OF ARTS AND SCIENCES. 171 



" A labial consonant seldom appreciated is the aspirate (or spirant) 

 of B, the German w and Spanish h in certain words, as Cordoba, Ha- 

 lana, a sound confounded with their v (also a Spanish sound) by the 

 English. Mr. Ellis spent a year in Germany without discovering the 

 difference, (which he now admits, but usually disregards in his ex- 

 amples of German,) — a singular fact in a professed phonetician, when 

 the most unlettered part of the population of Pennsylvania speaking 

 German and English make the proper distinction. Even the savage 

 aborigines inhabiting the frontiers of Mexico and the United States, 

 pronounce the sound in question with perfect accuracy in words taken 

 from the Spanish, as that for horse, which, in German characters, is 

 kawajo, the I of the original being dropped in Mexican Spanish, as in 

 corresponding French words. This aspirate b occurs naturally in 

 Weko and Konza. The Spanish grammarians have an imperfect idea 

 of this sound, which they insist is a b, because it is not the labio- 

 dental V. This language doubtless takes it from the Latin, of which, 

 as well as of Greek, it seems to be the digamma (j), the small letter 

 of which was probably two gammas (y/), from which resulted w, 

 which character, under this view, has not arisen from the repetition of 

 V. A critical review of the Greek orthography of Latin names, 

 and the reverse, will, I think, confirm the views here taken. I would 

 represent this sound by b surmounted by ". 



" The corresponding aspirate of P (undoubtedly the Greek phi) is 

 heard in German when/ follows p, as in pfropf, kopftoeh, dampfboot. 



" S is a post-aspirate of t, the theoretical aspirate of which is inter- 

 mediate in power to tlieta and s. The lene of th (in this) is to be 

 looked for in the Spanish d, in those cases where it is believed to have 

 the power of sonant th. The purely English notation sh (which is in 

 no case etymologic) is founded in error; as this consonant does not 

 bear the relation to s which the annexed h indicates in other cases. 

 Sh belongs to a more posterior contact than s, and there is an allied 

 sound formed still farther back, with the jaw more open. It is the 

 Arabic sad, (Hebrew tsadai,) mistaken by Mr. Ellis for an ordinary 

 s strongly hissed, overlooked in his Polish and Russian, and in San- 

 scrit mistaken for German ch (p. 56) and for sh (p. 133). I have 

 heard it in Polish, Chinese, and Nadako. 



" The French j (the sonant of sh) I have heard in but two 

 American languages, the Konza and Wyandot. 



" Sonant s (as in roses, has, is, his, loise) should not be represent- 



