This appeal is from the decision and judgment of the United States Customs Court, First Division, 66 Cust. Ct. 202, 324 F. Supp. 1113, C.D. 4193 (1971), sustaining appellee’s protest against the classification of “clinical dextran” under TSUS item 439.50 as synthetic drugs and ordering reliquidaition thereof under TSUS item 439.30 as advanced natural drugs. We reverse.
The Merchandise
The importation at bar consists of “clinical dextran,” which is produced from “native dextran,” as it is called, by a hydrolysis process in which hydrochloric acid serves as a catalyst. Native dextran is produced by bacteria; it consists of a mixture of polymeric chains of various lengths made up from a “glucose unit” having the following structural formula:
Its empirical formula is (OgHio05)h, wherein n represents the number of units in any particular chain. According to appellee’s principal witness, Dr. Kirsti Granath of Sweden, commercially produced native dextran includes chains of “All possible sizes from a few chains of a *198few glucose units until units of millions in one chain.”1 She further testified that the hydrolysis is designed to produce chains of the proper length for medicinal use, having an average molecular weight of 70 to 75 thousand and a range in weight between 25 and 120 thousand.2 About 10% of the native dextran is naturally of the weight used as clinical dextran, but there is no commercially feasible physical or mechanical way to separate it from the rest.
The TSUS Items Involved
Schedule 4, Part 3, Subpart C. — Other Drugs, reads, in relevant part:
Drugs, not provided for in Subpart A or B of this part:
Natural drugs, crude or advanced:
439.10 Crude
439.30 Advanced
439.50 Other, including synthetic drugs
The Customs Court in sustaining the protest classified the subject importations, under TSUS item 439.30 as an advanced natural drug. The Government contends that it should be classified, as it originally was, under TSUS item 439.50 as a synthetic drug. The relevant TSUS headnotes are quoted in extenso in the opinion of the Customs Court; because of the view we take of the case, the only issue which we need discuss is whether the hydrolysis process used in converting the native dextran to clinical dextran made a change in the “molecular structure” of the native dextran.3 If it did, the Government prevails; if it did not, appellee prevails.
Opinion Below
The point of departure for the lower court’s opinion in this case was its previous opinion in Chemirad Corp. v. United States, 65 Cust. Ct. 137, C.D. 4067 (1st Div. 1970), and it is therefore necessary to discuss the opinion in that case before turning to the lower court’s *199opinion in this. In Ohemimd the Customs Court held that a mixture of polymeric chains of various lengths and possessing various degrees of branching, but all made up from the same repeating unit, was a single compound rather than a mixture of compounds. It stated that it did so because,
While there are indications in the proof that the activity of the polymer is dependent to a certain extent on the complexity of the structure, there is insufficient evidence to indicate that the importation consists of commingled units each with distinctly identifiable and distinguishable characteristics. It may be that future research will reveal that the individual and different lengths of molecules of the same polymers are as distinct and different as salt and water, in which ease their combination will be known as a mixture in accordance with the above definition [Headnotes 3(a) and (b) of Schedule 4], Presently, however, such is not the case and the evidence of record does not permit such a conclusion.
In the present case, the court reasoned that, if a mixture of polymeric chains all built up from the same repeating unit is a single chemical compound regardless of variations in length and branching of the molecular chains, then “the cutting of a polymer molecule so that the long chain is transformed into shorter chains does not make it a different chemical compound * * *.” Putting that in the statutory language, the court held that:
The only difference between “native dextran” as produced by the bacteria and the final product of “clinical” dextran, is the length of the molecular chain. Their chemical constitution and molecular structure are identical. [Emphasis oursJ
The lower court discussed three of our cases, which we discuss below, interpreting paragraph 5 of the Tariff Act of 1930, a general provision for medicinal preparations, and paragraph 34 thereof, which provided for natural, uncompounded drugs advanced in condition. It distinguished all three on the ground that the chemical reaction involved in each produced a change in the molecular structure of the starting material. In the first two of these cases we held that the drug involved had been “compounded” — i.e., that it had been synthesized from a. naturally occurring starting material by one or more chemical reactions. In the third, we did not reach that question.
In Chemical Specialties Co. v. United States, 43 CCPA 93, C.A.D. 614 (1956), the imported 21-acetoxy pregnenolone was produced from a particular kind of yam by “a series of [unidentified] chemical steps” culminating in the acetylation of 21-hydroxy pregnenolone to produce the imported compound. The lower court here focused on the substitution of the acetate group for the hydroxy group, stating that it was-“clearly a change in molecular structure * * *.”
In Organo, Inc. v. United States, 50 CCPA 17, C.A.D. 812 (1963), the molecule occurring in nature was broken down by hydrolysis into *200the imported polymer and a protein molecule. The lower court characterized that process as “sever [ing] one distinct molecule from another * *
In Sandoz, Inc. v. United States, 57 CCPA 44, 418 F.2d 1396, C.A.D. 974 (1969), the majority specifically refrained from expressing an opinion concerning whether the imported digitoxin was “uncompounded.” However, the facts were that the molecule occurring in nature 4 was broken down by enzyme-induced hydrolysis into digitoxin, glucose, and acetic acid, and the Customs Court purported to “dis-tingush” that case on the ground that “the disqualifying chemical reactions therein were reactions which broke down a complex molecule into substances of different chemical identity.” In this case, the catalytic hydrolysis serves to break down naturally occurring dextran molecule into a plurality of smaller dextran molecules.
