UraBI BB are H wSEm SSI ■■ CskVK 5SB g gg j hmGh RnDD § n ■9SJSB Wm w!MBtv» KB Ken raPUHffll IHHS3 r^fccccroWOTftWfi Glass PgiQl Book .E- 2a _ /3 /* ' art CHINA'S PLACE IN PHILOLOGY. CHINA'S PLACE IN PHILOLOGY AN ATTEMPT TO SHOW THAT THE LANGUAGES OF EUROPE AND ASIA HAVE A COMMON ORIGIN. BY JOSEPH EDKINS, B.A., of the London Missionary Society, Peking ; Honorary Member of the Asiatic Societies of London and Shanghai, and of the Ethnological Society of France. LONDON: TRUBNER & CO., 8 and 60, PATERNOSTER ROW. 1871. All rights reserved. Tzo\ E.Z n " AND THE WHOLE EARTH WAS OF ONE LANGUAGE, AND OF one speech." — Genesis xi. 1. ' l GOD HATH MADE OF ONE BLOOD ALL NATIONS OF MEN FOR TO DWELL ON ALL THE FACE OF THE EARTH, AND HATH DETER- MINED THE TIMES BEFORE APPOINTED, AND THE BOUNDS OF THEIR HABITATION." Acts XVli. 26. 'AAA' 5 [xkv AlOiowas nereKiaOe rtiXod' iovras, AldioiraSy ro\ SixBa Sefiaiarai 4, " to rub," " handle," are from the same root, and it appears in Chinese as Ma, " to handle," " touch," and Mo, " to brandish the arms." This therefore is the most probable explanation of the word Magi, the common appellation of the Persian magicians. In the same ancient Chinese records the arts of weaving and of working in metals are mentioned. The "Tribute of Yu" 2 says of Tsing cheu in Shantung, "Its articles of tribute were salt, fine grass-cloth, and the productions of the sea of various kinds, with silk, fine hemp, lead, pine-wood, and strange stones. The Lai barbarians are shepherds. They brought in their baskets silk from the mountain mulberry." They seem at that early period to have been acquainted with all the ordinary metals. The date of the "Tribute of Yu" is given by the Chinese b.c 2205. From South- eastern and Western China came, as tribute to the emperor of that time, gold, silver, copper, iron, tusks, 1 Perhaps this tone of disparagement may be taken as an indication of later composition. It occurs in a long speech attributed to the sage Yi Yin. 2 Legge's " Shoo king," Part I., p. 102, slightly altered. KINGS WERE PRIESTS. 9 hides, feathers, cinnabar, timber, and various fabrics of flax, hemp, and hair. In the time of Joshua, B.C. 1450, Babylonish garments were conveyed to Judaea (Josh. vii.). The Persian race in Persia Proper, and as colonists in Turkestan, have always manufactured elegant woollen carpets. The Chinese ordinary word to "weave" is ^ Chi, old sound Teh; Latin, texo, " weave." To " build " is gg Chu, old sound Tok. Its being placed under the bamboo-class symbol sug- gests that it was the custom to make bamboo hedges. They interlaced thin bamboo stems and used stouter ones as posts. The same word was applied to brick, earthen, and stone walls : for instance, the walls of cities. The Greeks called a carpenter re/crow. It is highly probable that the kings of Nineveh acted as priests. " As in Egypt, they may have been regarded as the representatives on earth of the deity, receiving their power directly from the gods, and being the organ of communication between them and their subjects." 1 In China there is no doubt on the point that the emperor has always borne a sacred character, and acted as a priest between God and the people, praying for them in times of distress, and acknow- ledging guilt on his own and their behalf. The tombs of the Chinese emperors are remarkable. They are vast conical mounds of earth from a quarter of a mile to a mile in circuit. A long arched passage 1 Layard's " Nineveh." 10 through brick- work leads up to the mound door. Over the passage is the monument inscribed with the title of the emperor. The hall in front, where worship is per- formed, is magnificent in size and appearance. Before this is another smaller hall. The tomb entrance, halls, courts, gates, and boundary walls are all on a large and complete scale. The tombs of the Lydian kings were something of the same kind. a The remains of that of Alyattes still stand near Sardis. The sepul- chral chamber is surmounted by a lofty pile, and so far it is like the pyramids, but as the pile is a mound it is more like the tumuli or barrows of the western world. The basement consists of immense blocks of stone, above which is a heap of earth, surmounted by five pillars carved with inscriptions. The ground plan is a circle three-quarters of a mile round, a little larger than the great pyramid. The sepulchral chamber in the centre of the tumulus is eleven feet long, by eight feet broad, and seven feet high." See Raw- linson, quoted in P. Smith's History. Several of the ancient Chinese emperors, fabulous and historic, have funeral mounds assigned to them by tradition. That of Yu, the great engineering emperor, is near JNmgpo. That of Fu hi is near Kai feng fu. The tomb of Yau is in Shansi, and that of Shau hau in Shantung. The identi- fication of the tombs of these ancient princes cannot be relied upon without excavations. But the custom of bury- ing the emperors in vast sepulchral mounds of earth is PERMANENCE OF CHINESE CIVILIZATION. 11 thus shown to be a custom as old as that of erecting large terrace altars for sacrifices, as already described. To suppose that the Chinese originated indepen- dently the arts and usages to which allusion has now been made, isjx) assign. iwj^J^ginui3n,gaJa„ t a_jmany" branched civilization which is one in its main features. At this stage of archseological inquiry in Europe it is preferable, when we accept the conclusion now generally arrived at, that it was the Cushites, the brothers of the Egyptians, who commenced and de- veloped the Babylonian civilization, to proceed to class the Chinese with them. The likeness found to exist in practical bent, in the arts of life, and in all the solid elements of the old-world regime is sufficient to justify this step. If the Chinese did not bring with them to their new country all the arts mentioned, at least they came away with the same sort of mind and the same instinctive impulses. With a perseverance and enthusiasm which insured success, they laboured tri- umphantly for science and for the arts. More fortunate than the inhabitants of Babylon and Thebes, they have never seen the wreck of their institutions or the ex- tinction of their national existence. In this they were favoured by their isolated position and the compact mass of their immense population. "No Indo-European races approached them. The aborigines they found in the country, and the races that occupied Tartary, Tibet, and the Birmese peninsula, have always been inferior to 12 china's place in philology. themselves. "When vanquished and subdued by Tartar races, they taught their conquerors the Chinese civil- ization, and when they became enervated by it, easily drove them back to their native wilds. With a wise foresight, two centuries before the Christian era, they abandoned feudalism, and adopted the centralization system of government, which they have ever since retained. When merchants brought them paper, and probably ink, of Greek manufacture from the West in the Han dynasty, they at once began to make them for themselves. The cumbersome bamboo tablets and coarse paint which were formerly used, they exchanged for wolf's-hair pencils and Indian ink, the modern implements of writing. They gave up war chariots, as did our ancestors, and commenced the use of cannon in place of catapults and battering rams. They discovered the properties of the loadstone, and probably applied it to navigation in the Indian Ocean several centuries before the mariner's compass was thought of in Europe. It was by these and such like improvements on their old institutions that the Chinese have kept pace with the ages, and prevented the fabric of their ancestral civilization from crumbling to irremediable decay. CHAPTER II. Comparison with Western Asia continued. — Resemblances in Philosophy and Religion. — Numerical Philosophy. — The Nine Categories op the Hung Fan. — Measures. — Practical Genius. — Astrology. — Cycles. — Early Religion op the World. — Monotheism and Burnt Sacrifices in Genesis, Job, and the Shu King. — Sabeanism. — Angels, — Evil Spirits. — Chinese Burnt Offerings to Shang Ti. — Worship of Spirits and of Visible Nature. — These Customs were brought from the West. — Worship of Ancestors in Temples. The numerous and very remarkable resemblances found to exist between the ancient Chinese philosophy and religion and those of Western Asia constitute a powerful proof of early connexion. There are many and very detailed allusions in the Chinese " Shu king," the most important of the classics, to the philosophy and religion current among the people in the second millennium before the Christian era. The philosophy was in one aspect numerical. The five elements are alluded to as the five energies. Hing (old sound, Gang), " to walk," "to act," may be trans- lated " elemental activities." They are water, fire, metal, wood, and earth, or the five powers supposed to inhere in these substances. Then we meet with the 14 five relationships, namely, those of prince and subject, father and son, husband and wife, elder and younger brother, and the bond of friendship. The following extract, somewhat altered, from Legge's " Shoo king," p. 79, will illustrate the usage of the numbers four and five, etc , in common phrases : — " The emperor Shun said to Yu, You, my ministers, are my legs and arms, my eyes and ears. I wish to help and protect my people. You assist me. I wish to proclaim the powerful efficacy of my government through the" four quarters. You act for me. I wish to see the emblematic figures of the ancients : the sun, moon, and stars (illumination), the mountains (security), dragons (variety), and pheasants (beauty), painted on the upper garment, the tiger of the ancestral temple (filial piety), the aquatic grass (purity), fire (bright- ness), rice (the support of life), the hatchet (legal decision), and the symbol of discrimination, consisting of two representations of the character £, hi, placed back to back, thus, 5H embroidered on the lower garment. They should be figured with five colours, splendidly distributed among the five colours for the imperial robes. It is for you to adjust them plainly. I wish to hear the six pipes, the five sounds, the eight kinds of musical instruments, and the seven begin- nings, in order that poems, made according to the scale of five sounds, may go forth from the Court and be brought in from the people. Hear this." THE HUNG FAN. 15 In the Hung Fan 1 is found the most comprehensive statement on the old numerical philosophy to be met with in any ancient book. It is said to have been received by Wu wang, b.c. 1100, from Ki tsze, who informed him that Heaven gave it to Ta yu, b.c 2200, as a reward for his success in subduing the inundations of the rivers, and that the orderly arrangement of the moralities and social relations might thereby be com- pleted. THE NINE CATEGORIES OF THE HUNG FAN. I. Five Elemental Energies. water moistens and goes down salt fire blazes and ascends bitter wood crooked and straight sour metal obeys and changes acrid earth sowing and reaping sweet II. Five Human Actions. expression respectful venerable qualities speaking persuasive order seeing clear prudence hearing intelligent deliberation thinking profound wisdom III. Eight Departments of Government. food commodities sacrifices works instruction crime ■ guests the army IV. Five Registers of Time. years months days stars calendar V. The emperor's perfection in virtue, or himself attaining the summit of virtue 1 Legge's "Shoo king," p. 320. 16 china's place in philology. VI. The Three Virtues. uprightness times of peace prevailing by firmness times of violence and resistance j or _, e £ es . erve prevailing by mildness times of harmony and compliance j °l- n Ji! ?Jf n intelligent VII. Investigation of Doubts by the Tortoise and Diviner' 's grass. Tortoise {Chan), Tam. i^'Z^L^ ^ ^^ ^^ v " ( connexion, crossing. Diviner's grass (Pw), Pole. Two marks : solidity, repentance. VIII. Five Natural Indications. rain sunshine warmth cold times IX. Five Kinds of Happiness and Six of Misery. long life riches health and peace love of virtue submissively accomplishing to the end the will of heaven accidental death sickness grief poverty wickedness weakness. The Pa kwa, or system of whole and broken strokes in groups of three, arranged octangularly, was a set of symbols intended to represent a very ancient philoso- phy, consisting partly of physics, partly of morality, and partly of divination. It is the basis of the " Book of Changes," the time-honoured text-book of the masculine and feminine or dual philosophy. There are two other schemes of strokes - and lines, called the "Ho t'u" and the "Lo shu," maps fabled to have come, the one out of the Yellow River, and the other, the Lo, one of its Honan tributaries. But none of these can be compared in value with the Nine Categories of the Hung Fan, if it be desired to see at one view the forms of ancient Chinese thought. In the sphere of physics, the sages of this nation saw THE HUNG FAN. 17 five poivers moving through heaven an m w a $ 34 china's place in philology. are allied to the Siamese. They reside in the south- west of Kwei cheu province. The Lo lo, a very old and extended tribe, in the north-west of the same province, are connected with the Burmese. The first personal pronoun in these dialects varies between ye, ku, and nau ; while the second is very frequently meu and meng, so that they are cut off from any Indo- European or Tartar connexion. Logan has conferred a great service on philology by his division of the Himalaic languages into two branches, eastern and western. He states that the eastern or Mon Anam branch has some radical pecu- liarities in structure, and has been deeply influenced, first, by the Dravidian family, and, secondly, by the Chinese. The area of this branch is Cochin China, Pegu, Siam, and Cambodia. Farther south, at the peninsula of Malacca, it meets the Malay, which con- stitutes the type of the Australasian and Polynesian languages. The western Himalaic branch retains the same characters in Tibet, India, and Ultra- India, and is more Scythic than the eastern. Most of the migrations of races have been in the direction of radii from the common centre where the first human pair was created, and where the first gene- rations of their posterity lived. Along one radius came the Dravidian races, and after them the Hindoos, by way of the Punjaub into India. The Eastern and Western Himalaic peoples, after traversing Tibet, CHINESE AREA. 35 Dassed along the valley of the Brahmaputra into Ultra-India ; settling not only in that peninsula, but Drobably also in Southern China, where the Miau, Lo lo, Nung, Yau, and other tribes are their descen- lants. The Chinese, taking a more northerly route, lame along the lands watered by the Turkestan rivers ill they reached the north-western corner of China Proper. They met with the Jung in Western China, ;he modern Sz c'hwen. These people have left descen- lants in the Nung, one of the most celebrated branches )f the Miau. The equivalent of the Mandarin initial T is in old Chinese N and JVL In 778 B.C., the Jung were powerful enough to kill an emperor 1 at the sapital, which was then in Southern Shen si. They vere soon afterwards driven back. At present one of ;he tribes bordering on Yunnan is called Nu i, or ;he Nu barbarians. Here the same name occurs, but without the final ng. The Chinese also met in their earliest wanderings other sections of the Himalaic nigration in Hunan, viz., the three Miau tribes, and ;he Lai and Nung in North-eastern China. The old lames in China of rivers, tribes, and mountains are jut one word, and appear to have been all mono- .yllabic. From this it may be inferred that the various aborigines all spoke monosyllabic languages. The present spoken language of the Chinese, as ised over two-thirds of China, is called Kwan hica, and 1 Legge's "Shoo king," p. 615. 36 china's place in philology. by Europeans, the Mandarin dialect. The term Man- darin is of Portuguese origin and means commander. Indispensable as it seems, it is a name which cannot be defended, except on the ground of convenience. The northern Mandarin is spoken in the capital and in the four north-eastern provinces : Chili, Shantung, Shansi, and Honan. It has also spread itself through Manchuria and parts of Mongolia by colonization. The Manchus in Girin and Kwantung form but a tithe of the population, and have long forgotten their native tongue. The northern Mandarin is also spoken partially in Shensi and Hupe. The western Mandarin is spoken in Kansu and Sz c'hwen, Kweicheu and Yunnan, and partially in Shensi and Hupe. Portions of Kwangsi and Hunan also belong to its area. In the modern Mandarin language the old sonant initials g, d, b } v, z, have disappeared. But they are retained in the dictionaries of the book language. So also the final letters k, t, p, once abundant at the end of syllables, have been entirely lost over the whole of Northern and Western China. The northern and western Mandarin are differen- tiated principally by the Ju sheng tone class. The immense group of words, amounting to nearly a fourth of the vocabulary, belonging to this tone class are in the northern Mandarin irregularly distributed among the other four classes. In the western, on the other THE OLD MIDDLE DIALECT. 37 hand, they have all gone to swell the Hia ping class, which has thus come to embrace about a third of the entire vocabulary. The southern Mandarin, which retains the Ju sheng as a distinct class, prevails at Nanking, in the north part of Kiangsu and Anhwei, and partially in Hunan. Its area is a belt of varying width, extending from the ocean at the mouth of the Yang tsz kiang and the old' mouth of the Yellow River to Chang sha in Hunan. The old middle dialect is spoken at Sucheu, Shang- hai, Hangcheu, and Mngpo, and has the distinctive characteristic of possessing the old thirty-six initials and four tones as used in the syllabic spelling. Kan gin's Dictionary and the native tonic dictionaries all register an ancient pronunciation, which, so far as the initials and the medials are concerned, is best represented at present by the old middle dialect. Its area embraces Chekiang and the southern part of Kiangsu. It then proceeds westward through Anhwei and Kiangsi into Hunan, where, near the boundary of Sz c'hwen, it meets the western Mandarin. This dialect is invaluable for the study of the old Chinese language. A knowledge of its peculiarities renders the syllabic spelling, now eleven or twelve cen- turies old, perfectly available ; and thus the sounds of all characters may be known as they existed before the language underwent that great organic change which produced the Mandarin dialect in its three-fold form. 38 The assistance derived from the old middle dialect for research into the ancient Chinese language needs to be supplemented by the southern dialects, which are also, especially in regard to their final letters, of great archaeological value. The lost finals, m, k, t, p, are retained with almost perfect uniformity in the dialects of Canton, Chaucheu, and Amoy. The dialects of Fucheu and Hweicheu, and that called the Hakka, are less valuable in research, being situated on a line of transition. The relation of French to Latin resembles that of Mandarin to the Canton and Amoy dialects. Am, " dark," Latin umbra, " shade," has become an in Mandarin, as mum, " his," has become son. Kot, "to cut," has become ko in Mandarin, as gladius has become glaive, and traditor, trditre. The root kot, " to cut," appears in gladius, with the sonants g and d, instead of the surds k and t. It is also found in coedo, "to cut," culter, "a knife,' , and the English cut. The I inserted in gladius and culter is dropped in the French couteau. The Japanese call " a sword," katana. The Mongols say hadahu, " to cut," " to reap." The Tamul- speaking people of Southern India say katti for " a knife." The Hebrew word for " to cut off branches " was gadang, J^ll In the case of a wide-spead root like this, found in so many families, it is certainly no slight advantage to have the ancient form well preserved in the south-eastern dialects of China. JAPANESE AREA. 39 The Japanese language, spoken and written, is much mixed with Chinese. The Chinese language, literature, and customs were introduced there in or about the first, fifth, and seventh centuries of the Christian era. In addition to many thousand Chinese words, introduced with the contemporary pronuncia- tion and still kept unaltered in the language, the native vocabulary of words is also very extensive. The first Chinese immigration was probably Tauist, and perhaps chiefly intended for the propagation of religious opinions; but it spread also the Confucian literature and morality, and gave the Japanese the alphabet of fifty letters which they still use. The temples and habits of life and thought of the Sinto priesthood resemble those of the ancient Chinese Tauists of the Han dynasty, who did not use images. During four or five centuries before the arrival of the Buddhists, a.d. 400, the influence of China in Japan continued, and this was the period when the Sinto system, with its numerous Kami, " spirits " or " gods," became consolidated. From a.d. 400, during the intro- duction of the Go won, "Wei pronunciation," and To won, " Tang pronunciation," there was an immigration of Buddhist priests of various Chinese schools. They aided in continuing that powerful impulse which ended in the establishment of a complete system of Chinese instruction throughout Japan, and the universal pro- fession there of the Buddhist faith. From this time 40 china's place in philology. every youth learned the language of Confucius at school, and the Colloquial Chinese of the period became mixed with the national language to a most remark- able extent, for the ordinary purposes of life as well as for the exigencies of the scholar. Eut in regard to the Japanese native idiom and vocabulary, what is it ? It bears a manifest resem- blance to the Mongol. The root takes polysyllabic suffixes and vowel prefixes in both languages. The verb is placed at the end of the sentence, and is pre- ceded by its object. The case particles are syllabic suffixes attached to nouns. In Akari wo tomoshi, 1 " to light a lamp or candle," wo is the case suffix for the objective case. Akari, "a light," is the Mongol gerel with a vowel prefix, in Chinese kwang. Tomoshi is the Chinese tiem, "to light," "kindle," with verbal suffix oshi. The Japanese, in regard to pronouns and sub- stantive verbs, is more like the Chinese than the Mongol, but in respect of syntax and polysyllabic derivation, it is manifestly like the Mongol, Manchu, and Turkish. It is then Turanian, but it does not bear so close an appearance of kindred to the Tartar languages as they do to each other. Their having in common the first personal pronoun and substantive verb in b and m, links these three modes of speech together as first cousins, while the Japanese, Corean, and Tamul languages, from the want of 1 See Hepburn's "Japanese Dictionary." COREAN AREA. 41 these prominent features, are but as second or third cousins. Hence, for the convenient classification of the Tur- anian system, three sub-families are required : — 1. The Tartar, comprising Mongol, Manchu, Turkish, etc. 2. The Japanese, embracing Japanese, Aino, and Corean. 3. The Dravidian, including Tamul, Telugu, etc. That the Corean language should be placed in close family relationship with the Japanese cannot be doubted, when it is remembered that there is in it no trace of the favourite Tartar and Indo-European pronoun and substantive verb in b and m y and that it resembles the Mongol and Japanese in placing the verb at the end of the sentence, immediately following its object, and in adding to the roots polysyllabic suffixes. For the sen- tence " this room has two windows," the Coreans say i k'utul, " this room," t'ul c'hang isir, " two windows has." The pronoun i, "this," is in Mongol ene, "this," in Chinese «, " that." K'utul may be the Mongol ger, "house," and Chinese Ma, ke, "home." T'ul reminds us of the Persian du and English two. C'hang is borrowed from the Chinese c'hwang, " window." Isir is probably the Chinese yeu, " to have," with suffix sir. The Japanese ware shiranai, " I do not know," where ware is " I " and nai is " not," may be compared with the Corean na, "I," aji, "know," mothar, "not." 42 china's place in philology. Like the Japanese, the Coreans study Chinese litera- ture, and mix Chinese words with their own in the common intercourse of life. An immigration of Chinese Buddhists, continuing for several centuries, communi- cated to the language a large Chinese element. The introduction of French words like adieu into English may be adduced as an example of the same kind of in- fluence on our own language. The Chinese sacred books are read in schools throughout Corea, and the doctrines of Confucius inculcated. The Corean alphabet made for them by the Buddhists on a Tibetan or Sanscrit model, is now used to write the mixed languages as at present spoken. The Aino language spoken on Yesso has the Japanese polysyllabic formation and laws of position, and is without the substantive verbs and personal pronouns in m and b. It is therefore a Turanian language, and is to be classed with the Japanese branch. The best type of the Tartar sub-family of the Turanian languages is apparently the Mongol. The Turks have always been much mixed with the Persians, who early occupied Bactria. That country, indeed, is spoken of in the Zendavesta as the original home of the Arian religion. Though called Turkestan by our geographers, it was Persian before it was Turkish, and its Persian population are the Tadjiks of the present day, and the °fc $£ Ta shih, 1 old sound Da zhik, of Chinese historians. 1 In the Chinese dynastic histories, the Arab conquests are attributed to MONGOL AND TURKISH AREA. 43 They pressed over the passes of the T'sung ling chain, called by the Turks Mustag, "Ice Mountains," into Chinese Turkestan ; here they became mixed with the Wigur Turks, as at Bokhara with Usbeks, Turcomans, and other races. The result has been that the Turks of Yarkand, Cashgar, and Bokhara, as well as those of Constantinople, have assumed more of the Indo- European appearance than is seen in the Mongols Or the Manchus. This is true also of the Mahommedans who have crowded into North China during the Sung, Yuen, and Ming dynasties. This numerous class, coming, as their traditions say, from Bokhara and the other Turkish cities, have very much of the European head and phy- siognomy — their deep and horizontal eyes, prominent nose, with a tendency to a vertical facial angle, and to the growth of whiskers, bespeak western descent, and help to give them, among the surrounding Chinese, a characteristic and easily recognized physiognomy. This mingling of Turanian and European features of race has affected the Turkish language. The Mahommedan religion has also added many Arabic words which have been adopted into the Turkish, both of Constantinople and of Yarkand, with the other cities of Chinese Turkestan. The word Adam for " man," and ruh for " spirit," are used in the easternmost Turkish cities. A Bucharian vocabulary, translated by Klaproth from the Ta shih. This is through an error in their information. They did not learn the true name of the Arabs till more recently. 44 Chinese, and printed in the Asia Polyglotta, is entirely Persian. It is called in Chinese the language of the Hwei hwei or Mahommedans, who during the Ming dynasty appear to have been identified by the Chinese with the Persians, in regard to language, religion, and race. The Turkish is consequently so permeated by the Persian and the Semitic element introduced by religion, that it can scarcely be considered the best type of the Turanian languages ; especially is this true, because the relative pronoun, otherwise foreign to the Turanian family, is found in Turkish in its Persian form, and may best be regarded as borrowed from that language. The Persian influence on Turkish extends even beyond the boundaries of Mahommedanism, into Siberian dialects. In Castren's vocabulary of Turkish dialects in Siberia, Kudai, the Persian word for the Supreme Being, often identified with our term God, and the German Gott, is employed for "heaven" and for "God." Our word foot appears as put and but, which are quite Indo-European, the Persian being pal. The Mongol, therefore, may be viewed as a better Turanian type. It occupies scattered sections of that great belt of land which stretches from near the mouth of the Amoor to the banks of the Yolga, and from the Kokonor lake to the Alta'i mountains. In its eastern extension it meets with Tungus tribes and Chinese colonies of agriculturists, some of whom, near the banks of the Amoor, learn to speak better Mongol than MANCHU AREA. 45 they do Chinese. The Buriat Mongols, east of the Baikal Sea, are also conterminous in area with Tungus tribes. West of the Gobi Desert the Mongols are mixed with a Turkish population, the descendants of the ancient Wigurs, and with various other tribes of the same race in Turkestan and European Eussia. To the south-west they come in contact with the Tibetans, and to the south-east with the Chinese. The Mongol language occupies the centre of the Turanian area so far as Tartary is concerned, and became a written language about five centuries ago, when, in the Yuen dynasty, it was -necessary for the fierce nomades of the great central plateau of Asia to accommodate themselves to the usages of civilized countries and commence the formation of a literature. They adopted the alphabet already in use among the "Wigur Turks and which had been given them by the Nestorian missionaries. Thus the present Mongol and Manchu alphabet (for the Manchus took theirs from the Mongols) was derived from the Syriac, through the missionary zeal of the Nestorian communities in "Western Asia. The Manchu language is spoken on the lower course of the Amoor by tribes under Chinese and Russian domination. In the Greek church mission, recently established there, the Manchu translation of the New Testament, made at Peking about 1805 by Lipoptsoff, is found to be intelligible and useful. This is the version published by the British and Foreign Bible 46 china's place in philology. Society. In the Chinese province of Hei lung kiang, north of Girin, the Manchu language would seem to have lost ground and to have contributed to the Mon- gol area, for the Chinese colonists there speak Mongol fluently. In the provinces of Girin and in sea-board Manchuria Chinese is the common speech. If we would look elsewhere for spoken Manchu, it must be among the Tungus tribes of Siberia, found scattered at various localities east of the Baikal. In Peking Manchu is spoken as a Court language, and learnt for that purpose from teachers. It is also extensively written as a documentary language. Numerous helps exist for the study of it in the form of translations, dictionaries, and phrase books, published at Peking. The study of the language is maintained in all the Manchu garrisons in the eighteen provinces of China Proper, and in Mongolia. A syllabary is used of about 1,000 syllables. Where the Mongol writing was deficient in the power of distinguishing sounds, the Manchu has added special marks, so that the mode of writing indicates the pronunciation satisfactorily, which is far from being true of the Mongol. The Tibetan, perhaps the most convenient type of the Himalaic languages, has been well opened to obser- vation by the Dictionaries and Grammars of Csoma de Koros, Schmidt, and others. These two grammarians have not, however, considered the tones, which in a monosyllabic language become of special importance. TIBETAN AREA. 47 Georgi's notice of the Tibetan tones is only sufficient to show that they are of the same nature as the Chinese. We have not yet any comparative lists of common words in the Anamitic, Siamese, Burmese, Tibetan, and Chinese languages made with reference to their intonations, by help of which the general laws of tones for all this widespread family might be investigated. The Tibetan language spreads from Ladak, the most northerly of our British Indian possessions, to Sz c'hwen, where it meets the Chinese area. Its eastern member is the Si Fan dialect. The nomade Mongols also occupy Eastern Tibet, and are there mixed irregu- larly with the Si fans. Crossing the Himalayas we find the Dravidian area occupying hill districts in Northern India and the plains and mountains of the South. Among the lan- guages of this family, the Tamul is the best to use as a type. It is spoken by ten millions of people, extending on the east coast from Cape Comorin to a point eighty miles north-west of Madras, and on the west coast from Cape Comorin to Trivandrum. 1 The Dravidian family is cut off from its relatives in Tartary and Tibet by the intrusion of a broad belt of the Indo-European area. The Arian invasion of India is supposed to have taken place about 2,000 years before the Christian era. Those who came into India 1 Pope's " Tamil Handbook." 48 china's place in philology. at that time spread the Sanscrit tongue, which was followed by the Pracrit, over three-fourths of India, and gave origin to the numerous group of languages known as Bengali, Sindi, Hindustani, Gruzarati, Urdu, and Marathi. The superior energy of the Indo-European race enabled it to conquer wherever it found a home. Europe and Asia Minor, Persia and Bactria, were all subdued and occupied by this powerful branch of the human family. Their home extended from Samarcand to Lisbon, and from Calcutta to the land of Thor, and the multiplied experiences of so wide a region tended to excite in their intellectual development a proportionate richness and variety. The gift of imagination was awakened in this race by residence in mountain scenery and around inland seas. They wandered far, they grew up amidst the most beautiful and varying landscapes. Their homes were among the great mountain chains of the world: the ^ Himalayas, the Bolor Tag and Mustag of Bactria, the Caucasus, Mount Taurus, the mountains of Thessaly, and the Alps and Apennines. Their earliest navi- gators traversed the Black Sea and the Caspin, thv,° Archipelago and the Adriatic. Hence the ?jirit of freedom and the irrepressible sense of poetr, the tendency to speculation and the keen appetite for science, that have always characterized this race. All INDO-EUROPEAN AREA. 49 other races, except the Semitic, are comparatively wanting in these splendid gifts, which make the Indo-European nations the very flower and crown of humanity. A natural love for variety of experience, difficult travelling, and new scenes, led the earliest colonies of this favoured race to choose their homes where the eye and the hand, the mind and the body, should be exer- cised in due proportion, and thus the human species be rbrought higher up on the ladder of progress. The result we see in the wonderful expansion of philosophy, science, and literature among many nations of this race, ancient and modern, which has made Europe what it now is. The elder branch of the Caucasian race, the Semitic, occupying Syria, Mesopotamia, Assyria, Babylonia Judaea, and Arabia, was destined to do more for the religious culture of the race than any other linguistic family. The religious, moral, and spiritual impress on the European races had a parallel in the earlier lin- guistic influence which it appears to have exercised. The superior ease and fluency of European speech, compared with that of Eastern Asia, comes partly from the r^itive pronoun and partly from the liberty allowed i in the construction of sentences. Both the relative, and the freedom used in the position of verbs and words belonging to the other parts of speech, probably come from the influence of Hebrew and its 50 cognate languages. On the Chinese side of the Hima- layas, and of the Persian- and Russian-speaking area, the laws of position in sentences are fixed, and there is, properly speaking, no relative pronoun. It was not, then, from this side that the Arians, the youngest of races, derived their freedom in syntax, leading to a beauty and expansion in poetic expression which are inconceivable to the less imaginative races. These characteristics, with the genders of nouns and the voices of verbs, came from the influence exercised by the combined Hamite and Semitic races on the early language of the world. The Semites were always neighbours to the Hellenes and the Persians. The influence of old Turanian languages on the formation of the Indo-European system was favoured to an equal degree by geographical contiguity. Colo- nies belonging to this stock were sprinkled over Western Asia in many localities; and in the Persian area, Iran and Turan from the dawn of history stood in close contact to each other and in hostile attitude. The polysyllabic development of the Arian lan- guages, their case and tense suffixes, together with such vestiges as they retain of a law relegating the verb rigidly to the end of the sentence, are the effect of Turanian influence. CHAPTER IY. On the Primeval Language.— It was Monosyllabic— Examples. — Pronouns. — Laws of Position. — Laws or Ehythmus. — Pro- nominal Eoots also Verbs. — Closed Syllables, a Proof of Man's Continental Origin. — Early "Use of the final M. — 9 Names of Animals. — Divine Origin of Language. Without venturing to discuss, except very cur- sorily, the origin of language, I shall here first attempt to mark out some of the common elements existing in the speech of all nations which seem to belong to a primeval language older even than the Chinese and the Egyptian. The mother from whom all existing dia- lects have been born may possibly be revealed to our view by carefully rejecting all new elements and retaining what appear to be universal. That it was monosyllabic is deducible from the fact, that in all the families, from the Indo-European upwards, the roots are monosyllables. The words separation and departure, for example, are traced to the Latin Part in pars, partis. The r is lost sight of in the Sanscrit bheda, " dividing," bhedita, " divided," bhinna, "separated." It occurs in a dissyllabic form in the Hebrew badak, " split " (Latin fidit), and badad and 52 china's place in philology. badal, " divided," and without a third consonant in the Hebrew bad, "separation." The Chinese is Bit, Pit, " separate," " other." Our words rotation and radiation are traced to rota, "a wheel," and radius, from the same root, Bad or Bot, German rad, " wheel," Sanscrit lut, " roll about." In Tamul we find urutchi, " roundness," and urul, " a carriage- wheel." The Chinese call " a wheel " Lun, and many round things, as " a stove," " a cottage," " a skull," " a reed," are known as Lu, where a final t has been lost. Musical pipes they call Lut. The same idea of roundness is found, more or less remotely, in the English rod, reed, oar, row, round. The German ruder, " oar," rudern, " to row," compared with the Greek eretmos, "oar," eresso, "row," Latin remus, "oar," remigo, " row," throw light on the origin of the word oar, and enable us to trace it to the same root with the others. The Chinese Lu, for Lut, is a scull, such as is used in China for propelling a boat by stern action. When we have arrived in such investigations at the monosyllabic root Lut, Bad, Bot, our progress is en- tirely checked, and we are left to conclude that the primitive speech of man was monosyllabic, and con- tained in it such widespread roots as the two just given. Father and mother must be admitted without hesita- tion into the primeval vocabulary of the human family, for though some nations, as the Mongols, appear to PRIMEVAL MONOSYLLABLE. 53 want them, nine-tenths of the inhabitants of Europe and Asia agree in their use. Of course they must be accommodated to the necessities of infancy by cutting off the second syllable of the. English form and changing the initial / into p, or, still better, into b. The Turks in saying Baba for " father " are more primitive than any. The Semites in saying ab and em for " father " and " mother " gratified a tendency to prefix vowels. In the Chinese fu, mu, we have the newest form of what was a few centuries ago Bo, Mo. The claims of brother to a place in the primeval vocabulary are quite hopeless. It appears to be un- known in Asia beyond the Indian and Persian area. There is more hope for sister than for brother. It may perhaps be recognized in the Chinese Tsie for an older Si or So, used in the sense of elder sister. The names of number differ so widely in the various Asiatic languages that they are not to be expected to be Very ancient. Of the pronouns, a and nga or ga for the first person, u, mi, and yu for the second, and i, gi, hi for the third, may claim a very high antiquity ; for their widespread use through the linguistic families is a palpable and striking fact. For the first person we find the old Chinese ffi nga and ^ yo or o, of which the latter, being without an initial consonant, suggests that the ng was prefixed afterwards. Another prefix consisting of a or e made 54 the Sanscrit ah am and Latin ego. The m final was a suffix also found in the Chinese ^ am, a dialect form for "I" The Hebrews added nochi, anochi, "I" The Arabs said ana, and the Egyptians anok. There appear to be very few languages in any part of the world that do not in their pronominal forms betray the presence of this root. The same is not true of the pronoun in m, which is almost entirely limited to the Indo-European and Tartar languages. Bi, men, "I," and manai, " my," are as common in Tartary as me, mein, metis, e/-to? in Europe, but there is not a trace of them to be found in China, Tibet, or Japan. Quite as little are they known in the Dravidian area or in the islands of the Eastern seas. The Mongol riding into Peking on his camel, says, manai bic l hig, for "my book," and the Manchu student learns from his in- structor, manai bit-he, in the same sense, while the German, in a region 100° of longitude further west, says, mein Buch. But these words, in their European or Tartar form, are alike foreign to the Chinese ear, and to that of all the races, Arab or Hebrew, Tamul, Corean, Tibetan, Burmese, or Malay. While, there- fore, the pronoun a, nga, Jcau, or go, for our English I, represents the primeval pronoun of the first person with great probability, the root in m and b, with its correlate substantive verb, be, bin, futurus, fume, in Mongol, amoi, bolhu, bolmoi, can be traced no farther back than the Turanian family in its Tartar branch, NATURAL ORDER OF WORDS. 