174 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 34 



in the presence of alkali. On liyclrolyzing the methylated cellulose 

 with weak acid they obtained chiefly 2, 3, 6 trimethyl glucose which 

 may be represented 



CHoOH 



HC 



HCOH 



HOCH 



CH2OCH3 



I 

 HC 



6 



HCOH 

 Hi- 



HCOH 



I 

 CH3OCH 



I 

 HCOCH3 



HC 



O 



2, 3, 6 trimethyl glucose 



HO 



HO 



Glucose 



Since there are two free hydroxyl groups in the trimethyl glucose, 

 it is apparent that these could not have been present in the original 

 cellulose molecule but must have been produced during the hydrol- 

 ysis. The methylated ones are the ones which were originally present 

 in the cellulose and since the constitution of trimethyl glucose is 

 known, it folloAvs that in cellulose there must be three and only three 

 hydroxyl groups, one of which is a primary hydroxyl group and the 

 other two secondary. Since the empirical composition of cellulose is 

 CgHioOs and that of glucose CgHioOg, it is obvious that cellulose is 

 made up of anhydro-glucose units. 



This statement, however, would hardly suggest the exceedingly 

 complex character of the cellulose molecule. Although externally 

 cellulose and its derivatives always appear as amorphous or colloidal 

 materials, nevertheless the X-ray has shown them to possess definite 

 crystalline characteristics. The unit cell responsible for the crystal- 

 line properties is believed to consist of four glucose residues. It is 

 now believed that these unit cells are bound together by primary 

 valences into long chains of what might be termed anhydro-glucose 

 residues, and that these primary valence chains are in turn bound 

 together by secondary valence forces to form the micelle. This basic 

 idea has been extended and elaborated to account for the structure 

 of the cellulose fiber, including its high longitudinal strength, the 

 orientation about the long axis of the fiber, and the presence of the 

 outer layer of fibrils in the cellulose fibers from wood. 



Thanks to recent investigations by Staudinger, by Freudenberg, 

 and by Stamm, we even have some conception of the molecular 

 weight of cellulose. While these figures do not agree, nevertheless, 

 they at least indicate that the molecular weight is very high. Stau- 



