40 SECTIONAL ADDRESSES. 



permit o[ these formulee being given in full, but the general form will 

 be evident fi'om the simplest example: — 



f ^ 1 



— CH— 0— UH • CHOH • CHOH • CH • CHOH • CH.,OH 



i i I « 1 



CH— 0— CH • CHOH ■ CHOH • CH • CHOH • CH.2OH 



o I , ^1 



OH— 0— CH ■ CHOH • CHOH • CH • CHOH • CHoOH XIII. 



I 

 -CH 



I , O i 



CH— O— CH • CHOH • CHOH • CH • CHOH • CH^OH 



I i ^ 1 



CH,— 0— CH • CHOH • CHOH • CH • CHOH -CHoOH 



Variations are introduced by lengthening the sugar chains until finally 

 a molecule is obtained with eighteen glucose residues (CiosHinaOgi). Tho 

 feature common to all these glucosidic formulae is that the hydroxyl 

 groups are not symmetrically distributed in the glucose residues. The 

 simplest structure suggested by Hess would give by our processes five 

 molecules of 2,3,5,6-tetramethyl glucose and one molecule of glucose, 

 while his more elaborate molecules would also result in these compounds 

 together with trimethyl glucose. The reactions of trimethyl cellulose 

 exclude all formulae of this nature. 



Starch. 



[With Mr. John Macdonalb, M.A., B.Sc] 



Turning to the problem of the constitution of starch, we encounter 

 very much the same difficulties as have already been referred to under 

 the heading of cellulose. The importance of the compound, its mani- 

 fold technical applications, and the special appeal its study makes to 

 the biologist have alike combined to produce a voluminous and some- 

 what scattered literature. If, however, we eliminate unsupported or 

 contradictory results the following evidence as to structure emerges 

 Starch, when purified from constituents containing nitrogen and phos- 

 phorus, possesses the formula (CeHioO.^)^' ^^^ the molecule consists 

 entirely of glucose units. Further, three hydroxyl groups are present 

 for every six carbon atoms, but, as in the case of cellulose, this does 

 not necessarily imply that each glucose residue contains three unsub- 

 stituted hydroxyl groups, or that their distribution is symmetrical. The 

 primary reaction of starch, which must be accommodated by a structural 

 formula, is the production of maltose by the action of diastase. 



The first essential is the identification of the unit of which the starch 

 molecule is composed, and here, as in the case of other polysaccharides, 

 molecular weight determinations give results of doubtful value. At the 

 present time there is little tendency to regard starch as a highly complex 

 glucoside in which a large number of hexose residues are mutually con- 

 densed together, and the view prevails that the polysaccharide is derived 

 from a comparatively simple anhydro-sugar by profound polymerisation. 

 Attention may be focussed on three formulae based on such ideas. 



