The low yield of XX is partly accounted for if XIX reacts further with ethylmagnesium bromide giving XXI-A or XXI-B. XXI-A seems the more probable of the two possible structures on the basis of the probable relative acidities of the hydrogen atoms involved and the tendency for maximum dispersal of any charge associated with the-MgBr moieties. The feasibility of a reaction of this kind The conversion of isopropyl#-tolyl sulfone