ON ETHNOLOGY. 
343 
The terrainaiion of the preterit, ildvi, and of the future (tbo), bear the 
greatest resemblance to the corresponding Latin forms (bam and bo), and it 
tnight be possible to adopt the origin, assigned to these forms by Prof. Bopp, 
also for the Bengali, so that the terminations of these two tenses, small and 
insignificant as they are, might be shown to contain not only the remains ol 
personal pronouns, but also those of an obliterated auxiliary verb. As to 
the future, there is externally not any difference between the pronunciation 
of the Latin bo and Bengali va (pronounced bo)\ and die Latin bam might be 
compared with the termination of the Bengali preterit Idm, when we re- 
mcniber that instances are not wanting of the semivowel I having taken in 
other languages the place of an original v. Thus Sanscrit stiap, to sleep, 
German tcklaf ; Sanscrit vad, to say, Gothic lath-dn. 
I prefer liowever another explanation of this tense, which is more in 
accordance with die development wliich the form of the past tense has taken 
in Pracrit as well as in other modem languages of India. There is a 
change of letters which is of very frequent occurrence in the popular dia¬ 
lects of India, I mean die traasition of the dental t into a lingual ff> which, 
according to its pronunciation, may often be represented by / or r. Adopt¬ 
ing this theory, Bengali forms, like karlldm, I did, would contain a past par¬ 
ticiple with an active signification (Sanscrit Bengali tari/o), fol¬ 
lowed by the secotitJary personal terminations, lhat tliis is the real ori^n 
of these fonns may he put above every contradiction by u comparison with 
the Mahratta preterit, which, as Professor Lassen has shown, has preserved 
even the tliree genders of the participle, saying to held, he did, U keli, she 
did, Icm keleih, it did, and even in the first person, mi kelom, I did (masc.), 
mi icclem, I did (fern.), mi kelom (neuter), This being the case, and seeing 
that ill Pracrit already all the original luatorical tenses arc altogether lost, 
1 think that even the Pracrit termination of the preterite la or (a, either in 
an absolute or neuter statu, or followed by a masculine termination, may be 
taken for a corrupted form of a past participle, for we see that in Pracrit 
the termination ta is already changed into da, and that sometimes the t is 
entirely supiiresaed, as osuria for osarida. 
The conditional seutnB to contain in the I of its terminations the remains 
of u present participial form, to whicli the secondary personal terminations 
are added hi the same manner as in other clearly periphrastieal formations, 
of which we shall Imve to speak directly. This tciiac is also interesting on 
account of its having pr«M;rvud, in the second person singular, the final s, 
which is dropped in the correspoiiduig form of the imperfect. 
Besides the conjugation by means of the simple terminations, the Bengali 
language lias yet two other conjugations, which are periphraBtical in the 
proper sense of the word, although here also Ute two component parts are 
more intimately allied tlian in periphrastieal formations of the Latin, (like 
factus sum,) or of the French (likeyV <«i< faUyj'aiJ'aii). These conjugations, 
which furnish some tenses only, are tbrmed, the one by adding the auxiliary 
verb to the participle, the other by adding it to the past adverb. The 
auxiliary verb is achi, 1 uin, used in the present aud preterite only. 
This auxiliary verb Achi, the &atuu as tlie Sanscrit asmi, asi, asli, Doric e/ip/, 
iooi, iiTi, Latin jmjw, es, tsl, cannot be regarded as a primitive root. Lan¬ 
guage, representing us it does the images of all tilings or actions, which by the 
energy of their iiuprcaaion upon the mind are able to excite an idea (elcos, 
plionic image), has not nnd cannot hove a word which expresses the abstract 
and lifeless notion of mere .aftirniation or existence. Bat as the development 
of the mind advances, step by step, with that of language, and vice versa, we 
