The Differentiation of Hemicelluloses - ACS Publications - American

cellulose” in tlie Forest Products Laboratory methods of analysis. A second part ..... makers of swords, hatchets, and other devices placed these ma...
0 downloads 0 Views 766KB Size
1190

INDUSTRIAL AND ENGINEERING CHEMISTRY

(3) Eckart, O., Seifensieder Ztg., 54, 82-3, 103 (1927). (4) Elsdon. “Edible Oils and Fats,” pp. 215-16, Benn, 1926. ( 5 ) Ibid., p. 218. (6) Glick, B. W., Cotton Oil Press, 4 (lo), 41-2 (1921). (7) Jamieson and Baughman, Ibid., 7 ( 5 ) , 29-30 (1923)

Vol. 24, No. 10

(8) Malowan, J., J. Oil & Fat Ind., 4, 127-30 (1927). (9) Sohwartze, E. W., Ibid., 3. 173-8 (1926). RECEIVED September 9,1932. Presented before the Division of Agricultural and Food Chemistry a t the S2nd Meeting of the American Chemical Society, Buffalo, N. Y.,Bugust 31 to September 4, 1931.

The Differentiation of Hemicelluloses L. F. HAWLEY,Forest Products Laboratory, Madison, Wis.,

AND

A. GEOFFREYNORMAN,

University of Wisconsin, Madison, Wis.

T

H E g e n e r a l conception Hemicelluloses fall into two groups: ( 1 ) pecialb the pob’uronide group, /hose incrusting substances not closely associated S O t h a t t h e r e s u l t s of c o n of the characteristics of ventional analytical methods t h e hemicelluloses with the cellulose, and containing in practically may be interpreted cornot changed greatly since has this group of plant c o n s t i t u e n t s all cases a uronic acid, and (2) those very intirectly in terms of actual conmutely associated with the cellulose and never stituents. was first described by Schulze (36). This worker, who gave containing a uronic acid. These two groups are AIETHOD OF DIFFERENTIATION the generic name, considered accurately separated in the Cross and Bevan An alkali extract of wood or the h e m i c e l l u l o s e s cellulose determination, provided that there is no of other p l a n t m a t e r i a l conrelated to cellulose in chemical Pretreatment Of the with either acid Or tsins hemicelluloses of two disc o n s t i t u t i o n , although less alkali before chlorination. The name “polytinct types, or rather, from two resistant to acid hydrolyzing uronide,” suggested by Candlin and Schryver, different sources. The major a g e n t s . F u r t h e r , he was of to all those in the Jr,?t group part will c o n s i s t of the inthe opinion that the hemicellushould be crusting material from the cell l o s e s w e r e intermediate subcontaining uronic acid, and for thc second a new wall and is largely polyuronide stances in the development of term “celldosan” is proposed* in nature; this material is idencellulose itself. S c h u 1z e Dretified only as “pentosans not in pared such compounds from a large number of plant materials and thought them poly- cellulose” in tlie Forest Products Laboratory methods of saccharide in nature. Arabinose, xylose, and galactose n-ere analysis. A second part d l come from the polysaccharide the sugars usually concerned; his preparations mostly con- that is very intimately associated with the cellulose itself and tained more than one sugar unit. Except for their relative is normally isolated with it in the Cross and Bevan cellulose ease of hydrolysis with acid, the most characteristic property fraction. Perhaps the most direct evidence of a dual source of this group is their solubility in dilute alkali, and it is by occurs in the work of Hawley and Campbell (11), who anaextraction with alkali that they are usually obtained. Unless lyzed wood both before and after treatment with dilute alkali, the treatment is drastic and prolonged, no change appears to finding that both the Cross and Bevan cellulose and the take place in the polysaccharide, which may then be obtained “pentosans not in cellulose” had been diminished by this in a crude condition on acidification and addition of alcohol treatment. Quite mild treatment was sufficient to remove a to the extract. Further purification is accomplished by certain amount of the associated polysaccharide from the resolution and reprecipitation. Until recently, the view was Cross and Revan cellulose. Further evidence is supplied by held that the substances obtained in this way were true the work of Miller and Swanson (25) on the graded hydrolysis hexosans or pentosans or, more frequently, hexopentosans, of wood with boiling hydrochloric acid. Additional support containing both anhydrohexose and anhydropentose units. is given by the high p- and y-cellulose content of Cross and During the past few years, however, i t has been found that the Bevan cellulose from woods, and especially by the presence majority of the hemicelluloses thus prepared are not true of mannan and pentosan in the p- and y-cellulose fractions polysaccharides in the strictest sense of this term, but often (38)* There is no direct published evidence that the alkalicontain also considerable amounts of uronic acids (4, 28, 3034, 37). Glucuronic and galacturonic acids are those com- soluble material from the Cross and Bevan cellulose of wood monly found, though mannuronic acid has been obtained is precipitated by the methods commonly used for preparing from certain marine algae (2, 26). The widespread occur- the polyuronides, but there is indirect evidence in the large rence of the uronic acids in several types of plant constituents amounts of p-cellulose usually found. Qualitative tests on is now recognized. Pectin contains nearly 80 per cent of this point kindly made by Ritter have shown that a dilute galacturonic acid, and several gums and mucilages have also alkali extract of the Cross and Bevan cellulose from white been shown to contain a high percentage of uronic acids spruce wood will give considerable quantities of a precipitate on acidification followed by the addition of alcohol. ( 3 , 27, 29). Candlin and Schryver (4) suggested the use of The alkali solubility of the polysaccharide associated with the term “polyuronide” for compounds that contain uronic acids in combination with sugars. The term has not yet cellulose in the Cross and Bevan cellulose fraction has been come into general use but would be of considerable value in unquestionably demonstrated for certain straws, esparto connection with the composite group of hemicelluloses. grass, bamboo, and similar plants. Moreover, it has been Hemicelluloses containing uronic acids linked to anhydro- shown for these materials that this polysaccharide can be sugar units may rightly be described as polyuronides, although obtained from the extracts in the way usually employed for substances other than hemicelluloses (pectin and gums, for the precipitation of hemicelluloses (9, 14, 15, 21-23, 28, 41). An alkali extract, therefore, as ordinarily employed for the example) are also polyuronides. The purpose of this review is to examine and correlate preparation of hemicelluloses, is likely t o contain compounds recent work on hemicelluloses from many sources, and es- from these two sources-the cell-wall incrusting material,

