Experimental SectionThe compounds were prepared and purified by methods which have been described! The nitrosobenzenes used in the present investigation had the following physical properties: 2,6-dibromo-, mp 134-135" (lit! mp 135-136'), e040.54 (763 mp); 2,6-dibromo-4-methyl-, mp 136-138" (lit! mp 136.5-138'), €0 42.57 (758 mp); 2,6-di-bromo-4-chlor3-, mp 110-111' (lit! mp 110-111'), eo 42.75 (773 m p ) ; 2,4,6-tribromo-, mp 122-124" (lit? mp 122-123'), co 43.48 (772 mp) [lit .4 co 43.5 (775 mp)].Infrared spectra of the crystalline dimers were run as Nujol mulls and KBr or KI pellets. The spectrometer was a grating instrument with resolution of about 1-2 cm-l (Perkin-Elmer 337 G). Peak positions were reproducible to within f l cm-1 and were the same in Nujol mull or pressed disk.Spectrophotometric measurements were carried out and temperatures were controlled to within &0.05' of 25' as described previously.1 Stock solutions were made up at 25' and dilutions were made (also at 25") by means of calibrated pipets and volumetric flasks. Practice runs on nitrosobenzene itself (which does not dimerize in dilute solution) showed that Beer's law was obeyed closely on diluting stock solutions in this way, so that solvent losses due to evaporation and other dilution errors were not serious.Using experimental optical densities and concentrations, p!ots of D / C L us. D 2 / C L 2 were constructed and "best" straight lines were determined by least squares. The intercepts on the D / C L axes gave eo, and, from this, individual values of cy, the fraction of nitroso compound actually present in solution as monomer, could be calculated from experimental optical densities. Knowing cy, equilibrium constants were computed from the following relationship where C is the concentration of total nitroso compound, all figured as monomcsr.K O = 2C(cy2/1cy)