Compound Celluloses 163 



The authors have made (1883) the following determinations of 

 the constituents of these concretions : 



Inorganic constituents (ash) . 0-91 



Cellulose . . . . . .26-0 34-2 



1 Furfural i8'O p.ct 



Loss on boiling in 12 p.ct. HC1 (30 mins.) . 53-6 



In conclusion, we can only call attention to the desirability of 

 re-investigating the product, and, upon the evidence of close 

 similarity to the typical lignocellulose, of adopting, at the outset, 

 the general plan of investigation laid down for such compounds. 



The formation of a lignocellulose under such totally dif- . 

 ferent conditions from those which obtain in a flowering stem 

 is of especial significance in regard to the physiology of the 

 production of such compounds. 



THE LIGNOCELLULOSES OF CEREALS. Both in the straws of 

 cereals, and the seed envelopes of the grain, there is a typical 

 and characteristic process of lignification. With the formation 

 of quinone-like bodies, as in jute, there is associated the produc- 

 tion in the tissue of a large quantity of pentosan derivatives. 



The composition of brewer? grains has been carefully inves- 

 tigated by Schulze and Tollens (Landw. Vers.-Stat. 40, 367), 

 and an abstract of their results is given in Section III. p. 259. 

 From the more recent results of Tollens this material has been 

 found to yield 16-03 p.ct. furfural, corresponding to 26-93 P- ct 

 of pentosan. A considerable proportion of the pentosan con- 

 stituents may be directly hydrolysed to pentaglucose ; on the 

 other hand, a not inconsiderable proportion is so intimately 

 united to the cellulose as to resist hydrolytic treatments of 

 some severity. The lignone constituent was not specially 



1 The furfural was estimated by the colorimetric method of comparison 

 with a standard solution of furfural (V. Meyer, Bed. Ber. n, 1870), the 

 only method available at the time. As a specimen of jute similarly investi- 

 gated gave 10-6 p.ct. furfural, it is probable that the above determination 

 is 2-3 p.ct. in excess of the true number. 



