This appeal is from the decision of the Patent and Trademark Office (PTO) Board *1350of Appeals (board) affirming the final rejection of claim 3, the sole claim in application serial No. 110,599, filed January 28, 1971, entitled “Water Insoluble Nitrogen-Containing Polyols.”1 We reverse.
The Invention
Appellants’ invention is directed to a water-insoluble polyol (a poly-hydroxyl compound) which has sufficient self-catalytic activity to react, without the need of extraneous catalysts, with organic polyisocyanates to form rigid polyurethane foams. It is asserted that the foams produced from the polyols of the present invention are characterized by greater ease of fire retardancy and good dimensional strength when extraneous fire retardants are employed. Claim 3, the sole claim on appeal, reads as follows:
3. A water-insoluble polyol having the property of self-catalyzing reaction with organic polyisocyanates to form rigid polyurethane foam said polyol having the formula
[[Image here]]
As will become evident below, it is essential that the two reactions required to produce the claimed polyol be described. In the first reaction, which can generally be described as a Mannich reaction,2 para-nonylphenol, diethanolamine, and formaldehyde are reacted in a molar ratio of 1:2:2; the predominant component of the resulting MRP will be a pentol (a compound with five hydroxyl groups). The second reaction, which appellants denominate as a propoxylation reaction, involves reacting propylene oxide with the MRP in a molar ratio of 3:1. The predominant product of this reaction will be the claimed compound, viz., a pentol with three degrees of propoxylation.3
Prosecution History and the Rejection
The examiner finally rejected claim 3 on three separate grounds: (1) under 35 U.S.C. § 112, second paragraph, as being indefinite; (2) under 35 U.S.C. § 103, as being obvious over Bruson et al., U.S. Patent No. 2,998,452, issued August 29, 1961; and (3) under 35 U.S.C. § 102(b), as being anticipated by Edwards et al., U.S. Patent No. 3,297,597, issued January 10, 1967, on an application filed May 23, 1966; the grandparent of this application is serial No. 288,474, filed June 17, 1963, which is the same application from which appellants’ application originated (i. e., appellants’ grandparent application). With respect to the § 102(b) rejection, the examiner was of the opinion that neither the grandparent nor parent applications provided a written description (35 U.S.C. § 112, first paragraph) of the claimed compound; consequently, appellants’ claim that they were entitled to their earlier filing dates under 35 U.S.C. § 120 was denied. By restricting appellants to their actual filing date (January 28, 1971), Edwards et al. (which issued January 10, 1967) constituted a statutory bar.
*1351The board reversed the first two grounds of rejection but affirmed the § 102(b) rejection. Since the parent application was filed less than one year after the Edwards et al. patent issued, the board correctly concluded that if appellants were entitled to rely on its filing date, Edwards et al. (same inventive entity as appellants) would be removed as a reference under § 102(b). In view of appellants’ concession that Edwards et al. did, in fact, disclose the claimed polyol, the sole remaining issue was whether the parent application provided a written description (35 U.S.C. § 112, first paragraph) of the claimed polyol. Concluding that it did not, the board stated:
[Tjhere is no description in said parent of the invention claimed in the present case. The only disclosure of nonyl phenol (required by claim 3 as the phenolic component) appears . . . with about twenty-five other phenols. While it might be obvious to combine [this] disclosure . . . and Examples III or VIII, [that is to say] select nonyl phenol out of a list of twenty-five phenols, and then combine with propylene oxide in an amount sufficient to obtain the pentol of claim 3, we cannot agree that such selection and combination is equivalent to a “written description” of the claimed invention.
We note that a recent line of CCPA cases have significantly tightened up on the application of the “written description” requirement of 35 USC 112, first paragraph; see In re Ruschig, 54 CCPA 1551, 379 F.2d 990, 154 USPQ 118 (1967); Fields et al v. Conover et al, 58 CCPA 1366, 433 [443] F.2d 1386, 170 USPQ 276, 279-80 (1971) and In re Smith, 458 F.2d 1389, 173 USPQ 679, 683 (CCPA 1972). [Emphasis in original.]
