Opinion for the court filed by Chief Judge MICHEL. Circuit Judge DYK dissents.
MICHEL, Chief Judge.McNeil-PPC, Inc. appeals rejections of claims as obvious. The Director of the U.S. Patent and Trademark Office (“PTO”) seeks dismissal of McNeil’s appeal as untimely. Because we determine that McNeil’s appeal is timely, we reach the merits and, because the Board of Patent Appeals and Interferences findings about the disclosures of a prior art patent application are not supported by substantial evidence, we reverse the claim rejections.
I. BACKGROUND
McNeil owns U.S. Patent No. 6,310,269, which claims a tampon for feminine hygiene with a solid fiber core in which the core is denser than the radially projecting “ribs” and the ribs are narrower at their bases than their distal ends, as shown in Figure 4 of the patent:
*1395[[Image here]]
McNeil requested that the PTO reexamine its '269 patent based on an unexamined Japanese application by Tetsu Sasaki, No. 55-168330 (“Sasaki”). Sasaki discloses
*1396[[Image here]]
maMng a tampon blank by stitching together layers of material (see Figure 6, above), and then molding (as shown in Figure 7, above) the blank into a finished tampon (see Figure 8, above). In light of Sasaki, the examiner rejected claims 1 and 3 as anticipated and claim 4 as obvious.
McNeil appealed to the Board, arguing that Sasaki did not disclose the relative densities or coarseness of the core and ribs on the tampon. McNeil also disputed that Sasaki disclosed ribs which were narrower at the base than at the proximal end. Because McNeil argued claims 1, 3, and 4 together, the Board selected claim 1 as representative. The Board affirmed the examiner’s rejection. Ex parte McNeil-PPG, Inc., No.2007-3158, 2007 WL 3325022 (B.P.A.I. Nov. 8, 2007). It specifically found that “Sasaki reasonably appears to depict a tampon having a generally cylindrical absorbent portion with a generally cylindrical compressed solid fiber core from which longitudinal ribs extend radially outward.” Id. at *3.
McNeil filed a request for rehearing, objecting to the Board making findings of fact in the first instance and arguing against the rejections on the merits. The Board denied the request.
The typed date on the face of the Board order denying McNeil’s request for rehearing is “May 30, 2008.” The mailing sheet for the order is dated June 2, 2008. The online “Transaction History” for the reexam has two entries for May 30, 2008: “Mail BPAI Decision on Reconsideration— Denied” and “Dec on Reconsideration— Denied.” J.A. 320. The image file wrapper (“IFW”),1 which is also available on*1397line, lists the “Mail Room Date” of the decision as June 2,2008. J.A. 296.
McNeil filed a notice of appeal on August 1, 2008. The Director responded that the appeal was untimely because it was filed more than sixty days after the Board made its final decision on May 30, 2008.2
II. DISCUSSION
A.
We address first, as we must, whether McNeil’s appeal is timely. Compliance with the PTO rules regarding the time for an appeal is required by statute, and while the Director can grant extensions of time, we cannot. See In re Reese, 53 C.C.P.A. 1099, 359 F.2d 462, 463 (1966) (per curiam). The parties agree that whether the Board decided McNeil’s motion for reconsideration on May 30 or June 2 determines whether McNeil’s appeal is timely.
Congress gave the Director some authority to set the time for filing an appeal:
When an appeal is taken to the United States Court of Appeals for the Federal Circuit, the appellant shall file in the Patent and Trademark Office a written notice of appeal directed to the Director, within such time after the date of the decision from which the appeal is taken as the Director prescribes, but in no case less than 60 days after that date.
35 U.S.C. § 142 (emphasis added). The regulations promulgated under this statute provide, in pertinent part:
The time for filing the notice of appeal to the U.S. Court of Appeals for the Federal Circuit (§ 1.302) or for cornmencing a civil action (§ 1.303) is two months from the date of the decision of the Board of Patent Appeals and Interferences. If a request for rehearing or reconsideration of the decision is filed within the time period provided under § 11.52(a), § 11.79(a), or § 11.127(d) of this title, the time for filing an appeal or commencing a civil action shall expire two months after action on the request.
37 C.F.R. § 1.304(a)(1) (emphasis added). Elsewhere, this regulation makes clear that the “two months” language is to simplify the calculation of the due date, making the due date the same date as the decision date but two months later (with certain exceptions, such as for the short month of February), as opposed to calculating exactly sixty days. Id. § 1.304(b). There is also a provision extending deadlines that fall on weekends or holidays, but none of the dates at issue on appeal are weekends or holidays. Id.