October, 1932

INDUSTRIAL AND ENGINEERING CHEMISTRY

1191

and the Cross and Bevan cellulose fraction. The f:wt that a iirabinose has never been obtained from a properly chlorinmixture is obtained is not particularly serious, because a ated cellulose. hlethylpentosans have been reported by hemicellulose preparation as isolated from wood or straw has many workers as present in small amounts. The evidence for never been considered a single chemical compound. O’Dwyer their presence, however, is still unsatisfactory, since the (32) effected a separation of a crude hemicellulose preparation methylfurfuraldehyde said to be obtained on hydrolysis with into two groups, slightly different in properties and consider- 12 per cent hydrochloric acid might actually be w-hydroxyably different in the proportions of sugars and uronic acids methylfurfuraldehyde, which is given in small quantities from present, by precipitation first with acid and then with acid hexosans (and therefore from cellulose itself) under the alcohol. More recently, Sorris and Preece (30) have ob- conditions of this determination. tained a separation into six fractions by various methods of Although the cellulose from all hardwoods examined conprecipitation. Whether, even then, the resulting prepara- tains varying amounts of pentosan, that from the softwoods tions are chemical entities is a matter of some doubt, though is always distinctly lower in furfuraldehyde-yielding material, the fact that certain of their fractions are not represented in I n this group, however, appreciable quantities of hexosan some materials examined tends to indicate that their scheme material are also found associated with the cellulose. of fractionation has its basis in fairly clearly defined differ- Schorger (34 examined quantitatively a large number of ences. S o others have attempted such a thorough separa- gymnosperms and showed mannan to be present in amounts tion, most investigators confining themselves to the two varying from 7.68 per cent in Monterey pine, downward. groups-material precipitated by acid, and material precipi- These determinations were made on the total wood, not on the tated by acid-alcohol-first suggested by O’Dwyer (32). Cross and Bevan cellulose from it, but subsequent work on Although recognizing that both these fractions are un- hydrolysis points to the Cross and Bevan cellulose as the doubtedly mixtures, it is still important to remember the source of this mannan. Schorger examined a number of harddouble source of such a preparation, and the fact that there woods a t the same time, in no instance identifying mannan. may be present in it some nonpolyuronide hemicellulose from Other workers, however, have reported small quantities of the Cross and Bevan cellulose fraction. This last fraction mannan in certain hardwoods. It would be interesting t o has always been considered to be strictly polysaccharide. in reexamine these hardwoods, since the mannan in them is nature-that is, to contain sugar units only. (Schwalbe and probably not associated with the cellulose. Of other hexosFeldtman (87’) have reported the presence of glucuronic acid -am, although small amounts of glucose have been obtained in a wood pulp, but even well-cooked pulps are not entirely in the hydrolysis of wood cellulose (8,38),galactose has never Cross and Bevan cellulose, usually containing appreciable been obtained from this fraction. To summarize, therefore, quantities of pentosans not in cellulose.) Ritter, in un- it appears that the polysaccharide associated with the cellupublished work on the chlorination of cellulose in sunlight, lose of woods is usually xylan, though in softwoods mannan has had occasion to determine uronic acids in normally may also be present to a considerable extent, while some chlorinated Cross and Bevan cellulose, and has obtained only glucosan has been found in both hardwoods and softwoods, such traces of carbon dioxide as are ordinarily given by all though