480 



SCIENCE 



[N. S. Vol. LIV. No. 1403 



case of the scientific investigations stimulated 

 by the mercerization reaction. Since 1850 (or 

 thereabouts) a number of investigators includ- 

 ing Gladstone,^ Vieweg^, Thiele,* Cross,^ and 

 others have assigned various formulas to the 

 compound (or compounds) that had been 

 formed between cellulose and sodium hydroxide 

 when concentrated alkali acted upon cotton. 

 The existence of such compounds was disputed 

 by Hubner and Teltscher" and later by Leigh- 

 ton.'^ From a hasty review it would appear that 

 the existence of a definite compound between 

 cellulose and sodiujn hydroxide had never been 

 demonstrated, and that alkali cellulose may 

 perhaps be attributed to adsorption phenomena. 

 Nevertheless, Leighton's work has not affected 

 our interpretation of the constitution of vis- 

 cose, which ri"esupposes a cellulose alcoholate, 

 (CoHsOiOISra or some similar compound) which 

 then reacts further with CS= to form at the out- 

 set of the " ripening " process sodlum-cellulose- 

 xanthogenate, which gradually hydrolyzes with 

 the loss of NaOH and CS2 until cellulose is 

 regenerated. It remains possible of course that 

 the xanthogenate reactions given in our texts 

 accurately represent the formation of viscose — 

 and yet, in the light of Leighton's investiga- 

 tions it is disconcerting to note the quiet as- 

 surance and certainty with which this explana- 

 tion of the xanthogenate reaction is generally 

 accepted. 



A far more striking example of the lack of 

 critic and indifference with which experimental 

 details are treated in our modern cellulose lit- 

 erature is to be found in the case of the hydro- 

 lysis of cellulose to glucose. Our literature 

 has been replete with confident statements that 

 within the limits of experimental error, cellu- 

 lose is quantitatively hydrolyzed to d-glucose: 



(C,H,.,0,)„ + nH,0 ^ nC„H,,0,. 



Irvine and Soutar,* however, have justly shown 

 2 J. Chem. Boc, 5, 17 (1853). 

 3Ber.,40, 3876 (1907). 

 i ChemiTcer-Ztg., S5, 610 (1901). 



5 ' ' Cellulose, ' ' p. 23. 



6 J. Soc. Chem. Ind., 28, 641 (1909). 



7 J. Physical Chem., SO, 32 (1916). 



8 J. Chem. Soc, 117, 1490 (1920). 



that this claim has always been made on the 

 grounds of questionable or incomplete experi- 

 mental evidence, and that in no case was dex- 

 trose or a dextrose derivative isolated in any 

 amount approaching the theoretical yield. 

 There is no object in reviewing the work of 

 Flechig," Schwalbe and Schultz,!" Willstatter 

 and Zechmeister,^^ or Ost and his co-workers.'^^ 

 Such a review would either show indirect evi- 

 dence or incomplete evidence regarding this 

 very fundamental reaction. It is only within 

 the past year that Irvine and Soutar them- 

 selves' have shown that the above equation is 

 substantially correct and that at least 85 per 

 cent, of dextrose is formed when cotton cellulose 

 is hydrolyzed. They failed to account for less 

 than 15 per cent, of the hydrolysis products. 

 Irvine's work is noteworthy in that he isolates 

 his compounds in a state of analytical purity. 

 His experiments are all quantitative and all of 

 his products are definitely identified. The 

 judgment and critique exercised throughout 

 this study are remarkable, and the research 

 must stand as a classical one. It presents a 

 marked contrast to the previous investigations 

 in the same field. It is furthermore interesting 

 to note that whenever the cellulose-dextrose re- 

 lationship has been brought into question, the 

 question has not been raised as the result of 

 some investigators' lack of critique, but be- 

 cause of certain reactions (like the bromo- 

 methyl furfural reaction of Fenton and Gost- 

 ling) which were themselves far from quanti- 

 tative, and the mechanism of which was not 

 fully understood. 



During the course of the myriad cellulose in- 

 vestigations that have crowded our literature, a 

 number of so-called " compounds " of cellulose 

 have been isolated and characterized. Let us 

 examine briefly the case of the " oxycelluloses," 

 compounds obtained by the oxidation of cellu- 

 lose. There is no necessity of reviewing the 

 methods of formation, or the properties of these 

 substances. If we accept Hibbert's view of the 

 constitution of cellulose, the oxidation of cellu- 



Z. physiolog. Chem., 7, 523 (1883). 



loBej-., 43, 913 (1910). 



"Ber., 46, 2401 (1913). 



12 Chem. Ztg., 34, 461 (1910). 



