and classified at various epochs, notably in 1836, 1844 and 1866. They vary in the several cantons, presenting matter well worthy of study in detail. M. Dareste has only lightly

touched on the subject. We will conclude, he remarks, in emphasizing Prof. Kovalefsky's remark that the criminal law of the Ossetes offers a perfect analogy with ancient IndoEuropean codes, and particularly with the ancient laws of Ireland recently published. All these monuments of the past illustrate and explain each other, and the points of comparison met with among the people of the Caucasus are all the more precious because they show us living institutions.

I need only say a few more words in conclusion. I am indebted to the present article for an acquaintance with its author, Prof. Kovalefsky, who has kindly read over my MS., and suggested two or three notes by way of elucidation. His knowledge of jurisprudence, of which he was for many years Professor at the University of Moscow, enables him to speak with great authority on all the customs of the semicivilized inhabitants of the Caucasus, among whom he has made several journeys. The results of his last year's travels are published in some pamphlets on the Pshaves and Khevsurs, and he has also communicated some results of his earlier observation in two articles published in the Vestnik Evropii.

ART. X.-The Languages spoken in the Zarafshan Valley in Russian Turkistan. By R. N. CUST, LL.D., M.R.A.S.

In the course of my reading preparatory to my proposed trip in September next to Orenberg and the Steppes of Central Asia, I came upon the valuable book by Dr. Radloff, "Aus Siberien," Leipzig, 1884. His account of the languages spoken in the Zarafshan Valley, of which the famous city of Samarkand is the capital, seems so important, that I have had it translated from the German for publication in the Journal. In Vol. XVIII. of the Journal, 1886, pp. 177-195, I communicated a paper on the Geographical Distribution of the Túrki Languages, but some points required elucidation, which are cleared up by Dr. Radloff's remarks. His statements with regard to the Tájik or Persian portion of the population of the Valley is very important, as being fresh, and no doubt accurate.

May 10th, 1888.

ROBERT N. CUST, Hon. Secretary.

The population of the Zarafshan Valley may be divided into two groups, according to its languages: 1stly, races speaking Túrki languages, and 2ndly, races speaking Persian. The former constitute the largest portion of the population, while the latter are scattered about in various parts. The Persian-speaking inhabitants as a rule go by the name of Tájik; even the uncultured country people and nomad tribes call the Persian language Tájik-til (Tájik language).

The Tájik generally only inhabit the towns, and only busy themselves with commerce and handicraft. They have

their origin partly from very old Persian emigrants or freed Persian slaves, who were sold in great numbers every year, by the Turkomans in the Khanates. The newer Persian inhabitants are generally called Iran, and are partly, although perhaps secretly, Shiah.

The chief seats of the Tájik (by which general name I here comprise all the Persians) are the cities of Khojend on the Syr Daria (which separates the Tájik town from the Uzbek town, inhabited nearly wholly by Tájik) and Samarkand. The inner town of Samarkand is almost solely inhabited by Tajik, and Persian is the prevailing language there. The Tájik fill the western gardens and the neighbourhood of Samarkand almost exclusively as well; however there are a few Iran villages, which make the silk-worm their chief industry. I have in vain endeavoured to find from the Tájik their tribal names, neither do they know anything about their early history.

As a remnant of the former population, I may mention the so-called Mountain-Tájik or Galcha, who inhabit districts more or less extensive. It seems as if these ancient inhabitants had saved themselves from the stream of new-comers by keeping to these high-lying districts. As far as I could tell, these great Tájik settlements are situated as follows: 1) one day's journey from Kokand on the way to Dauan, the following villages were pointed out to me, Schaidan, Babadurchan, and Yangas; 2) in the south-west mountains of Tashkend; 3) on the north border of the Kara-Tag; 4) along the upper course of the Zarafshan, east of Pentshikend, where the population was called Galcha and Kara Tegin. Unfortunately I could not visit any of these Persian settlements, so I cannot give any further details about them. However, as far as I can know, these people differ very little from the other inhabitants in their customs and ways of living.

The Túrki inhabitants of the Zarafshan Valley consist for the most part of Uzbek, with the exception of a few places on the Nurpai, where there are several important Arab settlements, but which have long succumbed to Túrki influence, and even speak their language.

The chief Uzbek tribes are by no means strictly separate from one another, but are considerably intermixed, always keeping together as a race. A fact of this is, that a great number of the Kiptchak names, which often, as one clearly sees, have been given after the names of small tribal fractions. However, I consider it superfluous to discuss this fictitious genealogy further; one thing I will yet mention. The Khan of Bokhara, who came from the Mangyt tribe, is said to sit on a felt cloth every time he ascends his throne, the four corners of which are held by delegates from the four branch tribes.

From what I have said, two languages, the Persian and Túrki, are spoken in the Zarafshan Valley, and in Transoxiana generally. I do not venture to give any decided opinion about the Persian; however it seems that the Persian spoken here differs very little from the written Persian. As to the Túrki languages, there are four here: the Kirghiz, the Kara-Kalpak, the Turkoman, and the Jagatai or Uzbek. The three former are closely related, while the Uzbek differs considerably from them; the latter alone is a literary language. Of course, throughout the large area which the Uzbek inhabit, there must be some variety in the dialects; however, generally speaking, this language may be taken as a whole; at any rate, the inhabitants of Bokhara and the Sarts of Turkistan understand each other well enough, which is not to be surprised at, when one considers the long and constant intercourse between the towns of Central Asia.

As to the purity of the Túrki language, it is in the steppes that it is the most purely spoken, where it has not yet been permeated by the civilization of Islam, the destroyer of language and of national spirit. The language of the Kirghiz is the least poisoned with Arabic and Persian words, and whatever foreign elements they may have taken up, they have completely assimilated. However, they have been invaded in many parts by foreigners, which fact is proved by the fact of the Kirghiz living in close proximity to the Sart.

As regards purity of language, next to the language of

the Kirghiz stands that of the Kara-Kalpak, and the Turkoman in the Nurata mountains, although the tribes have succumbed considerably to the Uzbek in all their social intercourse, and have therefore incorporated many loanwords.

The language of the Uzbek residents of the Zarafshan Valley is not nearly as pure as those which I have just mentioned. Arabic and Persian expressions are used a great deal, even by the non-educated. In the towns this language is the most disfigured, as it is considered good style to borrow foreign expressions. The higher the society, the more does the language get debased, so that to an outsider it sounds like a different language. Not only are an innumerable amount of foreign expressions used here, but the grammatical structure is changed. The harmony of names has been quite destroyed, and changed to please the foreigners. It seems as if learning itself had required this unnatural course, for the Mulla forces upon the reading scholars pronunciation contrary to that of Túrki, and severely denounces the correct intonation. Thus it comes about that the less learned people read according to the Mulla's instruction.

It seems as if learning had the object of eradicating the language of the people. The ordinary man, who reads and writes without being learned, does so in Túrki. However, as soon as he knows more, he turns his back in disgust on these reputed signs of ignorance, and gives himself up to the study of Persian. This half-educated man still writes in Túrki, as he is not yet sufficient master of the Persian. He only reads the Korán in Arabic, learns prayers, and works through a few Arabic books, provided with a translation. However, if he gets as far as the Arabic Grammar, and gets to know a little of Arabic, he then neglects the Persian, and gives all his time to Arabic, the aim of every learned man. The greater scholars generally only write in Persian, troubling themselves very little if the receiver of the epistle understands Persian or not. Very often this unfortunate man is obliged to find a Mulla first to translate the writing of his correspondent. All the official business