in relatively small amounts. carbohydrate material under the conditions of this determiI n plant materials other than woods, there is still less evination. Norman (28) obtained only 0.22 per cent carbon dence, since no such systematic analyses as those of woods dioxide from the Cross and Bevan cellulose from oat straw; have ever been made. In the specimens examined, the this was probably due in part to slight oxidation during the associated polysaccharide was a xylan. It is, however, not chlorination process, as demonstrated by Heuser and Stockigt meant to imply that the cellulose of all nontimber materials (19) and Hibbert and Parsons (20). It seems likely, there- contains only xylan, since the number of sources investigated fore, that the polysaccharide associated with the cellulose in is small. The xylan associated with esparto cellulose has the Cross and Revan fraction does not contain uronic acid. probably received most attention from investigators, though Accordingly, the polyuronides that have been isolated and the first well-authenticated xylan preparation was obtained studied are probably mixtures of polyuronides from the in- from chlorinated straw pulp (14). It was precipitated as a crusting cell-wall material, together with hemicellulose, free copper compound from an alkaline extract and subsequently from uronic acid, from the Cross and Bevan cellulose fraction. purified, yielding on analysis furfuraldehyde equivalent t o Additional studies of the polyuronide hemicelluloses should more than 95 per cent xylan. Some properties of this comtherefore be made on products uncontaminated with any pound were investigated by Heuser and Ruppel ( I 7 ) , Heuser material from the Cross and Bevan cellulose. It is possible and Schlosser (18),and Heuser and Jayme (16). A complete further to suggest the source of such a product-namely, the constitutional investigation of that from esparto cellulose liquors obtained from the chlorination and sulfite extraction was commenced by Irvine and Hirst (91) and completed by during the isolation of Cross and Bevan cellulose. Such a Hampton, Haworth, and Hirst (9). These workers esliquor, of course, contains also chlorinated lignin, but the tablished unquestionably the fact that xylan from this development of a method for the isolation of the polyuronide source is composed of 1,5-anhydro-fl-xylose units and differs should be possible. There is already some evidence (unpub- in configuration from cellulose only in that the terminal lished) that the polyuronides removed by this process and carbinol group is missing. The xylan has also been prepared present in this liquor are not hydrolyzed during the chlorina- from the cellulose of bamboo by Hess and Ludtke ( I S ) , from tion and extraction. corn seedlings by Link (22),and from young barley plants by As mentioned previously, no attempts have been made t o Norman (unpublished). The lack of a more general survey of the distribution of this investigate by direct preparation the nature of the polysaccharides associated with the Cross and Bevan cellulose of associated polysaccharide in nonwoody plants is regrettable. woods, but extraction and hydrolysis have indicated the sugar The scanty evidence available, however, indicates close units concerned. The standard procedure in analytical similarities between the Cross and Bevan cellulose fraction of practice is to determine the yield of furfuraldehyde from the such materials as cereal straws, esparto, and bamboo, and of Cross and Bevan cellulose, since pentose material is almost the hardwoods. Indeed, as Hawley (IO) has pointed out, the invariably present, often to a considerable extent. The relationship between the cellulose of such materials and that nature of the pentose has been investigated in many instances of the hardwoods is much closer than the relation between the and always found to be xylose, present, of course, as xylan. hardwoods and the softwoods. Mannans have not yet been

