have, under the circumstances at bar, any different meaning from the word “outer.” While, of course, in many respects the word “free” may, under circumstances different from those at bar, be far different in meaning from the word “outer,” it is our view that when applied to the facts at bar there could be no possible patentable distinction by reason of this fact between the claims in controversy and the counts of the interference, the remaining language of the claims being almost identical with the language of the counts as heretofore pointed out. We therefore approve this ground of rejection on the part of the board with respect to claims 15 and 16.
We next consider claims 5 and 6, of which claim 5 is illustrative and reads as follows:
5. An electrical device comprising a glass vessel, an aperture leading from said vessel and defined by a bollow boss on a wall thereof, and a metal contact member comprising a cylindrical portion, serving as a contact proper and adapted to connect with a lead-in conductor, and a liollow enlarged portion with sides flaring toward said vessel, said flaring sides engaging and fitting .over said boss, the engaged, sides of said, ,6oss being formed, so thick that the hole through the enlarged portion is constricted to close proximity with the associated lead-in conductor, in order to make a strong union, while allowing for the reception of said conductor, and the wall of said vessel extending outwardly directly from the edge of said flaring sides. [Italics ours.]
*1157We have italicized in claim 5 what we believe to be the critical limitation in both these claims, viz, that respecting the relationship between the aperture of the boss and the associated lead-in conductor. The Patent Office tribunals have held that the disclosure of the Kich-ardson parent patent was insufficient to support such a limitation, chiefly because figure 1 of the patent was not shown in cross section to illustrate the connection between, or the relationship between, the contact member, the aperture, and the lead-in conductor.
Appellant argues that when figure 1 of the parent patent is taken with figure 3 thereof, showing the hollow interior of the contact member, it is obvious to anyone skilled in the art what the relationship is between the lead-in conductor and the opening through the boss, but that in any event the claims do not claim the lead-in conductor fer se as a specific limitation. Appellant makes the further argument that claims 5 and 6 are almost verbatim copies of claims in his divisional application, referred to above, and that the fact that the Patent Office allowed entry of these claims by amendment in the divisional application constitutes an admission that they read on the parent case. (The same argument is advanced in connection with claims 1 to 4, inclusive, considered later herein, which claims were also copies of claims entered by amendment in the divisional application, but for reasons which ane obvious we do not think this argument is entitled to much weight.)
These claims, 5 and 6, were rejected by the tribunals below on the ground that they did not read upon appellant’s parent patent and were not patentable over the Wright patent No. 2,098,080. We agree with his ground of rejection because it seems clear to us that in the Bichardson parent patent the lead-in conductor is shown only reaching to the top of the thimble, and there is no showing as to what its relationship is with the aperture of the boss. In the instant application, appellant has clearly disclosed the relationship as being a continuation of the conductor into the hollow contact pin, in close proximity with the inside walls of the boss. We agree with the examiner, who said, in substance, that one is left to conjecture (it not being shown in the drawings or explained in the specification of the parent patent) just what the relationship was between the lead-in conductor and the interior of the boss. It is conceivable that a connection other than that disclosed in the present application might have been contemplated in the parent case. Appellant argues, in effect, that anyone skilled in the art could understand how this connection was made. But, it is not sufficient that one skilled in the art might know how to make the. connection. This court has so held in many cases. In Brand v. Thomas, 25 C. C. P. A. (Patents) 1053, 96 F. (2d) 301, 37 USPQ 505, where it was contended that one skilled in the art would *1158understand bow to construct the device in accordance with tbe count, we said:'
* * * Lack of clear disclosure is not supplied by a speculation as to wliat one skilled in the art might do or might not do if he followed the teaching of the inventor. The disclosure should be clearer than to suggest that one skilled in the art might construct the device in^a particular manner. * * * [Italics quoted.]
We therefore agree with the tribunals below in their holding as to claims 5 and 6.
The next group for our consideration consists of claims 7 to 10, inclusive. Appellant makes no contention that these claims read on his parent patent. Claim 7 seems to be sufficiently illustrative and reads as follows:
7. An electrical device comprising a vitreous vessel, an aperture leading from said vessel and defined by a hollow boss on a wall thereof, and a metal contact member comprising a cylindrical portion serving as a contact proper and formed hollow for the reception of a lead-in conductor, and a hollow enlarged portion with sides flaring toward said vessel and tapering in thickness to a sharp edge, said flaring sides engaging and fitting over said boss with the sharp edge portion sealed, in said, wall, the engaging sides of said boss being formed so thick that the hole through the enlarged portion is constructed to close proximity with the associated lead-in conductor, in order to make a strong union, while allowing for the reception of said conductor, and the wall of said veásel extending outwardly directly from the outer surface of said flaring sides. [Italics ours.]