Issue
The dispositive issue is whether appellants’ parent application, serial No. 682,560, filed November 13, 1967, complies with the written description requirement of 35 U.S.C. § 112, first paragraph, vis-á-vis the subject matter of the appealed claim; if it does, then the claim is entitled to the filing date of the parent application under 35 U.S.C. § 120, In re Smith, 458 F.2d 1389, 59 CCPA 1025, 173 USPQ 679 (CCPA 1972), and Edwards et al. is removed as a reference.
OPINION
While appellants argue that both the parent and grandparent applications provide an adequate written description of the claimed compound, in our view it is unnecessary to decide whether the grandparent application complies with the description requirement. Appellants filed their parent application within one year of the effective date (issue date) of the only reference— their own patent, and, as such, are within the one-year grace period allowed by § 102(b). Cf. In re Gibbs, 437 F.2d 486, 58 CCPA 901, 168 USPQ 578 (1971).
Turning to the parent application, appellants assert that it, by virtue of providing an adequate written description of the aforementioned reactions, provides an adequate written description of the claimed polyol. That these reactions will produce, as the predominant product, the claimed polyol, is not in dispute. The board, however, took the position that the parent does not provide an adequate description of the two reactions; specifically, that it provides neither direction for selecting, as the phenolic reactant, para-nonylphenol, nor direction for choosing a propylene oxide/MRP molar ratio of 3:1.
The function of the description requirement is to ensure that the inventor had possession, as of the filing date of the application relied on, of the specific subject matter later claimed by him. E. g., In re Blaser, 556 F.2d 534, 194 USPQ 122 (CCPA 1977); In re Wertheim, 541 F.2d 257, 191 USPQ 90 (CCPA 1976); In re Smith & Hubin, 481 F.2d 910, 178 USPQ 620 (CCPA 1973). To comply with the description requirement it is not necessary that the application describe the claimed invention in ip*1352sis verbis, In re Lukach, 442 F.2d 967, 58 CCPA 1233, 169 USPQ 795 (1971); all that is required is that it reasonably convey to persons skilled in the art that, as of the filing date thereof, the inventor had possession of the subject matter later claimed by him. See In re Driscoll, 562 F.2d 1245, 195 USPQ 434 (CCPA 1977). In the context of the present case, this translates into whether the parent application provides adequate direction which reasonably leads persons skilled in the art to the later claimed compound. See Flynn v. Eardley, 479 F.2d 1393, 178 USPQ 288 (CCPA 1973). By the very nature of this inquiry, each case turns on its own specific facts. See In re Driscoll, supra.
As the board apparently recognized, the description in the parent is not intrinsically defective merely because appellants chose to describe their claimed compound by the process of making it; our primary concern is whether the description requirement has been complied with, not the mode selected for compliance. Cf. In re Smith & Hubin, 481 F.2d at 914, 178 USPQ at 624. It is undisputed that the aforementioned reactions will inherently produce, as the predominant component, the claimed compound. Further, the parent discloses that:
Although it is within the scope of the present invention to separate the crude . [MRP] by conventional means into specific components or fractions, it is a feature of the present invention that the entire crude Mannich reaction product may be used as such without attempting to isolate the individual components thereof. [Emphasis added.]
The parent application, therefore, recognizes that, if desired, conventional means can be used to separate components of the MRP and, ostensibly, of the final product. While it is true, as stated in the dissenting opinion, that in the preferred embodiment the parent does not separate the components, this does not negate the express disclosure that such separation is “within the scope” of the parent invention; if such express language does not evidence “possession,” then nothing does. Thus, on the facts of this ease, an adequate description of the aforementioned reactions is, concomitantly, an adequate description of the claimed compound. This should not be construed as meaning that if an application adequately describes a process which, inherently, will produce a compound, then it necessarily adequately describes the compound: Each case must be decided on its own facts.