Opinion
Both of appellee’s witnesses testified that the hydrolysis involved here did not effect a structural change in the native dextran, although ■one went so far as to concede that, “of course,” it had effected a change in “the whole structure of a macromolecule * * However, he insisted that this was not a structural change in “the molecule” because the “Chemical structure of the units remain[s] unchanged except for the bridge between them” and that “The structure is the same because dextran of 100,000 molecular weight and dextran of 200,000 has exactly the same structure.” Appellant’s single witness did not dispute these statements, nor does appellant argue that judicial notice can ■somehow bridge any gap in its case left by lack of evidence, and we would thus be bound by this testimony if we thought it consisted of statements of fact. United States v. F. W. Myers & Co., 45 CCPA 48, 52, C.A.D. 671 (1958).
However, in our opinion the facts established by appellee’s evidence ■are that the polymers which are the starting material for the hydrolysis and the polymers which are the products of the hydrolysis are all composed of the same unit, the structure of which is unchanged by the hydrolysis (except for the units on either side of the point of cleavage). *201Whether or not these facts indicate that the clinical dextran molecules “have not had changes made in their molecular structure as found in nature” within the meaning of those words as found in Headnote 3 (a) to Part 3, Schedule 4, of the Tariff Schedules of the United States we take to be a question of law fully open to our review on appeal.
According to the United States Tariff Commission’s Tariff Glassifi-cation Study, Explanatory and Background Materials for Schedule 4,
Part 8 [of Schedule 4] brings together in systematic arrangement provisions from 20 paragraphs of the existing tariff schedules covering “nonbenzenoid” drugs and related products. * * *
The respective scopes of the provisions of paragraphs 5, 34, and 1969 are ambiguous and have been in issue on numerous occasions. The resulting administrative and judicial rulings have often been in conflict. A ruling in CAD 614 [Chemical Specialties Co. v. United States, supra] has had a settling effect on customs practices and it is the basis for the essential distinctions made in the provisions of headnote 3 between advanced natural drugs and synthetic drugs. [Id. at p. 85.]
Headnote 3(b) defines a “synthetic substance” as “a chemical compound made by the artificial combination of elements or radicals by any physical or chemical process,” and Headnote 3(a) specifies that “natural substances” are those which, inter alia, “have not had changes made in their molecular structure as found in nature.” We agree with the Customs Court that:
* * * the headnote definitions of “natural” and “synthetic” * » * [provide! a clear and complementary understanding of the dividing line between natural and synthetic substances which should serve well in interpreting those tariff provisions which rely on distinctions between natural and synthetic.
This distinction was derived from our opinion in the Ohemical Specialties case, and it was intended to be, we believe, the same one we drew there — i.e., a distinction between those substances which have and those substances which have not undergone chemical changes subsequent to their “natural” production.
The present case is apparently one of novel impression. As the lower court correctly pointed out, our three opinions, previously discussed,, interpreting paragraphs 5 and 34 of the Tariff Act of 1930 all dealt with chemical processes wherein the starting material differed from-the products by more than the number of identical units contained in each. However, we cannot agree with the lower court that this is a. difference compelling the conclusion that the starting material did not,, in the words of Headnote 3(a) of the statute, undergo a change in its-“molecular structure as found in nature.” As appellee’s witness-conceded:
If you speak of tbe structure of tbe chain molecules, the answer [to whether or not there has been a change in the molecular structure] of course, is yes, because you change the structure that you break down.
*202Moreover, in contrast to the situation obtaining in Chemirad Corp. v. United States, supra, there is an abundance of evidence in this case that the characteristics of the polymer are dependent on the length of its chains. If the molecular weight, which is a measure or indication of chain length, is too low, the polymer is quickly excreted by the kidneys and is therefore of little medical value, and, if the molecular weight is too high, the polymer is toxic to humans. The latter fact, obviously, is the reason for hydrolysing the native dextran in the first place. Under the circumstances, we think it unduly restrictive to say that, just because the empirical formula is the same but for the value of “n,” there is “no chemical difference” between the gigantic macromolecules produced by the bacteria5 and the many times smaller macromolecules which are medicinally useful. We therefore hold there has been change in the molecular structure as found in nature.
For the foregoing reasons, the judgment of the Customs Court is reversed.
Appellant’s witness testified that “U.S. Government specifications” were “approximately 75,000 plus or minus 15,000 * * *.” Since the molecular weight of each unit is 162, chains of clinical dextran would contain between 154 and 741 units each by appellee’s evidence and between 370 and 556 units each by appellant’s evidence. Appellee’s principal witness also testified that the average molecular weight of the native dextran manufactured by appellee is between 10 and 100 million, which means that the chains would contain between 62,000 and 620,000 units, on the average.
The relevant headnote indicates that a synthetic drug is one “made by the artificial combination of elements or radicals by any physical or chemical process * * (Emphasis supplied.) The parties seem to have assumed that this language encompasses drugs made by the artificial decomposition of larger molecules into smaller molecules. Compare Organon, Inc. v. United States and Sandoz, Inc. v. United States, both discussed infra.
The evidence there indicated that approximately 10 % of the commercially produced digitoxin was actually present in the growing leaf of the plant from which the drug was produced prior to the application thereto of the processes resulting in the production or the other 90%. However, as the Customs Court pointed out in its opinion below, the majority here “minimized” that fact, refusing to give it controlling significance. In any ■event, the facts of this ease are clearly distinguishable from the facts in Banüoz. There the naturally occurring digitoxin apparently could be separated from the leaves in which were found both it and the molecule which was the precursor of the other 90% of the ■imported digitoxin, whereas here there is no commercially feasible way of extracting the ■dextran of the proper weight from the native dextran prior to the hydrolysis.
The testimony of one of appellee’s witnesses was that they are “the biggest moleculels] In nature.”