I N D U S T R I A L A N D E N G I N E E R I I\; G C H E M I S T R Y

1192

found in the cellulose fraction of any group other than the gymnosperms. Thus far in this discussion no attention has been paid to the possibility of true hexosans, or pentosans, existing otherwise than in the Cross and Bevan cellulose fraction. Probably such substances, except in a few instances, are distinctly rare. Although the literature does contain many references to such polysaccharides, these must be discounted because the

1’01. 24, No. 10

woods, since the uronic content of the polyuronides of wood is quite low. To give an example, the highest yield of carbon dioxide from the uronic acids of a hardwood was obtained by Anderson ( 1 ) from catalpa; his results indicated 5.76 per cent of uronic acid anhydride. The furfuraldehyde yield from this quantity of uronic acid would be 0.96 per cent, which if calculated as pentosan would be 1.5 per cent, thus showing a quantitative error for the wood of 4.26 per cent; these percentages refer to the weight of the original wood. Softwoods seem to yield appreciably less carbon dioxide than hardwoods. In all instances the figures obtained are probably a little high, since a small amount of carbon dioxide is given by H y d r o l y r ~ b l e by dilute ac,d to carbohydrate units other than uronic acids. Cellulose preparations invariably seem to yield carbon dioxide in amounts up to 0.25 per cent, but it is not as yet possible to FOUND I N NOT F O U N D I N CROSS A N D B E V A N CROSS A N D B E V A N correct for this. The error caused in the analysis of woods by CELLULO5E FRACTION CELLULOSE F R l C T l G N calculating uronic acid furfuraldehyde as pentosan may, of course, be readily avoided by determining separately the NOT C O N T A I N I N G NOT C G N T A I N I N G CONTAINING URONIC a c i D URONIC A C I D URONIC A C I D uronic acid content of the material and then correcting the pentosan figure accordingly. Such a procedure is recommended for a more nearly exact and complete analysis of ELL U LUSA P€NTOSANS wood. I I FRAMEWORK SUBSTANCE There is another possible error to consider: With straws, MPNNAN XYLLlN an appreciable proportion of the polyuronide is accounted for GALACTAN MANNAN GLUCOSAN (7) ARABAN U R O N I C aciu by. hexose units, which would not be included by either ET C€T€RA or uronic acid determinations. With woods, furfuraldehyde FIGURE1. THE DIFFERENTIATION OF HEMICELLULOSES however, errors from this source are practically negligible. There is no direct evidence of the presence in wood of hemiubiquity of the uronic acids has only very recently been realcellulose hexosans, excepting those in the Cross and Bevan ized and their presence sought. However, there are certain cellulose of softwoods.’ In fact there is some good evidence well-authenticated instances in which hexosans, free from to the contrary. Ritter, whose report is as yet unpublished, uronic acids, have been found. The mannan of the ivory nut has shown that in both a hardwood and a softwood the total (Phytelephas macrocarpa) and of salep roots, and the galactan reducing value of the polysaccharide constituents when of agar-agar, are cases in point, but these materials are quantitatively hydrolyzed can be approximately accounted probably very different in physiological significance from the for by the Cross and Bevan cellulose plus the pentosans not polyuronides. in cellulose. In straws and similar materials, in which the INTERPRETATION OF SOME STANDaRD ANALYTICAL METHODS hexosan content of the polyuronides is appreciable, the situation is much more difficult and a wholly satisfactory method The effect of this new conception of the nature of the of estimation has not yet been given. Norman (28) had hemicelluloses on the interpretation of the results of‘ some recourse to an indirect and tedious method for the determinaanalytical methods, especially those employed for wood by tion of the polyuronides, but his method is open to some the Forest Products Laboratory, is important. The poly- criticism for reasons that will be pointed out later. Norris uronide group of hemicelluloses causes some deviations from and Preece (30) avoid this difficulty of the indeterminacy of accuracy unless the presence of uronic acids is specifically the hexoses by determining the polyuronides as a whole, using taken into account. The standard Forest Products Labora- direct preparation, and then subsequently analyzing them for tory methods determine three groups of polysaccharides: their constituent units. Such a procedure is far from simple (1) Cross and Bevan cellulose, (2) pentosans in Cross and and its accuracy is uncertain. Bevan cellulose, and (3) pentosans not in Cross and Revan Turning to other systems of analysis, that of Dore (6-7) cellulose. By an additional determination, either that for a- deserves attention. Devised especially for softwoods, it was cellulose or for hydrolysis number ( l a ) ,a fourth group, either intended to yield figures that could be summed up to 100 per p- and y-cellulose, or the more readily hydrolyzed portion of cent. Though he prepared his Cross and Bevan cellulose the Cross and Bevan cellulose, is sometimes obtained. As has fraction without previous treatment ( 6 ) , in general he did not just been shown, group 1 (the Cross and Bevan cellulose) determine the pentosan in the cellulose. I n his method, contains the “pure” cellulose together with associated poly- mannan was determined separately on a fresh sample of wood. saccharide. Since this associated polysaccharide does not He neglected to determine whether the mannan was associated contain uronic acid, the validity of this determination is with cellulose. The figure given for soluble pentosans was unaffected. Group 3 (pentosans not in Cross and Bevan found by determining the furfuraldehyde yield of the chlocellulose) is obtained from the difference between total rination liquid, Moreover, as he points out, hemicelluloses are pentosans and pentosans in Cross and Bevan cellulose, not completely covered in this scheme. Later ( 7 ) he modified pentosan being calculated from furfuraldehyde by employing his method with the intention of making it more readily the Krober factor. As pointed out by Norman (28), this applicable t o hardwoods, introducing a pre-chlorination involves the assumption that the furfuraldehyde is derived treatment with cold 5 per cent sodium hydroxide. This he only from pentose units, whereas in actuality a portion is found to reduce greatly the time of chlorination. He gave derived from the uronic acid units of the polyuronides. In 1 The presence of e-galactan in western larch is not an exception, since point of fact, the calculation employed a t present, in which is water-soluble and may be classed as an extractive rather than a hemithe polyuronides are expressed as pentosans, invariably gives this cellulose. The trace of galactose reported in other woods (up to 1.56 per low results, for while pentosans yield about 64.5 per cent cent in oak, 7) may possibly be due, at least In part, to the presence of furfuraldehyde, uronic acid anhydrides yield only 16.6 per galacturonic acid in the polyuronides, or may perhaps be due t o galactose cent. This inaccuracy, however, is not serious with most groups in them.