Example III, referred to by the board, discloses reacting phenol, diethanolamine, and formaldehyde in a molar ratio of 1:2:2; propylene oxide is then reacted with the resulting MRP in a molar ratio of 4.01:1. With respect to example III, we have noted that in their briefs, both appellants and the solicitor indicate that example III uses 3.6 moles of propylene oxide per mole of MRP; this is incorrect. Example III reacts 21.7 moles of propylene oxide with 5.41 moles of MRP, thus giving a molar ratio of 4.01:1. Original claim 2 of the parent application, which is part of the original disclosure, In re Gardner, 475 F.2d 1389, 177 USPQ 396, rehearing denied, 480 F.2d 879, 178 USPQ 149 (CCPA 1973), and to which the board made no reference, claims a polyol produced by reacting 1-7 moles of propylene oxide with one mole of the MRP of phenol or nonylphenol, an alkanolamine, and formaldehyde, reacted in a molar ratio of from 1:1:1 to about 1:3:3. The alkanolamine is selected from alkanolamines having the formulae:
[[Image here]]
where R is hydrogen or C1-C4 alkyl, R’ is hydrogen, C1-C4 alkyl or -(CHR)n-OH, and n is a positive integer having a value of 2 to 5. Diethanolamine is a species which falls within these generic formulae. Moreover, of the eight working examples in the parent which describe making various polyols, all eight use diethanolamine as a reactant. This provides adequate direction for selecting diethanolamine as the alkanolamine in claim 2. Thus, claim 2 recites a process for producing a genus of compounds which includes both the predominant compound pro*1353dueed by example III and the claimed compound. More importantly, by claiming the phenolic reactant as a Markush group consisting of phenol or nonylphenol, it is generally understood that appellants are asserting that these two members are alternatively usable for the purposes of the invention, and therefore, regardless of which is chosen, the resulting compound produced by the overall process will exhibit the disclosed utility. See In re Driscoll, supra; see generally In re Skoll, 523 F.2d 1392, 1397, 187 USPQ 481, 484-85 (CCPA 1975). We view this as providing adequate direction for those skilled in the art to substitute nonylphenol for phenol in example III. The solicitor concedes as much in his brief; to conclude otherwise would make the statement in In re Lukach, 442 F.2d at 969, 58 CCPA at 1235, 169 USPQ at 796, that “the invention claimed does not have to be described in ipsis verbis in order to satisfy the description requirement of § 112,” meaningless.
Turning to the propylene oxide/MRP mole ratio in the second reaction, the solicitor’s position is that, at best, the parent application describes the use of from 1 to 5 moles of propylene oxide per mole of MRP, and that “there is no specific teaching of . reducing the amount of propylene oxide to exactly 3 moles as required by the claimed compound.” At oral argument, appellants asserted that while a propylene oxide/MRP molar ratio of 3 would, of course, produce the claimed compound, a somewhat larger ratio, such as the 3.6 used in example III, would also produce the claimed compound. As was previously shown, example III does not disclose the use of a molar ratio of 3.6. Moreover, in their brief before us, appellants stated that a molar ratio of 3 is required to produce the claimed compound. Therefore, for purposes of this appeal, we will assume that exactly 3 moles of propylene oxide per .mole of MRP is required to produce the compound of appealed claim 3 (the board, in reversing the § 103 rejection over Bruson et al, and the solicitor also stated that 3 moles is required).
To determine whether the parent application provides adequate direction for using the required propylene oxide/MRP molar ratio, an understanding of the underlying reactions is essential. Broadly stated, the parent application discloses first reacting a phenolic compound with an alkanolamine and formaldehyde, in molar ratios of from 1:1:1 to about 1:3:3, to produce an MRP. The resulting MRP potentially can contain three different types of reactive positions: phenolic hydroxyl group, free amino hydrogen atom, and primary hydroxyl group. In the second reaction, alkylene oxide (propylene oxide is disclosed as being preferred) will react with any of these three positions. The molar ratio used in the first reaction will determine whether the MRP is a triol, pentol, etc.; the molar ratio used in the second reaction will determine the degrees of propoxylation of the final product.
