ed a
man. The old Finnish and Samoyede names for God--Ukko and Num--perhaps
belong to this stage of thought.]

[Footnote 106: See the account of the Faunus message in this book.]

[Footnote 107: The chief exception in Sanskrit is the Rajatarangini, a
chronicle of Kashmir composed in 1148 A.D. There are also a few
panegyrics of contemporary monarchs, such as the Harshacarita of Bana,
and some of the Puranas (especially the Matsya and Vayu) contain
historical material. See Vincent Smith, _Early History of India_, chap.
I, sect. II, and _Pargiter Dynasties of the Kali Age_. The Greek and
Roman accounts of Ancient India have been collected by McCrindle in six
volumes 1877-1901.]

[Footnote 108: The inscriptions of the Chola Kings however (c. 1000
A.D.) seem to boast of conquests to the East of India. See Coedes "Le
royaume de Crivijaya" in _B.E.F.E.O._ 1918]

[Footnote 109: Very different opinions have been held as to whether this
date should be approximately 1500 B.C. or 3000 B.C. The strong
resemblance of the hymns of the Rig Veda to those of the Avesta is in
favour of the less ancient date, but the date of the Gathas can hardly
be regarded as certain.]

[Footnote 110: Linguistically there seems to be two distinct divisions,
the Dravidians and the Munda (Kolarian).]

[Footnote 111: The affinity between the Dravidian and Ural-Altaic groups
of languages has often been suggested but has met with scepticism. Any
adequate treatment of this question demands a comparison of the earliest
forms known in both groups and as to this I have no pretension to speak.
But circumstances have led me to acquire at different times some
practical acquaintance with Turkish and Finnish as well as a slight
literary knowledge of Tamil and having these data I cannot help being
struck by the general similarity shown in the structure both of words
and of sentences (particularly the use of gerunds and the constructions
which replace relative sentences) and by some resemblances in
vocabulary. On the other hand the pronouns and consequently the
conjugation of verbs show remarkable differences. But the curious Brahui
language, which is classed as Dravidian, has negative forms in which
_pa_ is inserted into the verb, as in Yakut Turkish, e.g. Yakut
_bis-pa-ppin_, I do not cut; Brahui _khan-pa-ra_, I do not see. The
plural of nouns in Brahui uses the suffixes _k_ and _t_ which are found
in the Finnish group and in Hungarian.]

[Footnote 112: See t