THE LANGUAGES OF THE INDIAN ARCHIPELAGO. 211 
The 4th rule is one on which the preservation of simplicity 
and uniformity, and the avoidance of an extraordinary and 
inelegant typography, mainly depend, but it will not always 
be consistent with the extension of one system over many 
languages, because the sounds that are most common in one 
may not be so in another. While therefore the above rules 
may be always fully carried out in dictionaries, and works 
confined to particular languages, it will not be possible, in all 
cases, to do so in a system for a family of languages, without 
some deviations from the orthography that may be advisable 
for certain of its members. The system for the family, will 
always, however, be that for the majority of the members, 
because it will be drawn from their vocal character. 
The characters of the Roman alphabet furnish symbols for all 
the most common sounds of the known languages of the Archi¬ 
pelago, and the majority of the other sounds may be considered 
to be modifications and combinations of these. For the expres¬ 
sion of each kind of modification a fixed markshould be adopted, 
and as the degrees of modification are very great, and are not 
capable of exact expression save by the voice, each ol these 
supplementary marks should indicate not so much a rigidly 
determinate and unvarying sound, as the change in the sounds 
of the letters produced by certain definite actions of the vocal 
organs. Thus there should be a guttural sign to indicate a 
deeply guttural pronunciation of the gutturals, and a guttural 
one of the others which are susceptible of it. These signs 
should be of such a nature as to admit of being doubled to in¬ 
tensify their sounds, although this will only require to be 
resorted to in comparative philology, and should be reserved 
for cases of very marked excess above the ordinary degrees 
of strength. 
The following scheme, although recommended for adoption 
in vocabularies &c which may be sent for publication in this 
Journal, is entitled provisional, because it may be found 
advisable to modify it when our knowledge of the languages 
of the Archipelago becomes more extensive and profound. 
This however is not likely to be the case, as, while aiming 
chiefly at the expression of all the sounds of the principal 
languages, we have had in our view many of the others in 
different parts of the Archipelago. 
Vowels . 
The great difficulty occurs in the discrimination and repre 
sentation of the vowel sounds. We have been anxious, as far 
■! possible, to extend to the Archipelagic languages Sir 