The crux of the matter is what the phrase “the date of the decision,” which appears in both the statute and the regulation, means. The second sentence of the regulation quoted above actually uses different phrasing to set the time to appeal from decisions on requests for reconsideration (i.e., “action on the request”), but the intent is that the same sixty-day rule (and not some shorter time period) applies after a denial of a request for reconsideration. See Amendment of Patent and Trademark Rules Concerning Judicial Review of Decisions of the BPAI, 54 Fed.Reg. 29,548, 29,550 (July 13,1989).
*1398Unfortunately, there is little that indicates whether we should or must attribute any meaning to the “May 30, 2008” date which appears on the first page of the Board’s decision. The Director did not explain the Board’s internal procedure for issuing opinions or whether the June 2, 2008 mailing date reflects the first time the decision was released to the public. McNeil, however, presented a declaration from Jeffrey Nase, a now-retired member of the Board from 1996 to 2006, who had been employed by the PTO continuously before that going back to 1973. The Director did not challenge any of Nase’s factual assertions.
The following is a summary of Nase’s declaration: Historically, the date the PTO mailed a document was the date that triggered any response period. It is unclear why the opinion says “Decided: May 30, 2008” but was not mailed until June 2— perhaps the mail room was slow, or perhaps one of the members of the Board panel decided to revise the opinion or reconsider it over the weekend of May 31 and June 1, 2008. Before August 2006, the date decided was hand-stamped on the front of opinions; beginning only in August 2006 did the Board decision have the date decided typed on the front. Nase opines that if the PTO had intended this minor formatting change to have substantive effect, there would have been public notice (which there was not).
The PTO argues that both 35 U.S.C. § 142 and 37 C.F.R. § 1.304(a)(1) state that the time to appeal runs from “the date of decision,” and the opinion says “Decided: May 30, 2008,” so it would be contrary to the language of the statute and the regulation for us to deem the date of decision to be the date the order was mailed. While this argument has some superficial appeal, the inner workings of the agency are far more important than the PTO admits. Based on the evidence before us, it appears that only when an opinion is released to the public (or at least the parties) is it truly decided — until then, it is possible for the Board to decide to revise it.
The PTO’s position is supported by the entries on the “Transaction History” page indicating that the case was decided and the opinion mailed May 30. However, the parties apparently agree that the decision was not actually mailed on May 30, but on June 2, making the “Transaction History” inaccurate, and the IFW lists June 2 as the mail room date.
The PTO also relies on an unpublished order, Barbacid v. Brown, 223 Fed.Appx. 972 (Fed.Cir.2007), in which we held that one party to an interference, Barbacid, had filed an untimely appeal after the Board mailed its decision to the other party but not to Barbacid. Barbacid appealed as soon as he became aware of the Board’s decision, but we stated, “The time is not measured from the date of receipt of the Board’s decision but from the date of the decision itself.” Id. (emphasis added). We did not, however, specify what “the date of the decision” was, and Barbacid had appealed late enough that an extra day or three was not the difference between his appeal being timely or untimely — -unlike the situation faced by McNeil.
Because the “Transaction History” page appears inaccurate, and because Nase’s declaration provides the most plausible explanation for the conflicting evidence of when the Board took action on McNeil’s request for reconsideration and is corroborated by the IFW entry, we find that the Board issued its decision on McNeil’s request on June 2, 2008, and that the Board therefore decided this case on June 2, 2008, not May 30. McNeil’s appeal is therefore timely.
B.
Turning to the merits, we must determine whether substantial evidence *1399supports the Board’s determination that Sasaki discloses a tampon with the characteristics claimed by McNeil. See In re Graves, 69 F.3d 1147, 1151 (Fed.Cir.1995). McNeil’s claim 1 reads:
1. A tampon for feminine hygiene, comprising
a generally cylindrical absorbent portion having a generally cylindrical compressed, solid fibre core from which longitudinal ribs extend radially outward,
wherein each of the ribs has a proximal end attached to the fibre core, each of the ribs is compressed less than the fiber core, thereby having a coarser capillary structure than the fibre core,
each of the ribs is separated from adjacent ribs at the proximal end by an amount greater than such rib is separated from an adjacent longitudinal rib proximate the distal end.
(Formatting and emphasis added). The Board found that the underlined portions of claim 1 were disclosed by Sasaki, and McNeil challenges each of these findings on appeal.