It

I

NSI

October, 1932

I N D U S T R I A L A N D E N G I N E E R I N G C H E h3. I S T R Y

1193

figures to show that this pretreatment did not affect the yield (34) in his work on boxwood. This m-orker also differentiates of Cross and Bevan cellulose, though the results of other between free and combined polyuronides; the free are removed workers and also the relative amounts of CY-, p-, and y-cellu- through extraction with 4 per cent cold sodium hydroxide, lose fractions cited by him would hardly support this view. whereas, the combined require a hot alkali treatment for their He considered that the furfuraldehyde-yielding substances in solution. Whether this differentiation has its basis in actual woods may be divided into three classes: pentosans soluble in chemical differences is somewhat doubtful, since the cause, cold dilute alkali (corresponding to the polyuronide group); a t least in part, may be mechanical and physical difficulties the furfural-yielding constituents of cellulose (thought by him of extraction. This review makes clear the fact that some objections can to be oxycellulose); and a small amount of resistant pentosan removed during chlorination. In the light of recent work, be raised to practically all the methods of analysis and the objections t o Dore’s procedure for softwoods are similar evaluation used a t present. It is very desirable that each tq those already raised against the Forest Products Labora- method should be critically reexamined in the light of the tory methods, and may be met by the modifications suggested recent work on the nature of hemicelluloses. As far as is for those methods. K i t h hardwoods, however, the situation possible the theoretical basis of the various shortcomings has is more serious, and the alkaline pretreatment recommended been indicated, and means have been suggested by which the most serious ones may be avoided. A detailed survey of the probably invalidates the method, as mentioned previously. More recently a method of proximate analysis of plant distribution of the hemicelluloses in nontimber sources, materials has been put forward by Waksman and Stevens (40) which is a t present lacking, would undoubtedly assist in the and extensively used by Kaksman in his work on the de- better understanding of this complex group. composition of many types of plant tissue. Though possibly NOMEKCLATURE of value for comparative purposes, this scheme is far from satisfactory in several particulars, especially for the hemiThe Of the hemicelluloses be much celluloses, since it does not indicate correctly either the Of Candin and SchWVer’s term the Proper distribution of this group or the quantity present. Waksman and R~~~~~~(39) have in part admitted these shortcomings. “polyuronide.” More confusion than before will result, Lk&er preliminary extractions m-ith ether, water, and alcohol, however, if this new term is Wed loosely to cover the plant the tissue is subjected to hydrolysis for hours with per cent constituents previously called “hemicelluloses.” It has been hydrochloric acid, with the intention of converting the hemi- pointed out that two distinct major groups of hemicelluloses cellulases to sugar. The quantity of ‘LhemiCelluloSe” is then must be recognized, and that the term “polyuronide” is obtained by multiplying the reducing sugar in the extract by applicable only t o the one containing uronic acid units. The 0,9. Although it is true that the polyuronides wjll be hydro- other group, which has previously been referred to as polyby Miller and saccharide associated with Cross and Bevan cellulose, or the lyzed by this treatment, it is also true, as Swanson (25),that a portion of the polysaccharide associated readily hydrolyzable portion of Cross and Bevan cellulose, with the cellulose will also be hydrolyzed and removed. does not fall within the category of polyuronides, since, as has Further, the presence of uronic acids in the polyuronides been