55 from which it has gone over into the last great lin- guistic formation, the Indo-European. The structure of sentences in the primeval language, it may be reasonably concluded, was according to the order of nature. The nominative preceded the transi- tive verb, and the transitive verb preceded its object. The Chinese, the Hebrew, and the English here agree. It is the Turanian family that is chiefly at fault when tested by these laws. The Japanese, the Mongols, the Tamuls, and the speakers of Sanscrit, evidently follow- ing an older asus loquendi found in the contemporary Turanian speech, resolutely limit the verb to the last place in the sentence, and make the accusative precede it. This is extremely unnatural, and tends to restrict painfully the powers of human speech. Nature first names the actor, then the mode of his action, and finally the person or thing on whom or on which such action is performed. But Turanian speakers avoid this construction. Ching-gis hagan airiben t'umen k'umun alaba, " Grenghis Khan many ten thousands of men killed." The western branches of the Indo-European family refused to imitate the speakers of Sanscrit in their slavish adherence to this Turanian law, and succeeded in restoring the freedom of nature to our modern European modes of speech. So again, in placing the adjective before its substan- tive, the Chinese, English, and Turanian languages have a clear advantage over the Semitic, the eastern 56 china's place in philology. Himalaic, the Malay, and the Polynesian, which invert this order. The adjective naturally precedes the noun, as the mark of the species precedes that of the genus. We know a thing from its qualities. The " Bactrian camel " may be called the " Camel of Bactria," or " le chameau Bactrien." Of these, the first is the most natural, and is favoured by the greatest number of important languages and families. The second form, inverting the position of the words and connecting them by of, de, von, etc., adds greatly to the ease and variety of language. But it is almost exclusively European and Semitic. 1 The Sanscrit follows the Turanian and Chinese order in this respect, and thus it is shown that, although she may lay claim to be the model of the European languages in regard to her richly developed system of grammatical inflexions, she cannot be looked to as their mother in syntactical order. It is to the Semitic family that we must look for the origin of this inversion, and also for the introduction of the relative pronoun. The third form, " le chameau Bactrien," is not so much a peculiarity of any one family, as of languages occurring here and there in the area of various families. Its introduction has conferred no great advantage on language. 1 I suppose the post-position of the genitive and of the adjective to have been borrowed from the Semitic by the Polynesians, Siamese, and other races. PRIMEVAL PROSODY. 57 "We have now arrived at several approximate notions of a rudimentary kind with regard to the primeval language. 1. It was monosyllabic, and its syllabary, though containing no double consonants, had probably con- ^ sonant finals, as bid and loci. 2. Certain roots, verbal and pronominal, are so widely spread among the various linguistic families of Europe* and Asia, that a large portion of the primeval vocabulary may be expected to reveal itself as the reward of careful research. 3. The order of words in sentences was that of nominative and verb, verb and object, adjective and substantive, subject and predicate, species and genus. The common laws of position in the primeval language probably agreed with those of the Chinese, Greek, English, and some other languages in such sentences as Charles beat William, good man, this man is good, or this man good, fir tree. When two or more verbs occurred, the order was that of time. Our sentence, went near and killed him, would be " go near kill," or " go approach kill," and some device would be con- trived to represent past time. 4. The primeval language had probably a rudi- mentary tonic pronunciation. Variety in pitch, even tones, inflexions, pauses, accents, long and short quan- tity, belong more or less to all the tongues spoken The Greeks inflected the vowels of y v 58 certain syllables in their words. The Chinese do the same with their monosyllables, and so do all the neigh- bouring peoples on the west and south. The Hebrews had an elaborate system of accents. The Greek and Chinese inflexions exist in modern European languages, but without attachment to special words or syllables. Probably this last was the character of the primeval prosody. The speech of modern Europe, struggling for greater freedom, rebelled against the prosodial laws which prevailed in the old Indo-European, Semitic, and Chinese areas, and by a powerful instinct succeeded in recovering the primeval use of inflexions and accents. These aids to a natural, efficient, and graceful elocution should never become dialectic, or be tied to particular words. If language were what it ought to be, all local tones would cease, and those windings of the voice, simple or circumflex, which in England constitute the local habit of dialects, and in China are an element in particular words, would be limited to elocutionary uses. Thus language would be ennobled, the intercourse of men with each other would become refined, and the swiftly changing feelings of the heart would all have a suitable expression. Among the elements of the primeval language, capable of discovery by comparative philology, I omit the distinction between verbal and pronominal roots. All the pronouns seem to be used as verbs. It was when the eye of primitive man saw action that his PRONOMINAL BOOTS ALSO VERBS. 59 hand pointed to the moving object, and if his lips uttered a sound it was an imitation of the natural sound caused by the movement he witnessed. Speech became the instinctive imitation of natural sounds, and words were the names both of objects and actions. How then could the pronouns fail to be also verbs ? Thus, bad, "divide," "separate," "depart," was also in old Chinese used for " that," " he," and called pat or pit, $r now known as pi. Do, " to give," is the same word as that, das, etc. The Chinese locative case suffix chung, " middle," more anciently" tang, is, when employed as a verb, used in the senses "to strike in the middle," " to strike," " to undertake." As an adjective it means " middle," as in " the Middle King- V dom," and as a substantive it is the name of a " bell." A further proof is found in the fact that the instru- mental case suffixes are like others formed from pro- nouns, but they must from the nature of the case also be verbs. In fact, post-positions, like prepositions, are all verbs. All case suffixes, as well as case prefixes, may be explained, according to circumstances, either as demonstratives or as verbs. The nominative, possessive, and accusative case suffixes are most con- veniently explained as demonstratives. The case suf- fixes which express instrumentality, motion from, motion to, giving, and locality, are best considered as verbs. Should it be objected to this view that every verb v 60 would then become a pronoun, it may be answered that, for reasons not difficult to discover, the only verbs used as pronouns would be those that occurred most commonly, such as giving, going, coming, being, leaving, carrying. The early forms of such verbs as these by perpetual recurrence established themselves as pro- nouns ; e.g., the pronoun I, " he," is probably identical with 7, "to go." Such verbs as only find their way into conversation now and then would not become pronouns. It appears to have been an important feature in the primeval language that the syllabary had both open and closed vowels. Many modern languages have no closed syllables. They were rare in Sanscrit and are still more so in Japanese. It is susceptible of proof that the primeval syllabary was not one of this kind. Races occupying areas where enervation is in- duced by climate are liable to lose the final letters of their syllabaries. Nations that spread themselves over mountainous areas and cultivate hardy habits show less tendency to the disintegration of their roots. The absence of final consonants is the result of phonetic decay, or the addition of vowels through change in climate and in national habits, or through foreign influence, and other causes. Hence man must have been created in a temperate climate and in a continental locality. On the hypothesis that words were first formed from « ! 08BD n ii m.i i 1. ^l tlic imitation of natural sounds, it may be expected t licit both kinds of syllables will occur. Sounds ending in vowelfl and in consonants occur abundantly in nature, U is shown by the spelling of imitative words in our own language, >.;/., peewit, cue/wo, dingdong, hiss, hush, etc. There are other reasons why some words should terminate with certain consonants. Words ending in m and j), are usually expressive in Chinese of combination, closing, holding in the mouth, union, taste, containing, e.g., yap, " combine," /cam, "sweet," ijvm, " salt," gam, "hold in the mouth," "contain," k'am, "hollow," "deficient." The final letter seems in these and similar examples to indicate that the words where it is found are expressive of actions which are easily represented by the mouth opened or closed. Emptiness or deficiency would be fitly pictured by an open mouth, union by a mouth eliding. But the labial letters m and p, which would be brought into requisition on such occasions, would naturally be used, because the shaping of the lips in forming them was a not unlikely manner of expressing the ideas to be conveyed. In English gap and gape are nearly alike in sound. The labial p with which they terminate may be accounted for in the same way. Gap in old Chinese means "to combine," "press under the anus," "narrow," "a narrow pass through mountains," "books fastened together with two boards and straps," "the action of Bcissors.' 4 We may explain the filial p 62 as expressing the action of the lips, in imitating the act of pressure witnessed by the word-maker, when he first encountered the problem how to describe intelli- gently to his companion the events he had witnessed. The meaning of the root in English and Chinese coalesces when a narrow opening among mountains is in both languages called gap. The initial g will then be left to be accounted for on the principle of the imitation of natural sounds. Should a root once become established in use, the principle of association of ideas would explain the origin of a multitude of connected words. The adverb " back " is to be derived from the substantive " back." The Chinese word for " the back " is pak. In Kwan hwa, the modern pronunciation, it is pel. It has for derivative meanings " to carry on the back," " to repeat lessons " (because the Chinese pupil always turns his back to the schoolmaster while repeating his tasks), "to turn back," or a run away," "the north," "to disobey," etc. We find the some root in /3aard^o), fero, bear, porto, etc. For all these words, with a multitude more, one root bak, as we may judge from the Chinese analogy, would be approximately the original form, and it might be the imitation of sound. The finals k and g occur not seldom in words imitative of sounds, as flagellation, thwack, strike, and the Chinese p f ak to "strike gently," which is identical with the root of flagellum, plaga, tz}J) "bowl," 1% "bean." ^ "complete." $70 waw, " bind round." |^ www, "thigh bone," "that which bends." j| "round." [j] ^, "circle," gj| y«», "rhyme," 2 |J|| "fall." gj "cause." $Q ym, "marriage," »]^| yew, "smoke." None of these words, nor any of their derivatives, ever take initial consonants or undergo any alteration, except in vowels and an occasional change of the final n to its correlate t. We therefore conclude that 4,000 years ago these words, and others with the same phonetics, began with a vowel and ended with n. The Latin vocabulary furnishes us with annus, " a year," annulus, "ring," anima, "breath," "the soul;" the Greek has alcov, "an age," with its equivalent cevum. 1 As rudis, "rude," in Chinese »^ lu, lud, comes from rot, "round," viewed as unirapressible, so wan, "round," in Chinese is taken to mean Compare in Greek pvd{i6s, " rhyme," from rot, " round." PHONETICS AND THEIR DERIVATIVES. 77 The Latin v has the force of w, as in vinum, Greek, oivo. The Russian v, written b, and pronounced v, has in these comparisons of words also the force of w. We find vina, " cause," vyenetz, " crown," " coronet," " cloud of glory," vyenok, " garland," vyenchat, " to crown." The English wind, wend, and wander, appear to be of the same family. Wind is anima, "breath," ventus, "wind," Sanscrit, an, "blow." The "vine" is that which winds. In Latin it takes t for n, as in vitis. Further, the Mongol has undus, " root," " source," and the suffix dus may be compared with nus, in annus, " year," and the d in the English wind. The principle of adding syllabic suffixes and cognate letters is the same. In Sanscrit we have vdna, " pipe," vanada, " cloud," venu, " bamboo," vata, " circle." The letter v is in fact w. Further, in the Dravidian vocabulary we meet with the Tamul undei, " a round thing," " a ball," " pill." In Chinese $fc yuen, signifies " a cause," and jl yun, " cloud," doubtless the same words originally as y^ yuen, "origin," and yen, "smoke." The Chinese yin, " marriage," when compared with yuen, " draw with the hand," suggests the Latin unio and our unite. In Greek eVo?, " a year," eVo?, also " a year," belong to the same family of words, and perhaps iv, "in," is the same with the Chinese @ "cause," " because." If we take another example with m final, the light thrown on the primeval form of the syllabary will be 78 seen still more plainly. The Latin umbra consists of the root um, and a common suffix bra, which may be compared with ber in imber, "rain," brai in tenebrai, bris in salubris. We have in Hebrew emesh, " night," "darkness," where sh is a suffix, as in kafash, "to cover," " overwhelm," cognate with Jcafar, " to cover," "forgive," Chinese |g kai, "cover," old sound, as known from the phonetic, hap. The Latin um, Hebrew am, as it reads without the points, with the sense of " shade," " darkness," " night," are in Chinese flg an, old sound am, " dark," [SJ yin, om, " obscure," " the principle of darkness in the dual philosophy," Jg yen, am, "to shade," " cover," j|f yin, om, " eaves." In Chinese buildings the eaves project far enough to make a broad shadow. These coincidences are quite sufficient to show that in the Chinese primeval syllabary m was the final letter in this root, and that the initial was a vowel. We thus by multiplying our researches in all parts of the Chinese vocabulary, always adopting the old pronunciation registered in Kanghi's dictionary and other older lexicographical works, arrive at the fact that the final letters ng, n, m, k, t, p, with the vowels, were the final letters of the pronunciation in use when the characters were made. And though they are much disturbed in the Mandarin dialect, they are retained to this day with an approach to faultless regularity in the Canton and Amoy dialects. They are also found INITIALS. 79 in the Tibetan, Cochin-Chinese, and Siamese languages, all belonging to the allied Himalaic family. Having obtained this solid foundation of knowledge with regard to the final letters of the Chinese syllabary in use 4,000 years ago, we may proceed to inquire into the initials. Of these, the most certain are g, d, b y ng, n, m, I, z, dz, zh, and the vowels. Initials of the next stage of probability are the aspirates &', t\ p ( , Ps, the surds k, t, p, and the sibilants s, ts, sh. This difference in probability arises from the vestiges existing of an old law of change similar in part to Grimm's law, by which the sonants have always been throwing out words into the surd series. So numerous are the examples of this law, that it is open for consideration whether the surd series is not altogether made up of successive contri- butions from the sonants. Before giving examples of the sonant contributions to the surd series, let me premise that in the Amoy and Canton dialects, the surd and sonant series receive the name of upper and lower series, and are identical with the so-called four upper and lower tones. In other words, characters in each division are pronounced with special intonations of the voice, and thus distin- guished from the upper or surd division. Thus, £p bang, " even," . is at Amoy pieng in the lower first tone, while in the syllabic spelling of the dictionaries it is bang or biang in the first tone. In dialects having eight tones, words in the lower series may be trans- 80 ferred into the system of the dictionaries by changing the surd initials into sonants, and allowing the peculiar intonation to coalesce with that of the upper series. The word $* bang, "side," the English "bank," " bench," with its equipollent phonetics ~jf fang, "square," ffipang, "state," "kingdom," £p bang, "even," " peaceful," $fc bang, " side by side," etc., have together an extensive cluster of derivatives, some of which take b, and the rest p. The meanings are, " side," " even," " tie together," " tie," " impinge upon," " strike," " wings," " catalogue of names arranged side by side," " square," " anything square in shape," as a " territory," a " seal," etc., " edge," " mountain ridge," etc. Cor- responding words in European languages are impingo, " strike against," Trrjyvvfii,, " fix," pax, " peace," pack, bang, Jingo, fixi, etc. Two hard things brought into rapid contact caused a sound which primeval man heard as bang. Thus the peculiar phonal form of the root in the primitive syllabary of the world may have originated. It then came to mean "side," from the fact that the two portions of impinging matter remained side by side. Then the act of bringing them together and of holding them together, or of their coming together of themselves, were named with the same vowel and consonants. This gave rise to the words belonging to this family meaning "tie," "fasten," " fix." When evenness, physical or moral, had to be spoken of, the same root was used. SURDS DERIVED FROM SONANTS. 81 But how do we find them spelled in Old Chinese? Chiefly with b. Yet in part also with p. " Evenness," " impinging," " side," " bringing side by side," are all bang. To "assist," "squareness," a "territorial square," to " tie," a " wing," to " imitate," are all pang. The reason of this is evident. Language instinctively seeks to enlarge her bounds when they become cramped by an increase of words and of ideas. She aims to remove ambiguity by introducing differences in pronunciation between like sounding words. In the example given the words initiated by b are the older. Those in p are the newer. The obvious con- clusion is that p derives its origin from b, and that b is an older letter than p. The primeval syllabary did not need so many letters as are now in use. It started with b, and added p, p f , and / afterwards as they were needed. In the Mongol syllabary there is no p, p ( , or /. There is a fully developed p in the Indo-European and Semitic families. Hence the p may have sprung up contemporaneously in the Chinese and Indo-Euro- pean families after their separation. In both cases it was by a natural putting forth of creative strength on the part of language to increase its alphabet and its syllabary. It is thus that the preponderance of b over p in the Sanscrit and Hebrew vocabularies may be best accounted for. That Latin and Greek dictionaries devote so much larger a space to words in p and / than to words in b is an indication of recent origin in the vocabularies. 6 82 What is true of b in the old Chinese syllabary is true also of the other letters in the sonant and surd series. The sonants g, d, b, z, are the old letters ; the surds k, t, p, s, are more recent ; / and h seem to be the newest of all. In the Japanese transcription all Chinese words in h are written with k, while those which in modern Chinese commence with / are written with either b or h. But as h is the regular Japanese equivalent of the Chinese p, the weight of evidence is in favour of the statement that p and b were the old equivalents in all cases of the moderu Chinese /. If we carry back the inquiry another stage, p and b coalesce in the primeval and world-wide b. The Japanese, indeed, have an initial /; but as it is used N to write Chinese words in p as well as in /, it is probably a new letter. We should expect to find the name Buddha trans- cribed in old Chinese with something like exactness. We learn on investigation that the character ^ Fo, was anciently called But, as is shown by the syllabic spelling, in the Amoy pronunciation Put, and in the Japanese transcription Budzu. A few more examples are here appended. Among the sonants, ft bun, " divisions," " duties," j|| dan, " revolve in a circle," ^ bok, " return," Jf| gak, "learn," fljg ngang, u hard," $£ zung, "follow," jjlj bit, "other," |j| gun, "herd," sg ngu, "meet," J|| deng, " go up," have the following correlates in the TONES. 83 surd series, viz. : ft pun, " to divide " (Hebrew bin, Latin findo), $f tun, "revolve," "turn," English turn, 4t pok, "north," "back," gfc kak, "teach," pQ kong, "hard," "steel," ffc tsung, "let it be that," jjlj pit, " difference," % kun, " a body of troops," jg ku, "meet with," %£ teng, "go up." But it was not enough for language to add the surd letters to its acquisitions. The syllabary was still too contracted. "Words and ideas continued to multiply, and there was a scarcity of syllables to express them. The age of suffixes and prefixes had not yet arrived. It was too soon to think of dissyllables or polysyllables, of a prefixed s or an inserted r. Language in this time of need seized for the required service those flitting musical intonations of human speech which the orator uses to express decision, sarcasm, doubt, and interrogation. At this time there were in the Chinese vocabulary two great groups of words. Those ending with ng, n, m, and the vowels, formed one group, which we will call long in quantity. Those terminating in g, d, b, or k, t, p, formed another, in which the sound is shortened by the action of the final letters. They check the breath and bring the utterance to an abrupt conclusion. Hence these words become, for the pur- poses of tonic pronunciation, short in quantity. But final letters will drop off, through laziness in enunciation, through imitation of the defects of others, and from the circumstance that, when stress is laid by 84 china's place in philology. the speaker on some one element of sound, the other elements will suffer. "What did language do ? She did not resist change ? This she never does. She allowed new laws to enter, so that the inevitable changes might be kept under control. A third group of words was formed out of contributions from the other two. By the ancient poetry we learn that 3,000 years ago the words that could rhyme with each other formed three groups, which did not encroach on each other's limits. The new group was mainly composed of what is now called the Shang sheng tone class or second tone. The third, or K'ii sheng, was subsequently formed. The numerals were then pronounced yit, ni, sam, sat, ngo, lok, sit, pat, ku, zhip. Of these, sam, " three," was in the long tone, now become the first tone ; ngo, " five," and ku, " nine," in the new, or second tone ; ni, " two," doubtful; and the rest in the short tone. Of the ^.ye elements, kim, " metal," was first, mok, " wood," last, and shi, " water," ka, " fire," t'o, " earth," all in the newly-formed tone class. Fifteen hundred years passed away, and the Hindoo Buddhists were in China teaching the religion and sciences of India. The Chinese had never thought about the distinction between tones and letters, and when Bengal and Panjab pandits told them that sound was capable of analysis, and that tones must be distinguished from vowels and consonants, they listened incredulously. But the claims of the alpha- GROWTH OF THE TONE SYSTEM. 85 betical analysis were gradually allowed, and emperors appointed commissions to settle the sounds and con- struct dictionaries. Imperial pride condescended to learn the tone distinctions in a flattering sentence constructed by a courtier, which exemplified them in their order. 5c *3R S © ^ n ^ shing chit, " Heaven's son is holy and wise." The passage of 1,500 years had seen a new tone formed, the K'ii sheng. It consists of contributions from the second and fourth. Poetry at this time was made according to new laws. Not only the rhyming words were brought into subjection to the tones in groups of four ; but all the words of each line were made to conform to a complex harmonic scale, in the construction of which the tones formed the chief element. Another 1,500 years has passed away, and we now find that still greater changes have taken place than in the preceding period. The first tone class has been split in two. The old sonant initials have been ex- pelled, and their place supplied by surds and aspirates. The words of the fourth tone class, after losing all their final letters, have been distributed among the other classes, and the Chinese modern language has become more changed from the old type than any member of the monosyllabic family. 1 1 For a detailed account of these changes, see Mandarin Grammar, Part I. The principal step I have made in advance in the Chinese part 86 There have been three great periods of 1,500 years each. The first saw the earliest formation of the surd and aspirate series, with that of a triple tone system. The second witnessed an extensive dropping of the final letters k, t, p t and ng, and the growth of the tone system ending in the quadruple formation of the » dictionaries. The third period, perhaps the most revolutionary of all, saw the sonant initials, and the finals Jc, t, p, m, for ever dismissed, one of the primeval tone groups completely broken up, and the syllabic spelling of the Hindoo Buddhists thrown into chaotic confusion. All this may be taken as proof of the primitive character of the Chinese language. Had it inherited from the Turanian, Indo-European, or Semitic families, any of their peculiar tendencies to polysyllabic forma- tion, it would have had, historically, a very different development. But being itself of the first descent from the primeval mother of human speech, we can trace in it no later elements. Not the Egyptian nor the Hebrew nor the Sanscrit can compare with the Chinese in antiquity of type. They all have a more complex syllabary, and introduce appendages to the roots, which constitute an evidence of the comparative recency of their formation. of the investigation since the publication of that work, has been in the detection of the law by which the surd series has been regularly formed from the sonants, as illustrated above. SYNTAX. 87 If with these views alone before me, I should be inclined greatly to lengthen Chinese chronology; but the comparison of the ancient civilizations of China and Western Asia compels me to reduce the epoch of the commencement of Chinese isolation to very nearly that of accepted history. The similarity between old Chinese life and that depicted in the . Book of Genesis is so striking and so multiform, that it seems impossible to date the eastern migration of the Chinese earlier than a few centuries, at the most ten, before the time of Abraham. The laws of position in Chinese sentences are the same with those already given as belonging to the natural and primeval speech of man. The actor is mentioned before the action, and the verb before its object. The adjective precedes the substantive, and the specific noun the genus to which it belongs. The adverb precedes the verb, and the attribute the substance to which it is attached. The subject is first mentioned, then the copula, and lastly the pre- dicate. The only peculiarity to be here mentioned as not of natural and primeval growth is, that locative auxiliaries are made suffixes and not prefixes. " In a city," is more natural than " city in." The Chinese, however, prefer in their ancient and modern language to say the latter. Our phrase, "the world," is with them 5c T " heaven under." These locative post- positions are best explained as substantives. Hia is 88 " that which is below." The original force of such words was verbal. " To go down," is also Ma. As in the Turanian languages, so in Chinese, the verb became strongly substantival. Act became action. It is indeed the same in English. " Act," is a verb and a noun, and the mind learns to abstract the act from the actor, and look at it by itself. It is then spoken of as any other noun. Thus, c'heng net is translated " the city's interior " or " in the city." The word nei is nip, to " enter," the p being dropped. The modern form is ju. That which is entered is the interior. The language forming faculty performs the necessary transformation, and applies the name of the act enter to the inside of a city or house. It then becomes a locative suffix. All Chinese suffixes of this sort were originally verbs. So the other locatives shang, " above," hia, " below," t'sien, "before," heu, "behind," etc., were all verbs originally. As such their places would, before they assumed the locative character, be before their nouns. The germ of the Turanian and Indo-European sys- tem of declension appears here for the first time. What the Chinese did for the locative, the ancient communities, who founded those types of language, proceeded to do for the instrumental, ablative, and dative cases. All the case suffixes, whether locative, instrumental, or dative, were simply verbs robbed of ORIGIN OF THE POSSESSIVE CASE. 89 their activity and placed after nouns as signs of locality, direction, instrumentality, and so on, in order to facilitate the more speedy and convenient allocation of the objects of thought in the categories of space and time. The Chinese has also a sort of possessive case, the history of which is simple. In the earliest Chinese the possessive case was included in the law by which species precedes genus, subject precedes attribute, and the particular notion goes before the general notion. " Man's body " was jen shen. Soon one of the demon- stratives, ti, was used as a connective — \ ;£ B E3 Nin ti ngi.mok, "men's ears and eyes." There was originally no possessive force in this connective, an- ciently ;£ ti, now £$ ti. The possessive force was conveyed in the order of the words, in accordance with what may be regarded as a law in the primeval lan- guage from which the Chinese was derived. A hiatus is felt in the modern language if Wo tfhai sh'i, "my duties," is said for Wo ti c'hai sh'i. The Tibetian would perceive a similar hiatus. The remedy is found in the introduction of the particle ti. In the Shanghai dialect the particle used to fill the hiatus is ko. In Tibetan ki is employed. In Fu kien province, as in the Amoy dialect, e is the word. In all these cases the possessive force would be acquired subsequently. The origin of the possessive was simply a want felt, to make the sentence square, a rhythmical feeling which is not 90 contented until the laws of proportion are obeyed in language. It is the same feeling which prompts us to say " a long and happy reign," rather than " a happy and long reign," and which lies at the foundation of prosody. The order of verbs, when they represent two or more consecutive actions, is in Chinese that of time. This principle would be adopted from the primeval type. Thus, primeval man would say without any inversion, " Sit down eat food," in the language of command or of narration. The word down would be a verb, and thus three verbs would stand in juxtaposition before the solitary substantive food. The modern Chinese says Tso Ma c'Mfan, "sit down eat rice." The Semites were the first to introduce a conjunction and, as in Gren. xviii. 2, " And he lifted up eyes his, and saw, and behold three men standing by him, and [he] saw and ran to meet them." The words lifted, saw, behold, ran, are all introduced by and. The prepositions "by," 7^ ngal, and 7 le, "to," are originally verbs, the one meaning " to ascend," and the other " motion towards." The whole sentence thus consists of nouns, pronouns, and verbs, and the order in which the verbs stand is that of the time in which the actions symbolized by them took place. Not one of them is put out of its natural position. The order of time is the basis of the position of POSITION OF VERBS. 91 verbs in all languages. But it was subject to frequent inversion in the Hebrew, as in Gen. xx. 6, " And said, Sarah, laugh made to me God," for " Sarah said, God has made me laugh." The dative participle le before me is redundant. The verb laugh is placed before the verb make, and both stand before their nominatives. Such inversions do not appear in the Chinese lan- guage, which is unimaginative. The popular instinct is satisfied when it describes events in the order in which they took place, and could take no pleasure in those bold transpositions which delighted the Semite race. CHAPTER VI. The Semitic System Older than the Turanian ; Younger than the Chinese. — Triliteral Koots.— Insertions. — Suffixes. — Prefixes. — Growth of Inflexions. — Sex. — Personifications. — Syntax. — The Yerb placed First. — Post-position of Adjec- tive and of Genitive. — Post-position of Genitive borrowed by European Languages. — Semitic Relative and European Rela- tive COMPARED WITH THE CHINESE AND TURANIAN EQUIVALENT. There is no good reason to doubt the correctness of those views by which Gesenius and other Semitic philologists were led to seek affinities between the Indo-European system and that which formed the more peculiar object of their researches. The number of common roots found in these two systems is indeed very great. Thus, among the numerals, Shad, " one," in Chaldee seems to agree with the Greek heis, eh, " one," the Latin solus, and with the third among the common Chinese roots tan, yid, kit, all meaning " alone," or " one." The Chaldee shete, " two," becomes in the ordinal for thinyana, " second." The original dental initial t resumes its place instead of the favourite Hebrew sibilant sh, and points to an old connexion with duo. The very law which frequently changed t THE SEMITIC FAMILY OLDER THAN THE TURANIAN. 93 and d to sh or z or ts in Hebrew, prevailed in the Greek when tu, "thou," became cru, and still operates in German when tide becomes zeit. Where there are roots in common, there will also be found laws of change in common. But this is anticipating. Our task of comparison must for the present be rather limited to the linguistic systems of Eastern Asia. The Semitic family has older features than the Turanian, for in the progress towards a polysyllabic formation it has not gone far beyond the dissyllabic root. In the Turanian languages, words of four or five syllables are not uncommon. Another mark of superior antiquity in the Semitic system is the absence of case suffixes in the nouns and of temporal and model suffixes in the verbs. The earliest Semites bent their energy, unconsciously but surely, to the formation of a system of speech in which as much as possible should be done by prefixes, while the Turanians directed their language-forming power to the develop- ment of suffixes. Now, since the Semites never pre- fixed more than one syllable, while the Turanian instinct, by the creation of the polysyllabic suffix, has caused the upgrowth of immense lingual variety in the speech of more than half the area of Europe and Asia, the Semitic type must be regarded as less developed, and therefore more primitive, than the Turanian. When it is remembered that Mongol, Greek, and Sanscrit case suffixes are metamorphosed pronouns and 94 verbs put after instead of before their nouns, it must be admitted that the language- systems to which they belong are of recent origin. But where, as in Semitic speech it happens, the verb, which is required to do the duty of a case particle, becomes a preposition, and stands before its noun, we feel ourselves to be in the midst of speakers who retain closely the tradition of the earth's primeval language. No one will object to the statement that the Arabs have more primeval characteristics than the Grreeks. Their life, their customs, and their modes of thinking, bear the stamp of immense antiquity; and as is their life so is their language. Every language carries on it the impress of the genius of the people that formed it. If the Chinese type is the most conservative among families of languages, the Semitic comes next to it. It never went far beyond the primitive model transmitted by " the earth's gray fathers." The date of the formation of the Semitic type being thus shown to be older than that of the Aryan and Turanian families, it must now be proved that it is more recent than the Chinese, and that its origination constitutes the second great step in the progress of language. The most obvious point of contrast is in the triliteral roots. The ancient Chinese said for " happiness," pok, a root which has the connected meanings, " rich " and " vast." In Sanscrit we find bhaga, " good fortune," TRILITERAL ROOTS DERIVED FROM MONOSYLLABLES. 95 in Latin fortuna, in Greek ttXovtos, "rich," in Per- sian bakht, " rich," in Mongol boyin, " happiness," in Russian bogatie, "rich." The confusion between riches and happiness is easily accounted for. Among what people is it not common to make wealth the measure of happiness? In Hebrew the root occurs in barachi "to bless." Here we have a triliteral root brk. The vowels were not written by the early Phoeni- cians and Hebrews. We have, therefore, only the consonants to consider. An r has been inserted. There is in this nothing uncommon. The difference of an inserted r in the English word world as compared with the German Welt, does not render doubtful the \/ identification of these words. There is a root very widely spread in most languages. It is our English verb to cut. It is in Chinese kat gi], Latin ccedo, Mongol hadomoi, Japanese katana, " a sword," Tamul katti, " a knife." Gesenius says * that the syllable gad has in Hebrew the notion of cutting in common with gaz, as in gazaz, from which it is derived by the loss of the sibilant ; but on the other hand it may be traced still farther to the harder syllables Kats, Kash, Kas, Hhats, Hhaz, and, the sibi- lant disappearing, Kat, Kad, Hhat, Hhad. All these syllables have the sense of cutting. They appear as roots in the forms Gazaz, Katsats, Hhatsats, Kadad, Hhadad. To these may be added Gadah and Gadang 1 Lexicon Manuale, under Gadad. 96 china's place in philology. "When this great philologist proceeded to compare with the large family of words here cited the Latin ccedo and scindo, the Greek , the Persian chidan and khudan, and the English cut, it is evident that he regarded the triliteral form as the formal root, and the biliteral as the real. He was manifestly right in this, as the examples now given from the eastern Asiatic languages sufficiently show. But there can be but little doubt that he was wrong in assuming the priority of the s final to the t, and of the k initial to the g. 1 The Chinese syllabary shows that a sibilant final to a root syllable is an innovation, and the history of the changes of letters in that language renders it probable that the whole surd series is derived from the sonant. Hence we learn that the root gad changed its initial to the strong aspirate Hh or to the pure surd letter k. The final d became t or s or sh or ts. We need not be surprised if we often meet with an interchange between the dental t and the sibilant s. This may be illustrated by the second personal pronoun in t. This form for the second person does not occur in any families but the Semitic, the Indo-European, and the Tartar branch of the Turanian. It is firmly fixed in all these. The Mongols take the s form, ch'i or Psi, as do the Manchus when they say si, and the Turks when they say sen. The Greek av has followed them. The Sanscrit tuam, 1 Under the word D^&? shenayim, "two," Gesenius states that the primary form seems to be *0n, thus admitting the priority of the t sound. PREFIX SIBILANTS. 