The object of the work described was to discriminate between the two principal rival theorics concerning the polymerisat ion of 1.3-dioxacycloalkanes by anhydrous pcrchloric acid, the "Mainz theory" and the "Keele theory". By means of Jucrck.s's method for determining tertiary oxonium ions we found that in polymerisations under the driest conditions the concentration of these is negligibly small. This was done with 1.3-dioxolane (1). 4-methyl-l.3-dioxolane (4). and 1.3-dioxcpane (S), and the findings arc supported by determinations of the content of hydroxy groups of polymers prepared and "killed" under different conditions.
It follows that when the water content of the reaction mixtures is less than ca. 10-'M the propagation is principally by secondary oxonium ions, and that it must therefore go by the ring-expansion mechanism of Plrsch and W(>srcwnanti.
Water causes the formation of tcrtiary oxonium ions and this probably explains why previously other workers and oursclvcs had reported thcir presence, and sometimes their dominance, in reaction mixtures prepared under much less stringent conditions.
The recent theory of P c w c w k . which in~olves co-ordinated oxy-carbcnium ions. is also shown not to be applicable to the systems considered herc. ---The heuristic value of the ring-expansion theory is illustratcd briefly by reference to a new method of synthesising crown ethers from 1.3-dioxacycloalkanes. which arose from it.
ZUSAMMENFASSU N G :
Der Zweck der vorliegenden Arbeit war, cine Entscheidung zwischcn den ..Mainz"und ,,Kcele-Thcorien" uber die Polymerisation dcr I ,3-Dioxacycloalkane durch wasscrfrcie Pcrchlorslurc in Methylendichlorid herbeiiufiihrcn. Mittels dcr Juucks'schcn Methode zur Bestimmung von tertiPren Oxoniumioncn cntdeckten wir, dall deren Konzcntration verschwindcnd klein ist, wenn unter peinlichstem AusschluB von Feuchtigkeit polymerisicrt wird. Wir crzicltcn diese Resultate mit 1.3-Dioxolan (I), 4-~ethyl-1,3-dioxolan