324 FEEJEE GROUP. 
This of course can be considered only as an approximation, but I 
am inclined to believe it rather above than below the actual number 
of inhabitants. It will be perceived that I have set down no more than 
five thousand for the number of inhabitants of the interior, although 
there are a number of persons who believe that this portion of the 
large islands is densely peopled. But all my own observations tend to 
confirm me in the opinion, that there are very few inhabitants in the 
interior of these islands. The circumstances attending a residence 
there are so contrary to Feejee habits, that I cannot give credit to a 
statement so entirely at variance with what we find at the other 
Polynesian islands. The food that the natives most esteem, is gathered 
near the sea-shore and from the sea, and there is little probability that 
any persons would dwell in the interior unless compelled by necessity. 
The natives of the different islands are of various sizes: some have 
their forms more fully developed than others, as will have been seen. 
In the opinion of the white residents, the natives of Ovolau were thought 
to be of inferior size to those of the other islands; this, however, did 
not strike us particularly, and I was of opinion that they were a fair 
specimen of the natives of the group. Those who have Tonga blood 
are designated as the Vitonga, and are decidedly the best-looking 
natives that are met with. These are to be found more among the 
eastern islands than elsewhere, showing the effects of the intercourse. 
Our accounts of the language are derived from the missionaries, 
who are making great exertions to become thoroughly acquainted 
with it, in its different dialects, of which there are several in the 
group. They have found more than ordinary difficulty in bringing 
the language into a written shape, and have not yet fully completed the 
task. The characters they have employed for this purpose are the 
- Roman, and they have made such changes in the usual sounds of some 
of the letters, as are absolutely necessary to express the peculiar sounds 
of the Feejee tongue. The vowels are used generally to express the 
sounds they denote in the French language, except the broad sound of 
the a, which that letter is not always confined to; 6 is used to represent 
the sound m’b; c, that of the Greek 6; d is sounded n’d; g, n’g. Of 
all the letters, r and s retain most closely the sounds by which they are 
known to us; ¢ has a peculiar sound, partaking of th, and in some of 
the districts is not used at all. The sound of & is entirely wanting in 
the Somu-somu dialect, whilst it is much used and distinctly uttered in 
the others. 
In the Lakemba dialect they use the 7, sounded nja, which they 
derive from the Tongese. 
The following is the alphabet adopted by the missionaries. It con- 
