( 782 ) 



It is therefore not a inattei' of doiihl tliat a third isomer does 

 not exist hut that it is a mixture of the otiier two ^). 



Tlie two stereoisomerie metliylglueosides may Ije eonverted into tlie 

 corres|)ondiiig' pejitacetates and eoiiversely, the hitter into tiie former ■•^). 

 The «-glucoside correspomls witli the pentaeetate m.p. 112°, the 

 ,?-gliK'oside witii the pentaeetate m.p. 134°. It is therefore as well 

 to indicate the two [)entaeetates, respectively, with a and (?, as has 

 in fact been done in the said article of Behkknd and Roth. 



H. The mutual trausformation of the methylglucosides and pent- 

 acetates throws a new light on the phenomenon of the mnltirotation 

 of sugars. We may accept as the most probable explanation of this 

 phenomenon a mutual direct transformation of the two stereoisomerie 

 modilications which must exist according to Tollens' glncose formula. 

 The recently published ijivestigations of Fhankland Akmsthong^) and 

 of Behrend and Roth ') have furnished strong arguments in taxour 

 of this view. 



But it must be remembered that Tollens' formula does not express 

 the aldehydic properties of glucose; one is therefore inclined to 

 assume that in a glucose solution there must occur als(j molecules 

 in the aldehyde-form, or molecules which contain I more H, O with 

 the group H C (OH).^. One may then also come to the conclusion 

 that this hydrated aldehyde-form does not act as an intermediate 

 product in the transformation between the two stereoisomerie oxide 

 modifications "), but that we have here a complete analogy of what 

 takes place with the glucosides, namely that, although a direct trans- 

 formation takes place between the stereoisomers, there also occurs 

 a (piantitatively insignificant, secondary reaction involving addition 

 and eliminatioji of the solvent. These reactions will then all be in 

 equilibrium with each other. This point, I will also try to elucidate 

 experimentally "). 



^) This view is alsu held in a quite recently published article of Behrend 

 and Roth Ann. 331, 359. 



~) E. Fischer and E. Frankland Amstroxg. Ber. 34, 2885. 



•') Journ. Chem. Soc. 83, 1305. 1903. 



*) Ann. 331, 359. 



'-) Compare Martin Luwrv, Journ. Chem. Soc. 75, 212, (1899). 83, 13U. (1908). 



'') Several cliemisis (v. Lippmaxn, tlliemie d. Zuckerarlen 11, 130, 990. Ber. 29, 

 203, Trev, Z. f. phys. r.h. 18, 193, Simon, C. H. 132, 487), and myself (Ber. 28, 

 3081 (1895)) liave exprei^sed die opinion that with the three [y.\/,'a of + 106°, 

 + 53° and -|- 22 ,5 known foi- t^lucose, corresiiond three modilications, namely 

 two oxide forms and one aldehyde form. After llie research ol Mi. Ji;.\(ius on 

 the methylglucosides and the pentacetates it is jnactically certain thai the [ci]fj ot 



