PREFACE 



TO 



THE SECOND EDITION 



SOME of the published criticisms of the book appearing to 

 indicate a certain misunderstanding of its plan and purpose, 

 we think it opportune to further and more specifically explain. 



The chemistry of cellulose is necessarily a chemistry of 

 colloidal, uncrystallisable substance : hence the relationships 

 of any derivative to the parent substance is not established 

 by a comparison of composition and properties, but requires a 

 complete statistical account of the reaction. Even then it is 

 generally impossible to measure attendant changes of mole- 

 cular condition, or, to put it more precisely, changes of 

 dimensions of the reacting unit, accompanying the formation 

 of derivative compounds. 



It is consequently impossible in the present state of our 

 knowledge to write a chemistry of cellulose in the terms of 

 the thoroughly systematised branches of the science. We have 

 therefore adopted the plan of classifying the empirical subject- 



