chewing; half means five (foods), what the ancients called "the five right ones," viz. 1) gruel (, bread); 2) gruel of wheat or beans; 3) butter-cake (); 4) meat; 5) cakes ("rolls"); the five chewed foods are: 1) roots, 2) stalks, 3) leaves, 4) flowers, 5) fruit. If you eat the first five you may not eat the other five; but if you have eaten the second five, you may eat the first five. Milk and cream are not mentioned in both series; in the Vinaya there is even no special term for them; it is clear that they are not included. in the number of proper foods. If in some food, made of meal, the spoon will not enter, it is reckoned as cake; if in some dry meal food, mixed with water, traces of the fingers are seen, then it is reckoned among the five gulpings. Though I could not visit all the frontiers of the five Indias, that extend in each of the four quarters four hundred stations (yojana ?), yet by questions I know of the peculiarities of their gulping and chewing. In the northern country is sufficient flour, the western is rich in pancakes (cao), in Magadha is little flour but much rice, etc. They eat no garlic or salat in the five Indias, therefore they have no pain in the stomach nor obstructions. In the ten islands of the Southern Ocean the entertainments are more liberal still, often enduring for three days, chiefly if the king himself is host. But in China an entertainment (cai-fa) is held otherwise; there the host takes away the remnants, the clerics may not. Only when the host expressly bids them, they may do so with measure. After mid-day they read a short sūtra, and sometimes depart only before night. At parting they say "sādhu," and add (? they answer) to this, “a-nu-mo-to (anumoda).

In the northern countries with all the Hu, in Dukolo (Tokhara) and in the kingdom Suli, is yet this special custom, that the host first, with baldachins of different colours, makes offerings to the caityas, when all the troop go round (pradakshina), and one chosen from them says prayers. The form of the baldachins is described in the book "Description

1 The words are, of course, all I-tsing's own.

2 There is again a detailed description.

of the Western Countries" (Si-fan-tsi). But among the clergy are some that fulfil the dhūtas, beg alms, wear only three robes; if one should give them gold and silver, they throw it away like spittle; they hide their traces in inaccessible forests.

Note.-This means that even as late as the seventh century there were followers of those ancient Buddhists who had not adopted life in a monastery.

H. WENZEL.

VERSUCH EINES WÖRTERBUCHES DES TURK-DIALECT, VON DR. W. RADLOFF, OF THE RUSSIAN ACADEMY. 1st and 2nd Parts. St. Petersburg, 1889.

In the preface, in parallel Russian and German columns, the author tells us of the circumstances which led to, the object of, and his particular qualifications for, the compiling of this important and unique work. He resided many years at Vernoe in the employ of the State, and then moved to Kasán on the Volga, where he resided many years he resides at present at St. Petersburg, within the walls of the Academy: he had thus ample opportunities of informing himself of the different forms of language of the Turki Branch of the Ural-Altaic Family, and he has been labouring at the work since 1859, or thirty years, and, as the materials grew, he has made three distinct compilations: he has incorporated all the words contained in any of the works of previous authors. Even now he modestly describes his work, not as having any pretence to completeness, but as a "Versuch" or attempt, and yet it will consist of twenty to twenty-five Parts, each containing twenty sheets, or a total of many thousand quarto pages. It must be remembered, that it is a Comparative Dictionary, giving under each word the various forms, which are presented in each language, and every word has a distinct and independent entry. Each part costs one Rouble and twenty Kopeks.

The importance of this work cannot be over-estimated, and

1 西方記, a book unknown to us if it is not the same as the 西域記.

its appearance is most timely. Each sheet, as it passes through the Press, is submitted for the observations of Professor Ilminsky of Kasáu, Kasas in Sympheropol, Amirchanians, the well-known Bible-translator in Orenburg, Professor Budenz, in Buda-Pest, Kunes in Constantinople, Professor Baron von Rosen in St. Petersburg, and Professor Salemann, Librarian of the University of St. Peters. burg. Professor Vambéry of Buda-Pest, and Professor Pavet de Courteille of Paris, have also lent a helping hand.

Professor Radloff was good enough to present me with a copy of the two first fascicules at St. Petersburg, when I visited the Academy last September: it was peculiarly acceptable, as I had read a paper the previous week at the International Oriental Congress at Stockholm on the "Distribution of the Turki Branch of the Ural-Altaic Family of Languages," an effort to define accurately the LanguageFields of Central Asia, and this Comparative Dictionary with its accurate and carefully arranged word-store will greatly assist the inquiry.

The Osmanli-Turki, generally called Turkish, is but one, and the least interesting from a linguistic point of view, of a Branch, consisting of eight or nine languages, the features of which by the compilation of linguistic books, and of translations of the Bible, are becoming gradually known to

us.

Much still remains to be done, both as regards Dictionaries and Grammars, and Texts, and, as the whole of the Turki-speaking populations are slowly but certainly gravitating towards the Russian Empire, it is to Russian Scholars that we must look for the illustration of the phenomena of each language.

R. N. C.

JOURNAL

OF

THE ROYAL ASIATIC SOCIETY.

ART. V.-Early History of Kannada Literature. By B. LEWIS RICE, C.I.E., Bangalore.

AMONG the so-called Dravidian languages of Southern India none can boast of a higher antiquity in the cultivation of its literature than the Kannada or Karnâṭaka, commonly called Canarese by Europeans. And yet, while the sister languages Tamil and Telugu have their votaries, Kannada has received attention from but few, if any, among Oriental scholars. This neglect is no doubt partly due to its being principally spoken in Native States, whence it has come less into contact with Europeans, while the other languages form the media of official business through a large extent of British territory. An erroneous impression has, besides, been fostered by some writers, whose acquaintance with South Indian languages was probably chiefly confined to Tamil or Telugu, that these were in some way superior either in structure or in the contents of their literature to Kannada, a statement for which there is not the least foundation, and originating in the want of accurate information regarding the latter.

A few years ago some details regarding the earlier literature of the Kannada language, the result of researches for which opportunity arose in the course of work I had in hand, were published by me in the Society's Journal,1 but the fuller and more certain information I am now able to give,

1 See articles on The Poet Pampa, in Royal Asiatic Society's Journal for January, 1882, and Early Kannada Authors in the same Journal for July, 1883. VOL. XXII.-[NEW SERIES.]

17