shown, uronic acid units are absent. This group of have a ‘pecific name* The name provides another disturbing factor, since, as shown by Link substances and xiemann ( 2 4 ) , these acids are in part decarboxylated to “hemicellulose” itself applies accurately to this fraction, now and that the polysaccharides not closely associated with the give degradation products of unknown compos~t~on cellulose have been separated and provided with a satisfactory indefinite reducing power. A further and perhaps have suggested that the Indeedthe authors serious error is introduced by the calculation of the, hydrolyzed sugar obtained as hexose, whereas, in point of fact, from most term “hemicellulose” be retained and limited to this group material after except for the fact that it would still carry its old meaning to materials, i t is mainly pentose. The Persons for a long time. A new name is needed, and an hydrolysis is treated with 80 per cent sulfuric acid and, after dilution and boiling, the sugars given are determined and then appropriate one appears to be ‘Lcellu10sans.”2 This term calculated as glucose. This on multiplication by 0.9 is said indicates that this polysaccharide is composed of anhydroto give the cellulose content of the plant material. Although sugar units-hexosans O r pentosans- that are intimately this procedure is accuratefor cotton cellulose, In which no associated with the cellulose itself and isolated with it in the associated cellulosic polysaccharides are present, it must give Cross and Bevan cellulose fraction* the c e ~ ~ u ~ o s e This is not the place for a discussion of the relationship of a high result for many woods and cereal the cellulose to the cellulosans; the relationship, however, ,f which is accompanied bv considerable quantities of only a portion of which wo;ld be attacked by the preliminary is such that any quantitative determination of the cellulosans 2 per cent hydroch]oric acid treatment. These objections by methods involving either alkali-extraction or hydrolysis by render this system of analysis unsuitable when accurate acid must be on an arbitrary basis. The amount removed will depend on the experimental conditions, and there is no infomation as to the amountof hemicelluloses is required. N~~~~~ (28) recently examined the distribution of the very obvious break in the continuity of the reaction until a furfuraldehyde constituents of straws, and proposed a method portion of the pure cel’ulose has obviously been attacked. Although this method takes into The authors would retain the term “hemicellulose” as the for their account both the presence of uronic acids and hexose units in generic name for this whole group of alkali-soluble and readily the polyuronides, it is vitiated by the fact that he employed a hydrolyzable polysaccharides, even though certain workers, pretreatment with hot 1 per cent sodium hydroxide for 20 such as Hess and Heuser, have maintained, not without minutes before chlorjnation, following in that the original reason, that literally it is both incorrect and misleading. It is crossand B~~~~ procedure. since an alkaline extraction thought that by this differentiation into two major groupswas employed in obtaining the crude hemicellulose prepara- polpronides and cellulosans-the members of which differ in 2 Preece ($4) proposes the reintroduction of the old term “p-x;lan” for tion that is the basis of the calculation, the preparation was probably contaminated with a portion of the xylan from the the xylan associated with the cellulose of boxwood This term, however, cellulose fraction, and the figures for furfural&hyde not from is obviously unsuitable for the softwoods in which xylan is not the only associated polysaccharide present, and accordingly the new term “oellucellulose were thus increased‘ The former loosns.” which covers both hexosans and pentosans and mixtures of these, objection may be raised to the procedure suggested by Preece is preferable.