Applying this to example III, since the first reaction uses a molar ratio of 1:2:2 (phenol: diethanolamine: formaldehyde), the predominant MRP and the predominant final compound, like the claimed compound, will be a pentol; however, since example III uses a propylene oxide/MRP molar ratio of 4.01:1, the pentol will have four degrees of propoxylation, whereas the claimed pentol has three degrees of propoxylation. With respect to the degree of propoxylation, the parent application discloses that:
In accordance with the present invention, the Mannich reaction product is reacted with an alkylene oxide to provide the final polyol. The nitrogen present in the Mannich condensate [MRP] has sufficient catalytic activity to promote the reaction of one mol of the alkylene oxide with each free amino hydrogen atom and phenolic and primary hydroxyl group and no additional catalyst is needed . For example seven mols [the stoichiometric amount] of propylene oxide will add to the Mannich product prepared from a molar ratio of 1:3:3 of phenol, diethanolamine and formaldehyde to give a heptol
*1354It is, of course, possible to add less than one mol of alkylene oxide per free phenolic and primary hydroxyl group in the Mannich condensation product. The minimum desirable amount of alkylene oxide is one mol per free amino hydrogen atom and phenolic hydroxyl group . Generally, more than the minimum amount of alkylene oxide is used to obtain a product having a lower hydorxyl [sic, hydroxyl] number and lower viscosity. For example, a desirable product is that obtained by the addition of five mols of propylene oxide (rather than the maximum of seven or minimum of one) to the heptol obtained by the Mannich condensation of phenol, formaldehyde and diethanolamine in a molar ratio of 1:3:3. [Emphasis added.]
The first reaction in example III, as well as the first reaction required to produce the polyol of appealed claim 3, will produce, as the predominant MRP, a compound which has one phenolic hydroxyl group, no free amino hydrogen atoms, and four primary hydroxyl groups. With this in mind, we believe the above quoted disclosure will provide those skilled in the art with adequate direction for concluding that, in example III, from 1 (but preferably more than 1) to 5 moles of propylene oxide can be reacted with each mole of MRP and that, most importantly, the polyol produced will have the disclosed utility; ergo, it provides adequate direction for using three moles of propylene oxide in example III.
When viewed in the context of what the parent application actually describes, the PTO has, in effect, done nothing more than argue lack of literal support. The burden of showing that the claimed invention is not described in the application rests on the PTO in the first instance, and it is up to the PTO to give reasons why a description not in ipsis verbis is insufficient. In re Salem, 553 F.2d 676, 682, 193 U.S.P.Q. 513, 518 (CCPA 1977); In re Wertheim, 541 F.2d at 265, 191 U.S.P.Q. at 98. Stating, as the board did, that in a recent line of cases this court has “significantly tightened up” on the written description requirement, is no substitute for such reasons. Parenthetically, with respect to the board’s perception of this court’s past cases, suffice it to say that each case must be decided on its own facts, see, e. g., In re Driscoll, supra, and that precedential value of prior cases is, therefore, extremely limited.
In conclusion, we hold that, as a factual matter, the parent application, taken as a whole, reasonably leads persons skilled in the art to the reaction of para-nonylphenol, diethanolamine, and formaldehyde, in a molar ratio of 1:2:2, and to the reaction of propylene oxide with the resulting MRP, in a molar ratio of 3:1, and, concomitantly, to the claimed compound. Accordingly, since claim 3 is therefore entitled to the benefit of the filing date of the parent application, we reverse the § 102(b) rejection of this claim.
REVERSED.
. The application is a continuation-in-part of parent application serial No. 682,560, filed November 13, 1967, which, in turn, is a continuation-in-part (the examiner had required it to be denominated as such and appellants, while referring to it below as a continuation, have, in their brief before us, referred to it as a continuation-in-part) of grandparent application serial No. 288,474, filed June 17, 1963.
. The Mannich reaction can be generalized as the linking of a carbanion site (enolate or phenolate) with an aldehyde and an amine. The following is the general reaction scheme:
[[Image here]]
Hereinafter, the product of this reaction will generally be referred to as the MRP (Mannich reaction product).
. The product will, in reality, be a mixture of polyols each having various degrees of propoxylation; the predominant component will, however, be the claimed polyol.