THE POLYNESIAN ALPHABET. 



23 



These percentages are presented in the following table ; 



Table 2. 



It would yield no valuable results to pursue more extensively such 

 comparisons of an isolating and an analytic speech, yet there is an 

 interest in this simple exhibition of the extent and manner wherein 

 they differ. To sum up, we may note that for every 25 Samoan vowels 

 that speech makes use of but 17 consonants, whereas in our own speech 

 we employ 46 consonants to every 30 vowels, a striking illustration of 

 the difference between vocalic and consonantal speech. The simpler 

 language employs its two palatals almost half as much again as we 

 use our richer supply of palatals, even after the sacrifice of several 

 through disuse. When we examine the two series which we employ 

 with such beautiful precision we are struck with the lack of develop- 

 ment which is the characteristic of the beginning speech : the Samoan 

 employs his tongue but half as much as we, and his lips are but three 

 quarters as much occupied as are ours. 



Now when we look more closely into the column of Proto-Samoan 

 linguals and note the play of mutation we shall make an interesting 

 discovery. In the descendant languages the lingual pair r-l (in fact 

 the apparent pair is really a triplet because of my discovery of the 

 early existence of r-grasseye) hands down but one of its members at a 

 time ; some use I and some r, none uses both. Next we find a frequency 

 of mutation each way between the nasal and the semivowel, n-l and 

 l-n respectively. We therefore establish the first lingual closure at a 

 point equidistant from liquid and nasal. 



A second pair exists in reference to the aspirate and sibilant. The 

 languages of Nuclear Polynesia which have both are Tonga and Uvea ; 

 Samoa and Futuna have s only; Niue has h doing service for both. 

 When we leave Nuclear Polynesia and pass to the homes of Tongafiti 

 folk we find that, with the sole exception of little Manahiki, the sibilant 

 is an impossibility and is replaced consistently by the aspiration. 



So with the mute pair. Here the surd dominates; the sonant d 

 occurs consistently in Viti alone, and even there can not stand without 

 a preface of the nasal of the lingual series, nd. 



Thus we see that the tongue is used in pianissimo and fortissimo 

 expression, and that particularly in the former its superior flexibility 

 and the ease and beginning accuracy of its control enable the man to 

 produce at least two distinct sounds. Between the two limiting 

 extremes an intermediate sound has become possible, the sibilant; 



