1 
100 ON THE ORIGINAL INHABITANTS 
the older Prakrit phase."* While Prakrit is indefinite, Sans- 
krit is definite and becomes in consequence ossified and 
unchangeable. Eventually it loses its hold on the people, 
but remains the linguistic standard of the educated and the 
dialect of the learned. It supplies in its turn the material 
for a modern Prakrit, which may likewise contain some 
relics of the original Prakrit, but from which, as prior to 
Sanskrit, it must be distinguished. 
Applying these remarks to the special subject before us, 
it is not at all impossible that, as the Graudian Kanda has 
been changed in Sanskrit into Khanda, similarly the original 
Dravidian and ancient Prakrit word Palla has been already 
at an early date altered and become Palha and Pahia, which 
three different terms were then in use at one and the same 
time. Sanskrit prefers on the whole a form whose pronun- 
ciation is more difiicult than what satisfies the Dravidian 
languages. Some of these changes may have been made for 
reasons of which we are now ignorant. In support of my 
supposition that Palha or Pahla is a modification of Palla, 
I contend that a similar connection does apparently exist 
between the names Kalhana or Kahkina and Kalla ; between 
Balhana, BaUu,Balhika, Balhlka, Bdlhi, &c., or Bah/ana, BaliH, 
Bahlikd, Bahlllm, Bdhli, &c., and Balla ; between Bilham 
(Vllham) ox BihJam {Vihlnna) and Billa, [Villa); between 
Malhana or Mahlana and 2IaUa ; between iSi/haiia or ^ihhna 
and iSilla ; and between Sulhana, Suhlam or SuIIana and au 
original Sulla. The names ending in n like Balhana. Kal- 
hana, Malhana and Sulhana have some resemblance with 
those Dravidian names ending in anna, as Eilghanna, Nag- 
anna, &c. Of the change of double / iuto the change of 
Malldri into Mall/dri in Marathi afi^ords an example. 
^' For instance compare ^'r/A-aWsa ■with krika^^sii, piiroddsa vrith p\irolasa, 
ksMakn with ksudraka and bhaRaksa with bhadrdksa, in Professor A. Weber's 
Jiidischc Sli((fie>i, II, p. 87, note. 
I 