97 Persian tu and to, Latin ft*, 'German du, English thou, agree with the Hebrew atta, Arabic ant, and Egyptian entok in preferring t. We also learn that the first speakers of* the Semitic languages, in forcing the roots to assume a triliteral form, added as a third letter the consonants ng and h, or doubled the final letter when it happened to be d, ts, or %. There were similar laws of change attendant on the other letters of the Semitic alphabet where they occur. The second k, for instance, was added in mathaq, " was sweet," connected with the Sanscrit madhu, "honey." I now give examples to show that the phenomenon of a sibilant prefix, so common in the Sanscrit, and in the European languages, is also a favourite way of modifying the sound of a root among the Semites. The word saphak, "strike," is used 1 in the causative form in the sense of " strike a covenant," which is in Latin pepigit foedus, or in the completed form pactum. In Chinese p'ak is " to strike," and bang, in the modern form pHng, is a " proof," " evidence." In the verb saphak, " to strike," " to punish," there is a variation in the sibilant initial, samech being used for sin. The Hebrews also said for to " cleave," to " open," bakang, bakar, which meanings are expressed in Chinese by pHk. May it not be regarded as probable that 5 was prefixed to the biliteral root in p, k, just as we say 1 Gesenius, Lex. Man. in voc. Saphek. 7 98 smelt, and the Germans schmelzen, for to melt ? If so, then tsakhaq, "laugh," may be derived from kak, the root syllable of cachination, the German Kichern, and the Greek KayaCp. So shakab, "recline/' from kub, the root of cubo, and kvtttco, Mongol hebt'emoi, " lie down." So again, sagab, "to be high," from gab in gibeah, " a hill," and gabahh, " to be high." The word sabar, " to hope," derived from bar, a root meaning "to pierce," "scrutinize," as in the preposition per, and the verbs pierce, bore, may be compared with the Latin spero, " to hope." Tsadik, " just," will then be the same with the Greek S&ccuos and the Latin rectus, and agrees still more nearly with our own straight. By these and similar processes the primitive biliteral roots have become triliteral, and it was thus that the Semites pointed out the path of change to the more youthful Indo-Europeans. Finding among the two families similar laws of change, we assign to the Semitic system, on account of its more simple syllabary, a higher antiquity than to the Indo-European ; and so, when we compare the Semitic system with the Chinese, we must call the Chinese the older, because its roots are in a more rudimentary and primitive form. The Chinese ch'i, " straight," is in the oldest ascer- tainable pronunciation dik. The Tamul- speaking people say takuti, and the Mongols t'egshi. The Greeks used the root dik. The Latins changed it to rek. The English and Arabs prefixed s, and the Hebrews ts. EGYPTIAN SYLLABLE-EXAMPLES. 99 That the Hamitic and Semitic languages were closely connected is now generally admitted. Egyptian words show signs of a more modern form than corresponding Chinese words. I select a few examples 1 from "Egypt's Place in Universal History." g CHINESE OLD. NEW. MEANING. EGYPTIAN. mo ma hemp hma mo wu is not m or am pui ban fei p'an fly to sin, offend pai, pui ban put bak kit pak pu pe hi pei not white rejoice carry bu ubex haa fa, fai The tendency to assume a dissyllabic form is manifest in these words. The language of Ancient Egypt be- longs to a newer formation than the Chinese. When the structure of the Hebrew conjugations, the syllabic suffixes to express the dual and plural, and the pronominal suffixes to nouns, are examined, the advance of the Semitic system from the primeval monosyllabism towards the polysyllabic form becomes still more clear. For example, n is prefixed to make a passive and kith to form a middle voice. The prefix h makes the verb causative, as does the insertion of 1 These examples have been kindly corrected for me by a distinguished Egyptologist. * 100 go and gol in Mongol. The root of the verb to cause is in Chinese ho or kok, and this, as h grows out of h, may be the parent of both these forms. The root thus becomes lengthened into four or five letters and two or three syllables. The extensive use of I, r, s, and ts, as finals to the monosyllabic or biliteral root is another mark of more recent formation. These peculiar finals, entirely un- known in the ancient Chinese vocabulary, occur abun- dantly in the Hebrew, Turanian, and Indo-European syllabaries. The Chinese has I, but not r, in its alphabet, and the Japanese r, but not /. In modern Chinese r is struggling for recognition. In Mongol and Tamil I and r are fully developed, as in the Semitic and European systems. 1 They occur either as initials or as finals. The same is true of the sibilants s, ts, and sh. To the Semitic stock, therefore, should be assigned the honour of developing the syllabary of human speech in this direction. It was this system that first distinguished between I and r as initials, and added them, with s, sh, z, ts, to the list of final con- sonants. From them the Turanians took them during their ancient residence in South-western Asia, but subsequently to the time when they sent away the Japanese offshoot, and left it to pursue an independent existence in the far east of Asia ; for the inhabitants of that island- empire are very deficient in this part of 1 Not as initials in Tamil. GENDER. 101 their syllabary, and their language seems to be the oldest of the three Turanian systems. The vowels being represented by three letters in the earliest Semitic (that is, the Phoenician) alphabet, it is probable that when the ancestors of the Semites left the primeval stem of language, the vowels a, i, u, were sufficient for the needs of human speech at that time. Thus much for the Semitic syllabary. Another mark of advance to be now noted is the growth of the Semitic inflexions. Imagination was always powerful among the men of this race. It gave to the Old Testament in its poetical portions their metaphorical imagery, brilliant description, and rapid movement. This same gift was their inheritance long before the days of the prophets, at an earlier time during the formation of their languages. "We see its effects in the attribution of sex to the lifeless objects of nature. Cedar, gem, bunch of grapes, death, enemy, booh, were masculine. Pillar, egg, castle, intellect, year, sleep, were feminine. Some words, such as earth, fire, were masculine or feminine. The Chinese and Turanian languages know nothing of these distinctions, and hence we infer that this characteristic of the Sanscrit, Greek, and Latin tongues has been derived from the influence of the earlier Semitic type. The feminine was marked frequently by a special suflix, as by h, or th in Hebrew, and by a in Greek and Latin. 102 Among the personal pronouns, ani, the first, was the common property of the Hebrew man and woman, but in the second person a distinction commenced, and was also maintained in the third. In the verb also, when woman or any feminine objects were spoken to or spoken of, a special suffix was used. But in this the Indo-European system did not follow the Semitic ex- ample, preferring to express the distinctions of person by the pronominal suffixes, without giving attention to sex. The Semitic languages gain little by this laborious system of conjugating according to gender, and it has, therefore, nearly lost its place in language. The predisposition of the human mind for poetical and rhythmical expression leads to the introduction into language of many laws, which, on account of their burdensome nature, must ultimately be given up, and cannot be expected to continue their existence in newly formed linguistic families. To such laws the Semitic conjugation by gender must be referred. The dis- tinction of gender in pronouns has lived for a longer period, having lasted from the commencement of the Semitic age down to the modern English, the newest and freest form of Indo-European speech, which, while rejecting the distinction of gender in inanimate objects, has retained it in the personal pronouns he, she, it. One of the most striking phenomena in Semitic speech, the result, like the genders of nouns, of bold- ness in imagination, is the inversion noticeable in the POST- POSITION OF THE NOMINATIVE. 103 order of words. In the first verse of Genesis we read Breshith bara Elohim eth hashshamayim ve eth ha-aretz. " In the beginning created God the heavens and the earth." Why is the verb placed before its nominative ? It is in consequence of a law of inversion which it pleased the imaginative faculty to introduce. It was rendered possible by the previous formation of an objective case. The prefix eth being used to mark the object of the verb's action, there can be no con- fusion between the nominative and accusative, and it is, therefore, at the option of the speaker to place the actor before or after the verb, as he pleases. Guided by a poetic instinct, the Semite usually preferred to mention the verb before the actor. In so doing he departed from the old primeval law of human speech, still remaining in the Chinese and Turanian systems, and allowed the imagination to triumph over the logical faculty, according to which the nominative, as the first in nature and time, precedes its verb. Another instance of the effect of inversion is seen in examples where the verb stands first, the nomina- tive comes last, and the object is between them. Ki Yebiaka Yehova, " For shall bring thee Jehovah," instead of, "For Jehovah shall bring thee." This order is rendered possible by the object ka, " thee," the pro- nominal suffix to the verb, being always accusative, so that there can be no confusion between actor and object. 104 The laws by which the adjective follows the substan- tive, and the demonstrative pronoun its noun, are also caused by this tendency to inversion. The article came into existence opportunely to allow of this being conveniently done. The sentence, " This good land," is in Hebrew, Ha-aretz hattobah hazzoth. Ha, the definite article, is used three times. Tobah, "good/' follows aretz, " the earth," and zoth, " this," comes last. This law also meets us in the Malay and Polynesian languages, where, however, the article is wanting. A more important inversion perhaps than the pre- ceding is what may be called the post-position of the genitive, as in the Arabic zill Allah, " the shadow of Allah." The natural order is " Allah's shadow," as in all the languages east of Persia (including the Sanscrit), excepting the East Himalaic, Malay, and Polynesian systems. Our primeval ancestors, there can be little doubt, spoke of the possessor first, and then what he possessed. The Semitic imagination first seized the name of the object possessed and then that of the possessor. This caused what is called the "construct state." The first word had its vowel shortened, and the plural termination appeared in a clipped form. Thus, DH^T devarim, became ^l^* 7 ! divre, in the phrase divre hangam DJ7M H^l " words of the people." When in Greek we find the post-position of the genitive well established, as in aval; av&pwv, "king of men," and also remember the contiguity of the Greek POST-POSITION OF THE GENITIVE. 105 and Semitic areas and the ancient intermixing of the Phoenicians with the Hellenic race, it seems quite a natural supposition that the Greeks derived it from the Semites. The near neighbourhood of the Assyrian empire and civilization would aid powerfully in the introduction into the Greek language of this and other Semite idioms. The same strong and long-continued Semite influence caused its entrance into the Persian as in Mushk-i Khoten, " musk of Khoten." In English the two modes of arrangement are both in use, and this, as in other European tongues, adds much to the freedom, fluency, and variety which cha- racterize modern speech. Thus the Shorter Catechism commences, " What is the chief end of man ? " and says in the answer, " Man's chief end is to glorify God, and enjoy him for ever." If there had been any difference in intelligibility or propriety of use between " man's chief end," and the " chief end of man," a preference would have been shown here for one of these modes of speaking, to the exclusion of the other. In the English of the nineteenth century there are still no certain signs indicating that the Semitic mode of speech is coming near the end of its reign; and yet it is possible that the post-position of the genitive may pass into an archaism after no very long time. The Greeks said wo? Geov, for " Son of God." The Latins rather preferred to say Dei filius. The Sanscrit-speaking Arians could not transpose their 106 genitive, thus showing that they were under strong Turanian influence, and showed very little sign of Semite connexion. Of the remarkable inversion of order, which in the Turanian and Indo-European families led to the system of case suffixes, there is scarcely any trace in the Hebrew, except in the suffix ah, expressing motion towards a place. But we do not know what this ah was. In the case suffixes of the Sanscrit and Greek we find, or think we ought to find, metamorphosed demonstratives placed after their nouns. Perhaps we should rather say verbs metamorphosed. Looking for an old verbal equivalent to this suffix, we find the Chinese hiang |fij hung, kung, "towards." One of the greatest improvements in language due to the influence of the Semitic mind is the introduction of the relative pronoun. This pronoun is originally formed from the interrogative or demonstrative. In English the demonstrative that has acquired a relative force, and so it may be said of the interrogative who. In Hebrew, the relative pronoun *lB?tf asher, is not so easily accounted for. We find in Chinese an inter- rogative zhok, "who?" which appears in the modern form as shut, after dropping its final and changing its zh to sh, "We also have si, " this," and zhi, " this," both old words; shat, "what?" a dialect word; and shen, "what," or, in an older form, zhim, a Mandarin word. Gesenius prefers to derive asher from the ORIGIN OF THE RELATIVE. 107 primitive demonstrative in s, in Sanscrit sa, sas, English so, she, German sie, and finds the final r in our words there, der, er, etc. The old word ^f si, "this/' and its equivalent jfc t'si, " this," show that the ancient Chinese had the same sibilant demonstrative. But the Hebrew sh has in some words the value t in cognate dialects. Thus, U$ sham, "there," was tarn in Chaldee, Latin turn. Our word asher may therefore be a dis- guised form of the demonstrative in d, used in so many- languages and dialects, Indo-European and Chinese. Thus we have in old Chinese di, " this," and the same in Tibetan, equivalent to the Grerman der, die, das, and the English this and that. In Chaldee we find da min da, " this from that," reminding us of the Tibetan di, "this," and Malay dia, "he." We also meet with di in Chaldee for " who," " which," " that," and as a sign to connect a genitive with its prefixed nominative. The relative is a device for continuing a description without coming to a full stop, and it allows the speaker to proceed without being compelled to commence again with a repetition of the noun. Hence the demon- strative pronoun is taken for this service as the repre- sentative of the noun, and as most suited to undergo the change in meaning which is required by its new position. The Hebrews often omitted the relative, an indication that in the early stage their language was without it. V? $*-73 kol yesh lo, " all was his," that 108 is, "all that was his." They afterwards introduced asher to fill the gap, and make the sentence entirely coherent. The device was successful. They used for this object an obsolete demonstrative, asher, not needed for any other purpose. When the Semites introduced the relative, it was in accordance with the genius of their language, which seizes on the central idea and then describes it in detail. The second verse of the second chapter of Genesis reads, if translated according to the Hebrew order, " And finished Gfod on day the seventh work his which he did, and he rested on day the seventh from all work his which he did." The emphatic verbs finish and rest stand first. Bay precedes its adjective, seventh. Work precedes the relative clause describing it. The action if a verb and the nominative if a noun must in all cases stand out in their clear individuality first. Then the particulars follow, whether expressed by adjectives, by pronouns, or by the relative clause. Such was the mode of constructing sentences which was most agreeable to the Semitic imagination. The eastern Asiatic languages have been content to be guided by the logical faculty. The old Chinese would say, "Seventh day, God's work being completed, then he rested." 1 Here the 1 & B 1 * I t S I ,1 T ' sit nit zhiun 9 U kong tsiun nai kH sik, " Seventh day Supreme Ruler work completed then SEMITIC AND CHINESE RELATIVE COMPARED. 109 time is put first, because it is (viewed grammatically) a subordinate circumstance. The nominative stands first because the actor in the order of nature exists before the act. The verbs completed and rested take the order of time, and one nominative, God, serves for both. The order of nature allows of brief description. If this order is broken in upon, the penalty must be paid in tautologies and circumlocution. The contents of every relative clause are capable of being inserted as a subordinate clause in the principal sentence under the control of the nominative to that sentence. This insertion is what the Chinese make use of instead of a relative clause. The Mongol reads " God," uberon uileduksen idled, " self-done work," jirgogan edure t'egusgeged, " sixth day being finished/' dolodogar edure, " on the seventh day," uberon uileduksen uiles eche, " self-done work from," amorabai, "rested." 1 Here the principal verb, rested, stands last, according to the invariable law of the Turanian languages. The nominative, God, stands first, ruling the subordinate and the principal clause. This is the fixed order of clauses in Chinese and in the Turanian system. What in Hebrew would be a relative clause is here constructed in immediate connexion with stopped rested." From Translation of the Scriptures by Medhurst and others into Chinese. 1 From Ihe Translation of the Scriptures into Mongolian, by Messrs. Swan and Stallybrass. 110 the nominative by means of the possessive suffix attached to the reflexive pronoun self. The influence of Semite speech appears to have been less on Sanscrit than on the other Indo-European tongues. The post-position of the genitive is entirely foreign to Indian grammar, and it seems to make but sparing use of the relative. The Hindoos did not commonly by its means construct a new subordinate clause after the principal sentence. They placed it as a Chinese or Mongol would do in a clause by itself before the chief sentence. They were fond of antithesis, and introduced a demonstrative he to correspond with the relative. In Williams' Sanscrit Grammar, the following example is given. " What you have promised, that abide by." Tat pratijndtam tat pdlaya. The Chinese would say in their modern language, tsen mo shwo, tsen mo king, "how speak/' "how do," meaning, " as you have spoken so do." Here, tsen mo is an interrogative, "how ?" The Sanscrit yad, yali, is simply an old disused interrogative " who ?" " what ?" employed to perform the simpler duties of the relative according to the limited Hindoo conception of them. It is to the European languages that we must look for the examples of the full development of the relative, as a main help to the attainment of that fluency in narra- tive and accuracy in description for which they are distinguished. . CHAPTER YII. i The Himalaic Languages Younger than the Chinese; Older than the tjuranian. — eastern hlmalaic' branch.— siamese phonal System. — Cochin-Chinese Tones. — Chinese Natural Tones. — Vocabulary. — Syntax. — Western Himalaic Branch. — Tibetan Phonal System. — Tibetan and Hebrew Common Words. — Tibe- tan Tones. — Post-Position op Case Particles. — Derivatives. — Tibetan Verb. — Antiquity of the Tibetan Type. On approaching the Himalaic languages on the western side, we find ourselves in contact with a system of case suffixes for the first time. For these we look in vain in the Semitic family, and in Chinese they are limited to the locative case. The Tibetan race connects itself by monosyllabic structure and tones, as well as by a large number of identical words, with the Chinese. But by its system of case particles it is seen to approach to the Tartar and Indian languages. The Tibetan belongs to a system younger than the Chinese, because it places the substantive before the adjective, and the verb at the end of the sentence. In the same way it may be shown to be older than the Turanian family, because, though it strongly resembles that system in placing the case particles after their noun's, and the verbs at the end of the sentence, yet its monosyllabic character and system of tonic pronunciation cause it to approximate to the Chinese. 112 The existence of the case suffixes in the Tibetan language, and the circumstance that the verb is there uniformly found at the end of the sentence, are suffi- cient to justify us in ascribing to the Himalaic family to which it belongs a later origin than to the Semitic. The third great step in the development of human language was made, therefore, in the formation of this family. At the same time it must be kept in view that the Eastern and "Western Himalaic languages are diverse in several important respects. The Cochin- Chinese and Siamese languages have an order like and yet not like the Chinese in the combination of the prepositions with the nouns. All the case auxiliaries are prefixed, whether locative, instrumental, dative, or ablative. In the Chinese the locative auxiliaries follow, and the rest precede their nouns. In the Tibetan they are all suffixes. The Eastern and "Western branches of the Himalaic family thus appear to differ in character very materially, and a division is rendered inevitable. Yet their common tonic pronunciation, and their advance beyond the Chinese in the extended capacity of their syllabaries, may still be regarded as furnishing sufficient ground for retaining them in connexion as branches of one family. In the Cochin- Chinese and Siamese languages, which are the chief members of the Eastern Himalaic branch, an alphabetic series and syllabary exist, much re- SIAMESE PHONAL SYSTEM. 113 sembling the Chinese. The words are monosyllabic. The finals are in Cochin- Chinese, besides the vowels, k, t, p, ng, nh, n, m, and ch. Of these nh is a variation from ng and ch from k. An effort has been made to throw off some of these finals. We find nhot, " day," the Chinese nyit, also spelled ngai, where the t is lost. There is a limited use of r and / after the initials b and t. Thus, tron, blon, trot, blot, all mean " perfect," " whole," and are the same with the Chinese ^ t'siuen, formerly pronounced zien and clzien. In exchange for dz, dj, ch, ts, the Cochin- Chinese introduced gradually the initials tr and bl. They also developed the modern letters r and I out of the old I. The Siamese have done the same, and have also added /, as the modern Chinese have done, to the old alphabetic elements. No other member of the Himalaic family has the letter/. The area of this letter is also limited among the Tura- nian languages to Japan, Manchuria, and Turkestan. The Siamese have no sh, but they have, like all the members of the Himalaic family, an abundant supply of aspirated surds. Thus, k, t, and p, with an aspirate, are extremely common. These aspirated mutes exist in certain localities in Europe, and constitute a main peculiarity in the colloquial Irish pronunciation of the English language, but it is only in the speech of Eastern Asia that they have been made to take the part of distinct letters. From Jones's "Grammatical Notices of the Siamese Language," it would appear that 8 114 there lias been no change in the finals : k, d, b, ng, n, m, rule undisturbed as the favourite consonants for termin- ating all closed syllables. Perhaps d and b, which come in place of t and p, are of even greater antiquity than these last. They may be the vestiges of an era when the surds k, t, p, were still unknown as initials or finals, and when in the primeval alphabet, as now in the Tartar languages, the aspirates and sonants were the only representatives of the triple series known as gutturals, dentals, and labials. The tones are in the Siamese phonal system closely intertwined with the syllabary. The letters are divided into three series, high, middle, and low. The aspirates k', V, p', s, f, h, c'h, are pronounced in the upper and lower series, that is, for example, in a high and low do. The surds and sonants k, ch, t, p, d, b, are in the middle series, e.g., in the key of sol. The remainder, ng, n, m, I, r, 10, y, are in the lower do. The words being arranged on a scale with a triple pitch, of which the two intervals, taken together, vary from, perhaps, a half to an entire octave, the inflexions and even-tones, five in number, still remain to be applied to them. These consist of a slow even-tone, a circumflex, which is a curve of the voice, first down and then up, a slow falling, a quick rising, and a slow rising inflexion. 1 The English and French interrog- ative tone is the same as what is here called the quick 1 See Grammatica Linguae Tai, by Bishop Pallegoix. COCHIN-CHINESE TONES. 115 rising inflexion. In the sentence, " What ! not obey me?" the tone of what is the quick rising inflexion, and that of obey is not unfrequently the slow falling circumflex. The Cochin- Chinese tones are also arranged on a triple pitch, which we may again think of as upper do, sol, and lower do, remembering, however, that the breadth of the intervals and the general pitch of the voice depend on the habit of the individual and the state of his feelings. The tones in this language are like the Chinese, and are not distributed among aspirates and non-aspirates, as are the Siamese and Tibetan, but are themselves set in sol, and the lower and upper do. COCHIN-CHINESE TONES. 1 NATURAL CHARACTER. MUSICAL NOTATION. ENGLISH EQUIVALENT. 1. 2. 3. 4. 5. 6. quick even quick rising falling and rising ( cir <™n- \ flex slow even quick falling slow falling upper do sol, si sol, fa, la sol fa, mi lower do monotone interrogative satirical circumflex monotone interrogative done of remonstrance (tone of decision This system differs from the Chinese only in having a triple pitch, while the tones of Chinese dialects are, 1 Prof, des Michels, " Sur les Intonations chez les Annamites." 1869. 116 perhaps, usually content with a double key. The Chinese, also, often use the other circumflex, namely, that which is bent first upward and then downward. The Chinese, as having a greater variety of dialects, have, of course, a fuller development of tones than the sister races can be expected to possess. They have the quick and slow even-tone, the quick and slow rising inflexion, the quick and slow falling inflexion, and the circumflex of two kinds, first rising and then falling, or first falling and then rising, and each of these in slow or quick time. They may be represented by straight and curved lines, thus : — CHINESE NATURAL TONES. Even stroke quick and slow monotone { DO or n ci tr v°ed e,Straight \ ^ luick and slow falling slide { TTP on S rv°ed C ' StraiSht0r -^' ' " qui* and slow rising slide Curve down and up ' ' \ S quick and slow falling circumflex Curve up and down , ^ /^ ^\ quick and slow rising circumflex Each of these may be placed in a higher or lower pitch, and perhaps there may be an intermediate or triple pitch in some cases. 1 Each dialect selects from this set of tones as many as it requires. The least number of tones that any dialect 1 The intervals may be do, mi, sol, or mi, la, do, and for the double pitch do, sol, or la, do, which last is the interval in Peking for ordinary voices. EASTERN HIMALAIC VOCABULARY. 117 in China is known to use is four, as the Pekinese ; and the greatest nine, as in the Hok lo patois, in Canton province. The waves of the voice in these inflexions are better represented by curved lines than by the musical scale of modern Europe, because the sound of the inflexion is continuous and not broken up into quavers and crotchets. But to convey a correct idea of the vari- ation in pitch noticeable in tonic elocution, reference to the musical scale is highly useful. The vocabulary of the eastern Himalaic languages is in many respects like the Chinese. English fish honey earth clothing mother breath Chinese Cochin-Chinese Greek, Latin ngud ka mid mot da dat terra wei mo ao me vestis yrf\Tf\p k'ui k'oi halitus English | two three four five six fire fowl head Chinese Siamese shong sam song sam si si ngu ha lok hwa koi hok fai kai du how In Siamese h takes the place of the Chinese I and d. The Siamese I corresponds to the Chinese h, as in Chinese hit, "blood," Siamese luit ; Chinese hicang, " yellow," Siamese leuang ; Chinese hung, " rainbow," Siamese lung. The modern Chinese h corresponds to an older k, and will bear comparison with our western 118 china's place in philology. cruor, crudelis, gore, clot. 1 So hwang, " yellow," may be compared with our crocus. The word for "rainbow," hung, is doubtless a variant of hong, " a bow." The ng final of Chinese words sometimes corresponds to our western m. So here the Persian hemdn and Greek /cdfjL7TT(o, " to bend," are derived from the same root. The Siamese pronouns ku, " I," meung, " thou," and k'ea, "he," may be compared with the Chinese nga, "I," the Hainan, Kwangsi, and Kweicheu aboriginal word mu, " thou," and the Chinese gi, " he." The extension of the second personal pronoun mu over the area occupied by the Blue and White Miau of South China, the Hainan islanders and the Shan tribes of Burmah and Siam, 2 helps materially to connect these scattered dialects, stretching from the Gfulf of Siam, N. latitude 14°, to Kweicheu in China, K latitude 26°, into one system. The Western Himalaic languages, including the speech of the Lo lo in the Chinese province of Kweicheu, the Burmese and the Tibetians, do not appear to have this pronoun. They use instead of it for our thou, in the Lo lo dialect, kai, and in Tibetan, k'yed. 1 These words all branch from the primeval root kit, "coagulate," "join together." Hence our kith, catena, and the Chinese kit, "tie," and gin, "near." 2 See Bishop Bigandet's comparative vocabulary of Shan dialects in Logan's Journal of the Indian Archipelago. LAWS OF POSITION. 119 The Siamese, the T'ung tribe in Kwangsi, the White Miau in Kweicheu, the Li tribe of Hainan, and the Shans, all say for " I " and " my," ku, hau, or k'au. These are all merely variations from the widely spread root nga common to the Chinese and Tibetians, and occurring as aham, ego, ich, in Indo-European languages. The tnird personal pronoun k'ia, in Chinese gi, has nearly as wide an area. The White Miau of China say kwa for "he," the Tibetians ho. The Japanese say kono for " this." The Latin is Mc, and the English he. As an interrogative, the same root takes the form of " quis ?" " quid ?" " who ?" and " what ?" It is also extensively used in the Turanian and Semitic lan- guages, as in the Mongol k'en, " who ? " and the Hebrew hu, "he." The laws of position in the Eastern branch of the Himalaic family are very peculiar. In all the dialects, whether those of the Miau aborigines in South-western China, 1 or the Li in Hainan, the Cochin- Chinese or the Siamese, the adjective follows the substantive. It is the same in the Western branch. The Chinese, Mon- gols, Turks, and Hindoos, encircling these languages on all sides, place the adjective before the substantive. The Malays only form an exception. The Himalayan races have not then, in the adoption of this inversion, 1 My authorities for Miau dialects are the Chinese works Hing i fu chi, Kwangsi t'ung chi, and for Hainan a manuscript vocabulary by Robert Swinhoe, Esq. 120 china's place in philology. imitated any of their neighbours. Shall we trace this law to Semite influence, or attribute it to their own independent efforts to effect changes in the primeval type ? Perhaps the latter view may be most favourably received. But an early connexion with the Semites is not unlikely, certainly not impossible. On the other hand, the Eastern branch of this family is, in regard to the position of the locative case parti- cles, older than the oldest of its neighbours. The verbs which mark the cases of nouns are all found before their nouns, and very curiously we see the same principle in operation in the Semitic languages. The Tibetians and Tartars belong to more modern migrations, and at the very commencement of their independent linguistic existence they performed with decision and the most thorough success the feat of transferring the verb to the close of the sentence. This process in- cluded necessarily the post-position of all case particles. There can be no doubt that this Turanian idiom is new, and the Ultra-Indian idiom old. The geographical situation renders this conclusion inevitable. If also it be remembered that the tribes called in the oldest of the Chinese classics, the San Miau, 3 were the first known occupants of the Chinese area, it seems difficult to resist the conclusion that the Eastern Himalayan 1 The reign of Shun, b.c. 2255, in the Shu king, included the pacifica- tion of the San Miau, or three aboriginal tribes, among its chief historical events. EASTERN HIMALAIC SYNTAX. 121 races are older than the Chinese. For how can it be accounted for that the Chinese should have taken the first step in the post-position of the case particles, and that their southern neighbours should show no trace of a similar phenomenon, except on the supposition that in the early migrations from the west, the Ultra- Indians came first and the Chinese next? Yet they continued uncivilized till Buddhist teachers visited them from India and covered the peninsula with monastic institutions and Hindoo practices and beliefs. This was nearly 2,000 years ago. At about the same time, the light of Chinese ancient culture penetrated also among them, especially in the reign of Han Wu ti, B.C. 100. They never originated, like the Hindoos, a mighty kosmos of the imagination, nor, like the Chinese, a complete practical system of the arts of life. The vast Cambodian temples, with their long colonnades, now hidden in the glades of unfrequented forests, the shining kiosks of modern Bankok, the books of prayers written on the palm leaf, the invoca- tions to Grautama, and the ascetic and convent life, are all Hindoo. The agriculture, the usages of commerce, the mode of government, are all Chinese. Hence their languages have probably changed more than the Chinese. Speech owes its persistence to civilization. Ancient words are crystallized in litera- ture, even if they are dropped from their place in the familiar intercourse of men. Barbarous idioms alter 122 china's place in philology. rapidly. Laws of grammar, words, sounds, meanings,- accents, are in perpetual transition. Hence the novel aspect of much of the vocabulary of these races. Living as separate tribes, the language of each has undergone rapid changes. But through all the principles of grammatical structure and the outline of the phonal system appear to have retained their ancient features. The marks of primeval formation are most remarkable, and their consanguinity to the Chinese type is as un- deniable as is their likeness in lineaments to the common mother from which all languages sprang. We do not meet with any full representative of the Western branch of the Himalayan race till we arrive at Tibet and Burmah, and perhaps the Chinese Lo lo. The Karens are in a half-way position between the two branches. They place the adjective and the demonstrative pronoun after the substantive, and the case particles before the object, whose relations they define. The possessor precedes the object possessed, as in all the Eastern Asiatic languages. They have six tones and a strong attachment for vowel finals. All the consonant finals have been thrown off, except ng. Although in vocabulary they have borrowed much from the Burmese and Tibetan languages, 1 they cannot with these laws of position be rightly classed anywhere but ih the Eastern branch. What strikes the eye most remarkably in the Tibetan 1 Logan's Journal. TIBETAN PHONAL SYSTEM. 123 syllabary is the prefixed letters. The early speakers of this form of human speech, not having before them the idea of terminations, that happy device made use of by the founders of the polysyllabic languages, bent their strength unconsciously to v add letters at the beginnings of the roots. In so doing they remind us of the Semite system, which, in the conjugation of the verb, prefixes n to make a passive, m to make a parti- ciple and an agent, and h to render the verb causative. The favourite prefixes of the Tibetans are g> d, b, h, m, r, I, s. Csoma de Koros says, they are in modern speech seldom heard. Hence this effort to extend the mono- syllabic root at its beginning must be regarded as a failure. The letters thus ineffectually placed at the commencement of the words are k y g, d, 5, m, r, /, s, and h. They help to distinguish words having the same sound, and thus in the written Tibetan they serve a useful purpose. Two letters are also added occasionally at the end of the root, namely, s and h. Csoma de Koros says, mi-mams is pronounced as it is written, but when the r is not preceded by a word in close combination it is silent. When these prefixes and suffixes are cleared away from the word, it is reduced to the radical form. Mi is the substantive root man, and nam is the plural suffix. The six consonant finals of the Chinese language occur again in the Tibetan, but with a slight variation. The mute surds k, t, p, all occur in the sonant form 124 CHINA S PLACE IN PHILOLOGY. g, d, b. The Tibetian cannot shape his vocal organs so as to pronounce k, t, p, at the end of a syllable. This peculiarity is more fixed in the Tibetan than in the Siamese, where k, d, and b occur as finals. In this respect, therefore, the Tibetan has the palm of an- tiquity ; for, as before remarked, there is a strong probability that k, t, and p are newer letters than g, d, and b y and have been derived from them. The capacity of the Tibetan syllabary is much en- larged by the addition of the final^-r, I, s. By the separation of r and /, the Tibetan phonal system is shown to be more modern than the Chinese, which has only /, and to stand on the same footing with the Eastern Himalaic and Semitic systems. R is very abundant in Hebrew as a radical and a suffixed letter, and its extensive use in Tibetan warrants a sus- picion of ancient intercommunication between the two families. The Bod race left Western Asia later than the families lying more to the east, and would naturally remain in contact with a Semite population for a much longer period. In Genesis it is said that the sons of Ham were Cush and Mizraim, and Phut and Canaan . As Cush had eastern and western branches, so may Phut have had also, and he may be the race-father both of the Libyans and of the modern Tibetians, now spread over the whole of Tibet and Bootan in the Himalayas. That section of the armies of Grog and Magog mentioned in Ezekiel as belonging to the TIBETAN AND HEBREW. 125 Phuttian race may have been contributed by the Eastern branch. However this may be, the Bod race and the Semites are, at any rate, alike in their fondness for prefixing various letters to their words, and in adding r and I as suflpjxes. The Hebrews said gilgal, galgal, for a "wheel," and gulgoleth, for a "skull," from which comes Golgotha, the Aramaic original of the Latin Calvary in the Gospels. The Tibetians say kor, "a circle," Mor, "; wheel." The Chinese have gu, "a, ball," and kit, "a garland," "a hoop." To this root the Tibetians added r, the Semites /, and then farther west it took the forms circulus, kvkXos, kv\Iv$g). The Sanscrit chakra, "wheel," and chahrawat, "circular," are from the same root, by the common change from k to ch. The Hebrew, I^J? ngagur, " revolving," 73JJ "revolve," 7^ "round," may be also included, because the primitive value oiayin, the initial consonant, is ng or g. Among the letters the surd mutes k, t, p, are very weak. They scarcely fill a page each in the dictionary. The first k is the most important. The aspirated forms kh, th, ph, k', t\ p { , abound, as do the sonants g, dy b. Much the same law appears to exist in the Hebrew vocabulary. The two &'s together cover fifty- one pages, while the aspirate heth covers seventy, and g and ng together ninety- seven pages. The surd t, the aspirate th, and the sonant d, occupy respectively, 126 china's place in philology. five, forty, and twenty-three pages. The labial series includes,^ and/together, eighteen, and b seventy pages. Compare these results with the Sanscrit vocabulary, and it will be found that the tables are turned, k and its cognate ch occupy 108 pages, their aspirates eleven, and the sonants g and,/ fifty- seven. The dental series, t, d, and dh, has the numbers thirty-one, forty, twelve. The labial series p, b, and p', b e , has ninety-six, eleven, twenty-one. Here the surds have a clear superiority, and the influence of the aspirates has greatly diminished. These facts reveal the existence of a great general law, according to which the aspirates and surds grew out of the sonants. The older vocabularies, as the old Chinese, the Turanian, the Tibetan, and the Semitic, have a preponderance of sonate initials and finals, b, c, d. Then the limits of language were extended to satisfy the ever-increasing wants of the historical races and the advance of civilization, and the aspirates appeared, Jc { , V, p l , h l , A, 6, , /, with the surds k, t, p, h. These would spring up in some countries contemporaneously. In others, as in the Tibetan and Tartar languages, the aspirates grew into use alone, and the surds slowly followed. This law embraces the celebrated Grimm's law as one of its particulars. The reason why dip, deep, door, are in German taufen, tie/, and Thur, is that the English vocabulary is in this respect older than the German, and that the German has advanced one stage farther TIBETAN AND HEBREW COMMON WORDS. 