Tlic rcclncing po\vcrs of the stnrcli frnctions froin gronnd \vhent stnrcli inclicntu tlint fnr-rcncliing clc~~olyiiierisotioii of tho stnrcli miic~ronioleciile Iins tnkcn plncc clnring grinding in n bid1 niill ; tho niolec*nlnr \vciglits of the frnetinns liiivc been cnlcnlutecl froni the rcchieing power clntn. Cngroinitl wlient aturclics Iinve Iiccn sliown to contiiin R sninll qnnntity of \vntcr-soleblc nintcrinl of high rcclncing power.It lins long becii rccognisetl3 tlint i i kiio\rletl:rc of the reducing powers of stnrcli nnd of stiircli fractious would aid in tlic forniulution of structure.? for these bodies, but the significiince of the values obtiiinccl 1111s I)ecn liniitetl by the scnsitirity of tlic methods eniploycd t o tletcrniine the reducing power. For csnniplc, if tlic results of the investigntors wlio us~~cl the Folding nictliotl' nre to be tnkcn ns reliresentntive, then t.lie conclusion is tlint stnrcli lins no rcduciug power, \rhilst results obtiiiiied by the iilkiilinc Iiypoioditc method jnclicnte tlint stnrcli Iiiis n reduciiig power which, Iio\rcvcr, \vns considered to be so sindl ns to be negligilJle.l Tlicse couclusions were sho\vn to bc erroiieous by Ricliiirdson el ml.,?who in their invcstigiitions ninclc use of n niore sensitive nlknline copper reagent, iind conclrided $lint stnrch nnd its friictioiis possess definite rcduciug power wliicli increnses iis depolynicrisntion proccecls wl~en the stiircli is subjected to viirious types of trcatnierit (~y . , iiutoclaving or ncitl treiitnicnt). Furtlicriiiore, it wns deinoiistrntetl tlint the soluble fractions liiive greater reducing powers tlinn tlic corrcspondiiig iiisolul)le friictions, niid the niithors conclude, froni considerntioil of their results, tliiit reducing po\rer viirics inversely with the tlegrco of polyinerisntion. Siinilnr but less detniled results Iinvc nIso lieen obtiiinctl by otllbr investigntors for s t i i r c l~e s ,~~ for celluloses,l' nnd for pcctius.'The reducing po\vcrs of stnrclics obtiiincd liy Riclinrd-6011 nrc niucli sniiillcr tlinii would be expected if cuch of. the Hnwort.11 cliniu-lciigtlis of 21-30 a-glucopyrnnose units hiis u fully reactive iildclrytlic cnd-group3 (reducing power 6*67--8*33% R* niiritose, correspontlin~ ix*ith cliniii-lcngtlis of 30 and 2.4 a-glucopyrnnoso iniits) rind it is clcnr tliut tlw tcrniinnl nldeliytlic groups plny SOIIIC definite piirt in Lintling these units together. More rccently, this point lius been clucidntcd Iiy lIirst nnd Young,( and tlic rep~uting uiiits of 24-30 anhydroglucose residues sliown to be linked to one nnotlier though the terminnl oldeliydo group of one unit untl nu * An abbreviation of "roduciny powor of mnltoso." 013 group of nnotlicr unit, ~rllieh in turu is einiilnrly linked to ii third unit, nud so 011.If it is nssuiiied thnt the tcrininnl nldehydic group of cncli nincroniolccule (irrespective of its nioleculnr size) reinnins fully reactive, the reducing poivcr (expressed ns n pcrcentnge of the reducing power of ninltosc) will vnry inversely ns the niolecul...