agi. 
TRANSACTIONS OF THE SECTIONS. 139 
frequentative verb is formed by duplication; it is du-du. This gives (du-du) ku, 
analogous to (du) ku of the preceding line; and for this we have, in the great in- 
scription of Tiglath Pileser I. (du-mis) ku. In this instance, it is impossible to assign 
any value to mis but ‘‘twice.’’ The double Accadian form mi, mis, analogous to the 
Sanskrit dvi, dvis, and to the similar forms in Greek and Latin, is one argument out 
of many for the closer connexion of the Accadian with the Indo-European lan- 
guages than with the Semitic and Egyptian languages. The form in 7 occurs in all 
the languages ; the adverbial development in is is Japhetic, being common to both 
the Accadian and the Indo-European languages; but the adjectival declension is 
exclusively Indo-European. 
Having obtained the Accadian form from the inscriptions, the author investigated 
the other forms in the verbal pedigree by inference from this form and the Semitic 
and Indo-European forms. There can be no question that the two last radicals of 
the primitive Semitic root for “‘two” were?}. Before these some languages had {¥, 
and others [); the Arabs having their peculiar letter Tha. Again, there can be no 
question that the primitive Indo-European form had v or w, preceded by a dental, 
before i. Now m is interchangeable both with its fellow-nasal x and with its cog- 
nate labial » or w; whereas no direct interchange could exist between m and v or w. 
It necessarily follows that the primeval form had m before 7; and as a double change 
in passing from the primeval to the primitive Semitic is not admissible, and as zm is 
an unpronounceable combination, these two forms must have been ¢thmi and thni, the 
th being pronounced as in “‘thin.’”? The Japhetic form must have been thmi, and 
the primitive Indo-European thwi; but in the last two words the th may have been 
sounded as in “thy ;” and probably was so, as it is represented by d in the great 
majority of existing forms. 
This may serve as a specimen of the manner in which verbal pedigrees have been 
formed. A list is given further on of Accadian words with the primitive Indo-Euro- 
pean and Japhetic words, to which they and the known Indo-European words that 
correspond to them conduct us. It is necessary, however, to ascertain the original 
values of certain Semitic letters which occur in the Accadian words. It will be seen 
that the values which some of these letters had subsequently to the Babylonish cap- 
tivity, were very different from what they had when the Accadian language was first 
committed to writing. 
In the Achzmenian inscriptions Samech has the value 5, and Shin SH. The 
Sibilant in the name of Darius and the former of those in that of Hystaspes, fig. 1, 
are represented by Shin, and in the first kind of Persepolitan writing by SH; the 
latter of the Sibilants in the name of Hystaspes is represented by Samech, and by S. 
This is in conformity with received opinion; but when we go back to the Assyrian 
inscriptions, we find a very different state of things. There, in the Assyrian repre- 
sentations of foreign proper names, the Hebrew Shin is represented by Samech; 
although in the roots which were common to the two languages Samech corresponds 
to Samech, and Shin to Shin, no matter where the diacritical point of the Masoretes 
be placed. This is invariably the case where the Shin precedes a vowel or an ordi-« 
nary consonant, or terminates a word. Where, however, it precedes Kaph or Qoph, 
Shin and Samech seem to be used indiscriminately in Assyrian. To establish these 
facts, from which he inferred that Shin was always pronounced by the Assyrians as 
S, and that they used SK (which they represented by Samech) for the Hebrew SH, 
the author exhibited the following transcriptions of Hebrew names, occurring in Assy- 
rian inscriptions, with the vowels used by the Assyrians represented by points. 
The Masoretic pointing of the Hebrew he thought it useless to set down. 
770N TION NID NID 
ow T7O7N pow Palale) 
rab) oP) {pwr TYPOS or “WE 
wor) DOMNZ pwnt pvoi 
sel a yen qe or: yee 
_In all these Hebrew words the Shin is pointed to represent SH. When S was to 
be represented in a foreign name, Ezra, or some more recent editor, has invariably 
