^{page 175 note 1} I have been indebted for this assistance to Lieut. Eastwiek's English translation, communicated to me in manuscript.

^{page 175 note 2} I have indeed expressly stated in my examination of the Cuneiform vowels, a belief in the principle of interior vowel combination; but I had not at that time sufficiently examined the grammatical inflexions to be enabled to derive from the discovery its full phonetic value.

^{page 175 note 3} Professor Lassen terms the different grades tenues, aspirated tenues, and medials; but I prefer retaining the old nomenclature, which I have all along followed.

^{page 176 note 1} Hence we have the Greek &c.; the being, as I think, the true phonetic representative of the old Persian au.

^{page 178 note 1} Compare the Zend taokhma.

^{page 178 note 2} “A friend,” or “supporter,” probably from the Sans. “an arm,” with which are allied both the Pers. dush, “back,” and dost, “a friend.” I suspect also that the obscure daosha of the Zend is connected.

^{page 179 note 1} There is however much uncertainty attaching to this name, the Median and Babylonian copies evidently distinguishing it from the well-known title of Media. It may possibly denote the

^{page 180 note 1} The m in those languages opening on a vowel is certainly a labial.

^{page 180 note 2} I do not think, however, that the Cuneiform combines freely with the u. In the combination could not be avoided; but in paru for puras, the transposition of the vowel can only be explained by a repugnance to the groupe The only second example of it is in an imperfect name in the geographical list at Nakhsh-i-Rustam.

^{page 180 note 3} The orthography of this word, it must be remembered, is in Pehlevi and Parsi, kup.

^{page 182 note 1} Compare the Zend raocha.

^{page 182 note 2} As in the Sanskrit superlatives and

^{page 182 note 3} The conversion of the primitive vowel in the first four tenses is common to several of the Sanskrit conjugations.

^{page 182 note 4} The ush and aush of the old Persian, are of course equal to the us and os of the Sanskrit.

^{page 184 note 1} Respecting the terminations in ish and ush, I have already observed, that the sibilant is aspirated by the force of the preceding vowel, and is then retained in virtue of its aspiration.

^{page 184 note 2} There is also a difficulty with regard to the final r. From the Median and Babylonian orthography, I should certainly be disposed to read the name of the Babylonian king Nabukudrachar, and the Sans. should give us atar rather than atara; but on the other hand, according to the above classification, r is a surd, and if final should be elided.

^{page 184 note 3} I do not yet understand upon what principle the old Persian wrote pri instead of fri, (of pru I am doubtful); but I think I see that where an aspiration was necessary, and such an aspiration would disfigure the word, the language rejected the r and retained the p in its Bimple form; compare pati and palish, for prati and

^{page 186 note 1} I do not think it necessary to notice the anomalous compounds, or and