1194

INDUSTRIAL AND ENGINEERING CHEMISTRY

structure, distribution, and function, the situation will be distinctly clarified and confusion avoided. Figure 1 summarizes the basis of this differentiation. Furthermore, this classification of the hemicelluloses into two major groups gives an additional significance to the Cross and Bevan cellulose determination, since i t is only by the isolation of this fraction that the two groups can be separated and be determined quantitatively. Though the differentiation effected by the chlorination process was suspected and commented upon some years ago ( l a ) , the basis of it is now made clear. To be emphasized also is the necessity of determining Cross and Bevan cellulose in material that has previously been treated only by neutral solvents, and of rejecting for critical work the numerous proposed modifications that include any chemical treatment before chlorination, lest, as a result, the cellulosans be diminished from their true value. LITERATURE CITED Anderson, E., J . Biol. Chem., 91, 559 (1931). Bird, G. M., and Haas, P., Biochem. J., 25, 403 (1931). Butler, C. L., and Cretcher, L. H., J. Am. Chem. Soc., 51, 1519 (1929). Candlin, E. J., and Schryver, S. B., Proc. Roll. SOC.(London), B103. 365 (1928). Dore, W.H.,’J. IND. ENG.CHEM.,12, 472 (1920). Dore, W. H., Ibid., 12, 476 (1920). Dore, W. H., Ihid., 12, 984 (1920). Hagglund, Biochem. Z., 70, 416 (1915). Hampton, H. A,, Haworth, W. N., and Hirst, E. L., J. Chem. SOC.,1929, 1739. Hawley, L. F., J. Forestry, 29, 191 (1931). Hawley, L. F., and Campbell, W. G., IND. ENG. CHEM.,19, 742 (1927). Hawley, L. F., and Fleck, L. C., Ibid., 19, 850 (1927).

