ON ETHNOLOGY. 
341 
Except in the second, third, seventh and ninth class of Sanscrit verbs, and 
some ancient forms in the Veda, we find in all the other verbs the shortest 
form of the infiectional base in the second person singular imperative. 'I’hese 
four classes however comprise only between 150 to 200 roots, while of the 
six other classes the first comprises alone about 1000 roots. 
As to the second point, iiobody denies that nengoli, as u modern language, 
employs periphrasticul formations instead of the simpler forms of the ancient 
language; but when the form of the parueiple as well as of the auxiliary 
verb, arc of Sanscrit origin, I do not see why their cnmbmaiion into a peri- 
phrastical form should point to a barbarous origin. That the Bengalis em* 
ploy an aorist which denotes past, present and future time, is not at. all 
extraordinary, since we sec in many languages that when a new specific 
inflectional base has been assigned to the present and imperfect, the simple 
form represents generally the action of the verb only, without reference to 
any time, and is therefore called aorist. 
As to the next point, the negative verb in Bengali has nothing irregular 
in its formation. By the addition of verbal terminations almost every word 
may become in modern languages a verbal base, and Dr. Stevenson must be 
aware, that like the negative verb ndi, I am not, from nd, not, there is also 
an affirmative verb, ta{i, I am indeed, from va{a, indeed. The loss of the 
reduplicative syllabic in the perfect is sufficiently accounted for by the same 
occurrence in almost all the modern, and even some of the ancient branches 
of the lndo*Germanic family; and in sup|)o»ing an auxiliary verb like 
d'tdhalc to be the original form of the lerniinations of tlie past tense, like da 
or ta, Dr. Stevenson seems not to ho .aware of his quite being in accordance 
with Professor Bopp, only that the latter takes not the modern Marathi form, 
d'ldhnle, but the ancient and simple form dhd. 
An element, whicli might perhaps be called aristocratic, has exercised 
much influence in the personal terminations of the verb. The Bengali gram¬ 
marians pretend that there are two sorts of terminations equally employed 
for the singular and plural. One sort convey a kind of respectful meaning, 
the other has a contemptuous sense. This distinction is so generally adopted, 
not only by the graramarians but also by those who have w ritten in the lan¬ 
guage, that we do not dare to pursue any other method, although we are 
convinced that the forms which convey contempt are nothing else but the 
singular ones, while those which express respect are the plmml. To explain 
this distribution of the ancient forms, it is m)t sufficient for us to have re¬ 
course to iie analogy of modern languages, in some of which, in speaking 
with resiJcct of persons, we may perceive that a sense of superiority has been 
often attributed to the plural; for the peculiarity of Bengali consists not so 
much in a verb in the plural being connected with a substantive in the sin¬ 
gular, but that a verb in the singular is governed by .a substantive in the 
plural. J his peetdiarity does not admit of explanation, except on the ground 
that Bengali, to speak correctly, has properly no plural in declertsions except 
in the case of reasonable beiugs. It would therefore be impossible to em¬ 
ploy a third person plural, when we arc speaking of animals or of inanimate 
objects. We might say, the wise men think (think plural), but if we wished 
to express the idea ‘ that animals eat ’ (eat plund), we must say, the mass of 
animals eats (verbally, the animal-mass or animality eats) ; hence, in this 
way, the plural of the verb will always find iwelf united with the names of 
superior beings, and the singular with those of inferior beings, and thus the 
two numbers of the verb must assume, by little and little, the peculiar cha¬ 
racter of the substantives on which they are dependent. 
