B.— CHEMISTRY. 47 



formidable, as methylated fructoses of the y-type are liquids and give 

 no crystalline derivatives. By the following method, however, it has 

 been possible to obtain the necessary information, and it is no>w 

 established that only one form oi trimethyl fructose is produced from 

 inulin. 



A large quantity of trimethyl inulin was digested with methyl alcohol 

 containing hydrogen cldoride under conditions which effected: — 



(a) Depolymerisation, 



(b) hydrolysis, 



(c) condensatio'U tO' give trimethyl methylfructoside. 



The product was distilled in a high vacuum, and fractions were 

 abstracted at frequent intervals while the boiling-point remained 

 constant. All the fractions showed the same refractive, index and 

 specific rotation. Moreover, the speed of hydrolysis of the fructoside, 

 as indicated by polarimetic observations, was in each case the same, 

 and in each experiment the trimethyl fructose then formed showed 

 identical physical constants. Of greater importance is the fact that 

 all the specimens of trimethyl methyl-fructoside reacted in the same 

 way when dissolved in acetone containing hydrogen chloride. Under 

 these conditions trimethyl fructose monoacetone was formed, and here 

 again the speed of the reaction measured polarimetrically showed no 

 difference in any of the specimens. 



There can be no doubt, therefore, that inulin is an aggregate of 

 anhydro y. fructose residues and that each of the units is identical. 



As the exact structure of the methylated fructoses of the y-type is 

 not yet determined with certainty, it is needless tO' speculate here on the 

 manner in whicli these residues are united. The subject has been 

 engaging our attention fo^r a considerable time, but is complicated by 

 the readiness with which the methylated inulins undergo both 

 polymerisation and depolymerisation. 



It is not unlikely, bearing in mind the structure assigned to cellulose 

 and starch, that inulin is based on a tri-anhydro-y -fructose. 



The structural discussion which I have had the honour to lay before 

 you on one of the most important groups of natural compounds is 

 admittedly incomplete, and no claim is made that the formulae now 

 submitted are final. But they at least indicate a new development in 

 the chemistry of polysaccharides, and lines of further research are 

 opened out which promise in time to reveal the intimate constitution of 

 these substances. Much reliance has been placed on the validity of the 

 methylation process as a means of determining structure, but it has to 

 be remembered that most speculation of the kind on carbohydrates is 

 now based on results obtained by this one particular method. The 

 structure of glucosides, the nature of y-sugars, the constitution of 

 sucrose, maltose, and cellobiose, are all involved in cuiTent discussions 

 on this subject, and all are based on the properties of the simple 

 alkylated sugars. 



Numerous details, such as the specific reactions of the individual 

 hydroxyl groups in carbohydrate units, have still to be settled before 

 the further problem of the polymerisation of polysaccharides can be 



1022 F 



