To regard most Amoy 'changed tones' in the traditional sense as their underlying forms is already implicit in an attempt at reconstructing Proto-Min tones by the internal reconstruction method based mainly on the 'changed tones.' This view has also found a strong support in a typogeographical Study of modern Chinese dialects in conjunction with the Altaic linguistic structure which invaded into Chinese from the north and with She Tai and/or Austroasiatic contact observed throughout southern half of the Chinese speaking territory-a study which lead to the interpretation that the Amoy-type tone-sandhi is basically a syntagma-final modification of tones instead of the non-final alternation, We have pointed out in this study that there are two different types of tone sandhi in modern Chinese dialects, and the reliance on the non-final, so-called 'changed tones' in construing the underlying forms is a mere accident in the case of Amoy tones-what is more essential is to determine a series of underlying tones with which we should be able to account for the tonal phenomena related to the syntagma marking in Chinese in general. This should be, and in fact have been, done for any reasonable description of linguistic structures, and the case for Chinese can not be an exception.