127 than the English in the development of the surd initials. K, t, p, have grown out of g, d, and b, just as we have seen r and I, in the Semitic and Himalaic systems, grow out of an original / in the old Chinese. Our English / and th have grown out of a more ancien^ b and d. Father is in Hebrew ab, in old Chinese be, in Turkish baba, in Tibetan yab, in Latin pater, in German Vater. In the older syllabaries it was ba and ab, and here we see another lurking similarity existing between the Tibetan and the Semitic families meeting as they do in the use of ab, yab, " father." It is found with p in some southern Chinese dialects, and in Sanscrit, Greek, and Latin. Modern Chinese agrees with English and German in giving the / sound. New vocabularies have a pre- ponderance of surds, as old vocabularies delight in sonants. Grimm's law is not so much a circular law, as one of perpetual advancement. TABLE OF TIBETAN AND HEBREW COMMON WORDS. TIBETAN. HEBREW. ENGLISH. MISCELLANEOUS. yum em mother lang lakahh receive Greek \ay xdvco. lug rahhel {ewe) sheep log, side tselang rib, side Chinese lok, rib. rum rahham womb rab, exalted rab great ring rahhoq long, far Chinese dung. la r to lha, gods eloah God, angels langs, vapour, ghost ruahh spirit, wind rogs, friend rd-ang friend 128 In this list of words common to the Tibetan and Semitic vocabularies, the circumstance that r agrees usually with r, and I with I, is itself evidence that the two families grew up together in their early youth. The Tibetan, like the Mongol, Cochin- Chinese, and the Indian languages, has not to this day admitted / into its alphabet ; but sh y which is not used in Mongolia or Japan, has here, as in the Semitic and Chinese languages, a full development. The tones of the Tibetan language are mentioned by Georgi, but no allusion is made to them by De Koros or by Schmidt. They are attached, like the Siamese tones, to the letters of the alphabet in sets, and are, I believe, arranged in a scale of two elevations, as is usually the case in Chinese. Dr. Jaeschke, a missionary long resident in Ladak, and who has extended his researches into the various provincial dialects, informs me that the tones are limited to the central provinces. To compensate for the loss in colloquial pronunciation of letters recog- nized in the written language, the Lhasa dialect has introduced (1), an aspiration which may be symbolized by the forms dh, gh, bh, jh ; (2), a deep tone. The word chang, "beer," sounds high, short, and sharp. The deep tone is a lower slow rising inflexion. By comparing Dr. Jaeschke's remarks with my own notes on the Lhasa pronunciation, as taken from the lips of a native visiting Peking, I believe it is correct to state TONES OF THE TIBETAN LANGUAGE. 129 that surds are pronounced with the upper quick falling intonation, and sonants with the lower slow rising. Among the sonants, however, there has been a loss of g, j\ d, b. These are by the Lhasa people pronounced kh, &h i t'h, ph. The loss thus sustained is compensated for by the lower or deep tone. This change resembles that which takes place in China in passing from the old middle dialect to Mandarin or to the Hakka, when du, " map," becomes Pu with the aspirate. In both countries the sonant is the older, and the aspirate with its special tone the newer form. What caused the tones? I believe Dr. Jaeschke to be right in his view, that it was the loss of letters. When certain initial and final letters, faithfully pre- served in the book language and in the dialects of the western provinces of Tibet, became mute in the neighbourhood of Lhasa, the tones were affixed by an unconscious effort of language to maintain dis- tinctions between words that would be otherwise confounded. This hypothesis of the origin of tones agrees with that advocated long since in my " Gram- mar of the Chinese Colloquial Language." After the researches of Dr. Jaeschke, which show that the mute letters of the Lhasa dialect and of the written language, are all heard distinctly in the pro- nunciation of some of the frontiers, philologists must regard the written form of the Tibetan, with its 130 china's place in philology. troublesome compound letters, as^faithfully representing the old state of the language. A Semitic principle here appears working itself out in a somewhat exaggerated manner. The language made too great an effort to expand itself by prefixes and suffixes, and is now throwing them off, and gradually assuming the primeval monosyllabic form. The tonic element seems destined to extend itself in Tibetan, as it has done in Chinese. It is now in the Lhasa dialect doing the work which was formerly done by the difference between surd and sonant initials. The syllables kha and ga have assumed tones, and ga has changed to kha, so that they are now separated only by intonation. A native of Lhasa reads kha for ga, and intones the syllable. A subject of great interest in Tibetan is the post- position of the case particles. Excepting the locative case suffixes' of the old and new Chinese, there was no earlier family from which the Bod race could borrow this idea. It manifestly originated in the post-position of the verb. For it is more likely that the case particles should take their place after their nouns, as an instance of a general law which drew all the verbs into that position, than that they should first go there themselves, and then draw the other verbs after them. There is little difficulty in con- ceiving the way in which the locative case particles came, in the old Chinese, to occupy a position after POST- POSITION OF CASE PARTICLES, 131 their nouns. They are in fact, as explained in a previous chapter, treated as substantives following other substantives in the relation of the part to the whole. In the phrase tHen Ma, " the world," literally "heaven under," the word "under" is viewed as a noun, "that which is under." The possessive particle ck'i might be inserted, t'ien ch'i hia, showing that we are quite right in regarding the Chinese post-position of the locative as only an instance of the juxtaposition of substantives. The Chinese language cannot, therefore, explain the great inversion of the Tibetan and Turanian languages, according to which the verb with the case particles comes after the noun. Perhaps the best explanation is found in a general tendency of these races to collect the energy of linguistic expression at the end both of sentences and words. The boldness of the Semite imagination was caused by religious culture, the habit of meditating on the objects of the spiritual sphere, and the possession of the primeval revelation made in the antediluvian period, and handed down from age to age. Hence poetic laws control the language and literature of the Semites. They attribute life to inanimate things, and action to objects that are at rest. They filled the world around them, as they did their grammatical paradigms, with the dis- tinctions of gender. The Tibetians and Tartars are at the opposite pole. They are almost destitute of 132 imagination. The sun and moon, the river, the stone, the mountain, are to them simply what their names imply— " A yellow primrose was to him A yellow primrose, and no more." They take the world quietly. Things are to them before action. Personification is to them an absurdity. The effort required to look on the universe as animated with living forces is to them almost impossible. Their books are translated, their alphabets are borrowed, and they have learned the arts of life from their neighbours. They let go with facility the old Turanian religion, and took in the place of it the Buddhistic faith, a creation of the dreaming Hindoo. This pleases them because it teaches inactivity. The thought of Nirvana imparts to them consolation, because it consists of unbroken rest. The Tibetians have two substantive verbs, nyug and dod, which mean either " to sit," or " to be." Sitting is being. Races of active intellect do not form substantive verbs thus. In conformity with this predisposition to inactivity, they postpone the place of the verb in a sentence to the end. All the details are carefully completed before action commences. A nation with very little poetry will have an unpoetical language, for the child is father of the man. A language, the work CASE PARTICLES. 133 of a race in its childhood, will be found to resemble the literature which that race achieves in its maturity. So the Mongol and the Tibetian, in intro- ducing the principle of the post-position of the verb, have only done what we might expect from the dullness of their literary development. The case particles in Tibetan are few. There is a possessive, kyi, gi, gyi, hi, and yi. In Chinese dialects occur as possessives, ku at Shanghai, ge and e at Amoy. They are probably identical with the Tibetan and with the demonstrative roots ki, gi, i. An s appended to the possessive particles makes them instrumental, and the sense, " by means of," " by the use of," is thus conveyed. Among the dative case suffixes the commonest, la, may be the Semitic le, used as a dative prefix. The Tibetians may have borrowed it at some ancient period of contact, before the Persian race separated them from the Semite area, and before they migrated to their present locality. The case suffixes, expressive of motion towards, tu and du, as in lag-tu, " into the hand," Bod-du, "into Tibet," are probably the Chinese to, in Mandarin tau, " towards," " to." The Mongol corresponding case suffix is de. After a vowel ru is used by the Tibetians for tu and du. This I incline to think is changed from du. Thus, ring, "long," is in Old Chinese dung, in Mandarin &hang. 134 The locative suffix in is na or la, and the ablative nas or las. Such is the beginning of the declension of nouns, which expanded itself somewhat in the Turanian languages, and grew to its fullest dimensions in the Sanscrit. We have also in Tibetan the rudiments of the system of derivatives. The following forms are in use : Monosyllabic Suffixes : pa, ba, ma, po, bo, mo, ka, k'a, ga, nga, ge, nge, ni, p'o, mo, bu, hu, gu, ngu, nu. Dissyllabic Suffixes : papa, pama, papo, pamo, bapa, bapo, bama, bamo. Closed-syllable Suffixes : chig, zhig, chag, dag, nams. The various significations of these suffixes are as follows : Plural Suffixes : chag, dag, nam. Diminutives : gu, ngu, nu, bu, hu. Masculine : po, bo, pa, papa, papo. Feminine: ma, mo, pama, pamo. Agents or Verbal Substantives : po, ba (masc. or fern.). It is the tonic pronunciation which prevents derived words from becoming dissyllables and polysyllables. The inflexions attached to the root and the suffixes have a tendency to check the consolidation of the syllables into a unity. Yet this is in time overcome. In the Peking pronunciation of Chinese a suffix very frequently loses its tone and becomes de facto a part of the word which precedes it. 1 The verb forms its infinitive by appending r to pa 1 Mandarin Grammar. THE TIBETAN VERB. 135 or ba, as byed par (pronounced ched par), "to do." Byed pa is either a present participle or a verbal noun, " doing." Byed alone is an indicative present, "he does." In many cases verbs are placed in the indicative present by adding byed, "do," as an auxiliary, as in za . par byed, " he eats," gro par byed (pronounced t'o par ched), " he walks." Other auxiliaries, zhin pa, hdug, snang, are used with the same force. Yerbs are made preterite by affixing s. An auxiliary verb, hdug pa, "was," placed after a verb, changes it to the imperfect tense, as hong hdug pa, " he was coming." The future adds hgyur. In the form for the imperative we meet curiously with a Semitic peculiarity. The vowel a or e is changed to o. Za, " he eats," becomes zo, " eat." Sel, "he cures," becomes sol, "cure." In the Hebrew paradigm, katal, "he killed," becomes in the imperative ktol; and sabab, "he surrounded," becomes sob. To a change like this there is no parallel in Chinese or Mongol, and it is difficult to conceive any explanation but that of ancient Semitic connexion. A precative is formed by the suffix chig, zhig, or shig. This may be the Chinese root sik, "give," or shung, " reward." The Mongols in their imperative add a verb "to give" just in this way, — T'a naded helji ug, "you me for speak give,", that is, "be kind enough to speak for me." 136 CHINA S PLACE IN PHILOLOGY. The conditional suffix na is undoubtedly derived from the root nak, which now appears in Chinese as jo and ju, "if." Thus, byed na, "if you do." The changes in the prefixes of the Tibetan verb are due to a principle which was also at work, as before noticed, in the formation of the verb. To " call " is hgngs, in the indicative present, bkug in the preterite, dgug in the future, and k'ug in the imperative. The Chinese root is kok, " call," in Greek KaXeco, in English call. The prefix h frequently marks the present, b the preterite or future, and d or g the future. Another principle, to which attention should be drawn, is the change, with the moods and tenses, from sonant to surd and from aspirate to sonant, e.g. from g to k and from k ( to g. TIBETAN. CHINESE. Present. Tret. Future. Imper. New. Old. ht'ags btags btag t'og chi tek texo weave htogs btags gdags fogs Saxon tig, a tie tie hbigs P'ig dbig p'ig p'iau p'ok prick pierce hbyed p'ye dbye p'ye TrpaTTOo, irpayfjia do ht'sog btsogs btsog tsog siau sok seco, section cut hdzem bzem gzem zem schamen, shame shame hgegs bkag dgag kie kak check hinder hgebs ! bkab dgab k'ob kai kap tfu7rra>, Heb. kafar cover 1 This is a widely-extended root. The Chinese kap means " head," " covering," " coat of mail/' and " to cover." The Tibetan has k l ob, " a covering," and mgo, " head." The "Western languages have caput, Eaupt, head, KtQah-f), crab, cope, etc. ANTIQUITY OF THE TIBETAN TYPE. 137 But these principles, the first of prefixed augments and the second of the interchange of allied letters, have not been carried through the language, and they have failed to acquire the authority of irresistible law. This may have been owing to the want of strong will in the race speaking the language. Although characterized by this weakness, the principles here alluded to are deeply interesting as examples of very early efforts of the human race to conjugate their verbs in a way neither Semitic nor Indo-European. The geographical position occupied by the Tibetians indicates that their language may be expected to be a stepping-stone between the oldest and the newest types. The Chinese are on one side and the Persians on the other. But no early literature crystallized the language in its ancient form. How far it may have lost features which once belonged to it, it is now impossible with accuracy to determine. The antiquity of the Tibetan type, as compared with the Turanian and Indo-European, cannot be ques- tioned, when its monosyllabic character and stunted derivative system are properly considered. The only modern-looking feature is, indeed, the post-position of the verb and of the case particles, as already alluded to. The personal pronouns show that the long neighbourhood of Mongols, Turks, Hindoos, and Persians, has failed to have any effect on the Tibetar towards introducing into it their favourite word*?, lho£ 138 and me, be and become. The first personal pronoun in b or m, the second in t or s, and the substantive verb in b, are used over the whole vast extent of the Indo-European and Tartar area, but into no Tibetan or Chinese dialect have they ever forced their way. The long continuance of linguistic differences between races that have been living side by side for thousands of years is at least as remarkable as the mutual influence they exert on each other's vocabulary and grammar. In the Tibetan pronouns and substantive verbs we see a Chinese impress. Nga for "I," k'hyed for " thou," k'o for " he," with yin, yod, for " to be," "to have," reveal a cousinship with the countrymen of Confucius. They are apparently no other than the old Chinese words nga, " I," m, 1 " you," gi> " he," wei, "be," u, "have." 1 The common Western equivalent for the Chinese ni is h, g, or k'. CHAPTER VIII. The Triple-Branched Turanian Family : Japanese, Dra vidian, and Tartar. — First, the Japanese. — Japanese Syllabic Alphabet. — Common Roots in Japanese. — Formation of Com- pounds. — Case Particles. We now pass the boundary between the mono- syllabic and polysyllabic languages. The dividing line is a sharp one, which the traveller crosses from the region of tone systems and carefully-pronounced inflexions of the voice to the 'freedom of polysyllabic speech. He suddenly finds that he is where tonic laws have been thrown away, and all accented and inflected elocution has been transferred from the region of the syllabary and the vocabulary to that of the passions and the will. It is but a short distance from the Chinese city to the Mongol en- campment, but the change of scenery is great. An agricultural plain, studded with villages and clumps of trees, with all the signs of industry, is left at the foot of the chain of mountains, along which the Great Wall is built. These mountains must be ascended, and at the height of 2,000 feet commences the table- land, which has received the name of "the land of 140 china's place in philology. grass." Field labour suddenly comes to a termination, and everywhere are seen the marks of pastoral occu- pations. It is the land of the roaming deer, the patient camel, the vigorous ox, of tents and fleecy flocks, and droves of ponies; of vast plains without trees, and a limitless horizon, only varied by the undulations which this immense prairie has retained from the far distant time when it formed part of the bottom of the primeval ocean. Such is nature's own well-defined line of separation between the mono- syllabic and polysyllabic languages. But it is necessary to begin with an older stock than the Tartar. The Mongol and the Turk are much nearer to the Western type of language than are the far-off Japanese, nor apparently can the Indian Tamul compete successfully with the Japanese and the Corean for the prize of superior antiquity. In looking at the Japanese alphabet, with its forty- seven syllables, generally terminating with a vowel, we remark at once several limitations. The letters r and I are not separated. The Japanese use r, and the Chinese I, and these letters are employed only to commence a syllable. In Mongol both are used at the commencement and close of syllables. The Dra- vidian languages have a very full development of r and /. As a child whose mother-tongue is English learns to distinguish the other letters first, and r and I last, so it is in the comparative chronology of Ian- JAPANESE LANGUAGE. 141 guages. The distinction between r and / is a sign of late formation. Judged by this test, the Japanese and Chinese are older than their Western neighbours. The word mid, " honey," has final d in old Chinese, and in the Sanscrit it is madhu. In Hebrew we find mathckk, "was sweet," doubtless the same word, and here the final k is a Semitic addition. The Greeks had a wine called /juedv, "mead." The Turks and Mongols use I final, and change the initial m to b, saying bal. The Japanese have mits, and the Tamul madu. The Greeks and Latins appear to have followed the Turanians in the use of the final /, as in mel, /j,e\i, " honey," fiiXcacra, "bee." Here the Greek is more under Turanian influence than either the Sanscrit or Germanic branches of the Indo-European family. Also, the Tamul and Japanese both appear to be older than the Tartar subdivision of the Turanian family. Take another example. The Mongol gol, " river," is in Japanese kawa, in modern Chinese ho, and in old Chinese ga. The addition of I seems to have been made after the separation of the Tartar and Japanese races. The word for crow, Kopai;, in Latin corvus, is in Sanscrit kdka or karada. The Mongol is k'eriye, and the Japanese karasi. The Chinese have kwa, in the modern compound lankica, "crow," where lau means "old." The Chinese and Sanscrit forms indicate that r is an addition to the primeval root. The Hebrew form is 2°$, where ayin, 142 china's place in philology. as very frequently happens, represents k or g, and the word may read goreb. The r medial connects the Hebrew, second Sanscrit, Mongol, and Japanese forms in one group. The last addition, b in Hebrew, v in Latin, ks in Greek, h in German (Krahe), d in Sanscrit, ye in Mongol, si in Japanese, must, from its variety, have been made after the separation of the races. Thus, the Japanese, although to the east of China, are connected more closely with the Western than with the Chinese system. It may also be inferred that the Japanese brought r with them in their migration eastward, and the question then arises, whether the initial r of Western languages is older or younger than the Chinese /, to which it corresponds? The Old Chinese hit, "musical tubes in definite lengths, used for regulating weights and measures," agrees in idea with the Greek pvd/JLos ; and the Latin ritas of the same group corresponds to the Chinese li or lit, " ceremony." As I is easier for young children to utter than r (mothers tell me that they can say / a year and a half sooner than r), the palm of priority in the history of language should be accorded to I in this case; and thus the bulk of Western roots com- mencing with an initial r may with probability be supposed to have taken it in exchange for a more ancient /. Another peculiarity in the Japanese syllabary is, that the aspirates are wanting, If words cross the JAPANESE SYLLABIC ALPHABET. 143 sea to Japan, whether Chinese or Mongol, the aspirated letters, k', t', p' y become simple surds, namely, k, t, and h, or /. Insulation seems to be the cause of this change. The absence of sh, ch, zh, and j from the syllabary, gives it a very defective appearance, but this is one of the characteristics of some of the most important Turanian languages, and helps to establish the near kinship existing between them and the Japanese. Perhaps it should rather be said that these letters are used to a small extent. In Hepburn's very valuable Japanese Dictionary the syllables si, tsi, dzi, are written shi, chi, ji ; but this mode of writing, though doubtless convenient in some respects, is probably not so accurately descriptive of the real sound as the Dutch spelling. In explanation of the want of sh and its cognate letters, it may be men- tioned that in the Mongol and Tamul languages they are not found. The Mongol has indeed occasionally an sh, but it is only, like the same letter in Japanese, a modification of si. So the Mongol ch' is in fact a modified ts', and,; is a disguised d, as will be shown. The surds and sonants are by the Japanese con- sidered as so closely allied, that a short double stroke on the right hand is used to change k, t, and s ( , into g, d, and z. The letters /, b, p, are considered as one sound under three modifications. The double stroke denotes b, and a small circle p. Thus kami, " god," 144 china's place in philology. " spirit," becomes garni, in the combination onna garni, " a goddess." Here it is on account of a word pre- ceding it that k becomes g. That k and g were originally one letter seems likely also because the sounds of the Chinese language are by the Japanese written with extreme irregularity. Thus k and g and other pairs of cognate letters, carefully kept separate in Chinese dictionaries, are in the Japanese transcrip- tion much intermixed. The Chinese sin, " heart," is spelt sin or zin ; and zhin, " spirit," " divinity," " the genii," "marvellous," is in Japanese spelt sin or zin, as in zin riki, " marvellous strength " (in the native language, " kami no chikara ") ; while Japan is called sin koku, " kingdom of the genii " (in the native language, " kami no kuni "). This tendency to an interchange of surds and so- nants is probably due to the recent appearance of either the surds or the sonants. In the syllabary, it is the surd series that holds the place of honour, and it is therefore likely to be the older. When the Japanese, nearly 2,000 years ago, invented their alpha- bet, or rather borrowed it from China, they made no provision for g, d, b, or z. This was a later addition, dating from the time when Corean, Chinese, and Hindoo Buddhists propagated their religion in Japan. As an auxiliary proof, it may be mentioned that the Mongol egude, " door," appears in Japanese as kado ; yek'e, "great," as ikai; maihan, "a tent," as makuya, JAPANESE SYLLABIC ALPHABET. 145 though this word may be directly derived from the Chinese (mu, in the old form) niok, "a tent," in Japanese maku, a "curtain." 1 Why should & always occur? It is very likely that there was at that time no g, as there was no aspirated k. But jt is necessary to carry this inquiry further. The Mongols have g, d, b, and the aspirates, but no hard surd series. I suppose, therefore, that this was also the primitive condition of the Japanese phonal system. As the two races are alike in grammatical structure, and have many identical words, they may long ago have had the same sort of alphabet. The g may have become k after the progenitors of the Japanese passed to their island home, and subsequently g may have been again developed as a sub-division under k, or vice versa. The softness and simplicity of the Japanese syllabary, admitting no final consonant but n, and terminating all its forty- seven syllables by the five vowels a, % e, o, u, seem due to the mild and damp climate induced by its insular situation. Its syllables are predominantly Polynesian in form, but certainly not because of near connexion in race. The Polynesian islanders place their verbs before the objects on which their action is exerted, and their adjectives prefer to follow the nouns 1 Compare also Japanese katai, kataku, " hard," Mongol k'at'ago, and in the eastern dialect, hat'o. 10 146 china's place in philology. they qualify. 1 In Japanese the verb follows its accu- sative, and the adjective precedes its noun. It may be concluded then that, as Hoffmann pointed out in his notes to Donker Curtius' Japanese Grammar, 2 the family connexion of the Japanese language is with Manchu and Mongol. This being admitted, that difference in the syllabaries which consists in separat- ing final consonants from the first syllable, and causing them to form new syllables, should be attributed to the relaxing effect of sea air on the vocal organs. The Mongols can say gos, gol, gang, gar, yah, ed, beg. The Japanese will make dissyllables of all these, thus in- creasing the influence of the vowels at the expense of the consonants. The Chinese dok, "poison," becomes for instance dohu. In the present state of the Japanese syllabary, ng has taken the place of n final, but this has not affected the orthography. N is still written. The sibilants s, ts, and dz, also sometimes drop their vowel, and in actual pronunciation take their place as final letters. 3 The language of the Japanese had already become 1 Notes by T. Gulich, M.D., on the language of Ponape, one of the Caroline Islands. 2 Professor Max Muller, writing in 1861, has invested Prof. Boiler, of ' Vienna, with the honour of discovering that the Japanese language ought to be called Turanian. But the resemblance had several years before (1857) been perceived by the penetrating sagacity of the Dutch Professor, to whom we owe so many ingenious remarks on the Japanese language. 3 Hepburn notices some other final consonants, as m and p in certain positions. COMMON ROOTS IN JAPANESE. 147 polysyllabic when transferred from Corea to their islands, for a few Mongol words of three syllables occur in the vocabulary, e.g. kataku, " hard," Mongol k'at'ago. The root is the same with that of our word hard, and the German hart. In Sanscrit we meet with kaParay "hard," kat'ina, kdt'inya, "hardness," kat'a, "rock," Mongol k'ada, "rock." The first two syllables of a native Japanese word usually represent the monosyllabic root. Thus kit, in Chinese " to harden," " coagulate," " tie a knot," kin, " hard," " firm," occur in Japanese with long suffixes. Katamari is " to become hardened," katame, " harden," and as above adduced kataku, katai, " hard." So also the root zhut, in Chinese, sheu, " to give," " to receive," formerly distinguished by tone, the one taking the rising (shang), and the other the falling (k'ic) inflexion, but now amalgamated in the falling tone class, is found in the Japanese vocabulary, with the forms, sadzukaru, " to receive," and sadzukeru, " to bestow," or sadzkatta and sadzketa. The Chinese have ch'i, " to stop," in Japanese todomari, "to be stopped," todokori, "to im- pede," "stop," todome, "to stop," as in uma ivo todo- meru, "stop ahorse," todomerare, "to be stopped." The Chinese has lost a final t, which appears in dot, " to stand," "to tread upon," " rest the foot." The Chinese initial eh in all cases comes from t or d. Hence the root assumes the form dot, " to stop," and dat, " to stand." This is really the root of our " stand," the 148 china's place in philology. Latin sto, the Sanscrit st'ala, " stand," st'dna, " a place." The Tamul has tandu, " a stand," the Japanese also say tatsi and tatta, "to stand." The initial s was prefixed by the forefathers of the Indo- Europeans before the separation of their western and eastern branches, for they all have it. The primeval root was probably dad and dan. It may have originated from the noise of the foot striking the ground. Families of words closely allied are not wanting. Among them may be mentioned the Chinese ti or dad, "earth." Sanscrit dhdrd, Latin terra, Cochin- Chinese dat. The earth on which we stand receives its name from the verb "stand," and is a verbal noun, just as "inkstand," and "the grand stand" at a race course, receive their names for a similar reason. "We are now in a position to compare the Japanese roots with Chinese, Semitic, and Himalaic roots, and with those of the cognate Turanian languages. In doing so the Chinese initial h must be read k or g, ch must be read t or d, and / must be read p or b. Thus ho, " fire," is gal in Mongol and color in Latin, where the inserted I shows that a Turanian influence has been at work in the formation of the Indo- European polysyllable. The Greek /ccllco, "burn," and Grerman heiss, English hot, are connected, as also the Sanscrit kdrshanava, " hot," and the root kdsh, " shine." The Chinese word ho, " fire," was in the COMPARISON OF ROOTS. 149 time of the creation of the syllabic spelling, a.d. 500, pronounced ha. More anciently it was ka, and more anciently still ga, which is as far as the analogies of the connected languages will carry us. The Japanese have koge, " burn," " scorch," in Chinese k'au, " scorch/' Our word scorch, if the prefix s be removed and the ch changed for its ancient equivalent, k, appears to be the same word. The letter s, when prefixed to a consonant, never belongs to the root. The Japanese say for " fire " hi. This must for etymological use be changed to bi, or pi, as in the case of all words beginning in Japanese with h or /. It may then be compared with the verb aburi, " roast," Tamul pori, with the Chinese bun, " burn," the Greek Trvp, Latin comburo, and the English burn and fire. For " warm " the Japanese word is atatakai, and for " hot " atszku. The root is at, for the sibilant form of t in the latter example is accidental. We may compare it with the Persian atesh and the Hebrew &?{St esh. The etymological equivalent of the Hebrew sh is t, as in shor, "bull," taurus ; sham, "there," Chaldee tarn. The Persian final sh is thus seen to be a reduplication of the final t. Compare the Greek aiOco and Latin cestus, which Gesenius believed to be connected. This author proceeds to say that ^fttf ur, "light," belongs to the same family relationship. This is an extremely interesting identification, because the letter r occupies a frequent place among the 150 Semitic initials. For convenience of comparison with Chinese roots it should have the value d. For example, VfV\ rosh, "head," may be advantageously compared with the Chinese t'eu, "head," old form dut. The equivalent of sh being also t, the resemblance is complete. A fourth Japanese root for "fire" is yake, "to be on fire," "bake." The Chinese say yik, "flame," "fire," "light." The Sanscrit agni, "fire," and Latin ignis, are the same word. An example of a word common to the three branches of the Turanian family will help to show the con- nexion in which they stand to each other. Mune is in Japanese "the breast," and it is found compounded with many words; for example, muna gawara, "roof tiles," where the ridge of the roof is called mune from its resemblance to the chest. The Mongol has emun, " before," " front," " south," and the Chinese mien, " face." The Tamul people say mun for " before," "front." In Japanese omote is "before," "front," " the face," " outside." In Cochin- China the face is mat. A door is the front of a house, and in Chinese " door " is men, and with this seems to be connected the German Mund and the English mouth. The interchange of n and t is easily accounted for, they being allied letters. The final consonant k is found in the root of a family of words closely related to this one. The Japanese makai, " to face," " stand with the face towards," muki, " frontage," " exposure," is like PREFIXES AND SUFFIXES IN JAPANESE. 151 the Tamul mukam, " mouth," " face," and the Sanscrit mukha, "mouth," "face," "commencement," "first." In the preceding examples occur several prefixed vowels. They are very common in Mongol and Japanese. Thus "horse," which is ma in Chinese, is mori inbMongol, and uma in Japanese. The Manchus say morin. The prefixed vowel agrees in nature with the vowel of the root, as in omote, " before," ishi, " stone," Chinese zhah. If a vowel be appended to the final consonant of the root, when already thus augmented, our primeval monosyllable is already ex- tended to a trissy liable, and this without the addition of new words to make compounds. Thus " hone," to be read " bone," the German Bein, and English bone. We have the same suffixed e in kake, " to hang up," " hook on," in Chinese kwa or kak, and in English hook. The next step in additions to the root we may suppose to have been the appending of consonants. Thus from ma, " grind," in modern Chinese mo, in Latin mola, in English mill, is derived the Japanese maru, "circle." From kak, "black," came k ( ara in Mongol, and kuroi in Japanese, the final k being lost in both cases. The r and I do not mean anything. They are not abbreviated words. They are merely phonetic additions. The Mongols are content to add an r or I to their roots, without supplementing it by a vowel, as gar, "hand," i.e., the "holder." The 152 china's place in philology. Japanese prefer to add a vowel. Hence arose several syllabic suffixes in ordinary use for forming deriva- tives, and they gradually, as they grew in length, assumed distinctive characteristics as nominal, quali- tative, or verbal terminations. Thus e in hate, " an end," " to end," from a root bat, " to end," in Chinese, our word butt, and the French bout, does not dis- tinguish between parts of speech. So eshi and ashi in hateshi, "the end," and hatashi, "to end," are ap- pended to the same root without any mark of dif- ference between verb and noun. But in Mongol the suffix si or Psi marks nouns distinctly. The following derivatives occur to. the roots maru, " circle," and kuroi, " black " : — marui, "circular." kurai, "dark." marume, "make round." kurami, "grow dark." marushi, " round." kure, " darken." maruku, "round." kuroku, "black." rnarusa, "roundness." kurasa, "degree of darkness." maroi, "round." kuroshi, "black." mart, " a ball." *kurosa, "blackness." Of the suffixes here used only me, mi, have a decided verbal sense, and they are probably connected with the verb suffix meri, meru, mere, which is translated " becoming." Of the substantive suffixes, sa is the only one that seems to be exclusively used of nouns. The word siro, "white," takes the derivative forms shiroi, shiroku, shiroshi, " white," shiromi, " whiteD," FORMATION OF COMPOUNDS. 153 shirosa, " whiteness." The root is sit in Chinese, meaning " snow." In Mongol the t is lost and the suffix gan appended, the sibilant initial taking as a prefix t aspirated. The Manchu form is shay an. Shiromi, " to become white," is also used as a noun in the ( sense " white of an egg," and " whiteness," as in shiromiga aru, "it has whiteness," where ga marks the objective case, and aru is the substantive verb used possessively. Generally speaking, the final mi marks a verb ; oi, ui y ku, si, mark an adjective ; and sa, ru, a noun. But these distinctions are not strictly adhered to. Language is in the Japanese only approaching to accuracy of conception. It was in fact first in the Sanscrit that the parts of speech arrived at their full form, with accuracy of outline and suitable variety of expression. The Mongol conjugates the adverb as he does the verb, because language, in its ever- advancing development, has not yet reached the epoch of accurate grammatical distinctions. So it is in the Japanese derivatives. The terminations are wanting in sharp- ness of definition. This was for the first time attained in the Indo-European system, and even there the separate independence of the parts of speech is far from being complete. The next step in the progress of development is the formation of compounds. Ki, "a tree," becomes kiburi, "shape of a tree," from/wn or buri, "shape," "manner. 154 china's place in philology. Species precedes genus. This law of position is in- variable. Kado bi, "door-fire," is the name of the fire in front of a dead person's house to light his way to the next world. In Mongol compounds are not used without the intervention of the possessive suflix. In Tamul, however, they abound, as also in the Himalaic and Chinese languages. In Mongol, in- flexions have more power, and hence the genitive or accusative mark cannot be omitted, except where the case is one of simple apposition. Take the following example : English cowherd, cowkeeper, Japanese asikdi. Here kai is " keeper," and as a verb means "to keep." Mongol uk'erc'hi, from ukher, " cow," with the suffix c l hi, which is equivalent to our er in shipper, monser, chandler. They also say uk'eri sahikc'hi, " cowkeeper," where the verb sahihu, "to keep," governs the accu- sative in i, and takes itself the suflix of agency, giving it the form of a present participle. The Tamul has kopalar, " cowherds," and the Sanscrit gopa, where pa means " ruler," and may be compared with the Semitic Baal, "lord." The Greek /3ovtt)<; and Latin bubulcus are formed like the Mongol from words meaning " cow," with a suflix of agency. In pecoris custos the Latin order is strictly Turanian. The Mongol would say uk'erun ejen, literally " cow's lord." Take the common Latin word for " cow," wakka (i.e., vacca), and the resemblance is still more striking. The etymological value of the Latin v is always w or u, THE CASE PARTICLES. 155 as in volo, " to will." And it may also be asked, What is the Greek suffix of agency tt;?, as in hnroTqi, but the Turanian ch'i, of which the etymological value is si ? The letters s and t are convertible in Greek and Latin. Apposition of substantives preceded in the Turaniati languages the formation of the suffix of agency. The formation of compounds by apposition, as in Japanese, is an older principle than that by which in Mongol a derivative of agency is formed by a syllabic suffix. It has also been destined to achieve a longer lifetime. The derivative suffixes of agency in European languages have not the prevalence now that they had 2,000 years ago, and especially in the Germanic stock they show signs of approach to ex- tinction. Cavalief will become in English an obsolete word before horseman. The Manchus and Turks agree with the Mongols in the use of c'hi as the suffix of agency, but the Turks have also the form dji. Its origin may be in sak, " to make," in Chinese tsolc and tso, as in mutso, " carpenter," from mu, " wood." In proceeding to the case particles, it may be observed that they originated in the great Tibetan and Turanian inversion, found also in the Sanscrit, by which the verb and the demonstrative pronouns were transferred from their primeval position, before the noun, to the end of the sentence. Prepositions are verbs. The case suffixes of the Turanian and Indo- European languages are modified prepositions, and 156 CHINAS PLACE IN PHILOLOGY. originally verbs or pronouns. The post-position of the transitive verb took place first, and subsequently the verb roots of the case suffixes became changed and shortened in form, and now appear to the investi- gator as suffixes, more or less closely combined with the substantives to which they belong. The Japanese genitive no — as in hi no ha, " leaf of a tree," where ha, i.e., ba, " leaf," suggests a connexion with the Siamese bai, " leaf," and with folium and blatt — is in Manchu ni, and in Mongol sometimes nu, nai. We have beside this possessive, four others in China and its neighbour countries. They are, ti in Chinese, gi or go in Tibetan and the old middle dialect of China, i, e, or u in Mongol and the South Fukien dialect, and in or un in Mongol and Turkish. The Eastern Himalaic languages have adopted the Semitic inversion, and place the nominative before the genitive, as in Cochin- Chinese luai dau, "edge of knife," where dau is " knife." Our five possessive suffixes are all, let it be observed, in form demonstrative pronouns. No is the Chinese na, "that." Ti is di, "this." Gi is gi, "he." E, i, or u, is i, "he." In and un are other forms of the third personal pronoun. Here we may see, therefore, a confirmation of Bopp's view, that the Sanscrit genitive suffix sya is an old demonstrative pronoun and is equivalent to tyam and tyat, " that." He adds 1 1 Vergleichende Grammatik, yoii Bopp. § 194. CASE PARTICLES COMPARED. 