Each claim in this group has the limitation requiring that the sharp edge of the thimble be “sealed in,” that is, embedded in, the glass wall of the base, and this limitation clearly will not read on the parent disclosure, as the Patent Office tribunals have held. The examiner rejected all these claims as “clearly unpatentable” over Wright No. 2,098,080, and the board held that claim 7 was unpatentable over the said Wright patent and that claims 8, 9, and 10 actually read on that patent.
It seems to us that these claims, without too strained a construction, could perhaps be read on the Wright patent No. 2,098,080. In any event, whatever limitations of the claims might not, under a strict construction, be supported literally by the Wright disclosure are clearly not distinguishable from Wright in a patentable sense. For example, each of the claims has the limitation requiring that the contact member be “formed hollow for the reception of a lead-in conductor.” We note that the upper part of Wright’s prong member is formed hollow for the reception of a lead-in conductor, and it would surely not be patentable to make the contact member hollow throughout its length, as in appellant’s case. We therefore approve the holding of the board with respect to claims 7 to 10, inclusive.
*1159This leaves for our consideration claims 1 to 4, inclusive, and 11. Claim 1 may be taken as an example:
1. An electrical device comprising a glass vessel, an aperture leading from said vessel and defined by a hollow, tapered boss extending from a wall thereof, and a metal contact member secured to said boss, said member comprising a hollow, cylindrical port,ion closed at its outer end, shouldered at its inner end, serving as the contact proper, and adapted to receive a lead-in wire from said vessel through said aperture, and an enlarged portion extending toward said vessel from said shouldered portion, the sides of said enlarged portion decreasing in thickness from said shouldered portion to a sharp edge, said enlarged portion flaring toward said vessel, fitting over said tapered boss and sealed only to the exterior thereof, the sharp edge of said enlarged portion engaging said wall which is continued outwardly therefrom at an angle to said contact member so as to effect a gas-tight seal. [Italics ours.]
All the claims in this group have a limitation requiring that the sharp edge of the thimble engage or contact the wall of the glass base. While appellant has made numerous arguments as to this group, the chief point of disagreement between him and the Patent Office tribunals appears to be with respect to the proper construction of the term “wall.” The board and the examiner construed the downwardly bulging portion, i. e., that portion sloping downwardly at a slight angle from the horizontal portion of the wall and connecting with the boss over which the thimble fits, as a part of the boss, while appellant contends that it is part of the wall proper.
In order to understand the nice points involved, we think it essential that figure 4 of 'the instant application,, which, except for numerals, is identical with figure 4 of the parent case, should be here reproduced. [See p. 1180.]
It was the position of the examiner that the claims now under discussion (except claim 4) do not read on the parent case, the theory being that in the parent case the edge of the thimble is not shown to engage or contact the base wall. He rejected claim 1 on the Wright patent No. 2,098,080 and additionally, “If it be held that the claim is readable on the parent case, then the rejection is given” on the Wright patent No. 2,069,638 in view of Houskeeper or Kruh et al. Claims 2 and 3 were held by the examiner not to be patentable over Wright No. 2,069,638 in connection with Kruh et al., and then the examiner stated:
Furthermore, as all the counts lost to applicant are prior art, it is held these claims disclose nothing patentable over tlie same in view of the shoulder disclosure of No. 1,867,852 * * *.
As to claim 4, the examiner held that it was “for structure shown in the parent case,” but he rejected it as “unpatentable over Wright No. 1,967,852 in connection with” Houskeeper. It should be pointed out here that the filing date of Wright No. 1,967,852 was December 8,1933, whereas the application for the Richardson parent patent was filed *11602 months earlier, on October Y, 1933. The examiner rejected claim 11 on Wright No. 2,098,080 and further said that “If it be held that this claim 11 reads on the parent case, then it is rejected as met by Wright No. 2,069,638” in connection with Kruh et al., Houskeeper, and Madine.
The Board of Appeals approved the examiner’s holding that claim 1 does not read on the parent case and stated that “it is our view that
claim 1 is unpatentable over the Wright patent with which this application was in interference.” It disagreed with the rejection on the Wright patent No. 2,069,638 in view of Houskeeper or Kruh et al. As to claims 2 and 3 the board stated that “for the same reasons these claims are considered unpatentable to applicant,” and in its second decision on petition for reconsideration the board said that claims 2 and 3 “are considered properly rejected on the Wright patent No. 2,098,080.” While agreeing with the examiner that claim 4 was di*1161rected to the modification shown in figure 4, the board disagreed that the claim was directed to structure shown in the parent case and said that “claim 4 is held to be unpatentable over the Wright patent No. 2,098,080.” As to claim 11 the board said in its first decision that “this structure is clearly not disclosed in applicant’s prior patent,” and in its second decision it held that-the claim reads on the Wright patent No. 2,098,080.
In concluding its first decision the board, speaking of all the claims stated:
Except for tlie rejection on the Wright patent No. 2,069,638 in combination with certain other patents, it is considered that the rejection of the claims for the reasons given by the examiner and herein is proper and the same is accordingly affirmed.
