8 On the Power of the Letter p. [No. 9, new series. 



Supposing that a probable case has been made out to prove that 

 the letter p is a surd of the * t' family having some power differ- 

 ent from <s> and i— , it remains (a.) to account for its present popular 

 pronunciation; and then (b.) to endeavour to ascertain the parti- 

 cular power of this 't.' 



(a.) With regard to the former of these it is not unimportant 

 to remark that there appears to be a certain definite direction in 

 which many letters have a natural tendency to corrupt them- 

 selves : and it may be added by the way, that the rules of this 

 species of cacophony, which may probably be found to be appli- 

 cable to all languages, appear to be well worthy of investigation 

 with a view to assist in determining the etymology of words of 

 corrupted pronunciation whose corruption has been fossilized by 

 a corresponding spelling. — In the particular case of the aspirated 

 sound of a ' t' (soft 1 th') when it occurs in the middle of a word, 

 (the circumstances of our doubtful p), there is a considerable 

 tendency amongst uneducated persons and children to change that 

 sound into a trilled one. Instances of this might be pointed out 

 in the English language. In the southern Tamil country, unedu- 

 cated Native Christians frequently pronounce the iSjgir ' Father* of 

 the opening sentence of the Lord's prayer as ' Pira :' Guir^i i2>, 

 ' enough,' is very commonly pronounced ' Porum' by all classes : 

 and e&GG)<£ ' seed,' is almost universally pronounced ' Virei' in con- 

 versation : and so also of other similar words. 



Since, therefore, there is a tendency to corrupt the aspirated 

 sound of one ' t' into that of a trilled letter, it beco mes not impro- 

 bable that a' t' of another power may also have been subjected 

 to the same influence. The less easy the proper pronunciation of 

 such a letter, the less likely it would be to save itself from this 

 influence, especially when used by those whose lips and to ngue 

 make convenience rather than propriety their rule of utterance : 

 and our p, being more difficult of pronunciation than the would 

 be the more likely to fix itself in that convenient form of corrupt- 

 ed pronunciation towards which it naturally tends. 



There seem to be reasons for thinking that illustrations of the 

 position advocated in this paper may be gathered from the Telugu 

 and Canarese languages, in cases of words of the same original 