157 that, " in sya and tya are contained the two stems sa and ta, ' he/ with the relative stem ya> ' which.' " The Chinese ti, usually read ch'i, is in the ancient books used not only as a possessive particle, but in the sense of it or that after verbs, and also as a verb with the meaning of go to. The Greek and Latin genitives in i and u we may perhaps derive from the Mongol possessive u, affixed as a genitive ending to nouns closing with n. But I do not lay stress on this resemblance, for it is possible that u is here in fact nu, the Japanese genitive. The modern last Mongol dialect allows the possessive ne to be used with more nouns than the grammar of the book language would admit. The case of direction " towards " is in Japanese expressed by the suffix he or be and ye. The Tibetan has la, the Mongol de and dor, the Turkish ga and yeh, the Tamul ku. The Chinese has the verbs to, ti, or tau, the same as our " to," and gip or ki, " arrive at." In Greek 7r£$ov$e, "to the ground," agrees in form with the Mongol. Examples abound in Homer, as SofiovSe, "to the house." The Greeks afterwards preferred to prefix et?, "to," with an accusative. That is, as it appears to me, they were under Turanian influence while they used the suffix Be in the sense, "to a place," and emancipated themselves from it in this instance when they changed the suffix for the preposition. This took place soon after the time of 158 china's place in philology. Homer. In confirmation, it may be remarked that there is an aspirated form of the Mongol, namely, for, which corresponds to the Greek 6l, an old dative. Other Chinese verbs, which may be referred to in explanation of some of the forms now given, are wang, "go towards," hiang, "towards." W and y are inter- changeable initials, and the final ng is frequently dropped, as in the Chinese ta, " beat," anciently tang. The ancient equivalent of initial h is k. Hence these two verbs become ye and ka. The Japanese be, " to, "^suggests a connexion with the Greek irpos, "to," and irapa, "beside," "towards," etc. The word proximus, "nearest," is of the same family, and the Chinese bing, " unite," bang, " beside," bang, " to strike against," are probably related. Hoff- mann says be is the side or direction of a thing. The verbs heru and furu, mean " to pass from one place to another." This is undoubtedly the same word. The Chinese words for " unite," " union," " side," " neigh- bourhood," " collision," all tend to meet in an ultimate root bang, "strike against," derived probably from the noise of collision, and preserved in the familiar English expression, "bang the door." In the Japanese and Mongol languages, the final ng of Chinese roots is usually lost. Thus in kwang, " light " (at an older period keng), the ng is dropped, and the word re- appears in Japanese as karui and akari, and in Mongol as gerel. It was then by the Turanians that the ng CASE PARTICLES COMPARED. 159 was dropped and an r substituted. In this state the root was introduced into the Indo-European vocabu- lary, as in the German hell, "clear," and the Latin gloria and clarus, where / is inserted. Motion " from " or " by " a place or route, is ex- pressed- in Japanese by kara and yori. With the Mongols ec'he is the word. But this is etymologically ese. They also have yiar as in eguden yer, "by the gate." The corresponding case suffix in Turkish is den or dan, and in Manchu deri. The Chinese has the verbs yen, " take origin from," " let a man do as he thinks best," and dzung, " to follow," " obey." As prepositions these words appear with the sense " from." Yen (Japanese yori, Mongol yer) has lost a final k, for the character jfj yeu, is frequently used as a phonetic in words which to the present day retain final k, in Southern Chinese dialects, as dik, "flute." The old value then is ok, the ex of Greek and Latin. The other word, dzung, early lost ng, and appears com- monly in Chinese in the form dzi, the modern g ts'i. All the words having ts as their initial had anciently s or z. The old form of this verb is therefore zung or zu. The Mongol, having no z, adopted it in the form of se, and prefixed to it the vowel e. Another old Chinese verb, taking the sense of from as a preposition, is tang, " to strike," in modern pro- nunciation ta. This may be the source from which the Manchu and Turkish forms are derived. The Turkish 160 china's place in philology. den, dan, is found in the Greek 6ev, the old epic suffix for " from," as in ovpavoQev, " from heaven." The Japanese hara, "from," is compared by Hoff- mann to the High Grercnan her, " from that place to this." It may be derived from the Chinese verb k ( ai or k'i, " to open," " to begin," and hence in colloquial usage "to start from." The Japanese use ake, akeru, for " to open," and aki, aku, for " to be open," as in to wo ake, " open the door," where to, "door," is the same word, perhaps, as the Manchu duk'a, " city gate," and the Gfreek dvpos. The vowel a is a prefix, not radical. The locative particles used by the Japanese are ni, te, de, and nite. The Turkish has der, the Mongol de and dor, the Manchu de, ;the Tamul il and idattil. The Chinese have the prepositions yu, "at," "in," dzai or zai, " to be in " or " at." As local suffixes they use li or lai, net or nip, " within," cluing or tung, " in the midst of," or " within." Of these, nei is from nip, "to enter," in Modern Chinese ju; and chung is either "the middle," or "to strike the middle." The p and ng being dropped, most of the locative forms now given may be derived from these two locative auxiliaries. The Greek epic dative in 6l may, from its corre- spondence in form and sense with the Japanese and Mongol locative, be regarded as of Turanian origin. Thus, oinoQi, " at home," aXkoOi, " elsewhere," may be CASE PARTICLES COMPARED. 161 compared with the Mongol ger t l or or ger t'e, " in the house," where the usual de or dor becomes t'e or t'or, on account of r preceding ; and ober oron dor, or more colloquially, ore oron de, " at another place." To illustrate the Greek dative suffix 61, Bopp also cites evOa, eviavQa, "here," as compared with the ablatives evdev, " hence," ifiidev, " from the place where I am," forms which resemble the Turkish ablative, as above stated. He also refers 1 to the Sanscrit suffix dhas in adhas, " under," as connected with the Greek forms, and derives all of them from the demonstrative stem in t. I ask, will it not be more satisfactory to trace the forms in Be and Qi through the Mongol, as a modern type of the old Turanian language, to the Chinese §?ij tau, " to," used as a dative suffix, and the other word already mentioned, tung, " in the midst of," used as a locative ? The English to and the German zu have the dative force, as well as that of motion " towards " and " arriving at "; and the extension of the meaning of the word tau, to embrace a dative force, is no more than what we should expect when it became a post- position and was employed in case formation. This word tau, "to," has in Chinese and English the surd form. In Sanscrit and Mongol it appears with the sonant d, as also in the Latin ad, where a is perhaps a prefix for sound's sake. In Mongol and Greek it has the aspirated form, and in German it 1 Bopp. Zweite Ausgabe, § 223. 11 162 occurs as a sibilant, through the fondness of that language for the initial ts. The Turkish dah, "in," "at," "within the limit of space or possession of," 1 with n before it in kandah, "where," bundah, "here," much resembles the Japanese locatives te and nite. May not this be pointed to as the possible origin of the Sanscrit ablative in d ? The Turks say buradah, "here," where the r is a mere phonal extension of bu, "this." The final h, now silent, may possibly represent an old k or g, which would render the transition easy from the Chinese tang, "from." The Sanscrit kutas, "whence," 1 is not far from the Turkish kandan, " whence," especially when compared, as by Bopp, with the Greek iroOev, where n final replaces s. The Turkish kani, "where," contains the Japanese ni as its locative suffix. Compare it with the Sanscrit kadd, "where." The other Japanese locative is here used, d being the representative of the Mongol d in dor, da, and the Japanese t in te. Who doubts that the Japanese proceeded from the same part of the world from which the Hindoos proceeded ? If there be any one, the occurrence of resemblances such as this should cause him to pause. The Turkish for " when " is kachan, and the Mongol heje ; but the Mongol j represents d, and h is k ( , so that the Sanscrit and Mongol forms agree, except in the circumstance 1 Redhouse's English and Turkish Dictionary, p. 700. CASE PARTICLES COMPARED. 163 that the Mongol k is aspirated. Can we doubt that the period during which the Indo-Europeans lived beside the Turanians in Bactria, Persia, and Armenia, was fruitful in linguistic results ? Take another form. The Sanscrit kati, "how many," is in Mongol hedui, k'edui, or, in the modern colloquial, simply hedi. The Latin is quot. The Chinese original of these words, ki y " how many," is unaspirated, and has probably lost a final t. The instrumental case in Japanese is formed by the suffixes ni, nite, de, te, motte, the last of which is de- rived from the verb motsi, " to employ." The Mongol has her, yar, log a. In the Dravidian languages are found dl (Tamil), an (Sen-Tamil), im (Kannada), 1 in newer forms inda. The Chinese na, " to take," tan or twan, "to carry," pa, "to take in the hand" (in Mongol barihu, "to take in the hand"), and i, "take," jgl, are the roots of these forms. The Japanese ni as a loca- tive is derived from the Chinese nei, " within," and as an instrumental from na, originally nap. The Russians have the word nosisk, " to carry." The second word is tan, "to take up" or "carry." We also find tai, to "carry," or "lead," old form tak, the English take. "We also have tarn, "carry on the shoulders," and tang, " to undertake." Since se, " the back," seems to be derived by dropping n final from 1 Reise der Novara urn die Erde. Linguistischer Theil, von Dr. Friedrich Miiller. 1867. 164 senaka, "back," I suppose that the first of the four words is the root here sought for, and that the Japanese instrumental de and verb tori, "take," are the Chinese tan. The third Chinese instrumental verb is pa, Mongol barihu, " seize," Russian brat, "to seize." This originated the Mongol her, when the verb was placed after its noun by the Turanian inversion. Here too we find a probable origin for the Sanscrit instrumental suffixes bhydm, bhih, and the Latin bus. The Sanscrit suffix na I suppose to be the Chinese na, " carry." This seems to be a more natural way of accounting for it than to refer it to the pronominal root a, as Bopp does, a supposition which requires the insertion of n for euphony. The last Chinese verb to be considered is i, " take," " re- gard as," "use." It was much used in the style of the Chinese classics. Now it has given way to pa, tsiang, na, and tan. It affords a probable origin to the Mongol instrumental per or yar, and the Sanscrit suffixes ya, a, used in an instrumental sense. The Tungus suffix dji is probably the same with the Japanese de, by change of d to j. The Zend instru- mental is a, agreeing with the prevailing Sanscrit form. The Lithuanian instrumental suffix mi should be compared with the German mit and the Greek fiera. The Japanese instrumental motta and motsu are no other than this. Motsi, is " to hold," motsiyi ru, is CASE PARTICLES COMPARED. 165 "to use," " employ. " Bopp's derivation from the Sanscrit bis seems forced, but lie had so firm a con- viction that the Indo-European case suffixes are all to be derived from pronominal roots, that he neglected nearer and more probable analogies. It is, however, a remarkable fact that the Chinese instrumental verbs bear a close resemblance to the primeval pronouns. The old instrumental * is like the old pronoun i, " he," and the modern na, " take," is like the modern pro- noun na, "that," and in sound they are distinguished only by tones. The Japanese accusative suffix wo is like the Turkish yi and Mongol i, and is probably derived from the old pronoun i for the third person. The Chinese i usually comes from an older ui or wet, and the transition from wet to wo is not great. In Manchu the accusative ba reminds us of the Chinese pa, which is used to introduce the accusative, when in the col- loquial language the speaker desires to place it before the verb which ordinarily governs it, as in pa t'a sha liau, " he killed him," literally, " take him kill finished." Another accusative ending in Mongol is gi. This may be the Chinese third personal pronoun gi. Thus Bopp's view that the Indo-European accu- sative is of pronominal origin may receive confirmation from the formation of the Turanian accusative. The Tamil accusative in ei, the Telugu in ni, and the Tibetan in gi, appear to be all constructed in a similar manner. 166 The Japanese case suffix to has the sense "for the sake of," and "in conjunction with." In the former sense it agrees with the Mongol to'la, and with the Chinese Pi, " instead of." In the latter sense it agrees with the Mongol t'ai and Po; as in hamPo, "together," and the Chinese dung, "together." The Chinese have also dai and ivei, meaning " for the sake of," and "on account of." I suppose therefore the Japanese to to be a mixture of two words, which -are in Chinese t l i or dai, "for," "instead of," and dung, "with." The Japanese nominative ba is used like the Mongol inu, and as the nominative termination in Greek, Latin, and Sanscrit, was probably first used. "With- out doubt it is a metamorphosed third personal pro- noun. In the Turanian languages this suffix is not part of the word, but is a pronominal repetition of the nominative. Such too was the origin of the termination s in the Greek ol/co?, in Chinese ok, "house," in the demonstrative o? in Greek, and is in Latin, in Chinese gi and i. But the final has in these languages been taken into the word, and forms a part of it. The Turanian is the older, and the Indo- European the newer mode of doing the same thing. Let it not be said that the Turanian languages as now known are altogether too modern for the philo- logist to regard them as constituting a stepping-stone between the Indo-European system, and the venerable mother from whom all languages, eastern and western, CASE PARTICLES COMPARED. 167 have sprung. The Japanese writing, being 2,000 years old, secures to that language a claim to a very respect- able antiquity, and when the Dravidian languages are taken into account, that antiquity is greatly increased. As the three races, Tartar, Dravidian, and Mongol, have not been neighbours since 4,000 years ago, the approximate time of the Arian invasion of India, all common features existing in the three branches must have belonged to the old Turanian stock from which all of them proceeded. For example, the post- position of the verb, and the formation of cases had then already taken place. Among the case particles which resemble each other most closely are those which mark the accusative and conjunctive relations, viz., ei in Tamil and Mongol, and wo in Japanese for the accusative, and odu in Tamil, to in Japanese, and t'ei or loga in Mongol for the conjunctive, or, as De Castren calls it, the comitative case. For example, in Tamil pilleiyodu wandan, " he came with the child." 1 It also appears from this instance, that at that early time the final ng of the Chinese word dung, "together," was already thrown off in cognate languages, a phe- nomenon which occurs in the history of many Chinese words, such as wu, " not," formerly mo, and still earlier mong y as known from the fact that the character £ mong, was frequently used for it in the ancient Chinese books. 1 Pope's Tamil Handbook. CHAPTER IX. Second Division of the Turanian System. — The Dravidian Languages. — Proof that this Family is truly Turanian. — Common "Words. — Common Laws of Sound. — Surds and Sonants. — Deficiency in Sibilants. — Abundance of Liquids. — Syllables usually Open. — Derivation. — Comparative List of "Words. — The Verb. — The Passive Negation. — Tense Formation. — Dravidian Syntax. Professor Friedrich Muller Las expressed doubts respecting the Turanian character of the Dravidian languages. The proofs of this rest on a multitude of common roots, resemblances in alphabet and syllabary, identity of syntactical construction, and the similarity observable in their system of suffixes. First, the roots are the same. Thus, we find resem- blances like the following : COMMON TURANIAN ROOTS. TAMIL. MONGOL. JAPANESE. CHINESE. INDO-EUROPEAN. kal, foot k'ul kak to kick. talei, head tologai atama du, dud silei, stone c'hilagon isi zhag saxum. karam, hand gar Sans. kara. nay, dog nohai inu marei, rain boron ame mo, mist Pers. baran. kiragam, house ger ke casa. — Mnr;} degu ototo de 'ade\(p6s. ira, night yoru ya adigam, much dake TURANIAN ORIGIN OF THE DRAVIDIAN FAMILY. 169 TAMIL. MONGOL. JAPANESE. CHINESE. INDO-EUROPEAN. wegu, much yek'e okini teyilam, oil fossa ter, chariot t'ereg t'e dray, drag. pattini, hunger hidaru mar am, wood modo mok agam, s%n Ma ek'e aku ak wicked. kar, blackness k'ara kuroi kek caligo. andam, egg undug u6v. There is a law in all true Turanian languages, according to which the vowel of the root repeats itself in the prefixes and suffixes. If it does not repeat itself exactly, it takes the form of an allied vowel. In Mongol a and o are allied; e and u are also allied. The vowel i is doubtful. Hence the syllabic alphabets of Turanian languages. The vowels are regarded as inherent in the consonants. The consonants are essen- tial, and the vowels are secondary. The vowels only attained their full and individual importance in the Indo-European languages. The sounds of the Tamil and other languages of the Dravidian family are such as to confirm the fact of their Turanian origin. There is noticeable a de- ficiency in the development of the letters sh, ch, and the surd series generally. Thus k, t, p, become g, d, b, when they occur in the middle of a word. Analogy, Japanese, and Mongol, shows that the original sounds were g, d, b, which become surd at the beginning of words or when doubled. In the Japanese language, 170 china's place in philology. words in h or /, for instance, take b for h or /, when they follow another word. Hito, " man," becomes bito in certain cases. Thus hito bito means "men." Hoff- mann has shown how the Japanese h of this century- was, last century and previously, /, and that it really belonged anciently to the labial series. But it is necessary to go further than this, and to reduce the h in all cases to b } as its ancient form, as hatashi, " to complete," Chinese ba, bad, Mongol barahu. So in the plural kuni guni, "kingdoms," corresponding to the Manchu gurun, " kingdom," it is better to avoid being misled by the Japanese orthography, which make g a modification of k, and to regard k as being rather a modern modification of g. Here I use the word modern with a wide acceptation. The history of the Japanese alphabet shows that k and g have been divided in Japan for 1,500 years. The fact is that the Turanian ear formerly recognized no such dis- tinction. Nor does the Mongol of the present day. When naming his little tent images to his foreign visitor, he will call this one Kalin ejin or Galin ejin, " lord of fire," that one Shiggamuni Borhan (that is, Shaky amuni Buddha), and another Gesser San or Kesser Han, the hero who in Tartary takes the place of the seven champions of Christendom. It is nothing to him whether he attenuates his initials into k, t, and p, or thickens them into g, d, b. His language has not yet arrived at this stage. The Japanese have gone TAMIL SIBILANTS. 171 forward most successfully in the division of the surds from the sonants. The Tamil- speaking people are in a midway position. The Mongol has still to arrive at the consciousness of the distinction. But he makes use of aspirated surds as a substitute for the pure surds. He has a fully developed ¥ and P in his alphabet. It may be concluded by analogy that the Tamil k, t, p, have come out of g, d> b, and that the true ancient sound is heard when it occurs in the middle and end of words, e.g., ug, "to desire," Chinese yug or yuk, Sanscrit vag, Greek ev^ofjuac, English wish. Here the final g becomes k in modern Chinese, and sh in San- scrit and English, while in Greek it prefers the aspi- rated form. The importance of the Tamil is shown by this example, for with the intermediate form ug as a guide, there can be no just ground of hesitation in identifying the Chinese root having the initial y with the Indo-European root having the initial v, or, which is the same thing, w. The Tamil sibilants are very defective. When s is doubled or follows d or r, it is pronounced ch as in Charles. The analogy of the Mongol and Japanese languages shows that s is the true old sound. The Mongol has s and Ps, the latter of which is called c'h, but its value in comparative philology is simple s, as in c'hi, "thou," Manchu si, Greek av. The Mongol has no true /, the j in use being a modern corruption of d, as jirohe, " heart," the same as the Persian dil. 172 So jigahu, to " teach," to " point to," is the same as the Latin doceo, digitus, the English " teach," and "betoken," the German zeichnen, etc., and the Chinese chi or U, " to point," chi or dik, " straight." So the Tamil tagu, "to be just," and tagudi, "justice," have the same etymology. Nor has the Mongol an sh proper, the initial having this orthography being modified from si. The same occurs in Japanese, where si is pronounced like shi, and tsi like chi. Thus in Mongol c'hagan, " white," seems to be connected with c'hasa, " snow," but c'hasa is evidently connected with the Chinese sit, " snow," and snow is a substance which in all countries where the winter is cold, origi- nates adjectives indicating whiteness. The Manchu shanggien, " white," is apparently formed from the Mongol by dropping the initial t. After the letter sh had been thus introduced many Chinese words were perhaps borrowed, such as shu'min and shum, "deep," Chinese shim. I suppose, therefore, that the Tamil s, though sometimes pronounced ch, is really the s of Tartary and Japan. The three r's, two /'s, and three n's of Tamil reveal the existence of a principle that has been at work among all the Indian populations since the intrusion of the dominant Arian element. The Sanscrit lan- guage has among the vowels a long and short r and I, and an r and I at the bottom of the cerebral and dental t series, respectively. The sister language, ABUNDANCE OF LIQUIDS. 173 Zend, has one r and no /; and hence it may be concluded that this rich development of r and / took place in Sanscrit after the migration to India. It was, therefore, probably the effect of climate, for it characterizes the Dravidian languages as it does those of Sanscrit origin. Hot and moist climates induce luxury and softness of manners. The vowels and liquids then become extensively subdivided, while letters which in their enunciation require decision and physical energy suffer in proportion. The remarkable completeness of the Sanscrit alphabet, wanting only / among the consonants, and eu and u among the vowels, was due to the Arian race having first been located in a temperate region and afterwards migrating to a hot and moist one. That the Tamil and other Dravidian languages have, when compared with the Sanscrit, so poor an alphabet, is partly due to the fact, that the Turanian stock from which they sprang was itself poor. To this should be added, that deterioration had followed on their separation from it. A softening process deprived Dravidian speech of much of the pith and force which belonged to it at an earlier stage, when it was one with the Mongol and Japanese. Proof of this will now be given by adducing the deficiencies of the Tamil syllable. The syllable admits in modern Chinese of a prefixed t before the initials s and sh. This liberty is also 174 used in Mongol, and t becomes ts before i and u in Japanese. In Tamil, s becomes occasionally ch. In old Chinese the six final consonants by which a syllable could be closed were g f d, b; ng, n, and m. In Mongol the same rule prevails. The Japanese lan- guage restricts this law, and takes pleasure in changing the old monosyllable into a dissyllable. The Mongol went farther and added s, I, r, to the number of finals by which their syllables might be closed. The Tamil people are more like the Japanese than the Mongols in this respect, and give their syllables no consonantal letters with which to close them, except n, m, I, and r. The Telugu and Kannada languages know no finals to their syllables but the vowels, and they thus assume in regard to this feature a completely Polynesian aspect: The following examples of derivation in the Tamil language will at the same time show that the roots are found alike in the Chinese and in the European vocabulary. They have been chosen within the space of a very few pages in Dr. Winslow's Tamil Dictionary, and in a part where the identity of words is very easily detected, because the features of family resemblance have not been much defaced by the processes of secular corruption. In Chinese words three sounds are some- times given, the first modern, the second that of the dictionaries a.d. 500, the third that of the era (ac- cording to tradition) of the formation of the phonetic DERIVATION IN TAMIL. 175 characters, B.C. 2000. The Japanese h is replaced by its ancient equivalent b. Padi, " step of a ladder," padam, " foot," " road,' ? " metrical foot." Chinese pu, " step," bo, bod. Indo- European pada, foot, pes, passus, pace. The Tamil here uses as suffixes of derivation i, am. Para, " spread," " be diffused," paravu, " lay open," " spread," parambu, " to spread," " become diffused," " multiply." Japanese fure, bure, " publish," " pro- mulgate," haru, baru, "spread over," "extend," "dis- play." Chinese pei, bi, bid, " coverlid," " to spread over," "extend to." Indo-European bed, spread, pando, pateo, broad, breit. Tamil derivative syllables a, avu, ambu. Padar, " widen," " ramify," " extend," " pass," pa- dam, " path," i.e., " that by which we pass " or " pro- ceed," padavi, "road." Mongol badarahu, "to extend," badaral, "extension." Chinese fa, bat, "to expand," "go forth." Indo-European forth, path. Russian raz- brasivat, razbrosat, to " dissipate," " extend." Tamil derivative syllables ar, am, am. Paru, " to be increased," parambu, " to multiply," pattu, "fold." Betel nut in folds for guests. Hurdles in folds for folding cattle. Cloth either as spread or as folded. A " plait " or "doubling" of cloth. Japanese hida, bida, "fold," " plait," fata, buta, "two." Chinese pei, bi, bit, " double," " add as much again." Indo-European both, German beide, fold, plait, to boot, 176 china's place in philology. i.e., " add," freebooter, i.e., "one who wanders freely." (Here the sense of spreading is approached.) Pari, "to part," "separate," piri, "to part from," " separate," pddi, " part," " proportion," pddida, " dis- tribute," pddu, " sharing," pdtti, " division." Mongol buda, " group." Japanese hedate, bedate, " to separate from." Chinese,^, bit, "to separate," " other," fen, pun, " divide," pic, bu, bud, " division," " class," fen, bun, "a division." Indo-European pars, separo, Sans- crit bheda, "dividing," bhedita, "divided," bhinna, "separated." Padu, "to suffer," "to be acted on," "to perish," " die in battle." This word forms a passive when joined to the infinitive of active verbs. Japanese hate, bate, " end, to end," batashi, " finish." Mongol barahu, "finish." Chinese j%r, ba, bat, "to end," pel, bi, bit, "to be acted on " ; used as a sign of the passive, as in pei sha, " was killed," pai, " destroy," " be destroyed," ba, bad-, fa, "strike," "put down," "make to fall," bat. Indo-European bout, butt, patior, passus, beat, batuo. Pari, "burden," "load," "speed," pari, "to be heavy," "to feel heavy," "to be thick." Mongol bidugun, "thick," bidu gulig, "thickness." Chinese/^, bu, bud, "burden," "to bear.'' Indo-European bear, fero, (j)ep(o, porto, bahren, ffapvs, /3apo?, berden, speed, a7rev8a). Padam, "boiled rice," "eating," pddeyam, "pro- EXAMPLES OF DERIVATIVES IN TAMIL. 177 visions for a journey." Mongol bada, "food." Manchu but'a, "cooked rice." Chinese fan, ban, "cooked rice," "food." Indo-European food, feed, fodder, petayu (Russian " nourish"), bwyd (Welsh " food "). Padiv { y,, " stooping," " lying near the ground," pa- dukkam, " servility." Manchu budun, " vile." Mongol beg en or bog en, "low" (g is apparently part of the suffix, and d is probably dropped). Japanese hikui, bikui, " low." Chinese pel, pi, pid, " low." Indo- European bottom, base, fidOo?. Pal, "many," palam, "force," "strength," "fruit," " result," " profit," palan, " profit," " fruit." Japanese batashta, " to result," hodoshi, bodosi, " to give," " bestow." Mongol butogehu, " fulfil." Chinese pei, pi, pid, "add to," "give," "annex to," "benefit," "assist." Indo-European fructus, fruit, fortis, abundo. Paneiya, " old," " decayed," paneimei, " oldness," " decay." Japanese furui, burui, " old." Chinese pai, ba, bad, " to decay," " destroy," " decayed," fa, bad, " wearied," " worn out." Indo-European fatigatus, fetid, vraXaios, beaten (in the sense of " weary "). Para, " fly," " move quickly," " be dispersed." Japanese hashiri, basiri, " flee," " move fast." Chinese fei, pi, pid, " fly," po, pad, " scatter," " winnow." Indo-European aireipca, ireTavvviMt, bird, fly, Flugel, 7T€T6ivd, flee, fast. Padi, " resemblance," polu, " to be like," poll, " like- ness." Mongol buduPu, "likeness." Chinese pi, pe, 12 178 china's place in philology. ped, " compare." Russian podobie, " resemblance/' upodobknie, " comparison." These, and many other words like them in form, appear to have sprung from a very few roots, such as bad, bid, bud, which may easily have originated in the imitation of natural sounds. These sounds would be, for example : the noise of the foot in stepping, of a bird beginning to fly, of striking with a hatchet, or of a heavy object falling to the ground. The many sharp sounds heard in nature favour the opinion that closed syllables were common in the primeval syllabary. It is not likely that our first forefathers ended their words with vowels exclusively. The preceding examples show that a close com- parison of the vocabularies of the Turanian languages with the Chinese old vocabulary is likely to be most fruitful in results, Philology, indeed, has at hand no vocabulary of roots so complete and so ancient in form as that found in the Chinese dictionary. Among the additions to the root in Tamil are m and /, marking substantives, as kddam, "killing," kddal, " act of killing. ,, The m reminds us of the Semitic m, which is a demonstrative root, and is used to form participial substantives from verbs and also to mark the participle. The suffix / is the same with the Mongol suffix for verbal nouns. For instance, the Mongols say c'hidal, " strength," " ability," derived from c'hidahUf "to be able," by adding I. Chinese COMPOUNDS. 179 t'sai, "power," "riches," ze, zat. The root is also found in trie Sanscrit sattva, "vigour," "power," and sattra, "wealth," "sacrifice" (Chinese tsi, tse, sat, " sacrifice "). The vowels i and u, added to the root, are sounds, and nothing more. Thus, parru is " a grasp " and " to grasp." The same is true in Japanese. They are therefore used merely to make a second syllable, by giving a vowel to the final consonant of the root They show that language has an inherent tendency to become polysyllabic. The Tamil, like the Japanese, but more freely, makes compounds by annexing words to each other. Thus, from tarisi, " to see," are formed tarisanam, "vision," and sandarisanam, "the capacity to see all things in common," where the first syllable is the root sam, " all," " even," " common," Latin simul, English same, Greek dfia, Chinese t'sam, " equal," Tamul samam, " evenness," " sameness." From palam, " strength," is formed samabalam, " equal power." It is only in the Indo-European languages that we find a parallel development of compounds. But the pre- positions irapd, pro, super, etc., which make so promi- nent a figure as prefixes to words in Greek and Latin dictionaries, are not able to take the same position in any Turanian language, because their nature as verbs requires them to be placed last. This exception being made, there can be no doubt that the transition from 180 china's place in philology. the Turanian languages to the Indo-European system is, in regard to the formation of compounds, most easily made from the Dravidian branch. The Turkish and Mongol in regard to this feature afford no foot- hold for comparison, for in those languages the loose compounds which exist cannot be regarded as single words. The intervention of possessive and other par- ticles prevents the fusion of the two words into one. A glance at the verb will enable us to judge of the relation held by the three branches of the Turanian family to each other. The essential identity of the verb and noun is plainly taught by the Chinese and Turanian systems of languages. This identity is not in idea, but in sound. The framers of language did not confound action and thing, but they gave them the same name. Thus, dong in many languages expresses the sound of a bell. So anything that gives a ringing even sound, as also the sound itself, and any action that causes it, would be called dong. The Chinese say ta, old form tang, "to strike," chung, old form tang, "bell," dung, "copper." Of the Dravidian roots, Caldwell says, as quoted by F. Miiller, l " When case-signs are attached to a root, or when without the addition of case-signs, it is used as the nominative to a verb, it is regarded as a noun: the same root becomes a verb without any internal change or formative addition, when the 1 Eeise der Novara. Linguistischer Theil, p. 95. THE VERB. 181 signs of tense and the pronouns, or their terminal fragments, are suffixed to it." Thus, in Tamil occurs kuttu, a word which means either i ( union " or " to join." Changing the vowel, we meet with kattu, "a tie " and " to tie," " a fabrication " and " to fabricate," " a bundle " and " to bind." The Chinese kit means, in the same twofold manner, " a v tie " or " to tie," " to coagulate," " to solidify." But if we observe the same root in Mongol and in the Indo-European languages, a difference is perceptible. The Mongols say hada for "a rock," hat'ago, for the adjective " hard," hat'aho, for the neuter verb " to harden," and hadaho, " to make fast " (by hammering), hat'agaho, "to dry," "to harden," in a causative or transitive sense. The Japanese, like the Mongol, has advanced beyond the stage when the verb and noun were one. Thus, we find in the Japanese vocabulary katai, " hard," katameru, " con- geal," "harden," katasa, "hardness," kataku, "hard" (the same with the Mongol), katamaru, " become hardened," kata, "a mould," "shape," katatsi, "figure." Among the suffixes here observable, i and ku serve for the adjective, sa and a for the noun, maru for the verb in a passive or increscent sense, meru for the verb in a neuter or transitive sense. The naked root does not appear, nor do the verb and noun meet in any one form. Hence it may be concluded that the Tamil and 182 china's place in philology. Chinese types of language are both in this respect older than the Japanese and Mongol, while the English combination of the noun and verb in one form as " a tie," and " to tie," is a return to primeval usage after the language had been temporarily subject to the laws of derivation which reigned in the Sanscrit, Greek, and Latin. The Sanscrit vocabulary contains the forms kathora, "hard," kathoratd, "hardness," kathina, "hard," kathin- aid, "hardness," kdtha, "rock," kathina, "hardness," kddambara, " the skim of coagulated milk," kitaka, " harsh," Ml, to " bind," " fasten," or " nail " (/ for d), kila, " a nail," " pin " (Mongol hadagaso, " a nail* "), kuta, " a hammer." To these correspond the English hard, hardness, harden, where an r has crept in before the final letter. It was this system of terminations, beginning in the Turanian languages and culminating in the Sanscrit, Greek, and Latin, that obscured the original identity of the verb, noun, and adjective, separated with precision the parts of speech, and thus gave origin to Indo-European grammar in its broadest aspect. This vast superstructure of derivatives raised on the original basis of the monosyllabic roots is now, and has been for two thousand years, gradually crumb- ling away. The contrast between the Anglo-Saxon and the English in regard to the extent of the prin- ciple of derivative suffixes is a measure of the change THE VERB. 183 steadily advancing in the whole Indo-European world. If the human race should last long enough, we might expect, supposing that the present law of decay continues, to find the whole superimposed system of derivative forms swept away from language, leaving behind only the primeval vocabulary of monosyllabic roots. But the working of other laws, and the in- satiable craving of civilization for new words, will prevent this. The verb in Tamil appears as transitive, neuter, causative, passive, and negative. A neuter becomes transitive by doubling the final consonant of the root. Thus pogu, "to go away," becomes pokku, " to drive away." Professor F. Miiller has noticed that this is a Semitic peculiarity. I would add that it should be traced to the influence of the Semitic form of speech on the Turanian at an ancient period, when they were geographically contiguous, or mixed in one older system. In Hebrew the doubling of a consonant is intensitive or causative. The syllables pi, hi, and m are appended in Tamil to make causal verbs out of transitives. Thus my, "do," the Chinese tso, in the old form sak, "do," or "make," when pronounced seyvi, means "cause to do." From hal, "to learn," and kan, " to see," are formed karpi, " to teach," and kdnbi, "to show," i.e., "cause to see." The Mongol inserts the syllables ga and gol, to impart to verbs a trans- itive or causative sense. This may be the Chinese 184 china's place in philology. causative verb kiau, old form ko. The Japanese causa- tive suffix is se. Thus from miru, "see," and naru, "be/' are formed miser u, " cause to see," nasi, "make to be." This syllabic addition reminds us of the Chinese causative verb shi. In passing from Chinese to the Turanian languages, as before remarked, sh becomes s. The causative syllable in the Telugu and Kannada divisions of the Dravidian branch is 8U, which may be referred to the same origin as the Japanese. The Turks insert al or ar, to give the sense of "coming into the state of," and add d or t to add the causative signification, as kararmak, "to become black," karart- mak, "'to cause to become black." Here mak is the sign of the infinitive. The Sanscrit causal p, inserted before the causal suffix aya, as in the example given by Bopp, sthapayami, " I make to stand," may be referred to the Tamil causative p for a probable origin. 1 The great philologist just mentioned derives the c of doceo, " to teach," from the Sanscrit causal p, and finds the root in disco, ehdrjv and BiSdaKco. In Chinese we have the roots ti, " to know," in Mandarin ch'i, and dik, "to lead," tok, "to superintend," to, "lord." These words will furnish a natural and probable etymology for doceo, dux, ScSda/cco, the Persian ddnem, " I know," and other connected words. Bopp derives the p in rapio, "to snatch," from the same Sanscrit causal p, 1 The Manchu, also a Turanian language, has bu for a causative syllable, cholimbi, "carve," choHibumbi, "cause to carve." THE PASSIVE. 185 "in case rapio corresponds to the Sanscrit rdpdydmi, ' I make to give/ of which the root is rd, ' give/ formed from dd, ' give/ by a weakening of the d." When we have the word rob, " to plunder," in English and German, it seems to be useless to seek for any- other origin for the Latin rapio than the root rab or lab, "to take with the hand," "to take violently," "to receive/' etc. ; in Chinese nip, " take with the fingers," nap, " to take/' " bring," in Mandarin nie and na. x In the eastern Asiatic languages the passive is a derivative verb. Thus, in Tamil the suffix padu gives to verbs a passive sense. This is the same word with the Latin patior, " to suffer," and the Chinese bad, in Mandarin pel, which from the original meaning " to cover," "to spread," "reach to," etc., has taken the sense of " being acted upon," and so come to be used as a passive auxiliary. As such it is quite common in the modern speech of the country. In some dialects, as at Shanghai, its tone (4 or 8) indicates that a final d has been lost. 2 The Japanese changes final i to e to form a passive. Thus, umi, " to produce," becomes ume, "to be produced," "to be born"; yomi, "to read " (Chinese niem), becomes yome, "to be read." The 1 The original root for the causative p, Manchu bu, I think is the Chinese bad or pet, "to give," for in the Shanghai dialect peh for pet is used in a causative sense. Edkins's Shanghai Grammar, p. 140, Peh la ngu ¥iuh kwen sz, " he caused me to suffer a lawsuit." 