SIMILARITIES IN WORDS. 523 
than these, or they will not get common-sense people to give 
them any heed. It is true that, while on the one hand 
mere unsupported similarity is no proof of connection, on 
the other hand the widest surface dissimilarity is no proof 
_ that the words are not connected; but when a man tells us 
that words so dissimilar as quassur and kootera, or as pel 
and parkulu, are the same word, he is bound to give tangible 
proof; and the only satisfactory way of doing this is to pro- 
duce the intermediate forms. 
This method (the presentation of the intermediate forins) 
often connects the most dissimilar words, and proves it dis- 
putably thei positive identity. Take, for instance, the 
dissimilar words ndalinga and taia. Each is the Melanesian 
word for ‘ear,’ and the two words are one and the same. 
In Fiji, and elsewhere, we find ndalinga; in Fate, it be- 
comes talinga; in Motu, it drops / and ng, and appears as 
taia. This, however, should be written ta’i’a, the apostrophe 
representing a queer catch in the breath, the ghost of a 
departed letter. Take, again, the Fijian taba and bee. 
How unlike they are! And yet they are the same word. 
In one Fijian dialect the word is taba, in another it is tebe, 
in a third tebe is reversed and becomes bete, and a fourth 
drops the t, making the word be’e. But the most striking 
instance of these unlike identities is the word for “ banana,” 
which appears at one end of the line as phitim, and at the 
other as un. Phitim and un! Were ever two words more 
unlike? Nevertheless, they are one and the same. Phitim 
appearsin Wallace’s Vocabularies of the Malay Archipelago, 
where. also, we find phudi—both meaning banana. In 
Fiji the 7 (ph) changes to v, and the d is strengthened by n, 
the word being vundi; in Aurora, the initial v is dropped, 
and the word becomes undi. Finally, in. Duke of. York 
Island it discards its second syllable, and presents itself as 
Un. 
Surface resemblances, often fanciful or specially manu- 
factured to suit, have had much to do with the etymological 
guesses of the men of the sun-myths, who based their 
theories oni their conjectures as to the meaning of the names 
of the old gods, demigods, and héroes, and got the sun, the 
rosy dawn, the lightning, and the cloud out of the ancient 
myths by what Professor Tiele calls ‘mere philological jeux 
d’esprit.’ Sober-minded investigators, such as Dr. E. B. 
Tylor, and others, have risen up in rebellion against them, 
and even Mex Miiller himself has been brought to book. 
Of course, I do not mean to deny that there are real sun- 
myths; there are plenty of them; but if a man tells me 