The Board determined that Sasaki disclosed “a tampon wherein ‘each of the ribs is compressed less than the fiber core, thereby having a coarser capillary structure than the fibre core.’ ” McNeil, 2007 WL 3325022 at *4. It is not entirely clear, however, that the Board understood McNeil’s claim. In its initial decision, the Board stated, “We find that Sasaki’s Figures 8 and 10 reasonably appear to depict the ribs of its tampon as being compressed less at their distal ends than at their ends proximal to the drawstring at the center of the drawn fiber strips.” Id. at *3 (emphasis added). McNeil’s claim, however, states that the ribs are less dense than the core, not that the distal ends of the ribs are less dense than the proximal ends.
Furthermore, nothing in Sasaki indicates the relative compression of any portion of Sasaki’s tampon.3 Figure 6,
[[Image here]]
can be reasonably interpreted to show that *1400the junction of the three strips of absorbent material is compressed relative to the rest of the strips — Figure 6 shows the three strips as thinner at this junction than elsewhere. However, Figure 6 depicts a tampon blank, not a finished tampon. There is no evidence that this compression at the center of the tampon blank leads to a finished product with a core denser than the ribs. As shown in Figures 7 and 8, below,
[[Image here]]
the blank of Figure 6 is compressed into a cylindrical shape, and the PTO has not explained how this change in shape necessarily leaves the ribs less dense than the core. In its initial decision, the Board quoted a claim of Sasaki, which stated that “compressed masses [are] formed at equal intervals along the axial line direction on the outside circumferential surface of the tampon.” Id. at *3. However, Sasaki makes no comparison between these “compressed masses” and any core. In fact, it is not clear that Sasaki contains a “core” at all. Instead, Figure 8 is drawn in such a way that suggests Sasaki’s tampon is composed of six ribs and no core. There is not substantial evidence, indeed, no evidence, that Sasaki discloses ribs “compressed less than the fiber core” or “a generally cylindrical compressed, solid fibre core.”
In its appeal brief before the Board, McNeil stated that “two key elements of the tampon are ... absent from Sasaki, i.e., the relative rib compression and spacing from adjacent ribs.” Id. at *1. The PTO argues that this shows that McNeil waived arguments about Sasaki’s lack of a “generally cylindrical compressed, solid fibre core” and “coarser capillary structure than the fibre core.” The PTO’s reading of McNeil’s appeal brief is unduly parsimonious. McNeil went on to state that “Sasaki does not disclose, teach, or suggest a tampon, wherein ‘each of the ribs is compressed less than the fiber core, thereby having a coarser capillary structure than the fibre core,’ as required by claims 1 and 3.” J.A. 141 (emphasis added). Just as the Sasaki figures do not indicate the relative compression of the different portions of the tampon, the Sasaki figures completely lack any indication of the relative coarseness of different portions. And, as noted above, McNeil’s arguments regarding relative compression of the core and ribs encompass the issue of whether Sasaki dis*1401closes “a generally cylindrical compressed, solid fibre core.”
Lastly, turning to the issue of spacing of the ribs, Figure 8 shows a space between the bottommost ribs, and there is arguably some space shown between other ribs. However, because it is neither clear that Sasaki discloses a core nor which portions of Sasaki’s tampon the Board considered to be the ribs and which the Board considered to be the core, we cannot say that substantial evidence supports the Board’s determination that Sasaki discloses ribs separated from each other “at the proximal end by an amount greater than” than at “the distal end.”
The Board’s rejection of claim 1 as anticipated lacks substantial evidentiary support and is therefore reversed. Because claims 1, 3, and 4 stood or fell together, the rejections of claims 3 and 4 are also reversed.
CONCLUSION
For the reasons provided above, the decision of the Board is
REVERSED.
. The IFW system contains the official records of PTO proceedings. See Changes to *1397implement Electronic Maintenance of Official Patent Application Records, 68 Fed.Reg. 38,-611, 38613 (June 30, 2003) ("The Office plans to capture electronic images of all documents that form the record of patent application examination. These images will form the Official file of the application.”).
. We issued an order to show cause why McNeil's appeal should not be dismissed for lack of jurisdiction. After both parties responded, we ordered the parties to address the timeliness of McNeil’s appeal in their briefs on the merits.
. Figure 4 of McNeil's patent, in contrast, uses hatching to indicate the differing density between the core and rib portions of the tampon.