3 The lost letter might be d or b, but the phonetic shows it to have been d. 186 china's place in philology. Tungusian adds wum to the active form. Here m is a personal ending, and tvu is the addition, which may- be the verb ui, in Mongol meaning " to do," and in Chinese " to do " and " to be." The Mongol passive is formed by inserting da or de. The Turks insert il. Thus, in Tungusian silkim, " wash," silkiwum, " to be washed"; in Mongol abho, "take," abdaho, "be taken"; in Turkish deugurum, "strike," deugilurum, "be struck"; in Tamil adikkappadu, " to be struck"; in Mongol t'ugsedehu, " to be struck." l The Turanian verbs are negatived by the insertion or addition of the negative roots al, ne, ak, ma, com- mon in Semitic and European languages. The Tamil inserts a, as do the other Dravidian dialects. Its full form, says Caldwell, as quoted by F. Miiller, is al. The Telugu simply changes u to a, as in pampu, " to send," pampa, "not to send." The Tamil allows the usual terminations to follow. Seygindadu, " it makes," seyyddu, "it does not make." This negative may be identified with the Hebrew Stf al, "not." The Japanese negative element, says Hoffmann, is the sound n. From siri, "to know," is formed siranai, "I do not know." From ake, " to open," is formed akenu, " not to open." From yuki, " to go," are formed yuku na or na yuki so, "do not go." This negative may 1 The root dig, dug, "strike," occurring in three Turanian languages, is in Chinese tang; in English, with prefix s and insertion of r, strike; in Hebrew yj?ri "he struck," pjTJ "beat small' NEGATION. 187 be identified with the Latin non, ne. The Manchu negative is ak'o, " not," as in ak'o oho, " I have it not," ak'uc'hi, " if it is not," ak'ungge ak'o, " nothing is wanting," ak'ungge, " it is not." The Mongol ugei is the same word with the aspirate omitted. In the eastern Mongol w is inserted in this word, as in bada idesen ugwei, "I have not eaten food." Here bada is the English food, 1 ide the English eat, 2 and ugwei the Greek ovk and ov. It is marvellous that the roving inhabitants of the Tartar plains should be daily using words so familiar to the inhabitants of European coun- tries, but it is not the less true. These three negative roots, a, ne, ovk, have absolutely no representative in Chinese. Hence they must be regarded as of Semitic and Turanian origin, and the introduction of the last two into the Indo-European languages must be attributed to the influence of ancient union, or mutual influence and juxtaposition of the races. It is otherwise with the remainder of the negatives in the Turanian languages. They are chiefly words identical with Chinese negatives. Thus, the Chinese wu or mut, " do not," is in Mongol bu, as in bu bic'hi, " do not write," bu oro, " do not enter." The Turkish uses ma, as in olmah, " do not be," olmaz, " he is not." 1 Compare the Russian pitat, "to feed," German fiittern ; Russian pishc'ha, "food," German futter, English fodder. 2 Compare the Sanscrit anna, "food," ad, "eat," Latin edo, Greek iardlw, German essen, Russian yest, Greek eSw. 188 china's place in philology. The Tibetan has ma, mi, "not." The Chinese denial of existence is expressed by mo, mong, in Mandarin icu and mei yeu. The Greek (jltj, in /jlt) jevocro, " let it not be," is evidently connected with the same root. The two Chinese words are probably one in origin. In prohibitions they preferred the final t or d. In the denial of existence they chose the final ng. 1 Another Chinese root is pi, put, in Mandarin fei, pu, used to express contradiction, as in pu hau, " it is not good," pu lai, " he does not come." The Mongols say, boso, "it is not so," in the eastern dialect, bishi. In Europe this root appears in the Latin pereo and perch, perfidus and perjuria, in all which the prefix per has the sense of destruction or badness. It is also our own word bad. As a transitive verb it is in Chinese " to destroy " (fei) ; as a neuter, it is the substantive verb negatived; as an adjective, it denotes moral badness (fei lui, "bad people") ; as an adverb, it is pu, "not." The formation of the tenses preceded that of the personal endings. Thus, in the Tamil the form sey- gind-en, " I do," has for the mark of present time (gir) gind, and for the first person singular en. The first person singular in the aorist tense is sey-v-en, " I did." The perfect is sey-d-en, " I have done." In the older Turanian types, represented by dialects bordering on China, the personal endings do not occur. Thus, in 1 Mong, in Mandarin wang, is used for "to die" or "be destroyed." It is the Latin morior, Sanscrit mara, Persian mardan. TENSE FORMATION. 189 the eastern Mongol, "I kill," "thou killest," "he kills," are expressed by hi alana, &hi alana, Pere alana, while in Buriat-Mongol, spoken on the shores of the Baikal Sea, alanap, alanas, alana are used, where p is hi, "I," and s is c'hi, " thou." The distinction of masculine, feminine, and neuter is found in the personal endings of all the Dravidian dialects. Thus, in Tamil the third person singular of the perfect indicative is, masculine sey-d-dn, feminine sey-d-dl, neuter sey-d-adu. This peculiarity, being unknown in the languages of Tartary and Siberia, is best referred for its origin to ancient juxtaposition with Semite or Indo-European races. If traces of Semite influence occur in Dravidian speech, it may have been from the neighbourhood of some early people of that descent in Persia. The Cushite settle- ments stretched eastward along the sea- coast from Arabia to the mouths of the Indus, 1 and the language of the Cushites differed but little from that of the Semites. In Coptic and other Hamitic languages the distinction of masculine and feminine is still found in the personal endings of verbs. 2 When it is also remembered what striking indications of Semite in- 1 Compare the geographical names, Cutch, in the Gulf of Kutch, near Bombay; Gujerat; Katsh, the name of a Tibetan province; the Vale of " Cashmere," etc. 2 Reise der Novara. Linguistischer Theil. In old Egyptian, t marked the feminine. A neuter gender was unknown to Semitic or Hamitic grammar. The Arians probably introduced this distinction into language. 190 fluence are observable in the Tibetan language, it seems fair to conclude tbat the races which occupied the Persian area immediately before the Arian, being partly Semite and partly Cushite, imparted Semite elements to the Dravidian languages. The Dravidian tense marks are, present gindu, giru, in Tamil ; ta, te, in Kannada ; chu, tu, in Telugu. Perfect t, d, and i. Future L Aorist b, v, pp. On the origin of the marks for the present no light is thrown by reference to the Turkish ar or ur, as in korkaricm, korkarsen, korkar, "I fear," "thou fearest," "he fears," where um and sen denote the first and second persons and kork is the root ; nor to the Mongol moi, na, namas, in bi abomoi, "I take," t l a abomoi, "you take,-" t'ere abomoi, " he takes," in the eastern collo- quial Mongol abana, and in old books abonam. The forms for past time are, on the other hand, remarkably similar to those found in the connected languages. The Turkish preterite inserts d, as in •korkdum, "I have feared," where m is the first pro- noun — our own me. The Mongol gerunds have among them a form in d, which may easily have originated the indicative preterite in d. For example, in the eastern Mongol, noyin moran onad jidan beriji iheu hashigaran oroba, " the chief, mounted on his horse, and holding his spear, entered his great court-yard. , ' ) Noyin is "any chief." Moran is the second accusative of mori, "horse," and as such bears a possessive signification. TURANIAN GERUNDS. 191 Onihu is "to ride." Its gerunds are onad, onaman, onaj'i, which may be used in succession in a sentence composed of several clauses like the preceding. Jidan is the second accusative of jid, "spear/' Berihu is "to grasp with the hand," — the Chinese pa. Iheu, " great, " b is the Japanese okii, " great." Hashigar is " a palisaded enclosure." Oroba is the past indicative of orohu, " to enter." The form in d resembles our English participial form in ed, which agrees with the past tense indicative, as in, "he was mounted on his own horse," or, " he mounted his own horse." The Latin equitatus, or equo vectm, " mounted on a horse," are also equivalent. So is the Sanscrit participle in ta, as patita, " fallen," from pat, " fall." The third Mongol gerund in ji appears to be the same wide- spread form disguised by phonetic change, for the Mongol/ has d for its etymological equivalent. The Japanese gerund in te nearly agrees in form and use with the Mongol. I take an example from Hoffmann: Te wo aghetefito wo manekiyubu, "elevating his hand, he calls the people by signs." Te is " hand." Wo is the accusative case suffix. Aghete is the gerund of aghe, "lift up," the Chinese gu or kit, and Greek iyelpco. The Turanian prefixed vowel has, if this identification be correct, been retained by the Greeks from that time in hoar antiquity when the forefathers of the Japanese were next-door neighbours to the world-famous Hellenes. Fito wo is " man " or " men," 192 china's place in philology. in the accusative, and manehiyubu is a compound verb, consisting of maneki, " to beckon/ ' and yubu, " he calls." Hoffmann remarks that the suffix te means "at the time of," or " by means of," and is locative, modal, or instrumental. 1 The Japanese past tense takes the suffix ta, as in watakusiga mita, " I have seen," where watakusiga means " I," and has the honour perhaps of being the longest word in the world in use for that pronoun. Mita is the past of mi, "see," connected with the Chinese mok, " eye." I suppose that, as in English, the past tense is of later formation than the gerund form and founded upon it. So in Greek, the participle Xelircov preceded the imperfect or aorist eXnrov, and both are later than the Mongol infinitive or participle in n, with which they are connected by a distant relationship. So also eftovkevaa, " I coun- selled," and €(3ov\ev97]v, " I was advised," may be viewed as more recent forms, founded on the older jSovXevaas and fiovXevOel?. The Indo-European past tense in s, d, and n, is based on the participle, and this again upon the Turanian gerund. The Turanian intellect nominalizes the verb. Every verb is looked at as a substantive, and hence those parts of the conjugation which were first formed approach in their nature to the substantive. The Turanian in describing a succession of events gave to his verbs the forms of gerunds, and added to them, when needed, the case 1 Grammaire Japonaise, p. 177. PAST TENSE. 193 suffixes. Thus, each clause was a substantive. Yet, by the nature of the case, they retained a verbal energy. Time was an inherent element which was inseparable. The union of verb and noun in one word thus originated the participle of both the Turanian and Indo-European families. Then from this were derived certain indicative forms denoting past time. The origin of the past tense and past participle in d may be looked for, perhaps, in the ancient Chinese pronoun ti, "him," Mandarin chi. This word is used as a preposition, "to a place," and as a possessive particle. I take an example from M. Julien's Syntaxe Nouvelle de la Langue Chinoise : Kwo chi kiun, " the kingdom's prince." The Mongol use of the suffix de in the sense of "towards a place " is parallel. In the dialects of China the same word is used in the same way as the Mongol gerund. Tsu ts'i ping, tang tsang k l i tse, 1 " he has become a soldier and gone to fight." Tsu is to " do," " be," " act as," old form sale, in Tamil sey, " do." Ts'i is the common sign of the possessive, here used as a mark of the gerund. Ping is " soldier." Tang tsang is a compound verb, "to fight." K'i is " to go," and is put in the past by the last word tse, a particle fixing past time. In M. Julien's examples of the use of chi 2 may be seen, Wei shu sheng chi, " only millet grows," said of the barbarians of Tartary, whose country will not grow rice or wheat. 1 Shanghai Grammar, § 252. 2 Syntaxe Nouvelle, p. 75. 13 194 china's place in philology. At present the northern boundary of wheat cultivation passes at about 200 miles to the north of Peking. The word chi is here, says M. Julien, without signifi- cation. But may there not be here the commencement of a gerund formation like that seen in compounds formed with the word cho in Mandarin ? Thus, in Wo chan cho ti sh'i heu, "while I was standing/' wo is " I." Chan is " stand." Cho makes it a gerund. Ti is the possessive sign to the verb-noun, chan cho. Shi heu is a compound noun meaning "hour," "time." Looking at the use of chi in this way, and keeping in mind the Shanghai usage above adduced, the history of the gerund formation in d would become clear. The root ti appears in old Chinese literature, (1) with the meanings " this," "him," "towards," "go towards"; (2) with the pos- sessive sense, thus becoming a mere auxiliary particle ; (3) with a gerund-like signification, which comes out more distinctly in the dialects. 1 The other Dravidian perfect in i and the aorist in b or v are interesting from their striking resemblance to the Latin perfect in ui, vi, and imperfect in bam. In Mongol the ordinary past tense ends in ba when the root has a or o, and beu when the root has e or i, e.g., yababa, " he went," helebeu, " he said." The 1 If this is not the true origin of the Mongol gerund in d and ju, it may be possible to find it in the Chinese Mandarin gerund cho, meaning originally "to hit the mark," "strike," "take fire," etc. CONDITIONAL TENSE. 195 form in ba differs in nothing from the Latin imperfect, except that the Latin has proceeded to affix the mark of the personal pronoun, a stage which the older Mongol dialects have not reached. The Buriat- Mongols, however, have added b, s, t, to the three persons, thus making the identification complete. The Turkish and Persian languages, which have always been neighbours, both have m for the first person, as in the Persian imperfect budam, budi, bud, " I was," " thou wast," " he was," corresponding in Turkish to boldim, boldung, boldi. The Manchus also have a past tense in bi. In Japanese ba is used to serve as a suffix to the verb in a subordinate clause with the sense "when," or "as." This appears to be the same as the Mongol conditional suffix bel. Hoffmann gives the example : yama no ne kumo tsigiretaraba yagate fare, " should the clouds on the top of the mountain be dispersed, it becomes forthwith clear." Here no is the possessive, yama is " mountain," ne is " the summit," kumo is " cloud," tsigiretara is " disperse," yagate is "forthwith," and fare is "becomes clear." Ba is "should" or "if." As to the origin of ba and bel as conditional suffixes, or as signs of the imperfect indicative, there is perhaps nothing more probable than an ancient connexion with the Chinese pi, " to compare," and pei, " to give." The lost d of the latter of these words is recovered by comparing the Shanghai form peh for pet with the 196 china's place in philology. Japanese hodokoshi, " to give" (N.B. Japanese h=p or b). The Japanese for "compare" is haiszru, "to match," " equal." The Chinese, as at Shanghai, use both words in compounds, as in sung peh la ngu, "present it to me," literally " present give to me " ; slang pi, "mutually compare." The Tamil has these words with or without an initial o. Thus, oppanei is "simile," "parable," pol, " like as," oppdri, " comparison," oppi, " give," poll or oppu, " likeness," " congruity," poke or oppic, " to be like," " resemble," oppumei, " similitude," oppiwi, "to give," "deliver." The Chinese has also the aspirated words p'i, " a comparison," p'ei, " to match," " correspond to," and p'i, " a match," where the root is in all cases pHt. In the Indo-European languages, the Russian has podobie, "resemblance," and upodoblenie, "comparison," where the prefixed u is curiously like the o in the Tamil forms. The English h&spair, and the Latin par, "equal," and comparo, "to compare." The Latin paro, " prepare," is the Chinese bid, "prepare," in Mandarin pei. In Chinese there are also other members of this numerous family, namely, pi, " he," and pit, " other," already adduced in a previous chapter. The explanation now proposed of the conditional and past tense suffix in b is, that its original meaning was " resemble " and " give," and that it was appended as a verb, in juxtaposition with a preceding verb, as in the modern eastern Mongol helji og, " speak for me/' PAST TENSE. 197 where og means "give," and helji is the gerund of of helhn, "to speak." In Japanese and Mongol it became suppositive, and in Dravidian and Mongol preterite. In this state it passed over into the Latin, when tjje ancestors of the Romans were still in Asia, and in close connexion with the Turanians. "What is given is passed over to another. The very word past means transferred. Bopp has derived the Latin imperfect from the substantive verb fui, fore, but there is this objection to that view. The same suffix for the past tense exists in Dravidian languages which have not this substantive verb. The substantive verb in b first comes into view in the Tartar languages. The older branches of the Turanian family, the Japanese and Dravidian, have it not, nor do they contain any traces of the first personal pronoun in m, which is always found in the company of the substantive verb in b. The other Dravidian past in i — as in the Kannada aorist w and in the Tamil perfect in i — resembles the Latin ui and vi, in docui and amavi. Though it does not appear in the verb paradigms of the Mongol and Japanese languages, there is no diffi- culty in finding it in Chinese. It is the word i, " already." By analogy the old form of this word may have been i or wi. The Latin, Sanscrit, and Tamil v is the equivalent of the Chinese w. The word i, "already," is in Chinese used in the sense 198 china's place in philology. "past and gone" (lower second tone), and perhaps originated the final particle i, for which it is sometimes used. 1 It differs in nothing from the third personal pronoun i, except in tone : a quality which, as has been shown, may be treated as having been non- existent 4,000 years ago. The same word also means "other," "different." Hence the fundamental idea of it is " difference," in space, in person, or in time. Combined with jen, " man," it means " a man of another country," " a barbarian " (lower first tone). It is a noun, "difference," in the sentence, ta t'ung siau i, " great similarity and small difference." In i ti, " a different place," it is an adjective (lower third tone). As an adverb it means "again," and as such it is the word pronounced in Mandarin yen, but in the Shanghai dialect yi (lower third tone). As the third personal pronoun (upper first tone), it is still used in the south-eastern dialects. The Dravidian future in i or e is evidently identical with the Mongol future in ya, and these forms together constitute an old type from which the . Latin future in e and ie, as in regam, reges, and audiam, audies, may have been formed. Its origin may perhaps be discovered in the Chinese yau, which takes the old form ok, "wish," "desire." It is a common sign of the future in Mandarin- Chinese. The k was early 1 For an example of the use of i, " already," as a final particle in a predicative sentence, vide Syntaxe Nouvelle of M. Julien, p. 186. FUTURE TENSE. 199 lost in the colloquial language. The corresponding western word is volo, wo lien, will, fiovkofjuai, and perhaps wish. 1 That this identification is not unlikely to be correct may be shown by reference to the other Chinese signs of the future. Tsiang (old form siung) contains in it the sya, which is the Sanscrit sign of the future, and the s of the Greek and Latin future, as in fiovkevaa), " I will advise," OovXexxKov, " about to give counsel," and ero, "I will be" (r for s). Another sign of the future in Chinese is pit, in Mandarin pi. It means "certainty," " certainly." " It will certainly be so." The word is the same with the Latin fides, the Greek ttigtis, and the Hebrew Plp^l bata, "he trusted." This I suppose to be the source of the Latin future in bo, bis, bit, where bi marks the tense and o, s, t, the person, as in amabo, " I shall love." This affords a more natural explanation of the future tense formation than to derive it in the manner of Bopp from the substantive verb, fuisse, futurus, etc. The Latin future in r, as ero, amavero, etc., is coin- cident in a curious way with the Manchu future in ra, re, which again strikingly resembles the Mongol supine in ra, re. A supine is a sort of infinitive put in future time, and hence in English the supine and the infinitive are not distinguished. The Mongols use 1 Compare wash, in Chinese og, Mongol ogahu. Sh is a western equivalent for the old Chinese final g or k. In German, the inserted n in wunschen, "to wish," causes a difficulty in the identification. 200 china's place in philology. for the future both the present tense in moi or ne, and also the infinitive in hu. Thus, in Turanian gram- mar there is not a little mutual interchange between the present, the future, the supine, and the infinitive. Hence it should be regarded as open for consideration, whether (if yau, " wish," is not satisfactory) the Chinese substantive verb wet, " to be," " to do," and in the third tone "for," "for the sake of," may not be the source of the Mongol future in ya, and so of the Dravidian and Latin forms already adduced. This verb exists in Mongol in title, " an act," and uiledhu, " to do " ; and is probably the root of our western am, was, werden, est, esse, Sanscrit asti, Tamil iru, " to be," Japanese iru, oru, " to be," " to dwell." The syntax of the Dravidian languages is similar to that of Tartar and Japanese speech. This will be understood from some examples of Tamil sentences, taken from Pope's Handbook. " Open the door " is hadavu tira, where Jcadavu is the Mongol egude (or in modern vernacular tide) and the Japanese kado, " door." The Chinese equivalent is gud, in Mandarin hu. The verb tira, " to open," stands last. So in Mongol ude nehe, " open the door," where nehe is the Greek avoiye. An example of the participial construction is the following : nan paditta pddam, " the lesson which I have learned." Nan is "I." Paditta is the past parti- ciple of padi, " to learn." Pddam is a verbal noun from the same root. Compare in Mongol hi omsihu ne DRAVID1AN SYNTAX. 201 bichig, "the book I am reading.' ' Here omsihu is the infinitive or present participle "reading." It is in the possessive case, with ne to connect it with the following noun, " book." The past participle would be omsis { en, the other words remaining the same, and the meaning would be, " the books which I have read." The Chinese construction is similar, wo nien ti shu, " the book I am reading," or " which I have read." Here ti is the possessive (the verb nien, " read," being treated as a noun), and corresponds exactly to the Tamil td. The gerund construction will be perceived from the following instance : nadandu wandan, " walking he came," Mongol yabaju irebe. As j takes the place of d, the suffix ju is the same as the Tamil du and the Japanese te. The Chinese has the same order, tseu lai. The Indo-European languages invert the order, as in rj\0e, ^Xeircov, "he came seeing," rediit videns. Another example is kettu wasittu emidinan, "hearing, reading, he wrote." We should say, "he heard, read, and wrote." The Mongol would use one gerund in d, another in Ju, and then close with the indicative. There can be little doubt in regard to the pro- bability that the order of verbs in this Turanian construction rests on the older law ruling the order of verbs in Chinese, viz., that of succession in time. Hearing and reading precede writing. Walking pre- 202 china's place in thilology. cedes coming. After the Turanian period, when an indicative was fully formed, it was possible to transgress this order. The rich Indo-European verb paradigms allowed of verbs being easily distinguished from each other, and language was no longer obliged, in the interest of clearness, to maintain a strict adherence to the order of time in the arrangement of her verbs. A more complex example from Mongol will illustrate the syntax of an expanded sentence : Pere mande helsen ne uge hi mart'asen, "the words that he said to me I have forgotten." T'ere is "he." Mande is "to me." Helsen is the past participle of helhu, "to speak." It has the possessive particle ne. Uge is words. Bi mart'asen is, "I have forgotten," the participle being used as an indicative in the colloquial language. In the book language it would receive after it the substantive verb in the indicative, to complete its expression. The construction, with the participle, is here seen performing the duty afterwards assigned to the relative pronoun. Helsen ne uge is. a relative clause. This was, in the early state of language, rendered possible by the fact, that the verb was viewed predominantly by the Turanian mind as a substantive ; and, as such, the office of finding room in a sentence for the relative clauses of western languages was considered to belong to it in one of its cases, viz., the possessive. But the more lively DRA VIDIAN SYNTAX. 203 and energetic attributes of Semite language had in this respect greater influence on the Indo-European mind. The relative pronoun became the hinge on which the clauses of compound sentences could con- veniently turn, and the honour of accomplishing this duty was no longer assigned to the verb in the possessive. CHAPTER X. Third Division of the Turanian System. — Mongol as a Type of Tartar Languages. — An Old Turania in "Western Asia. — The Tartar Turanians come nearest to the Indo-Europeans. — System of Sound. — S and J for SS and D. — CH. for S. — Final NG dropped. — No F. — Seven Vowels. — Tone. — Acci- dence. — Substantive Verb and First Personal Pronoun. — Mongol Declension. — Pronouns. — The Mongol Verb Conju- gation. — A Mongol Verb. — Adverbial Suffixes. — Mongol Syntax. The great antiquity of the Mongolian type of language is manifest from its being found in several of its leading features in the Dravidian area. The historical events which have separated the branches of the great Turanian family furnish to us an approxi- mate chronology for the early stages of Turanian development. They point to a period anterior to the dispersion of the Indo-European families, when there was a primitive Turania in "Western Asia, from which the Japanese, Dravidian, and Tartar races proceeded. This time cannot be later than 2,000 years before the Christian era. At that time the Turanian verb had already its gerund, its past participle, and its TARTAR TURANIANS NEAREST TO INDO-EUROPEANS. 205 three indicative tenses, a scale of case suffixes, several polysyllabic derivatives, and a common syntax. The suffixes were attached more loosely to the root than in the Indo-European system. It could not have been otherwise. For the Turanian type stands mid- way between the monosyllable of China and the richly elaborated polysyllabism of modern Europe. The difference between the agglutinated and inflected lan- guages is simply a question of lower and higher development. Linguistic types come one out of another, like orders in architecture, or ages in geology. The Indo-European system rests on the Semitic and Turanian systems, as they do on the Chinese, and as the Chinese does upon the primitive speech of Western Asia. The special interest of the Mongolian type consists in the fact that it comes nearest of all the three Turanian branches to the Indo-European. As Iran and Turan stood opposite to each other with hostile front, but in close contiguity, in ancient Persian remembrance, so Arian and Turanian speech, in many respects varying, stand to each other in the closest proximity. Their remarkable resemblance consists mainly in the formation of tenses by suffixes and in the extensive use of the same substantive verbs and personal pronouns. The verb " to be," the first personal pronoun in m or b, and the second and third in s or t t are as widely extended in Tartary as they are 206 china's place in philology. in Europe, and they form an incontrovertible argument for common origin in language, race, and ideas. The same mental constitution which led the Tartar tribes to develope these roots in a declined and conjugated form, as the convenient expression of their ideas of existence and personality, led the Indo-European races to adopt them for the same use, instead of the more ancient substantive verbs and pronouns found in the Semitic, the Chinese, and the older Turanian lan- guages. That the Fins, Manchus, Mongols, and Turks should have borrowed this striking feature from the Indo-Europeans seems very improbable. It is worked thoroughly into the texture of their languages, and has nothing of the appearance of a foreign element. The Mongol and other Tartar languages have suf- fered less from phonetic decay than the Japanese and Dravidian branches of the same family, which have been exposed to the enervating effects of mild or hot climates. Hence there is found here a greater variety of sounds. Thus, the syllabary includes ng, n, m, g, d, b, I, r, s, among the finals. Of these, the last three are beyond the capacity of the Chinese vocal system, and they must be regarded as new. Thus, gol, "river," is formed by appending / to the Chinese root ga, in Mandarin ho. T'os, "opposite," is formed by dropping d in the old Chinese tod, " opposite," Mandarin tui, and adding s. The Tamil has edir, "to oppose." SYSTEM OF SOUND. 207 The Indo-European languages allow any letter to end a syllable. Thus, in English, in addition to the nine consonants by which the Mongols can close syllables, we have /, v, k, t, p, ch, dj, z, sh, and the surd and sonant th. This is an unmistakable proof of the advance in freedom which language has now made. In the Mongolic stage it had added three finals to the Chinese and Himalaic phonology. In Sanscrit the finals are ng, n, m, h, t, d, s, r; differing very slightly from the Mongol. In Latin, when we have repeated haud, aut, in, hie, collis, clam, f rater, multiplex, we seem to have exhausted the capacities of the syllabary, and have only eight final consonants, d, t, n, k, s, m, r, x, of which the last, x, is a compound of two others, k and s. It is only in the Gothic and Sclavonic speech that language assumed the power of ending syllables with whatever consonants it pleased. Among these two, the Gothic has more freedom than the Sclavonic, and probably there is no language in the world that can compare in this respect with the English. This, however, is a distinction which has been acquired only after long and patient waiting. Language passed from the monosyllabic stage into the Turanian, from this to the early or southern Indo-European type, and from that to the later or northern type of the same family, before venturing on so great a leap. In Semitic phonology, on the other hand, language, with characteristic boldness, claimed the privilege at 208 china's place in philology. a most ancient period of using as finals the sibilants and liquids, in addition to the mutes and nasals which were the finals of the primeval monosyllable. The unaspirated surds k, t, p, do not exist in Mongol or Manchu. These letters, as written by De Castren and Schmidt in their Grammars, represent aspirated surds. 1 They appear to have grown out of the sonants. Thus, t'ologai, " head," is in Chinese du, Mandarin t'eu. So also t'ola, "for," "on behalf of," is in Chinese t'ek and dak, in Mandarin t'i and tai, " instead of." The aspirated k has in the eastern Mongol, which is that spoken in the neighbourhood of Peking, become h, but k { is retained by the western and northern Mongols. The want of sh in Mongol, or at least its very sparing use, reminds the student of the Greek and Latin languages, which also lack this consonant. The coin- cidence can scarcely be regarded as accidental when the many remarkable resemblances in words between the Tartar languages and the Greek and Latin are kept in view. In Mongol k'umun or humun, "man," ere, "male," nehemoi, "to open," gar, "hand," dalai, "the sea," ebur, "horn," sara, "moon," nom, "sacred book," may be compared with homo, " man," fir, apprjv, " male," avouyw, " to open," x e W> " nan( l," 6aXaaaa, 1 What the rule is in Turkish I cannot in Peking learn with certainty. The influence of Arabic and Persian may have led to a change of the aspirates to the pure surds. CHINESE B AND T REPRESENTED IN MONGOL BY J. 209 " sea," ebur, " ivory/ ' aeXtfvr), " moon, ,, vo/jlos, "law." It is probable, therefore, that at some distant epoch a strong Turanian influence was exerted specially upon the Greek and Latin sections of the Indo-European family, subsequent to the separation of the Indo- Persian tribes from the common Aryan stock. The Chinese d and t, in Mandarin often ch, is in Mongol represented by j. Thus, ti, "to point to," di, "to -rule," "to cure," tok, "candle, "to shine," U, " decree," are in Mongol jahu, " to point," jasahn, "to rule," "correct," "cure,"/o/, "candle," "lamp," jerlig, " decree." In Greek and Latin the corre- sponding words are helKw^L, indico, rego, luceo, lux, lex. The law regulating these correspondences is, that words commencing with d and / in the Latin are found to agree in meaning with words whose initial is in modern Chinese ch, in Mongol j, and in old Chinese and Mongol d. As in the registered old sounds of the Chinese tonic dictionaries, dating from a.d. 400, some of these words are spelt with t, it may be reasoned that, as before stated, the old Chinese t probably came out of a more ancient d, for then it will not be surprising that I should be the Latin equivalent. The letters / and d have a known affinity for each other, and appear to be related, as son to mother. L has grown out of d, and so also has r, and thus has been caused an expansion of the alphabet in Sanscrit, in the Semitic family, and probably in 14 210 CHINA S PLACE IN PHILOLOGY. other ancient systems, following example. This will be seen by the PRIMITIVE IDEA, "TO POINT," DIK. CHINESE. SIGNIFICATION MONGOL. SANSCRIT. GREEK, LATIN. GER., ENG. ti, chi point jahu de'lKUVfil zeichnen. dik, ti lead daksha dexter token. di, da to rule jasahu dakshinat duco di, da to cure jasahu dux tok, tu oversee ejelehu digitus tu, chu lord ejen raj all rex le, li reason \6yos, lex reason. le,li to rule raj rego Eecht. ti decree jerlig decet law. ti will dticaios dik, chi straight riju rectus straight. A final k or g appears to have been lost from all the Chinese words where it is not marked in this list. In the Mongol ejen, "lord," n final is not part of the root. It disappears in declension, as in the plural ejid, and in the verb ejelehu, "to rule." In the Sanscrit daksha, "right (hand)," dakshindt, "south- ward," the double letter ksh has taken the place of k. The influence of religious ideas connected with the worship of light is perceptible in the east being regarded as the front and the south as the right. As in common roots the Indo-European r corresponds to the Chinese /, and the Indo-European I and d to the Chinese ch, the transition from d to I, observable in CH FOR S. FINAL NO DROPPED. 211 the Chinese examples now given, must in all proba- bility have taken place at a time anterior to the separation of the races, and when the forefathers of the Chinese and Indo-Europeans still spoke one lan- guage. The primeval root dig became doubled by the change of d to I ; and while dig and lig both remained in Chinese, they originated in the Indo-European languages two sets of derived words, one set with the initials d or I, and the other with an initial r. In the English example straight, s is prefixed to the root and r is inserted after the initial t. The same root occurs in the Semitic languages with the sibilant pre- fixed, as in the Hebrew tsadik, " just," Arabic sadik. The Mongol aspirated ts or ch is found to be the Chinese s, sh, ts, or t ( s. Thus, the words c'hitgur, " a demon/' c'hasa, " snow," c'hohom, " accurately," " altogether," c'hag, " time," c'hilagon, " stone," are in Chinese sut, suy, " an evil spirit," " to exercise demo- niacal influence," sit, "snow," sik, "all," "thoroughly," shi, "time," "hour," shig, "stone." The equivalent in Indo-European is s or sh, as in c'hadaho, "to be satisfied," satis ; c'hilagon, " a stone," saccum ; ch'i, " thou," gv. The change from c'h to P, in c'hi, "thou," t f a f "ye," is parallel to the change from the Greek av to the Latin tu, " thou." The final ng of many Chinese words is dropped in Mongol. Thus, solaraho, "to become weak," to, " dragon," gerel, " light," are the Chinese silng. 212 china's place in philology. ''loosen," lung, "dragon/' keng, kwang, "light." The same tendency perhaps lurks in the Sanscrit Rdhu, " the demon of eclipses," and the Latin gloria, where an I has crept in after the initial. The Japanese and Tamil agree with the Mongol in dropping ng final, as in the Japanese dhari, "light." The word morning, Morgen, may be derived from the Chinese many, " bright," in Mandarin ming, through the Mongol maragaPa, " to-morrow," spelled in books managar. Other common forms are magaPara, maragasL They all mean "to-morrow." There is a verb manahu, "to shine," " ascend like the sun." The want of / in Mongol suggests a close connexion in this part of the Turanian sound system with the Sanscrit and Greek. Its place is supplied by b. The vowels are seven. They are called by Schmidt a, e, i, o, u, o, u. These values answer for the western and northern dialects ; but for the eastern Mongol, spoken in the neigbourhood of Peking, and which has not been described by the Russian and Grerman gram- marians, the values are rather a, e, i, 6, o, o and u or u. The fourth is the English o in fond, the fifth and sixth are divided by tone, and the sound is the English o in bone. The seventh is sometimes the English oo in tool, and sometimes the French u. The distinction between the fifth and sixth vowels cannot be described in any other way than as a variation in pitch, the fifth being lower than the sixth. The eastern' Mongol A MONGOL TONE. 213 bears evident marks of being the most ancient of the dialects. It has no traces of the personal endings in the conjugation of verbs which occur in the Buriat dialect. This double tone, therefore, of the east Mongol syllabary must be regarded as a link of con- nexion with the Chinese and Himalaic systems. In learning the Chinese language the foreign student meets with the tones in the individual words to which they are attached. In the Tibetan, Siamese, and east Mongol, he meets with them in the syllabary. It is the same thing. If the Tibetan and Siamese were written with a separate character, half ideographic and half phonetic, for each word, the tone mark would be attached to the character in some such way as that in use among the Chinese. The difference between the fifth and sixth vowels of the Mongol syllabary would be expressed by saying that all words enunciated with the fifth vowel are in one tone class, and all words enunciated with the sixth vowel are in another. The existence of this double tone harmonizes with the view that the Mongol language rests on the Chinese as its basis. If a language came between them, it must have had a tone system, which would occupy a midway position between the Chinese system of tones and that of which the last vestiges are now slowly disappearing in the oldest and most easterly of the Mongol dialects. In the study of the Tartar languages, and the 214 dialects and languages of the same stock in European and Asiatic Russia, the occurrence of the substantive verb to be, and the first personal pronoun in m or b, is the most striking of all signs of kindred with the Indo-European family. The verb buhu, "to be," has a present boi, an imperfect bolai, a perfect buloge, a conditional bugesu, a potential boija and bubeja (?), three gerunds burun, bured (?), bugd'te le, and two infinitives buhu, buhwei. The root a, found in our auxiliary verb am, are, art, is also mixed with the auxiliary to be in a way resembling that to which we are accustomed in English. The parts are, a present amoi, an imperfect abai, a future aho or ayo, a conditional abasu, a potential amoija, a precative at'ogai, an imperative plural akt'on, two infinitives aho and ahwei, three gerunds aju, agad, and at'ala, a participle of agency ahc'hi, and a past participle ak'san. The first of these, bu, is the Sanscrit bhii, in the infinitive bhavitum, " to be," or " become " ; the Persian budan, "to be"; the English be; and in the Turkish bolmak, "become." In the Tartar languages the con- nexion of this root with the ordinary word for I, in Mongol bi, in Turkish ben, in eastern Turkish men, in Manchu bi, is manifest. The possessive of bi, " I," is manai, " mine." Thus, then, the English me in the accusative is the Mongol bi in the nominative, while the English possessive mine is the Mongol possessive SUBSTANTIVE VERB AND FIRST PERSONAL PRONOUN. 