Personal Pronouns.—Georgian (with Lazian, etc.) makes no distinction of genders in personal pronouns (as in general Georgian makes no distinction of genders). Herein Georgian corresponds perfectly to Sumerian. The personal pronouns in Sumerian are also very like to Georgian, though the same cannot be said of other pronouns. The 1st person singular is regularly me-e in Sumerian, and in Georgian me = I = Lazian ma, man = Mingrelian ma = Svanian mi. The 2nd person is in Sumerian za-e = Georgian šen = Mingrelian si = Lazian si = Svanian si. As to the 3rd person, it is most interesting that Georgian and the other languages of the Georgian group have no special pronoun to designate directly the 3rd person; in Sumerian it is the same. Both Sumerian and Georgian borrow the pronoun for the 3rd person from demonstrative pronouns: Sumerian ni (rectus) and na, (obliquas) for persons, and bi and ba for inanimate objects; Georgian is, igi, Mingrelian , ena, θena, Lazian ham, Svanian ada, ala. But in Georgian the root of the pronoun of the 3rd person appears in the genitive, dative, and other cases, and this root being m we can compare it with Sumerian demonstrative bi and ba. Indeed, we have in Georgian , is igi = he she, it, but in genitive m-is, dative m-as, etc.; in Mingrelian , ena, θena = he, she, it, but in genitive m-u-ši, etc. In Lazian as independent personal pronoun 3rd person, the demonstrative ham, = this, is used, but the pronominal nominative is m-u-q, genitive m-u-ši, etc. It must be remembered, moreover, that in Lazian the demonstrative pronouns have the root of the personal pronoun 3rd person m, and that is why they replace the personal pronoun 3rd person.

This collection of ancient Tibetan documents, of which I have been engaged in preparing an inventory under an arrangement sanctioned by the India Office, contains close on two thousand pieces, none of them probably of a later date than the ninth century A.D., and is certain to shed a flood of new light on Tibetan archæology, history, grammar, culture, religion, and folklore. Most of the documents were found at two sites, viz. Mīrān and Mazārtāgh. Mīrān is situated a few miles south of the westernmost end of the present Lop-nōr marshes, while Mazār-tāgh is a low barren ridge rising in the middle of the Taklamakān desert, on the left bank of the Khotan River. In his Ruins of Desert Cathay, vol. i, pp. 350 seqq., 439 seqq., Sir Aurel Stein has described the remains of the ruined fort which yielded these Tibetan records at the former site. For an account of the excavations which brought to light Tibetan and other documents in abundance from the refuse-layers adjoining the small ruined station on the Mazār-tāgh hill, vol. ii, pp. 417 seqq., may be consulted.

The conquest of the Persian and half of the Byzantine Empire by the Arabs, under the banner of Islam in the seventh century, was one of the most extraordinary events in the history of the world. On the one side were ranged the forces of two highly-organized military powers, Imperial New Rome and Imperial Persia, which for over three centuries had been engaged in constant conflict with each other. Although this necessarily tended to exhaust the material resources of the combatants, it would naturally be supposed that it must have given them military experience, and their leaders a training in generalship, adequate to enable them to face with confidence of victory enemies hitherto regarded with contempt as mere barbarians. On the other side we see hosts of men, reared in a country where the conditions of life have always been of the hardest and most precarious, divided by tribal feuds and secular hatreds, poorly armed, with no practice in warfare against disciplined foes, and with no allies to swell their legions. Yet from the beginning the progress of the Arabs was one of almost uninterrupted success.

Captain L. Molyneux-Seel has been good enough to send me a photograph of a Vannic inscription discovered by him built into the wall of a ruined Armenian chapel at Baghin, north of the Murad Su. Baghin is on the right bank of the Kighi Su, which falls into the Murad Su near Kharput, and it is about 50 miles north-east of the latter city and almost due north of Palu, where an inscription of Menuas has long been known to exist (No. XXXIII of my memoir). The new inscription is important, since it not only adds a new word, litiani, “a boundary-stone,” to the Vannic vocabulary, but it also fixes the western boundary-line of the Vannic kingdom in the reign of, Menuas.

There is no lack of numismatic evidence for the fact that the words Kūšān šāhān šāh and Kūšān šāh were used as personal attributes by certain princes reigning about the year A.D. 300. On the obverse of one of the coins in question Professor Marquart reads the legend: mazdēsn bagē Pērōzē (i) wažurg Kūšān šāh, “des mazdaverehrenden Gottes Pērōz, grossen Kōnigs der Kūšān.” M. Drouin translates the saine legend as follows: “le mazdéen, le divin Pérose, grand Kouchan, roi.”