o 



36 SECTIONAL ADDREHiSES. 



J)oiiliain"s work thus ga.ve the first oleiir eYideiice as to the hnkagc 

 of part of the- celhilose moleculo, aiul is one (yi tlie most; important 

 contributions made tO' the structural study oi' couijilex carljohydrates. 

 Cellulose must contain the unit 



V. 



— CH X 



CHOH 



I 

 CHOH 



Jh 



I 

 CH Y 



I 

 CH2OH 



but as an incompletely methylated cellulose was employed in the 

 hydrolysis the research left unexplained the nature of the residues X 

 and Y. 



The investigation was therefore continued with the object of com- 

 pleting the methylation of cellulose and providing answers to the follow- 

 ing questions : — 



(a) Does trimethyl cellulose give, on hydrolysis, a mixture of 

 methylated glucoses in which the average methoxyl con- 

 tent is three groups per 0^ unit? 



Or alternatively , 



(h) If trimetliyl cellulose gives trimethyl glucose alone, is this 

 sugar a single individual or a mixture of isomerides? 



The War interfered with the progress of the work, but other authors 

 have not hesitated to propose formulae for cellulose based on Denham's 

 results before he had an opportunity to complete his researches and 

 provide answers to the fundamental questions raised above. 



In consultation with Dr. Denham we have repeated his experiments, 

 and amplified them, so that we are now in a position to propose a 

 structure for the cellulose unit which is based on secure evidence. We 

 find that the exhaustive methylation of the polysaccharide, when re- 

 peated twenty times, gives a product containing 43.0 per cent, of meth- 

 oxyl in place of the 45.6 per cent, required for a trimethyl derivative. 

 The carbon and hydrogen values also agree with the foi'mula 

 (C6H,02(OMe)3).x' s-nd as the material preserved a fibrous structure 

 there seems little likelihood that profound molecular alteration had 

 taken place. The trimethyl cellulose was heated with a large excess 

 of methyl alcohol containing 1 per cent, of hydrogen chloride for fifty 

 hours at 125-130°. This treatment effected depolymerisation, hydro- 

 lysis, and conversion of the scission products into the corresponding 

 methylglucO'sides. These were distilled in a high vacuum, the total yield 

 obtained being 90 per cent, of the theoretical amount. The following 

 fractions were collected ; — 



