The solvolysis rates of 1-aryl-1-(p-methoxyphenyl)-2,2,2-trifluoroethyl and 1-aryl-1-(p-phenoxyphenyl)-2,2,2-trifluoroethyl bromides and chlorides were conductimetrically measured at 25.0 Β°C in 80% aqueous ethanol. The solvolysis rates of 1-(p-methylphenyl)-fixed series were also set up to analyze the substituent effect. The substituent effects on these Y-series of solvolyses were analyzed on the basis of the Yukawa-Tsuno equation. The pmethoxyphenyl-fixed series showed a linear correlation with r = Γ1.7 and r = 1.0 for the substituent range less reactive than p-Me including all meta substituents, and a discrete correlation with r = Γ3.9 and r = 1.26 for the strong p-donor class substituents more reactive than p-PhO. A similar bilinear correlation was obtained for the Y = p-PhO series. The partial correlations for the strong p-donor class substituents in both Y-series are nearly the same as the correlation r = Γ4.2 and r = 1.2 for symmetrically disubstituted (X = Y) series. These bilinear (or non-linear) correlations were explicable in terms of changes in the coplanarity of the two benzene rings depending upon the resonance capabilities of X and Y substituents. For strong p-donor class X substituents as strongly p-electron donating as the Y substituent, both aryl rings are equivalently twisted as in the symmetrical X = Y series, and for less electrondonating X substituents, the fixed p-methoxyphenyl (or p-phenoxyphenyl) ring remains in a coplanar conformation and X-phenyl is twisted more out of coplanarity. In the fixed Y = p-Me series, a significantly non-linear correlation was found, comprising three distinct correlations characteristic of three conformers, a conformer with the X-aryl coplanar and Y-aryl twisted, a symmetrical conformer with X being a weak or moderate p-electron donor and both aryls being equivalently twisted, and a conformer with Y-aryl coplanar and X-aryl twisted. The relative stabilities of these conformers were estimated by the ab initio MO optimization of the mono-p-methoxycarbenium ion, confirming the conclusions above.