215 manai, in the book form minu. It appears then that the English me and be are the same word, and that that in which the Mongol differs, namely, the con- vertibility of b and m, is derived from some Turanian language, the parent of the present Tartar languages. In the Indo-European family, m is appropriated to the pronoun and b to the verb. Hence their identification is not at first view obvious. In the Tartar languages, where b is used for the verb, and m and b are both in free use for the pronoun, the identification does not admit of doubt. The question then arises, what is the origin of the verb to be and the personal pronoun me? We have in Mongol and Manchu the word beye, " body," in Japanese mi. This Japanese term signifies both "self" and "body." We have also in Chinese mut (Mandarin wu) meaning " a thing." The Sanscrit matra, " matter in the abstract," and Latin materies, are by some derived from the word meaning "mother," in Sanscrit mata, in Latin mater. Further, we have in Mongol mun y " it is so," a strong affirmative, and in Hebrew amin, " certain." The Mongol and Japanese substantives furnish the ideas of self. The Chinese, Sanscrit, and Latin substantives contribute the notion of substance. The Hebrew and Mongol verbs add the conception of certainty. Why may not these ideas have met in the formation of one pronominal and substantive root, destined to pervade the languages of mankind from 216 Manchuria to Portugal, and from Calcutta to Finland ? That this root is not used in the Semitic, Dravidian, Tibetan, Chinese, or Japanese languages, either as a substantive verb or personal pronoun, affords a strong presumption that it was not originally either the one or the other. If this hypothesis be correct, the com- bination of ideas, which resulted in the growth of the substantive verb in b and the first personal pronoun in m, must have taken place in the language which originated the present Tartar dialects. The locality of this language was probably Western Asia, or Persia, or Bucharia, for only in one of these countries could it be in such convenient contiguity to the Aryan race as to allow of the engrafting of this fruitful germ into the mother-speech of that family. The other auxiliary verb amoi, I suppose to be the Chinese wet, " to do," " to be." This was, as we learn from the rhymes of the Shi king, anciently called wa. In Sanscrit it appears with a suffix s, which is retained in our was and were, in the last of which s is repre- sented by r. The present amoi seems to be formed from the root a by the addition of boi, the present tense of the substantive verb in b, with b altered to m. The imperfect abai cannot be derived in the same way, because, as before stated in the foregoing chapter, the Dravidian languages have this tense form, while they are without the substantive verb in b. We SUBSTANTIVE VERB. 217 may refer it rather to a Chinese root, pet or bed, " to give." The form in lai and that in loge may perhaps be derived from the Semitic le. The word /e, or al, means "towards," "to," "into," and is used to mark the dative case. The form el, with a vowel prefix, gives in a more marked manner the proper and physical sense ; and that with a short vowel suffix, le, is used for borrowed and metaphysical senses. 1 The Tibetan language has la for a dative case suffix. In the Shanghai dialect the same word is used as a dative case prefix, and with the force of a substantive verb in the locative, as in the sentence, I la a li, " where is he?" literally, "he at what place." Here the word la, translated by " at," has the force of " is at," that is, it is a substantive verb in the locative case. Its Mandarin representative is tsai, anciently %e. The Mongol imperfect in lai and preterite in loga may have been formed from the Semitic le and Shanghai la, by the intervention of a gerund usage, or, in other words, a predominant use of the verb as a substantive. For "he has come back" the Chinese say hwei lai liau, literally " return, come, finished." Three verbs are here in juxtaposition. Hwei, " return," is a gerund, and is translated into English by the present participle " returning." Lai, " come," is indicative, and is made past by the addition of the auxiliary liau, "past," 1 Gesenius' Lexicon Hebraicum. 218 " finished," which is a modern particle, formed from a verb, liau, "to destroy." The Mongol would say hairebe, "he returned," or haireji irebe, "returning he came." Put le in place of the gerund suffix ji 9 and the sense will be, " in returning he came." Then drop the last verb, " came," and the form in le or loga becomes a past indicative. So in modern Mongol, as spoken in Peking, sentences such as the following are in constant use. — T'ere nidenen jil yabaji, " he left last year," literally, "he last year left." The gerund form in ji is here used as a past indicative tense. It ought to be yababa. It is ungrammatical. l But language is always busied in making new forms, successfully or unsuccessfully. If the Mongol lan- guage needed a past indicative, it might easily be made from the gerund in ji or ju in this way. So we may suppose the preterite in loga, colloquially called lai, to have been formed. This is in harmony with the general principle, that tense and mood suffixes in the Turanian and Indo-European languages have been all formed from verbs viewed as nouns and used as gerunds. "When gerunds, participles, and infinitives had been formed, they became indicative in past, present, or future time by the simple process of drop- ping the following verb. This principle of tense and mood formation is at the opposite pole to that which exists in the Semitic languages. Thus, VJ"h )Vty : fiKT 1 The full form would add ie'hibe, "went," after the gerund yabaji. SUBSTANTIVE VERBS. 219 zoth asu vihheyu, "this do ye, and live." Here an imperative is used in both cases. To the Semite mind each verb was instinct with its own energy. He struggled to secure to each verb in a sentence its full activity, and therefore he connected them by the con- junction and. This device allowed them each to be indicative. This vital character of the verb has been usually retained in the English version of the Scrip- tures, as in the same example, " This do, and live ; for I fear God." (Gen. xlii. 18.) Luther has altered the Semitic mode of expression. He translates, Wollt ihr leben, so thut also ; denn ich furchte Gott, " Would you live, then so do ; for I fear God." He has two clauses, of which the conditional contains two verbs, wollt and leben, the latter with a Turanian suffix ; and the affirmative one verb in the imperative. Compare this with the Septuagint rendering, Tovto irocrja-are teal %r)ai(T0€, tov Oebv yap iyeb (pofiovficu, "This do ye, and ye shall live ; for I fear, God." The imperative and future are here employed. The Greek is only second to the English in its capability of imitating the freedom and energy of the Semitic verb. How different is the Mongol — T'a her amit'o baihwain t'ola egoni weiladok't'on hemebesu, bi ber Borhan ec'he aiyomoi, " Ye, for the sake of life, should do this. I fear God." Ber is a particle which marks the nominative t'a, "ye." Amit'o is " living," t'o being equivalent to the English suffix ing in living. To' la, "for," governs the infinitive 220 baihu, " to have/' " to be," in the genitive. Egoni is " it," the final i marking the accusative. WeiladokH'on is the plural imperative of weiledehu, " to do." Seme- besu is the conditional mood of hemehu, " to say," here used as a particle. Bi, her, " I." Borhan, " Buddha," is the term used for God in the Mongol version of the Scriptures. Aiyomoi is the present indicative of aiyohu, " to fear," governing the noun Borhan by the intervention of ec'he, " from." The Latin vereor, " to fear," which is the same word, sometimes governs the genitive. The Greek alBioficu, " to fear," " reverence," " be ashamed," retains the old final d, which has been changed into r in the Latin vereor. The Chinese wet, " to fear," and the Mongol aiyomoi, have both lost the final d. Beside the verb buhu? " be," there are also two im- portant auxiliaries, baihu, "have," and bolhu, "become," " arrive at perfection." Bolhu seems to come from bolai, the imperfect of buhu, by simply treating it as a root and adding to it the usual suffixes. Thus, bu is " be "; bol is " arrived at being." In a similar way, werden, "to become," seems to be derived from war, ware, by appending d, the sign of the past, which would give it the sense " already come into being." To werd was added the infinitive suffix en, and with it all the suffixes usual in the paradigm. Add to bol the causative ga, and we have bolgahu, " cause to become." Perhaps baihu, "to have," used to assert positive FIRST PERSONAL PRONOUN. 221 existence, is formed in a like way from the past tense, abai, of aho, "to be." But it may also possibly be connected with, the Chinese verb po, or pok, " to hold," " hold in the arms or hands," etc. On following the substantive verb root in b into the pronoun, we find it used in some parts only of the declension. As in English, I and me combine to make up the declension, so in Mongol bi and minu form the nominative and genitive singular, while na, another root, forms the dative naded, the ablative nadas, the instrumental nadar, the comitative nadale, and the substitutionary nadat'a* The root nad or na I believe to be the Chinese nga, in Mandarin wo, and to be identical with the western ego, aham, ich, I, and the Hebrew anochi. The Tamil uses this root throughout its declension, as nominative nan, "I," accusative ennei, instrumental enndl, dative enakku, ablative ennil, genitive en, locative ennil. The Tungus and Turkish use the root b or m throughout the declension. The Mongol plural bida, " we," is carried through all the cases. The root na does not appear at all in the plural. The suffix da occurs also in the plural of nouns not infrequently. The Turkish plural is biz, "we," where z, we can scarcely doubt, is a changed form of d, as we have found the Mongol j to be derived regularly from d. The same plural occurs in the Hebrew aboth, "fathers," from ab, "father." 222 china's place in philology. But t in the Indo-European languages occurs not seldom for a more recent s. The Latin tu, " thou," is older than the Greek out off receives as its name the name of the action which cuts it off. The word jfil kiai, "bound;: has almost certainly the same origin, and hafl V 232 china's place in philology. in the same manner. It means " that which is cut off." The Hebrew katsah signifies both to "cut off" and to "end." The Mongol hijagar, "boundary," retains the lost d in the modern j. The phonetic §| having anciently the same final d, we may then be allowed to regard as identical with the Chinese ki the Mongol heden, and the Latin quot. The origin of the Chinese ki, "how many?" may thus be seen to resemble that of the German compound wie viel and the English how many. Ki, how, wie, quis> ttoctol, are the same word. The Chinese, Mongol, Latin, and Greek forms added da or ta, "many," dropping the vowel, and in the instance of the Greek changing t into s. The Germans appended viel (the Greek iroXkoi), and the English the word many. The source of the interrogative element in the Chinese word is either }fc gi, "he," or ga, "what," {$ , in Mandarin ch'i and ho. Of these, I suppose the demon- strative to be the earlier, and the interrogative to be formed from it. The Turkish has kanda, "where?" kirn, "who?" kih, " that," " for," " who," kach, " how many ? " kachan, "when?" "We learn from these forms that the re in irore and the do in quando, "when," are the Turanian locative suffix de, " in " a place. The other Turkish form, kani, is probably formed from the root nip, ni, " in," " the inside," as in the Chinese wei ti, " inner land," kung nei, " in the palace." MONGOL ORDINAL NUMERALS. 233 Perhaps the n in the English when} and German wann may have been derived from the same root. The m of kirn may be the Chinese and Semitic inter- rogative ma, " what ?" Kill reminds us of the Hebrew ^p hi, "for," originally a demonstrative. "While the cardinal numbers in the Mongol are very different (from those of China or of the Indo-European languages, there is in the syllabic addition for the or- dinals a remarkable resemblance. The Chinese prefixes de (Mandarin ti). The Mongol adds dogar. The Sanscrit adds tit/a, the Latin tus, the Greek to" yuen, " cause," net or nip, " within." I prefer to regard yin as the true root, and identify it with our preposition in, iv. Gerund in ju. Old form du. Chinese pronoun and sign of genitive ti. In Mandarin ch'i. Identical with the English past tense in ed. Gerund in d. Schmidt calls it past, but it is little, more a past tense than the gerund in ju. Probably it is of the same origin. Gerund in man. A colloquial form, JBada idemen irebeu, "after taking food he came." Gerund in tola and sagara. Colloquial sara. They limit the verb in " time." Tala is " until," and sara, "during the time of." Nar onatalei helchibeu, "he conversed till the sun set." Origin, the Chinese to or tau. English to and till. Sara probably originates in the Chinese dze, " at," " to be in," or " at." Example, uder uder ireser baina, " daily he is in the habit of coming." Gerund in Itei. A sort of passive gerund. The future participle passive of western grammar, e.g. amandus, " deserving to be loved." Examples, bicheltei, " deserving to be written," icheltei, " worthy to be gone to," ene chichig ujeltei, " this flower is worth seeing." MONGOL VERB. 239 Formation : the I in bichel is a derivative suffix, form- ing a verbal noun. It may be originally la, " to." The syllable tei is an adjectival suffix, and must be referred to the pronominal root ta, " that." The Latin dus in amandus may in the same way be viewed as demonstrative. Supine, in ra, re. In Manchu and Latin, re is the infinitive suffix. In English the same preposition, "to," marks both the infinitive and the supine. We may, therefore, without hesitation, identify the supine of the Mongol written language with the Manchu and Latin infinitive. The colloquial supine in Eastern Mongolia is the infinitive construct in hwei, e.g. ujihwei iehibeu, "he went to see." The infinitive ends in ho, hu, and hwei. The Turanian conception of the verb being intensely sub- stantive, the infinitive is regularly declined as a noun. Origin : the Chinese pronoun gi, Latin hie, English he. The form varies, as in ho or hu, according as the vowel of the root is a, o, or e, u. It may be called the free infinitive. The form in hwei, or hoi, is the in- finitive construct, and is used in declension, and as a supine, e.g. holda hwen t'ola garaba, "he is gone out to sell." Here the suffix t'ola, " for the sake of," follows the infinitive in the genitive. Participle. A present in gchi or chi. As yabokchi, " going," " he who goes." It is used profusely for all classes of agents. Origin: demonstrative in s. In 240 Chinese t's'i, si. In Sanscrit, sah. There is also a past participle in gsan, san. For example, yabasen, " gone," in the book-form yabagsan. Origin: Chinese zeng, " already," in Mandarin Pseng or san, " scatter,'* "separate." For the negative conjugation there is a form in I, as in holdal ugwei irebeu, " not having sold it he came," that is, "he came without having sold it." There is also a past negative participle in ge, as in iregedei, " he has not come." Mongol Adverb. In the grammar of the Turanian languages, the verb, substantive, adverb, and conjunction, are imperfectly- distinguished. It was in the Indo-European system that adverbs and conjunctions first became indeclinable, and the verb began to lose its character as a noun. It was only by gradual steps that the eight parts of speech could arrive at the point of clear separation from each other. Mongol grammar presents us with a multitude of adverbs and conjunctions in the form of nouns and verbs. Much light is thrown by this part of Turanian accidence on the adverbial forms common in European speech. To show this, the following case suffixes of Mongol adverbs will be a sufficient proof. Locative suffix dor, "in," "at." Eastern colloquial de. Ende, " here," tende, " there." As ende is good book Mongol, the colloquial de may be fully as old as MONGOL ADVERB. 241 the ordinary book locative dor. Greek oIkoOl, "at home/' rj&Oi,, "in the morning," ev6a, "here." English yonder^ Suffixes to express motion " towards," dor and de. In Chinese tau and ti. Mandarin chl. Ende, "hither," t'ende, " thither," hande, " whither." Here the coinci- dence with the English ther is remarkable. The Anglo-Saxon forms are hider, pider, hvider, " hither," "thither," "whither." The old Norse forms are he"%ra, pa?6ra, hvert. The old Greeks used Se, as in akaSe, "to the sea," Oavcurov Be, "to death." In evOdBe, " thither," " hither," we have the locative suffix dor in 6 a, and that of motion towards in Be. In aXkoae, " to another place," a sibilant has taken the place of d. Suffix to express motion from. Mongol e&he, collo- quial ese. Hanasa, " whence," enese, " hence," tendese, " thence." The book forms are hamigasa, " whither ?" t'ende ec'he, " from thence," ende ec'he, " hence." The Manchu has c'hi as the suffix for "from," and the Turkish dan. English whence, hence, thence. Can the English have retained the suffix ce by tradition from an old Turanian language ? This question is difficult to answer, because the Anglo-Saxon forms were hvonan, henan, ponan. It may have been through the Danes, for the old Norse had hvaftan, he^San, pa&an, for "whence," "hence," "thence." Latham says, 1 "The ce in ' hence/ ' whence/ ' thence/ has still to be satis- 1 The English Language, vol. ii., p. 320. 16 242 factorily explained. The old English is whenn-es, thenn-es." The old Norse Kan and Greek 6ev being Turanian, may not the English ce be inherited from a Danish dialect, which has not transmitted a literature, and thus also be Turanian ? The Turkish locative suffix dah is the same as the Mongol dor. The Sanscrit atra, "here," tatra, "there," kutra, " there," have nearly the Mongol form. Instead of following Bopp in tracing the origin of the suffix tra to the comparative suffix taken instrumentally, I would suggest that it is better to see in it a Turanian suffix dor, as now explained. Compare the Latin citra, intra, and (without the r) quando. The Greeks said evda and ivravOa, "here," and avr60i, "in the place where he was," the old Hindoos kadd, " when ?" tadd, "then," and yadd, "when." The Zend had had ( a t "here," the Slavish kogda, "when ?" and togda, "then." The Mongol has heje, " when," and this is equivalent to hede. The Greek has ore, rore, " when," " then." The suffix in all these forms may perhaps be traced to one origin. It is ultimately a demonstrative and interro- gative pronoun, and is the same with the Turanian locative in dah and dor. "With the forms when, ivannt quum, before us, there seems no reason to look else- where. Bopp, however, finds, as he thinks, in the da of kadd, a contraction from diva, " by day." l Perhaps the forms here, there, dar, thar, her, hvar, etc., have 1 Vergleichende Grammatik, § 423. MONGOL ADVERB. 243 this source also. The t may be omitted and the r left. The Sclavonic and Eussian gdye retains the radical ga, "what," in the initial g, and the Turanian suffix dor in dye. The suffix appears in podii, "under," mezhdu, "between." The Greek derivative suffix Bov, indicat- ing the manner of an action, is probably of the same origin : ^va^avBov, " openly," avrocr^eBov, " near at hand." This Bov is often Ba in Homer, as in OB. III. 221. Ov yap irca Idov a>Se dzobs ava(pavSa (piKevvras, 'fls Keivij) amcpavSa TrapicTTaro IlaWas '&9i}vi). "From a place" is so frequently in western languages expressed by dan, or equivalent forms, that we are com- pelled to regard the Turkish ablative suffix in dan as in this instance preserving a very important old Tu- ranian type. The Greek iroOev, " whence," corresponds to the Sanscrit kutas, and we may regard the Sanscrit s as altered from an older n. The Latin has coelitus, which Bopp identifies with svargatas, " from heaven." He also finds in the Sclavonic suffix du, "from," the Armenian ti, and the Gothic thro, variants from the same primary form. The common ancient suffix for "from," in the Anglo- Saxon and German was nana or nan. Latham quotes from Grimm the Old High German hivanana, Old Saxon hwanan, Anglo-Saxon hwonan, all meaning " whence." The equivalents for " thence " and " hence " are simi- larly formed. We find in one of the Dravidian 244 china's place in philology. languages an ablative suffix which may explain this nan. The Malay alim has ilninna for the ablative, as in mala-y-ilninna, "from the mountain/ ' where mala means "mountain." We have not the opportunity of examining old types of the Turanian family. We must await the decipherment of Persian cuneiform inscrip- tions for further light on the subject of these remark- able resemblances between the adverbial suffixes of the Turanian and Indo-European languages. The Dra- vidian case suffixes may perhaps be regarded as having been in use for at least two thousand years, for the Tamil writing is based on the Devanagari of the monuments. Hence the Dra vidian languages were the first of the Turanian family to be committed to writing out of Persia. They were written before the Japanese or the Mongol. Any Dravidian case suffixes, therefore, which happen to agree in form with those of European languages, may easily be of very great antiquity. Some examples of Mongol syntax will be here given with parallel examples from the Chinese language. Adjectives precede their substantives, and adverbs their verbs. Mongol alt' en gerel, " golden light," Chinese kin kwang ; Mongol saihan yaba, " walk carefully," Chinese hau hau er ti tseu ; Mongol hamt'o echine, "we will go together," Chinese t'ung k'ii or i k'icai er c l hu. The nominative begins a sentence. Then comes the object of the verb. The verb stands last. Mongol hi teri alaba, "I killed him," Chinese wo sha Hau t'a. MONGOL SYNTAX. 245 The Chinese verb precedes its object. But the Chinese order is not like the Mongol invariable. If an auxili- ary particle be employed, the verb may stand last. Thus pa, " to take hold of," may be used to vary the order. Wo pa t'a sha liau, literally, " I taking him killed finished." This is something like the inaccurate English, " I took and killed him." Adjectives may stand in the predicate without a substantive verb, and when a comparison is made, they may take a comparative or superlative force without its being necessary to prefix adverbs. Mongol uge bugdege sain, " his words are all good," Chinese hiva tu hau; Mongol oseg bugdege t'odorahai, "the letters are all in their right places," Chinese ts'i tu wen t'o. The law of arrangement in the two languages is pre- cisely the same. Mongol hoyer yagomanu doVora ene sain, literally, " two things amidst this good," that is, " of the two things this is the better, Chinese Hang yang tung si che ko hau, literally, " two kinds things this one good." The comparative force is conveyed in the same manner in both languages, and that by position. The duplication of words, to give a plural to nouns, and to denote succession in time and place, occurs frequently in Mongol and Chinese. Mongol nig nig ere hamoge ireksen, " one by one all came," Chinese yi ko yi ko tu lai liau. Literally, "one" (numerative), "one" (numerative), "all come finished." The Mongol suffix 246 china's place in philology. ere or er or yar, is frequently appended to nouns and adjectives to make adverbs. It is probably the source from which the Latin adverbs in er, as libenter, in- stanter, have derived their last syllable. Mongol c'hag c'hag wei wei, " time after time," " generation after generation." Chinese sh'i sh'i tai tai, " age after age," " generation after generation." Compare in Latin quis- quis, " whosoever." These and other peculiarities show that a remarkable resemblance exists between the syntax of the Chinese and of the Mongol languages. The Tibetan in placing the adjective after its noun goes away further from the Chinese than does the Mongol. In the conjugation of the verb, and the absence of gender, the Mongol is nearer to the Chinese than the Tibetan, which pre- fers Semitic analogies. Thus in the order of succession perceptible among verbs when standing in juxtaposi- tion, there is a clear likeness to the Chinese. The order is that of time. Mongol t ( ere haireju ireksen, "he is come back," Chinese t'a hwei lai Kau. T'ere, " he," is the same word as t'a with the suffix r. Hwei, " return," is the same word as haireju, to which r was first affixed, and thenyw, the mark of the gerund. Lai, "come," is perhaps the same word as ireksen, the past participle of irehu, " come," here used as a past indicative. CHAPTER XI. Malayo-Polynesian.— The Malay the Type of a Distinct Family. Alphabet and Syllable. — Polynesian Syllable Based on the old Chinese Syllable. — Effect of Marine Climate on the Malayo-Polynesian Syllable.— Continental Origin of the Polynesians. — Connexion of Siamese and Malay. — Post- Position of the Adjective and Genitive. — Pronouns.— Case Particles. — Semitic Principles. — Chinese Influence on Polynesia. — Pronouns. — Verbal Directives. — Comparison. — Arithmetic. — American Languages. — Their Mixed Character. — Three Elements of American Population. — Polynesian Civilized Immigration. At the extreme south-east of the continent of Asia the Malay and Polynesian area presses upon that of the Himalaic races in the peninsula of Malacca, and meets the Chinese in Formosa. The Malaysian and Polynesian system presents to view some remarkable points of resemblance to the Chinese and Himalaic types. It is on this account that a brief chapter on this system is here introduced. Crawfurd has condemned the opinion of Marsden, Wilhelm von Humboldt, and Sir S. Raffles, that the islands of the Indian Ocean and South Sea, from Madagascar to Easter Island, are peopled by a single race. He remarks that the population of these islands 248 china's place in philology. consists of brown men with lank hair (Malay), of sooty men with woolly hair, and of brown men with frizzled hair. The first of these three, the Malay race, extends over the Sandwich, the Fejee, the Society, and the Friendly Islands, with the Malayan peninsula and most of the islands of the Asiatic archipelago. The Malay language cannot be regarded as Indo- European, because, as F. Miiller has shown in a criticism on the view held by Bopp, it forms deriva- tives by prefixes, and not by suffixes. From tidor, "to sleep/' is formed, by means of the prefix per, the word per-tidor-an, " a bed." In the Tagala dialect of the Philippines, from guntin, "shears," is formed, by the insertion of um, the word g-um-untin, "to cut with shears." Max Miiller inclines to regard the Malay as a Tu- ranian language, and as especially allied to the Siamese. But there are some strong objections to this very extended use of the word Turanian. To class the Siamese with the Mongol and Japanese is inconvenient, because it is a monosyllabic language with tones, and like the Chinese places the verb before its object. The word Turanian can be suitably limited to languages which form derivatives by polysyllabic suffixes, make use of case endings, place the verb at the end of the sentence, and have a certain system of rules for the use of vowels. It is better to regard the Malay as the type of a MALAY ALPHABET AND SYLLABLE. 249 separate family, as is done by F. Muller. The agree- ment between the Malay and the Siamese is indeed remarkable. The adjective follows the substantive, the genitive follows the nominative, and the demonstrative pronoun follows its noun in both languages. The per- sonal pronouns are also alike. But the non-existence of tones in the Malay, its polysyllabic character, and its entirely new series of numerals forbid our classi- fying it as one with any member of the Himalaic family. The alphabet of the Malay family is rich in letters, and in this respect resembles the Himalaic, and old Chinese, except in the want of aspirates. A Dravidian influence is visible in the cerebral series t, d, n. The surd series k, t, p, s, is found both in the Malay and in the eastern or Polynesian group, but the sonants g, d, b, of the western branch (the Malay), are wanting in the eastern. There is a resemblance to the triple-branched Turanian system in the use of s and the want of sh, and to the Japanese and Dravidian divisions in the absence of the aspirated forms of k, t, and p. The simplicity of the Malayo-Polynesian syllable shows the antiquity of the system to which it belongs. The initial consonant is usually single, and is never followed by another consonant, except sometimes by r. An initial sp, st, for example, would be impossible. In the western division of these languages, k, t, p, ng, n, m 250 (as in old Chinese), terminate syllables. Also s, h, r, and / (which is not true of the old Chinese), are some- times found at the end of syllables. The dissyllabic character of the roots in Malay reminds us of the Semitic system. "All monosyllabic roots, with the exception of some pronominal stems and particles, are shortened from dissyllables. All words of more than two syllables have become so by phonal additions to the dissyllabic base." 1 The possibility of terminating syllables with con- sonants extends eastward to the Caroline Islands. In the dialect of Ponape, sixty degrees east of Penang, and in nearly the same latitude, syllables are closed by consonants, as in the Malay. 2 In the East India Islands consonants are allowed to close syllables, and the letters used are the same which close syllables in the Turanian and old Chinese systems. The Polynesian dialects extend south-east from the eastern termination of the Caroline Islands for seventy degrees. Here the syllables are never closed by consonants. The peculiarities in the formation of the syllable in eastern Asia are adhered to in this respect through about half the longitudinal extent of the immense island group, which reaches from Sumatra to Tahiti. Through the Australian dialects the eastern Asiatic system is still 1 F. Muller, p. 324. 2 Grammatical Notes on the Language of Ponape, by L. H. Gulick, M.D., Missionary on that island. MALAY PHONETIC SYSTEM. 251 adhered to, but the final consonants are limited to ng, n, m, I, r. Australia then has, by a process of decay, lost the finals k, t, p. In the great cone of islands whose apex is twenty degrees east of Tahiti, and whose base is planted in the one case on the mainland of Australia, in the province of Queensland, and in the other on Ponape, in the Caroline Archipelago, the final consonants have all been lost from the syllable. In the East India Islands the finals k, t, p, are used in addi- tion to those of Australia, and the resemblance to the eastern Asiatic syllable there becomes complete. The same contrast exists in this respect between Malaysia and Polynesia as between Mongolia and Japan. The Japanese, living in a soft and luxurious climate, have dropped the final consonants, which in the cold and bracing climate of the Gobi plateau have been retained by their Mongolian cousins. So, also, the Malay syllable bears the same relation to that of Australia that the old Chinese syllable does to the modern. The Malay system admits k> t, p, at the end of syllables as well as ng, n, m; and this is also true of the old Chinese system still retained in the dialects of Amoy and Canton. The Australian system, like the modern Mandarin of China, at the end of its syllables only allows nasals or the letter r. 1 In view of these facts, it may be concluded that the old Chinese closed syllable, with the finals Jc, t, p, 1 F. Miiller, p. 247. 252 CHINA S PLACE IN THILOLOGY. ng y n, ?n, lies at the basis of, and was formerly found in, all the languages of Austral- Asia and the South Seas. Further, the additional finals, I, s, there existing, are such as occur in the Himalaic and Turanian systems. The want of final consonants in any of the Oceanic dialects may be accounted for by phonetic decay. They may have been simply dropped, or they may have taken vowels after them, and so become initial consonants to supplemental syllables. In addition to the question of the finals, there is also the question of the initials. Neither the Australian nor the Polynesian dialects have the letters g, d, b. Yet they have sounds something like them, which, after careful consideration, the missionaries and others busied in collecting data respecting the native lan- guages usually agree to write k, t, p. In the Malay region only do the letters g } d, and b occur in their full distinctness. The conclusion again seems forced upon us, that secular decay has wrought destruction in the alphabets of the more distant dialects, while the Malay, more recent in the time of its migration from the continent, has better preserved its ancient form. As in China it is only in the old middle dialect that the sonant series of the old language is well retained, so this relic of the primeval language of mankind finds a refuge in the Malay area when abandoned by all the more southern and eastern modes of speech. All this is in full harmony with the view CONNEXION OF SIAMESE AND MALAY. 253 that the Malay, and other Oceanic races of the same sisterhood, proceeded from Asia south-eastward, just as the Chinese (who drove the Miau tribes before them into the mountains of Kweicheu), and other races of Eastern Asia, all show signs of western origin. The Polynesian and Australian alphabets, now predomi- nantly surd, were originally, as it would appear, sonant, but the Malays left the continent with the double series of letters found in Hebrew and old Chinese. It seems premature for F. Miiller to say, " So much remains certain, and will never by the most brilliant and most trenchant reasonings be disproved: the Malayo-Polynesians are connected with no Asiatic people." In the discussion which has been originated by Max Miiller's views on the intimate connexion existing between the Siamese and Malay languages, — and in which Pott and F. Miiller have placed themselves in opposition to that philologist, — it seems to me that reason is on the side of the Oxford professor. The resemblance is in many respects most marked. Both languages are clear of all trace of the great Turanian inversion, by which the verb is placed at the end of the sentence, and in this they are at one with the Chinese and Semitic systems. Consequently the case- marks are prefixes in Malay, as in the Siamese and its sister dialects of the eastern Himalaic family. Thus, in Malay the order is as in English, disabrang sungei, 254 "beyond the river," buka pintu, "open the door," diJantei, "on the floor" (diatas, "upon," lantei, "floor^). The absence of the distinction of gender and number in nouns places the Malay in agreement with the Siamese, Chinese, and other monosyllabic languages. Thus, orang Halayu is "a Malay man," or "the Malays"; orang being "man" or "men," just as in Chinese ta Jen is " a great man " or " the great men." The same principle underlies the Turanian languages, as in Mongol: ende nei himn ho aina ("here's man all fear "), " the people of this place are all afraid." Here hwun, in the written language Jc'umun (Latin homo), is plural, though constantly used in the singular. That in such a case the regular plural form ending in d is not used is proof that the root without a suffix is, as in Chinese, either plural or singular. So in the Hebrew rW niK£ WISH Hhamesh meoth shana, "500 years," shana is in the singular, although two plurals exist, viz., shanoth and shanim. Even in English some nouns are undefined in regard to number, e.g., fish, which is singular or plural. But such examples are exceptional. A remarkable resemblance of the Malay to the Siamese and other Himalaic tongues lies in the post- position of the adjective. This principle characterizes all the Himalaic and Polynesian languages, and goes far to cut them off from any thoroughly intimate connexion with the Chinese and Turanian systems. The Semites placed the adjective after its noun, and POST-POSITION OF THE GENITIVE. 255 they once occupied Persia. Persia is the western neighbour of Tibet. May not this post-position of the adjective have passed from the Semites to the Tibetians, Siamese, Malays, and Polynesians ? Or did the Semites, at some date anterior to the Aryan conquest of Persia, borrow this peculiarity from the ancestors of the Polynesians ? In the parallel principle, the post-position of the genitive, the Tebetians have, under Turanian influence as it would seem, gone out of the line. But with this exception, the Semitic, Himalaic, Malay, and Poly- nesian systems, all agree in placing the genitive after its noun, that is, the possessor after that which is possessed. Thus the same powerful Semite influence which introduced this idiom into European languages has also made itself felt in all the eastern Himalaic languages, and in the Oceanic archipelago eastward to the Sandwich Islands, and south to New Zealand. Another very strong proof of consanguinity between the Siamese and the Malay is found in the pronouns. The three personal pronouns are in Siamese, ku, meung, mon, in Khamti, kau, man, man, in Malay, ku, mu, na. The Chinese nga, "I," appears in Chinese dialects under the forms gwa (Amoy), ngu, nu (Kiangsu), wo (Man- darin). We are not therefore surprised to find it nad in Mongol, ku in Siamese, ego in Greek and Latin, ku in Malay, natoi in Australian, hau in New Zealand, wau in Hawaii. 256 The Malay pronoun for the second person is mu. It is found among the Miau tribes in south-western China in the form mu, and among the Li tribes of the Hainan mountains under that of mow. The origin of this pro- nominal form for the second person, which is found only in the eastern Himalaic and Malay area, and does not extend into Polynesia, may be traced with great probability to an honorific use of the third personal pronoun in m. This pronoun is found in Siamese under the form mon, in Hainan as pun, in the Miau dialect as men. These forms all mean "he." In the Chinese language, the indefinite pronoun meu, "a certain person," is probably the same word. The Chinese and Semitic interrogative pronoun ma, "what?" may also be referred perhaps to the same root, for as the relative has often grown out of the interrogative, so the interrogative has quite as fre- quently grown out of the demonstrative. Thus the order of origin would be in Latin hie, quis ? qui, and in English he, who? who. The Malay pronoun for the third person is na. This we may identify with the common Chinese demonstra- tive na, "that," "which?" and with the Siamese demonstratives ni, "this," non, "that." The Malay demonstrative " this " is mm. This close similarity in the personal pronouns be- tween the Malay and Siamese does not extend to the Polynesian dialects, nor to all those of the Malay area. SIAMESE AND MALAY PRONOUNS. 257 The first person in k (ku, ko, ki) is found indeed in all the Malay dialects, including that of Madagascar. It also prevails in the form ahau and ku in the Tonga language and that of New Zealand. The second person in m is used in Borneo, Java, and the Philippine Islands, but not in the more distant members of the Malay group ; nor is it anywhere employed in Australia or Polynesia. Thus much I gather from the examples collected by Professor F. Miiller, p. 342. The argument from identity in pronouns is much stronger than F. Miiller allows. The example he adduces to show that it is of little worth (p. 278), is that of the existence of similar pronouns in the Ural- altai 1 and Indo-European families. But the identity of the pronouns in these two linguistic stems is a strong support to Professor Max Miiller's view. In the Tartar and Indo-European families, as has been shown, the striking resemblance noticeable in the pronouns is also found in the substantive verb, in the adverbial case suffixes, in the tense suffixes, in the gerundial and participial forms, in the signs for the plural, and in a large number of common roots. Hence, when the philological inquirer finds the pro- nouns identical, he may expect to discover other agreements revealing themselves on examination. The 1 De Castren, the Finnish philologist, proposed this term for the Tartar, Siberian, Finnish, Esthonian, and Hungarian languages. 17 258 existence of a second personal pronoun in m, over an area of 25° in longitude and 35° in latitude in south-eastern Asia, is parallel, on a smaller scale, to the existence of the first and second personal pronouns in m or b, and t or s, over the Ural-altai and Indo- European area, and affords good ground for expecting that many other fundamental similarities will be found to exist. The law of position for case particles is similar in the Chinese, Semitic, Siamese, and Malay. Preposi- tions are used for the purpose of indicating case. " To a place " is in Chinese tau, to, ti, chi, Semitic la, Siamese p'eni, Malay datan, Tibetan la, Mongol de. " From a place " is in Chinese zi, zung, or ts'i, t'sung, Semitic min, Siamese de, Malay deri, Turkish dan. " With " is in Chinese dung, t'ung, Semitic DJ7, gim (Latin cum), fiNi, eth (English with. Compare the sense of " with " in " withstand " with the meaning " against," which is the not uncommon force of the Hebrew eth). Siamese kab (connected with the Chinese gip, "to arrive at," "along with"), Malay dengan. " In " is in Chinese tsai, and as a suffix, li, tung, chung, Semitic be, Siamese mai, Malay di (in some dialects ri), Mongol dot 1 or a, " Towards " is in Chinese hiang, and in Malay ka. "By means of" is in Chinese i, tsiang, tan, yung, Siamese dwa, Malay ulih, oleh. SEMITIC PRINCIPLES IN MALAY. 259 Out of these six instances, there are five in which the Chinese and Malay approach each other, viz., " to," " with," " in," " towards," and " by means of." There are three instances of agreement of the Malay with the Siamese, "from," "in," and "by means of." The Chinese initial h is to be regarded as k, and ch as t The Chinese / often comes from an earlier d. The final ng is commonly lost, and n occasionally. The paucity of instances in which the Malay and Siamese approach each other in the use of prepositions is probably owing to a want of the means of com- parison. Jones^ "Grammatical Notices of the Siamese Language " is very brief, and contains few words. 