228 Transactio7is of the Royal Society of South Africa. 
it fit all the observations within the experimental error, which is usually not 
more than one unit in the wave-length reading. 
In any case, my theory in Part I, that it is molecular volume that is the 
main factor in colour- change, must be abandoned. It is obvious that when 
such different groups as methyl, isopropyl, chlorine, bromine and iodine 
produce almost the same [effect, the explanation must be of an unusual 
character. The only thing these groups have in common is monovalency or 
an electron, but this is common to all groups substituting a benzene ring in 
the ordinary manner, and is therefore not in itself characteristic enough 
for an explanation. 
