Measurements of electron concentrations by the method of cyclotron resonance, and of no.-therm.I iron line intensities in emission, as a fanction of composition and dilation in the reaction zos~e qf low-pressure acetylene-oxygen--ar,qon mixtures, reveal a marked Mmilarity in behaviottr of these tu'. quan$itiea. On this and other circumstantial evidence, the possible causes of the attom,tlou.~' IDreversals van be narrowed down to two, neither of which can, however, be era, girl, red ohjectio.-[r~,ยข until further experintental evidence reveals in even more detail the nature of certain of the vยข,wtioJ, steps proposed. One of these schemes involves postulating the &imolecular production of an ยข.reitrd species Y* by the meclranism also responsible for electron production. Species Y* mu.~t he ,~uffieieml/! energetic, and be present in sufficient quantity, to produce excited metal atom M* b!! etterff.q tratt~.fi, i uqth M. The other---4, more specific sehenae--involves reaction betu'een M + and e to produce M*. the emitted radiation being discrete. Although there i.~ e~'perimental evidence for all the import..t steps involved in this second scheme, this schente, too. can fail if the cross sections ,[or e,*pture of .by M+ prove to be too small. Schemes so far postulated involving termoleeular reactiotv~ ttrv elimimtted as not satisfying the ea'perimental ob,~ervations pre.~ented.