was also 10-4 M in leuco crystal violet and 2 X 10~1 234 M in toluenesulfonic acid. From the amount of crystal violet dye obtained as determined spectrally, and from the reaction stoichiometry (1 mol of dye per biimidazole dissociated) and quantum yield of unity,9 both experimentally determined, radical concentration immediately after the flash was obtained.Rates of Triarylimidazolyl Radical Dimerization.-A Perkin-Elmer "Spectrachord" Model 400A was set up in a dark room to record absorbance (at of each radical) vs. time. A solution of biimidazole in carefully purified benzene was irradiated until a steady state was established and absorbance was recorded for several half-lives and determined eventual!)-at "t = ® ." From the usual second-order expression, the slope of 1/A. plotted vs. t gave k/e (cell length = 1 cm) reported in Table I. From these values k was calculated.Extreme care was necessary to free solvents of impurities which react with radicals. Normally, three distillations of reagent grade solvent sufficed. Even with carefully purified solvents, radical dimerization was not the exclusive fate of radicals in the media studied. A small but detectable deviation from strict second-order was apparent sometime during the second half-life4 and A" was slightly greater than A measured prior to irradiation. Rate constants were taken from the first half-life.Rates in 50:50 vol methanol-benzene were measured both in the spectrometer and in the flash photolysis apparatus used to determine extinction coefficients. An oscillographic trace of absorbance vs. time was photographed.