Vol. 24. No. 10

Hess. K., and Ludtke, M., Ann., 466, 23 (1928). Heuser, E.. J. prakt. Chem., 104, 259 (1922). Heuser, E., and Haug. A., 2. angeur. Chem., 31,99, 103 (1918). Heuser, E., and Jayme, G., J . prakt. Chem., 105, 232, 283 (1923). Heuser, E., and Ruppel, W., Ber., 55, 2084 (1922). Heuser, E., and Schlosser, P., Ibid., 56, 392 (1923). Heuser, E., and Stockigt, F., Cellulosechem., 3, 61 (1922). Hibbert, H., and Parsons, J. L., J. SOC.Chem. I n d . , 44, 4731’ (1925). Irvine, J. C., and Hirst, E. L., J. Chem. SOC.,1924, 15. Link, K. P., J . Am. Chem. Soc., 51, 2506 (1929). Link, K. P., Ibid., 52, 2091 (1930). Link, K. P., and Niemann, C., Ibid., 52, 2474 (1930). Miller, R. N., and Swanson, W. H., IND. ENG.CHEY.,17, 8$3 (1925). Nelson, W. L., and Cretcher, L. H., J. Am. Chem. SOC.,51, 1914 (1929). Norman. A. G.. Biocha. J.. 23. 524 (1929). . , Norman; A. G.,’Ibid., 23, 135i (1929). Norman, A. G., Ibid., 25, 200 (1931). Norris, F. W., and Preece, I. A., Ibid., 24, 59 (1930). O’Dwyer, M. H., Ibid., 17, 501 (1923). O’Dwyer, M. H., Ihid., 20, 656 (1926). Preece, I. A., Ibid., 24, 973 (1930). Preece, I. A., Ibid., 25, 1304 (1931). Schorger, A. W., J. IND. ENG. CHEM.,9, 748 (1917). Schulze, E . , 2. physiol. Chem., 16, 387 (1892); 19, 38 (1S94). Schwalbe, C. G., and Feldtman, G. A., Ber., 58B, 1534 (1925). Sherrard, E. C., and Blanco, G. W., IND. ENG.CHEW,15, 611, 1166 (1923). Waksman. S. A.. and Reuseer. H. W.. Soil Sci., 33, 135 (1932). Waksman; S. A.; and Stevens; K. R., IND.ENQ. CHEM.; Anal Ed., 2, 167 (1930). Ziemiecka, J., PoZish Agr. Forestal Ann., 25, 1 (1931). RECEIVED April 2, 1932. Presented before the Division of Cellulose Chemistry at the 83rd Meeting of the American Chemical Society, New Orleans. La., March 28 to April 1, 1932

United States Trade-Mark Law and Practice for the Chemist ROBERTE. SADTLER, 412 W. Pine St., Selinsgrove, Pa.

T

HE use of trade-marks is thought to have started many centuries ago when few people were able to read or write. The marks a t that time usually comprised some figure, such as a cross within a circle, a cat’s head, or some simple device that could be easily remembered. The makers of swords, hatchets, and other devices placed these marks upon their products in order that the users might know how to identify them and thus be sure of getting a hatchet of similar quality when they next needed one. WHATIs

A

TRADE-MARK?

United States law does not define a trade-mark. The United States Courts have been reluctant in defining a trademark. One of the most satisfactory definitions which has been given is as follows: A trademark is a distinctive word, emblem, symbol, or device, or a combination of these, used on goods actually sold in commerce to indicate or identify the manufacturer or seller of the goods. *

A trade-mark must not be descriptive, because any person has a right to use descriptive terms in connection with his product, and no one has a right to withdraw such terms from public use and maintain a monopoly upon them. Further-

more, a mere phonetic spelling of a descriptive term may not be employed as a trade-mark. On the other hand, a trademark must not be misdescriptive, because such a term would tend to deceive the public. The use of the mark “Syrup of Figs” for a laxative which contained no appreciable amount of fig juice was held to be deceptive by the Supreme Court in Clinton E. Worden & Company us. The California Fig Syrup Company (187 U. S. 516). However, words that are merely suggestive may be employed as trade-marks, and commercial people are usually strongly in favor of this form of mark. A color alone cannot be used as a trade-mark. For instance some of the gasoline companies have colored their gas and have attemptea to register the color as the trademark. The applications were refused on the grounds that mere color, aside from some particular symbol or design, such as, for example, a circle, square, triangle, cross, or star, cannot constitute a valid trademark. This was decided in the case of I n re General Petroleum Corporation of California (49 Fed. Rep. [2nd]966).

How Is OWNERSHIP SECURED? In the United States and in many other countries the owner of a trademark is the first person who places the trademark