436 C. Hansch and R. M. Muir 



tuents in the auxins. In radical attack substituents have only slight 

 effect on positions rneta to them. In the unsubstitutecl acids one 

 would expect phenoxyacetic acid to be more susceptible to radical 

 attack than phenylacetic acid and theiefore to be more active as an 

 auxin if radical attack were involved. Again, just the reverse is 

 found. Molecular orbital calculations (10) have also indicated that 

 radical attack seems less likely for the benzoic acids than attack by an 

 electron pair. 



If radical attack and electrophilic attack on the ring do not seem 

 to offer an explanation for the effect of substituents on auxin rings, 

 what evidence is there to support nucleophilic attack? Our early A\ork 

 (14) with the phenylacetic, indoleacetic, and phenoxyacetic acids in- 

 dicated that whenever both ortlio positions were substituted with 

 halogen, alkyl, or alkoxy groups, inactive molecules always resulted. 

 This, plus the relative activating effect of substituents, led us to focus 

 attention on reaction at an ortho position by an electron-rich reagent. 

 This view was modified (27) when it was discovered that a-2,6-di- 

 chlorophenoxypropionic acids were active (28). It seems most likely 

 that a para position must be involved in this reaction since introduc- 

 tion of a group in the para position always destroys the activity of a- 

 2,6-dichlorophenoxypropionic acids in the Avena test. Some of the 

 2,4,6-trisubstituted phenoxyacetic acids show weak activity in the slit 

 pea test. The phenylacetic acids seem to follow the same pattern (30), 

 the 2,6-disubstitutecl acids being active and the 2,4,6-trisubstituted 

 ones being inactive. A very interesting and significant difference is 

 apparent in the requirements for activity with these two structures. 

 2,6-Dichlorophenylacetic acid is quite active, while 2,6-dichlorophen- 

 oxyacetic acid is at best very weakly active. If one assumes two- 

 point reaction by means of the carboxyl and the 4 position of these 

 molecules with a single substrate molecule, a logical explanation is 

 possible. In the phenylacetic acids the two hydrogens on the methyl- 

 ene group interfere with the ortlio halogens, locking the methylene 

 group so that the carboxyl group may be held rather rigidly over the 

 ring near the 4 position. In the phenoxyacetic acids, the relatively 

 small oxygen atom holds the methylene group out far enough so that 

 interference between the methylene hydrogens and the two ortho 

 halogens is slight, and the whole side chain has freer motion. Intro- 

 duction of the a-methyl group thus increases the biological activity 

 because hydrogens on the methyl group would help lock the side 

 chain so that the carboxyl group would be more rigidly held for 

 two-point contact involving the 4 position. All of the data at hand, 

 therefore, would indicate that a specific atom in the ring is involved 

 in the two-point reaction. 



If we assume then that normally an ortho position in the phenoxy- 