Thus we have a general affirmance of the examiner’s rejection of the claims now under discussion (1 to 4, inclusive, and 11) for reasons given by the examiner, with the exceptions above indicated. In substance, the position of the board appears to be that none of these claims will read on the parent case and that they are not patentable over the Wright patent No. 2,098,080.
The holding, therefore, of the Patent Office tribunals with respect to the claims now under consideration is that the disclosure of the parent case, as illustrated by figure 4, supra, does not support the limitations of the claims, such as that found in claim 1, “the sharp edge of said enlarged portion engaging said wall.”
It will be observed from figure 4 that numeral 15' represents the bottom wall of the glass cup or lamp bulb base 10' (shown in fragment) and that the skirt or thimble 29 extends up on the boss 81 to the sloping conical portion 40, which extends outwardly from the edge of the thimble and connects with the horizontal portion of the wall 15'. It is the holding of the Patent Office tribunals that the thimble engages only the boss 31 and not the wall. In other words, they have held that the conical portion 40 is a part of the boss 31 and not a part of the wall 15/ On the other hand, it is the contention of the appellant that the sloping portion 40 above and attached to the edge of the thimble is a part of the wall; that the wall, instead of being horizontal throughout its area, slopes downwardly around the apertures at the portion 40 and ends in the boss 31.
It is our view that the sloping portion 40 is a part of the bottom wall of the glass cup, and that the mere fact that it slants downwardly and ends in the boss does not prevent the claims from reading on figure 4. The Patent Office tribunals have given no satisfactory reasons why the portion 40 should be construed as a part of the boss rather’ than of the wall, whereas appellant, on the other hand, points to the *1162following quotation from the specification of his parent patent in support of his position:
It is an object of tbe present invention to simplify tlie method of attaching contacts to a ~buVb wall.
Another object of the invention is to provide a method whereby a more uniform and effective union between a metallic edge and a glass wall is attained. Another object of the invention is to provide a method whereby a plurality of contacts may be simultaneously sealed to a bulb wail. [Italics not quoted.]
Since the board rejected the claims now under discussion upon the theory that they did not read upon the parent case, and since we differ with the board' in this regard, we do not have its views as to whether or not the claims are unpatentable over the issues of the interference. Perhaps its language “it is our view that claim 1 is unpatentable over the Wright patent with which this application was in interference” is a holding that, or could be construed as a holding that, in any event, that claim is unpatentable over the issues of the interference. The language used would seem to have the meaning that the Wright patent, since it was held that the claims did not read on the Eichardson parent case, is to be regarded as prior art. Furthermore, appellant, in his reason of appeal No. 21 (which was probably assigned in an abundance of caution) would seem to suggest that the general language used by the board might be interpreted as meaning that the claims now under consideration were unpatentable over the issues of the interference.
However, in view of our disagreement with the holding of the board that the claims under discussion do not read upon appellant’s parent case, and in view of the fact that the board reached its conclusions on that theory, we think it proper to reverse as to these claims and remand the case to the board with instructions to clarify its decision in this regard in order to permit the question to be squarely raised as to whether or not, in view of our holding that appellant’s disclosure in his parent patent supports these claims, they should be rejected under any doctrine relating to the disposition of claims upon the termination of an interference.
It may be said at this point that we have found no case where the exact facts involved in the case were present. Here we have a situation where the appellant lost priority in an interference writh a patentee, involving counts which contained a limitation relating to the embedding of the edge of the skirt or thimble within the glass of the base wall — a limitation * which was not disclosed in appellant’s parent patent — and then a rejection by the Patent Office tribunals of claims which, we hold, read upon the parent case, on the ground that they are not inventive over the Wright patent with which the instant application was in interference.
*1163It is to be recalled that Wright, in the interference, made no application for reissue of his patent so as to include claims directed to the same subject-matter as the instant claims for the purpose of interference. Whether or not he might have done so, under all the circumstances then existing, we, of course, are not prepared to say. Appellant urges, however, and we are not disposed to disagree with him, that under the existing practice in the Patent Office there was no possible way in which he could secure a contest of priority with the Wright patent with respect to the particular subject-matter of these claims, as to which, the present record would seem to indicate, he was senior. Query: was it the intent of the board to hold that, having lost claims of the character above described and under the circumstances stated, appellant should be denied, on the ground of éstoppel, claims as to which, upon the record, he is shown to be the first inventor, even though it be held that the claims are not inventive over the subject-matter of the interference and even though it be held that they are of the same scope as, or of narrower scope than, the counts of the interference?
It follows from the foregoing that the decision of the board is affirmed as to claims 5, 6, 7, 8, 9, 10, 15, and 16, but reversed as to claims 1, 2, 3, 4, and 11, and the case is remanded for the purposes stated.