THE LANGUAGES CF THE INDIAN ARCHIPELAGO 217 
phisms, under which letters and words originally identical now 
shew themselves in the different members of the group. The 
first and lowest step, that of ascertaining the actual sounds of 
each language, is thus directly connected with the highest 
branch of the ethnology of the Archipelago. When examin¬ 
ing the vocal character of the separate languages, and ascer¬ 
taining the laws of transmutation, we shall find necessary 
a much more delicate balancing of particular elementary 
sounds, and a more complete reference of each to its physiolo¬ 
gical origin, than will suffice for our present purpose. 
In the languages of the Archipelago many vowel sounds 
occur which are different from those in the more current Eu¬ 
ropean languages. The latter contain every pure vowel sound 
and several compound ones, but the same letter represents dif¬ 
ferent sounds, and this is the case in the English language to 
an extraordinary and absurd extent. The three pure elementary 
vowels i, a and u will receive the Italian sounds, being those 
most generally prevalent in Europe, and which English writers 
on eastern languages adopt. All other vowels may be con¬ 
sidered compounds of these, and the only mode of representing 
them is by availing ourselves of the Roman composite as 
well as simple vowels, and expressing all other double and 
maltiple vocalic combinations by them. Where one of these 
Roman sounds preponderates, we must use the vowel which 
stands for it, and when we come to consider the vocalic trans¬ 
mutations, we must have recourse to description for more 
accurate explanations. To give a single example. The in¬ 
herent vowel sound in the Javanese consonants is neither a 
nor o. It is a combination of both, the latter being itself a 
vowel of two elements. As the o preponderates in this sound, 
we use it as the nearest approximation to its correct expressi¬ 
on. Mr Crawfurd uses a , de Groot a with a small o over 
it, a useful make shift which we may occasionally adopt, but 
which is too clumsy to become part of a general system of 
orthography. It may be- difficult for those who have never 
heard the sound, to conceive how the two vowels can melt into 
a third which is a perfect vowel and not a dipthong. Such 
is the case however^ and there are numerous other instances 
of a similar kind in the different Archipelagic languages. 
Some of these will be noticed in the elucidations annexed to 
our table of vowel sounds, and to these remarks we refer for 
all that we think it necessary at present to indicate respecting 
their organic characteristics, the only sound basis for the 
study of the vowels. The ear and the imitative power of the 
voice differ so much in different individuals that a reliance on 
