156 Cellulose 



way, if we find associated with a tissue an excreted product of 

 general constitutional resemblance thereto, we cannot avoid 

 the suggestion of genetic relationship. The suggestion be- 

 comes an hypothesis upon which investigation can proceed. 

 The lignocelluloses, for instance, afford many indications of 

 such relationship to the tannins. In the jute fibre, tannins are 

 always present in small quantity ; the characteristic R. hexene 

 groups of the lignocellulose occupy a definite and close rela- 

 tionship to the trihydric phenol, pyrogallol, to which many of 

 the tannins stand in direct constitutional relationship ; and we 

 have described an instance of * spontaneous ' transformation of 

 the lignocellulose into a substance having the essential charac- 

 teristics of the tannin group. The general discussion of this 

 question belongs to a later section of our subject (see p. 179). 

 It is introduced here in order to show that we cannot attempt 

 to formulate a molecule of a lignocellulose on the lines of a 

 carbon compound of ascertainable molecular weight and such 

 relationships of its constituent groups as are sharply defined 

 and verifiable by synthesis. It is true that when attacked in 

 detail the lignocellulose is resolved into well-differentiated 

 groups, which may be regarded with reservation as consti- 

 tuents of the parent molecule ; the reservation being that 

 unless and until the lines of cleavage are proved to be in- 

 variable, we cannot consolidate the results of various direc- 

 tions of resolution into a homogeneous view of the parent 

 substance. 



We will now point out how far the problem is solved by the 

 evidence available. 



(i) THE LIGNOCELLULOSE A HOMOGENEOUS COMPOUND 

 RATHER THAN A MIXTURE. The evidence for this conclusion 

 is as follows (a) Physiological : general uniformity in composi- 

 tion and reactions ; does not vary with age of fibre (i.e. from root 



