so strange a diversity of opinions and customs that they never agree with one another.1
An idolater will not eat bread nor drink water in a house belonging to
anyone of a different caste from his own, unless it be more noble and
more exalted than his; thus they can all eat and drink in the houses of
the Brahmans,2 which are open to all the world. Among these
idolaters a caste is, so to speak, what a tribe was among the Jews, and
although it is commonly believed that there are seventy-two of these
castes, I have ascertained from the most accomplished of their priests
that they can be reduced to four principal castes from which all others
derive their origin.3
The
first caste is that of the Brahmans, the successors of the ancient
Brachmanes or philosophers of India, who specially studied astrology.
Their ancient books, in the reading of which the Brahmans generally
occupy themselves, still exist and they are so skilled in their
observations that they do not make a mistake of a minute in foretelling
eclipses of the sun and moon. In order to preserve this science among
themselves, they have a kind of university in a town called
1 This has ever been the strength of those who have conquered India.
*
Bramines in the original. Brahmans' houses are certainly not now open
to all the world ; the very reverse is the case. The accuracy of this
statement, even in Tavernier's time, may be doubted. True as it is that
a man of lower caste may eat from the hand of a Brahman, a Brahman has,
himself, to guard against defilement by contact with men of lower
caste. Dubois, who allows that Brahmans permit no stranger to enter
their kitchens, was admitted into their houses by Brahmans (Hindu Manners, Customs, and Ceremonies, ed. 1906. Introd. x. 181, 201 f.) and Mrs. Stevenson was admitted to eat in Brahman houses in Kathiawar (Bites of the Twice-born, 240). But these cases are exceptional.
3 The number of castes at the present day is enormous ; see the list in the Census Report, India, 1911,
vol. i, Part 2, pp. 177 ff., and Sir E. A. Gait's analysis of the
Statistics in Part 1, ch. xi. He remarks that earlier writers
'.accepted his view that the classes [Tavernier's four castes] had
gradually developed into castes. It has, however, been shown by Senart
and others that the division into castes has no direct relation with
the division into classes. The castes came into existence
independently, without regard to the classes.' The best short summary
of the question will be found in Sir E. A. Gait's article ' Caste' in
Hastings, Encyclopaedia of Religion and Ethics, vol. iii, p. 230 ff.
