440 C. Hansch and R. M. Muir 



better illustration of tiie greater reactivity for two-point attachment 

 of the a position in naphthalene over that of the (3 comes from the 

 work of Luckwill and W^oodcock (21). These workers confirmed the 

 earlier report (15) that l,3-dichloro-2-naphthoxyacetic acid is com- 

 })letely inactive, as would be expected. The superiority of the a posi- 

 tion over the (3 position for two-point attachment was shown by com- 

 paring l-chloro-2-naphthoxyacetic and 3-chloro-2-naphthoxyacetic 

 acids. The latter isomer, where a attachment is possible, is highly 

 active, while the former compound, where p attachment is demanded, 

 is almost inert. The interesting observation that introduction of 

 chlorine in the 8 position of 1-naphthaleneacetic acid greatly lowers 

 the activity of this molecule (15), can also be interpreted in terms of 

 resonance stabilization of a reaction intermediate. Thus 8 substitu- 

 tion blocks a reaction and forces reaction at the less active (3 (2) posi- 

 tion. One would expect the 8 position in 1-naphthaleneacetic acid 

 to be more favorably placed for reaction than the 2 position. The 

 reasoning behind this is that if reaction occurs at the 8 position, one 

 can consider the structure to be related to cinnamic acid, while if re- 

 action occurs at the 2 position, one must consider the reaction with 

 respect to the side chain to be more like that of phenylacetic acid. 

 C/Vcinnamic acid has been shown to be at least five times as reactive 

 as phenylacetic acid in the pea test (19). Similar reasoning can be 

 used to rationalize the difference in activity between 1- and 2-naph- 

 thaleneacetic acids. Also the fact that partial hydrogenation of the 

 naphthalene ring leads to lower activity (36) may be interpreted as 

 decreasing the ability of the molecule to stabilize an intermediate 

 through charge delocalization. The observation (31) that indole-2- 

 acetic acid is much less active than the isomeric indole-3-acetic acid 

 again lends support to the above hypothesis. Reaction at the 3 posi- 

 tion would not permit as effective charge delocalization as woidd re- 

 action at tlie 2 position. It is noteworthy that delocalization of a 

 negative chaige with reaction at the 3 position would be even less 

 favorable than delocalization of a positive charge. That the 2 position 

 in the indole series is more favorable for reaction than the 4, is indi- 

 cated by the fact that 2-methylindole-3-acetic acid is less active than 

 lAA, while 4-chloroindole-3-acetic acid is more active than lAA (26). 

 Again such reasoning can be used to rationalize the difference in 

 activity found with 2-thianaphtheneacetic and 3-thianaphtheneacetic 

 acids (18). 



Although the overwhelming preponderance of evidence supports 

 our i^vo-point attachment hypothesis, two important molecules which 

 do not fit neatly into place are 2,4-dichloro-6-fluorophenoxyacetic 

 add anil 3,5-dichloro-2-pyridoxyacetic acid (37). In the latter mole- 



