B.— f'HEMISTRV. 35 



is to ascertain if the aggregate (CeHioO,),, is composed eiiliiely of 

 glucose units, as even this fuudainental point has been disputed. To 

 obtain the necessary evidence is not so simple as might api)ear, and 

 many of tlio' proofs ' which have been offered do not carry conviction 

 to those familiar with the detailed chemistry of the simple sugare. By 

 converting cellulose into the triacetate and thereafter decomposing this 

 product so as to give methylglucoside, we have recently obtained data 

 which leave no doubt that glucose alone is the basis of cellulose.'* The 

 yield of pure crystalline methylglucoside thus isolated amounts to more 

 than 95 per cent, of the theoretical yieFd calculated on tlie basis of the 

 equation : — 



(CeHioOs). + X H,0 > X CsHjoOc. 



This result applies only to cotton cellulose, and further discussion 

 is therefore restricted to tliis particular variety. Disregarding stmic- 

 tures which are not based upon glucose, the numerous formula' pro- 

 posed for cellulose may be approximately divided into two classes : — 



1. Constitutions modelled on that of the glucosides, involving the 

 addition of numerous glucosei residues by mutual condensation. 

 According to this view, cellulose consists of large molecules. 



2. The unit of cellulose may be regarded as a simple anhydi-o- 

 glucose, CsHioO,;, higlJy polymerised. 



As pointed out, the situation alters almost from day to day, but for 

 the moment a compromise between the above classes is supported, and 

 some authorities prefer to regard cellulose as a simple anhydro-w- 

 saocharide (where n is a small multiple) polymerised in unknown 

 numbers. 



Twelve years ago, after developing the methylation process into a 

 trustworthy method for determining the linkages of sugar complexes, 

 we turned our attention to the constitution of cellulose. The work 

 was undertaken by Dr. W. S. Denham,* who, using methyl sulphate 

 and sodium hydroxide as the alkylating reagents, ohtained a methylated 

 cellulose in which the methoxyl content was 25 per cent. This value 

 is lower than that required for a dirnethyl cellulose (32.6 per cent.), 

 and it followed that, on hydrolysing the product, a mixture of methyl- 

 ated glucoses resulted. From the mixture one definite sugar was 

 isolated, and this Denham^ proved to be 2,3,6-trimethyl glucose (IV.), 

 which was then isolated for the first time. 



IV. 



-CHOH 



I 

 CHOCH, 



O 



CHOCH, 



GH 



I 

 CHOH 



I 

 CHjOCH, 