1 The influence of the Semitic family extends, in regard to laws of position, into Malay and the Oceanic dialects to the eastward of the Malay Archipelago, but in regard to roots it seems to stop with Tibet. So the Mongols have some Semitic principles, as the plural in d, but very few Semitic words. In addition to the post-position of the adjective and the genitive in Malay, that of the demonstrative pro- noun constitutes another striking feature. This recalls the favourite Hebrew idiom, which places the demon- strative with the article after the noun, e.g., hammakom hahu, " that place." The article ha is here prefixed to makom, "place," and to hu, "he." The Malays say, Pikulkan peti ini, " carry that box." Kan is the tran- 1 Pallegoix' works are copious, but I have not access to them in Peking. 260 sitive or causative suffix to the verb pikul, " carry." Ini is the demonstrative pronoun " that." This idiom is in both languages only a particular case of the post- position of the adjective. The repetition of the article in Hebrew indicates that the order of the words is in such cases not the natural one. For, otherwise, why is the article repeated ? It may, then, be concluded that in the order of nature the adjective precedes its substantive ; and when the converse takes place, there is an inversion of the natural order. The Semitic principles occurring in the Malay tongue have been adverted to, while its resemblance to the Siamese has been more fully described. I shall now illustrate the connexion of the Polynesian family with Chinese, making use of the dialect of Ponape, in the Caroline group, as described by Dr. Grulick. The gender of nouns is distinguished by the use of special words attached to the nouns. In Chinese these words stand first. In Polynesia they come after. In regard to the number of nouns, it is in Chinese and the Polynesian languages known from the context, e.g., by that of the accompanying pronoun. In the Ponape dialect certain numerative particles are used with nouns. Thus, men follows animated objects, tun is used with bunches of fruit, urn with yams and bananas, pot with plants, sticks, and canoes. The same principle exists in Chinese and in Siamese. CHINESE AND POLYNESIAN CLASS- WORDS. 261 Thus, the Siamese say, luk reu sang k'on, " two boat- men." Here sang, "two," is the Chinese shwang, " a pair." K'on is the nnmerative for " men." Luk reu is " boatmen." Reu is " boat." This in Chinese would be shut sheu Hang ho, " water hands two." In the combination Hang ko, "two," ko is the numerative of " men," shui sheu is " sailors." The numerative is necessary after numerals by a common linguistic law. The law of position is, how- ever, somewhat different in the examples. The Chinese say " water hands," and place the specific term before the generic. The Siamese and Polynesians prefer to say "men of the boat." In English we can speak in either way, but the order of nature is to place the specific word first, and there is something artificial about the inversion. When we say " sea birds," we adopt a mode of speech in genuine accordance with the spirit of our language. "Birds of the sea," on the other hand, is an expression belonging to a borrowed poetical vocabulary which is ultimately Semitic. The Polynesian languages have a double, series of some pronouns. When in addressing a person the speaker includes himself with the person addressed under one pronominal designation, it is called the inclusive pronoun. The Ponape dialect has a dual pronoun kita, "we," which is inclusive. So in northern Chinese tsa-men, "we," is distinguished from 262 wo-men, "we," by the circumstance that tsa-men in- cludes the person addressed, while wo-men does not. The origin of this inclusive pronoun for the first person is in Chinese probably the reflexive ts'i, in old Chinese zi, and in Latin se. The Chinese write it Ug tsa. This form is compounded of keu, " mouth," (referring to its being a common locution), and ts'i, " self," indicating that the makers of this modern logograph felt that this was the etymology of the word. We may suppose kita to be the other Chinese reflexive pronoun hi. The Chinese of books has no inclusive pronoun, as distinct from the ordinary per- sonal pronoun, but it may have existed in an ancient unwritten colloquial, and may have descended to the Polynesians from a common source. The Polynesian personal pronouns agree nearly with those of China. Of the first I have already spoken. The second is in Hawaii oe, in Tonga hoe, in Ponape kowe, in New Zealand koe. These I take to be the Chinese ni, u thou," "you." Old Chinese has f|f nu and U| ngi, and the initial ng is easily interchangeable with k and g, as in the Turkish ugli, " son," Chinese ngi, Mongol k'u, "begun." Hence the Polynesian form in k is accounted for. But ng as an initial is often dropped, as in the Chinese wo, " I," from ngo, the Hebrew Ayin from an original ng or g, etc. Thus the Hawaian oe, " thou," is also explained. With regard to the third personal pronoun, i 3 ya, and na, are the CHINESE AND POLYNESIAN PRONOUNS. 263 prevailing forms. They agree with the Chinese i, "he," "that," and na, "that." We find, therefore, the European pronouns Ego, ich, vos, you, is, Hie, existing, not only in China, but also in the most remote Polynesian languages, at a distance from England of half the circumference of the globe, and yet capable of recognition with the help of the connecting link supplied by the old Chinese. The verbal directives in the Ponape dialect are another example of strong Chinese influence. F. Muller has not mentioned them in his otherwise full and valuable notices of Polynesian grammar. Nor are they referred to in Jones's "Notices of Siamese." It seems to me that they must exist in all the Polynesian dialects, as in that of Ponape. We have them in English in such expressions as go up, go down, go in, go out, where they are adverbs following verbs, and limiting the direction of the action in space. Hence the name verbal directives. In Chinese they are verbs in apposition. Tso hia is " sit down." Tso hia lai is also to " sit down," and consists of three verbs in apposition arranged in the order of time, thus " sit- down-come." Take another example, tseu tsin lai, " walk in," or " he walks in," literally, " walk- enter- come." Here the law of arrangement according to time is manifest. We may expect, then, to find verbs in all the English adverbs which are connected with verbs in this way. Thus " through," in the expression 264 china's place in philology. " go through," in Grerman durch, is in Chinese t c ok or feu, "to pierce through." The Chinese say of a soaking rain that it has Ma Peu liau, "fallen thoroughly." Here t'eu means that it has penetrated the soil to the full depth required by the farmer. We find in the Ponape dialect the following pre- positions and adverbs used as space directives after verbs. La, "from," ta, "upwards," to, "downwards," we, "away from," i, "going off," long, "in," ung, "to," Jung, " from,'* pena, " together," pqjung, " separate." Thus wa la, " take from," tau ta, " climb up," ko ti, " come down," ko to, " come hither," ko long, " go in," ko ive, " go away," ko ung, " go to," ko jung, " go from," ko- pena, " go together," ko pajung, " go separate." Among these words may be noticed the Chinese zung or ts'ung, "from." The initial j is pronounced in Ponape like dj or sh, and is hard to write down. The word to, "down," is the Chinese toi, te, or ti, the Mongol dotai, " downwards," and the English " down." The Chinese shang, " above," " up," is not improbably the Ponape ta, "upwards," for in Cochin- Chinese t is the common equivalent of the Chinese sh. Thus in Morrone's Lexicon Cochin- Sinense the Chinese sound shing is detected in thua, " to conquer," " to remain over," " to abound." The Chinese k ( e shang, " travel- ling merchants, " is the basis of the disguised kach thua, having the same sense. Now we may naturally SEMITIC PRINCIPLES IN COMPARISON. 265 expect to find that in a matter of this kind, what is true of the eastern Himalaic languages will be true of Polynesian languages. They will bear a similar re- lation to the Chinese. I may add that among the verb auxiliaries in the Ponape dialect is the causative prefix ka, kau, or ko, which corresponds with the Chinese ko* or kiau, " to cause," used commonly as a causative in modern dialects, and identical probably with the Latin causa. Such clear marks of consanguinity between the Chinese and Polynesian languages must be taken as proof, in opposition to F. Miiller, that there is no room to doubt their coming from one source. The laws of position and a common vocabulary connect the speech of the Pacific islanders with that of Siam, Cochin- China, and China. Where the law of position in the Himalaic type differs from that of China, Polynesia connects herself with Himalaya, and here, as it appears to me, is seen the action of a Semitic principle. It is worthy of remark that the Hebrew mode of comparing by the use of the preposition " from," min, is parallel to that of Ponape. In the Hebrew &y]fo pinp " sweeter than honey," |j& min, " from," is inserted after mathok, "sweet." The Ponapean says, met kajalel jung meteu, " this beautiful from that," in correct English, "this is more beautiful than that." The same idiom is found in the Tartar languages, as 266 CHINA S PLACE IN PHILOLOGY. in colloquial Mongol enese sain, "better than this," or literally, "this from good." Here the preposition becomes a post-position by the Turanian inversion, and sain, " is good," stands last as being the predicate. There appears to be little ground for doubting that the Semitic idiom is the older, and that both the Turanian and. Polynesian have sprung from it. So also our English comparative degree, formed with than, must be referred to the same origin. This little word, which has long gone seeking in vain for a plausible parentage, is no other than the Turkish suffix dan, "from." Latham says, "than is a variety of then; the notions of order, sequence, and comparison being allied." If so, then the final n of both words is probably the Turkish post-position ni in kani, " where." This means "in," "within," "at," and is like the Chinese net, nip, "within," as before remarked. It may have been originally a demonstrative pronoun. So the English as, Gferman als, and Greek «?, are perhaps the Mongol ese, which means " from " and " than." The Turanian form is asa or ese, according as the vowel in the noun is in the series a, o, b, or in the series e, u, u. In Sanscrit the root appears as the demonstrative sa, without the prefix. In Chinese it is demonstrative tsi, si, reflexive dzi, or prepositional dzi, "from." Here also it is without the prefixed vowel. In Latin the reflexive se also occurs without the vowel. The Turanian influence has been strong POLYNESIAN ARITHMETIC. 267 upon the Teutonic and Gothic portion of the Indo- European family, and has left its trace in the vowel initial of as anMals. The English z in as agrees with the old Chinese zi. The written Anglo-Saxon was swa. Anything nearer to it I cannot find in Yernon's Guide. ( The unwritten dialects, if known, would throw light on the form. There can be little doubt that the Persian ez, "from," is the same word. We have then the sonant form z in English, Persian, and old Chinese, and the surd s in German and Mongol. The modern Chinese form is ts'i. It appears then that in the comparison of adjectives the Ponapean dialect follows very widely spread continental models. So far from being a savage race originally, the Ponapeans, as their language shows, are an offshoot from the continent. In addition to the above instances of linguistic connexion with Asia, which might be easily increased by comparing, for instance, the demon- strative pronoun en, " this," with the Mongol ene, " this," it may be added that the Ponapeans count to ten, but beyond that number they become bewildered. Thus, ngavi is with them " ten of yams," but " one hundred of eggs or cocoa-nuts"; apuki," one hundred," (the Chinese pak) is "one hundred of men, trees, or yams," but "1,000 of eggs, cocoa-nuts, or stones." After centuries of isolation, Oceanic islanders lose the command of high numbers, and their value fluctuates or becomes lower in value. Thus, the Chinese man 268 (wan), "10,000/' retains its value among the natives of Samoa and Tonga, but when it reaches the Sandwich Islands it has already sunk to the value " 4000," and in New Zealand- it means "1000." 1 F. Muller, after comparing the names of number from one to ten of the Malay and Polynesian lan- guages, says, " From the comparison of the foregoing names of number, we plainly see that, widely as the languages which use them are separated from each other, they branched off at a time when the speakers could count at least to a hundred. This is certainly a proof of the not limited intellectual gifts and early development of these peoples." (Page 287.) I would go a step further, and say that this fact, regarding the numbers 100 and 10,000, proves deterior- ation. The Polynesians could formerly use a decimal arithmetic. Whether they have adopted a quaternary or quinary arithmetic, it is probably on account of long-continued isolation, which tends to produce bar- barism. The Australian tribes have already exhausted the arithmetical faculty when they have arrived at four and five. The word kauwul-kauwul means with them either "five" or "very many." With another 1 Samoa and Tonga lie between the Sandwich Islands and Xew Zealand, and, if the migration of the Polynesian islanders proceeded regularly by way of the Malayan archipelago, would be populated much sooner than those two more remote localities. In F. Muller's triple grouping of the languages, as the Malay, Polynesian, and Black-race groups, the islands mentioned all belong to the second. AMERICAN LANGUAGES. THEIR MIXED CHARACTER. 269 tribe punku, " four," is also " many," and punhu kalan, " five," is " very many." Their ancestors when they left Asia could probably all count to ten. Are not the ten fingers the proper foundation of arithmetic ? All human races would still practise it but for the degrading effects of long- continued isolation. "Where the arithmetical faculty is weak, the names of number easily and rapidly change. The multipli- cation table would be soon lost to civilization if left in the hands of the dunces. It is the bright in intellect that preserve society from lapsing into barbarism, for they transmit to coming generations the treasured discoveries of the past. Among Oceanic islanders degradation is inevitable until they are visited by the light of Christian civilization. But easy as it is to lose the names of number, and especially those of high numbers, it is not likely that the traces of ancient knowledge will entirely disappear. Yestiges wanting in one island will be found to exist in another, and a wide recension may be expected to restore, piece by piece, the image of the buried past. The languages of the American continent form a portion of the field to be investigated before the position and relations of the Polynesian system can be accurately determined. As Turanian languages border on North America at Behring's Strait, so the Polynesian dialects approach both North and South America by the ocean. In the valuable collection of 270 china's place in philology. Lord's Prayers in more than 600 languages and dia- lects, published by the Imperial Printing Office at Vienna, I have searched for dialects which by their syntax might be recognized as exclusively Turanian or exclusively Polynesian. None occur. The princi- ples of arrangement are so mixed and so evenly balanced that the principles of both families seem to be everywhere in operation. For example, in the Delaware language, alluded to in Cooper's romance, " The Last of the Mohicans," while it has case suffixes and the genitive before the nominative (Turanian), it has, on the other hand, the verb before the accusative and makes use of many prepositions (Polynesian). In the language of the Dacotahs, between the Mississippi and the Rocky Mountains, prefixes (Polynesian) pre- dominate over suffixes (Turanian). Among the Central American languages the Mexican is important. In no cal " my house," i cat, " his house," the order is Chinese and Turanian, as are the roots. In the Sandwich Islands, and other parts of Polynesia, hale means " house " also, but there the pos- sessive pronoun must follow its word. In Nicaragua the adjective rigidly follows its noun, which is a decidedly Polynesian feature. The language of the Incas in Peru, in having the adjective after the substantive, is Polynesian, but in having case suffixes and the verb at the end, is Turanian. Thus, they said Mango Capac, while we POLYNESIAN CIVILIZED IMMIGRATION. 271 should say in English King Mango; and nocaicuman, "to us," where man is "to," and is the dative case suffix. In the language of the Caribs, whom the discoverer of the American continent found in the West India Islands, there are case suffixes, and "the verb pre- cedes the accusative. They distinguish the elder and younger brother by different words, as is done in all the Polynesian and Turanian languages. Their speech is classed with the South American division of Indian languages. We are warranted by these linguistic data in con- cluding that there was a Polynesian immigration from the ocean, and a Turanian immigration by the Aleutian Islands, and by Iceland" and Greenland, which united to form the population of the American continent. The influx of ocean tribes would be favoured by the former existence of extensive lands in the Pacific, now submerged. Chinese tradition speaks of a chain of large kingdoms stretching from Japan to California, through which Buddhism was zealously propagated. These notices, belonging to the fifth century of our era, should not be forgotten, though it is not safe to build much upon them. The Polynesian element was the more civilized, and to this must be attributed the main influence in the production of the civilization of the Aztecs and Incas. The Turanian element was the more simple, and to 272 china's place in philology. this may be ascribed the doctrine of the Great Spirit, and the other religious views of the less civilized tribes of North America. The Polynesian element prevailed most on the western shores of the continent. The forms of science and art, national polity and belief, found there by the Spaniards, agree best with those of Southern Asia. The Turanian mould of thought and belief extended itself rather along the northern and eastern portions of the continent, and exists among the Siberian tribes in a similar way. The modern Polynesians residing on a thousand isolated points scattered over the ocean, have lost the civilization they once possessed, and have not been able, on account of their insular position, to advance in the intellectual sphere, as did the Aztecs and Tncas, but their religious and mythological traditions point to India and Western Asia as their source. The tradition of a deluge and an ark follows the line of Semitic principles of language through the mountain homes of the Karens to the ocean, and proceeds by the Sandwich and other islands in the Pacific to Mexico. The belief in the divinity of serpents exists in the Fiji Islands, as it formerly did in the land of Montezuma. This is both Hindu and Babylonian, and seems to have sprung out of the narrative of the Fall in the Book of Gfenesis. Cycles in time terminated by a catastrophe are almost necessarily to be regarded as of Hindu or Chaldean origin. The Mexican belief in the Age of the Earth POLYNESIAN CIVILIZED IMMIGRATION. 273 (corresponding to the Satyayuga of Hindostan, and extending to 5206 years), of Fire, of Tempests, of Water, and of the present Age, 1 may be best traced to India and Babylon. Mr. Hardwick says in regard to the American traditions of the Deluge, " So nume- rous, and so extremely arbitrary, are the points in which those legends are now found to have approached the sacred story, that some affinity between the two is generally recognized, except where an archaeologist or schoolman is incorrigibly blinded by his love of system-building. Even the divines of Germany, beneath whose shadow every kind of mythic theory has sprung up with rank luxuriance, seem to have been almost reconciled to a belief that the traditions now and formerly current in America respecting some great deluge must have all been carried over from the old Continent.' ' As the proof from language proceeds side by side with that from historical and religious tradition, we are driven to the conclusion that the Polynesian and American races are Post-Diluvian, and of the same ancestry with ourselves. " Ought we not/' says l A. von Humboldt, "to recognize the traces of .a common origin wherever the cosmological ideas and first tra- ditions of peoples offer striking analogies even in unimportant matters ? " 1 Hardwick' s Christ and other Masters, part iii., p. 160. 1 Hardwick cites this passage in p. 164 from " Vues des Cordilleres." i 18 CHAPTER XII. The Sanscrit Language. — Sanscrit Eichness in Forms. — Its Principles of Development Based on Older Systems. — Alphabet. — Syllable. — Prefix of S. — Insertion of R and L. — Polysyllabic Word. — Declension. — Case Suffixes. — Plural. — Gender. — Comparison of Adjectives.— Pronouns. — Deriva- tive Verbs. — Personal Endings. — Tense Marks. — Potential and Conditional Mood. — Infinitive. — Participle. — Auxiliary Verbs. — Adverbial Suffixes. — Prepositions. — Compounds. — Laws of Position. — Zend Syntax. In passing to the Indo-European languages, the Sanscrit first claims attention. The remarkable com- pleteness of its grammatical forms has attracted the admiration of philologists. The same analytical genius which aided Panini in the arrangement of Indian grammar, many centuries after it attained its perfection as a language, aided his forefathers un- consciously in its gradual formation. The peculiar intellectual attributes of a nation are first recognized in the germ in their language, and afterwards in the fruit in their literature. Languages are rich, noble, and worthy of study in close proportion to the political and literary development of the people that speak them. The merit of Sanscrit consists in its richness in forms, and its orderly development. SANSCRIT ALPHABET. 275 The origin of the peculiar principles of the Sanscrit grammar must be looked for in the families of language which existed previously. Such signs of Semitic in- fluence as appear in Sanscrit may be due to an ancient residence in Armenia, or somewhere in that region, when they were neighbours to the Semites. The traditions of Sanscrit and Zend literature point to an old national home in Bucharia. Here the tribes that spoke these languages were in proximity to Turanian races, and on the south with the occupants of Persia and Aflghanistan, at that time probably speaking a Semite language. But as there was an ancient Turanian occupation of Asia Minor, the original Sanscrit type would also easily gather Turanian elements during a possible older residence west of the Caspian. Alphabet. The peculiar double development of the t series may be ascribed to Dravidian influence. The dental series, t, P 9 d, d ( , n, is that which Sanscrit has in common with western languages and those of Eastern Asia. The cerebral series, t, /', d, d', n, links the Sanscrit with the Tamil and its sister-dialects. Let it be considered that in the Tamil there are a dental, a palatal, and a cerebral n, and a dental, palatal, and cerebral r; that there are three t sounds, and two / sounds. As these varieties do not exist north of the 276 Himalaya mountains, they may be supposed to be due partly to climate, and to have existed already in Dravidian languages before the speakers of Sanscrit entered the Indian peninsula. The aspirated k, t, p, ch, may be traced to a Tartar origin. These letters in Mongol, Manchu, and Turkish are always aspirated. It is in that part of the world the normal way of pronouncing them. An unaspirated t would there be counted as d. Thus, in Manchu writing a dot on the right changes an aspirated k into g. The aspirated series, gh, dh, bh, jh, has perhaps been originated by the Hindoos, from an unconscious tendency to make the sonants as complete as the surds. The unaspirated surd series, k } t, p, ch, seems to have been formed by the common ancestors of the Indo-European languages from the older series, g, d, b. The Mongol gar, "hand," is in Sanscrit kara, and in Greek %dp. Thus Gfrimm's law is the Indo- European expression of a wider law embracing all the Asiatic families, by which unaspirated and aspirated surds are both formed from an older sonant series existing in Turanian, Semitic, and old Chinese. No family has ever been so creative in politics, in literature, in the arts, and in language, as the Indo- European. It was suitable that they should start on their wonderful career with a more perfect alphabet than had hitherto satisfied the wants of nations. The SANSCRIT ALPHABET. 277 Turanian alphabet was deficient in surd sounds. The Indo-Europeans developed them by the exercise of a powerful instinct, and thus succeeded in so widening the bounds of the alphabet as to adapt it for embracing the vast variety of new grammatical forms, and new names of things and actions, which Sanscrit, Greek, Latin, grerman, and English require. In this they appear to have been assisted by the Semites, who at a very ancient period added k, p, and t to the still older by g, and d. The Semites, however, never arrived at the • evolution of so copious an alphabet as their younger brothers, the descendants of Japheth. The Sanscrit ch corresponds to the Chinese h. In the Indo-European languages generally k has shown a tendency to change into ch. In Italian Cicero became Chichero. In English /cvpt,afcrj, or kirche, became church. In Russian castus, " pure/' is chisto. This law of change, belonging to all the languages, must have commenced before the separation into dialects. It does not affect the eastern Asiatic languages. Yery recently, however, it has made its appearance in Chinese. Thus, in Northern and "Western China king, "to honour," is now pronounced ching. The law of change which usually corresponds in Eastern Asia to that of k to ch in Europe, is that of t to ch. This exists alike in Chinese and in the Turanian languages. Example. Chitra, " paint," " wonder," Chinese him, git, "paint," k% gi, "to wonder," kicai, hat, "strange." 278 china's place in philology. The Sanscrit v corresponds to the Chinese w. In Chinese, for instance, Jg wet, wat, is " to place rollnd, ,, "inclose." .The round covering of a cart, a tent, a curtain, and a low circumscribing wall, are called wei. The Tamil has vat tarn, " circle," " revolution in an orbit," " halo." The Sanscrit has vad, " surround," vada, " circle " ; Latin verto, " turn," volvo, " revolve." The consonant /, wanting in Sanscrit, was probably also unknown to the old Turanian language, on which it was based, for it is not found in Mongol, or in the old Chinese. The Syllable. The inherency of a in all consonants having no other vowel mark coming after them means that the Sanscrit- speaking people lost the habit of ending a syllable with a consonant. There need be no hesitation on this account in ascribing to the mother- tongue of the Indo-European family a syllabary of which one characteristic was the possession of final consonants. The Sanscrit roots are represented by the Indian grammarians as ending in many instances with con- sonants. Also many syllables actually end with the consonants n, m, t, d, k, r, etc. Hence the law of Sanscrit grammar here referred to is not strictly true. It is certain, however, that the tendency of ancient Hindoo pronunciation was towards vowel endings, just as is found to be the case with the Japanese syllabary SANSCRIT SYLLABLE. 279 as compared with the Mongol. This may be due to the enervation consequent on change to a warmer climate. The Greeks and Latins had much fewer consonantal endings than the English and Germans now use. The Semites and Turanians agreed in introducing r, l y s, among the final consonants of their syllabaries. They were followed by the speakers of Sanscrit. When words belonging to these languages are compared with the Chinese roots, such finals seem to be phonetic additions rather than changed finals. The Chinese he, " black," in the old language kek, appears in Sanscrit as k&la and kdka, and in Mongol as hara or k'ara. The / and r are here introduced in place of a lost k. But it would be improper to say that k had become meta- morphosed into either of these letters. With t and d final the case is different. These letters have a natural affinity for r and /, and interchange is not uncommon. Thus, the Cochin- Chinese dat, "earth," suggests that there was once an appendage consisting of t or d to the Chinese jjj ti, in the old language da. The Sanscrit form is dhara and the Latin terra. The Hebrew arets and the English earth seem to be con- nected by a change from d to r, and the prefixing of the vowel a. The r of the Sanscrit and Latin forms may be changed from the old final d or t. That the connexion between the Chinese syllable and the Indo-European syllable is to be brought to 280 china's place in philology. light through intermediate Turanian links cannot be doubted, when it is observed that these final r's and l's, coming in place of the Chinese t, occur not infrequently in Indo-European and Tartar languages. Chinese sat, " scatter," " sow," Mongol sargigolhu, "scatter," Latin sero, "sow," Tamil sidaru, "scatter." One of the most striking differences observable between the Sanscrit and Mongol syllable is the pre- fixing of s to other consonants in the former. As this is a permanent feature in all the Indo-European languages, it must have originated before their separa- tion. It was probably an intensitive. The Semitic roots which appear to have received s or sh as a prefix modify their sense so as to be in harmony with the idea that the sibilant was intensitive. East of the Sanscrit and Persian there are absolutely no examples of an s prefixed to the root. In comparing western words, it is necessary, therefore, to strip them first of this appendage. Thus, stand, sto, umj/ju, in Sanscrit st'al, " to stand," st'&na, " a place," "situation," may be be referred to the Chinese equi- valent by removing s. The final t of the Chinese word |i) dat, " to tread upon," is found in the Sanscrit st'ita, " steady," and in the English stead, steady. The Arabic has ddsa, " tread," and the Hebrew nathan, " to place," where the n does not seem to be radical. Compare in Tamil tandu, tdl, "stand," Tibetan ten, " to halt." The root tat or dad, tan or dan, is probably INSERTION OF R AND L. 281 imitated from the natural sound of the foot striking the ground. Another example is jgj c'hu, old form tok, " to pierce," stechen, stick, stigo, stingo, sting, ari^co, Anglo- Saxon stechen, "to stick in," sticcels, German stachel, English stickle. The Sanscrit is stak. 1 A change of almost equal importance, as adding greatly to the number of syllables, was the introduction of r and I between the initial consonant and the vowel. Thus, krit, "to cut," krishna, "black," kri, "to do," in Mongol hadahu, "to reap," hara, "black," hihu, "to do." Compare cut, cutter, ccedo, Hebrew gadang, JH3 an( ^ ^3pj Tamil katti, "knife," and for kri, "do," the Chinese hing, old form gang, "to do," "to go," recollecting that the loss of final ng is a common circumstance in Chinese words. As an example of the insertion of I may be mentioned kapdla, " skull," also karpara, Latin calva, Sclavonic glava and golova, "head," Gferman Kopf, Haupt, "head." We have dropped the p in our word head, but the German restores it to view. Greek /cecfraXrj, Latin caput. The Chinese is tip kap, "head of a series," " shell of a tortoise," " coat of mail," " a cover," " to 1 Compare ap.vpva, myrrh, darben, " starve," nose, sneeze, pike, spike, as examples where the prefix of s has been so recent that it exists in some languages and is wanting in others. Observe also that sh is prefixed in Sanscrit and German, while English, Latin, and Greek refuse to admit it. Schmerzen, smart, Schmidt, smith. Compare amartts, "bitter," miide, " toil." 282 cover." In this last sense the Sanscrit has kub and kubh, "to cover," which may be compared with the Greek KpvTrrco, "hide," and KokvirTG), "cover." In Mongol we find hobc'Ms, " clothing," and habhan, " a covering." The Hebrew has !%$, "he covered," " expiated," and the Arabic ghufran, " pardon," ghayb, "hidden." The occurrence of KaXvirra) with a vowel preceding the inserted I shows how the syllabary may acquire a new extension. The monosyllabic root thus becomes dissyllabic without either prefix or suffix. Instances, however, of this sort of extension among European roots are comparatively rare. Another mode of extending the primitive syllable is to insert r and / before the final consonant, as in karpara, " the skull," from the root kap ; kart, " to cut," from the root cut. Compare the English work with the Latin ago, actus. The Polysyllabic Word. The monosyllable needed to be lengthened and endowed with a more perfect and beautiful form. Just as among the works of the Creator are found first ferns and mosses, and afterwards grasses and trees and all the rich variety of flowering plants, so the plain and unattractive words of the most ancient men were destined to expand into the ever-changing abundance and beauty of the Indo-European vocabulary. FORMATION OF THE POLYSYLLABLE. 283 With the expansion of the monosyllabic root into a polysyllable by prefixes, suffixes, and inserted letters, the subject of derivation is inseparably connected. Take an example from the Sanscrit vocabulary. The old English quoth is in Sanscrit kat 1 , "to speak." From this is formed kat'aka, " a speaker," by appending a demonstrative pronoun ka, the English he, and the Chinese gi, "he," ku, "that." The Chinese hwa, " words," " to speak," is the same word, the old form being gat. In Mongol helhu, "to speak," takes c'hi in place of hu, to express the agent. In helchi, "the speaker," or "he who speaks," the syllable c'hi is also a demonstrative, the Chinese t's'i, "this," and the Sanscrit sa, " he." Another suffix which presents itself is n, as in kat'ana, " saying," a neuter noun. In Mongol we have helen, occurring as one of those substantive forms of the verb which we call infinitive or gerund. The same suffix meets us in the participles, as in karin, " a doer," from kri ; gh&tin, "a killer," from han; sdyin, "a sleeper," from si 1 (Chinese shut). In these cases the word in n is either a noun of agency or a present participle. The English participle in ing, formerly in (Latham's English Language), limits itself to the sense of a participle and infinitive, leaving the expression of agency to the suffix r, as in lover, 2 loving. 1 "Williams's Sanscrit Grammar. 2 The suffix r for agency may be changed from s, as was is called in the west of England war. It may then be regarded as the demonstrative in s. 284 Other forms from kat are kat l angkatHka, "an interro- gator," kat'angkat'ikafra, " question," kat'angkatHta, " questioner," kat'aniya, " that may be told," kat'anta, "inquiry," kat' a, "word," "tale," kat'dnurdga, "atten- tive to what is said," kat'ika, "story-teller," kat'ita, ^ Among them the suffix ta or td, used of an agent or participially, is found in the Chinese gfjf ti, %j che (old form ta), and in the Mongol gerund or past participle heled. The Sanscrit chitra, "painting," "to paint," forms chitraka, "a painter." This word is lengthened into chitrakara and chitrakdra, both meaning "painter." Chitratala is "painted like a floor." Chitralikh and chitrakrit mean " painter." Chitragata is " painted." Chitrala is "variegated." Chitralekhd is a " picture." Compare with this family of words the Chinese hicei or gat, " to paint," the Tibetan skud, " to smear," " to mark," kud-pa-po, "a marker," "painter," kus-pa, " smeared," and the Russian chertit, " to paint," cherta, " a line," ocherk, " a line." The suffixes rag a, rege, are quite common in Mongol. So also are Jig, al, el, del, ga. Thus, t'erege, " a cart," is formed from the Chinese c ( he, formerly t'e ; c'hic'higlig, " a garden," is formed from tfhic'hig, " a flower"; ujel, "a mode of viewing things," comes from ujiliu, "to see"; sigudel, "judgment," comes from siguhu, " to judge." THE FORMATION OF CASES. 285 The Tamil derivatives from kdtu, " to kill," are kdtakam, kdtam, "killing," kdtakan, "a killer," kdtal, " act of killing," kdtei, " killing." It appears, then, that the Sanscrit derivative nouns are formed by appending syllables which bear a strong resemblance to similar syllables in Mongol. Forms are, however, more numerous in Sanscrit, which admits compounds, than in Mongol, which does not. Thus, chitrakrit is formed from kri, "make," joined with chitra, "painting." Declension. Case. The Turanian languages had formed cases of nouns before they were known in the Indo-European family. All the best Turanian types have them. The Sanscrit shows a more close kinship with its Turanian cousins in this respect than any other Indo-European language, because it does not use prepositions at all to express the relations of nouns to each other. The words for from, to, in, out, by, etc., come after the noun, as they do in all true members of the Turanian family. The other Indo-European languages use these prepositions plenti- fully before their nouns. The Sanscrit has come, therefore, more fully under the control of Turanian principles than any other member of the family. Yet a distinction remains to the Sanscrit which forbids our classing it among Turanian languages. It uses pre- positions copiously as inseparable prefixes to roots, just 286 as did the Greeks and Latins. But it is contrary to the nature of the Turanian system to do this. The resemblances noticeable between the Sanscrit case suffixes and those of the Turanian system have already been examined. Obvious as they are, it would be wrong to say that the only influence at work in the formation of the declension was the Turanian. The Semitic system has had an effect of its own peculiar kind. It has given genders to the nouns and perhaps the accusative case in m. It has also added a dual number. The letter m plays an important part in Semitic grammar. It serves to form a plural im for the masculine gender, and is then a suffix. It is also a prefix to denote participles in Piel, Hiphil, etc. It marks an infinitive or supine in Numbers x. 2, fc$*1p]b ? " to call/' said of the use to which the silver trumpets, ordered to be made by Moses, were to be applied — " to call the assembly.'' Then it is further used as a prefix in verbal nouns, as mishpat, "judgment," moda]}, " acquaintance," from shaphat, " to judge," and yadcfy, "to know." It is also met with in the dual, where dyim is used instead of the plural im. As a common interrogative in Hebrew, ma would, it is likely, be originally demonstrative, and in that state it might originate the Dravidian plural suffix mar and the Sanscrit accusative in m, as well as the Semitic plural suffix and the participial prefix just CASE SUFFIXES. 287 described. This explanation of the Sanscrit accusative is the more probable, seeing that neuter nouns take am in the nominative, as well as in the accusative ; and in Tamil and Mongol 1 m is a very common suffix to nouns, and makes a plural in Tamil. Bopp refers all case suffixes to a pronominal origin, and points to the pronoun imau, "these two," ime, " these," as the source of the accusative ending in m. His view of the origin of the cases appears to me to be wanting in convincing evidence in some respects. Thus, the instrumental and some other suffixes must, if viewed under the light of Chinese grammar, be regarded as true verbs. Bopp, however, was not willing to allow them to be other than pronouns. I believe them to have been both. The following are reasons for this opinion. First, it is more natural when motion towards or from, making use of or giving to, have to be spoken of, to employ verbs to express these ideas. They are really verbs, and no word could easily be employed to describe them without its having a verb sense. Secondly, if pronouns are employed as dative, instrumental, and ablative case suffixes, it should be allowed that, since they are used with such a force, they have already a signification as verbs. Thirdly, the Chinese demonstratives agree in form with certain common verbs meaning " follow," " give to," " carry," "bring," "do," "be." 1 Compare hugjim, " music," the Chinese gak ; also the pronouns t'im, yim, " that sort of," u this sort of." 288 china's place in philology. {£ zung, "follow," "from." g zi, "from," "self," "spontaneously." flfc t'rn, $f si, "this," % t'si, "give," ft s^, "give." g^ ^^ "that," ^ wap, "carry," "capture." fp e, " he," £J[ e, " make use of," " other," eo, ibam, ivit, " go." ft pi, " he," " that," pet, " give," " go away," " another." % Pa, " other," " he," " to draw," " drag." f£ i, " that," jg «?et, " be," " become," " do," "action," "that." % zhi, "this," "is," "be." The ideas of existence, transitive action, self, other, carrying, following, moving, are all mixed in confusion in these words. 1 Probably the verb sense was the earliest, for to this a name would be most easily applied. The notion of the demonstrative pronoun would be a little more abstract, and therefore less easy for primitive man to grasp. He would see motion. He would hear a sound. The motion would be named from the sound. Thus the verb would first obtain a name. Early names for "walk," "move," "go," "carry," would thus come into use. With a small stock of verbs primitive man would be prepared to fix on his demon- strative and other pronouns. The name of an action 1 For some criticisms on Bopp's views on this subject, see article on Language in the English Cyclopsedia. NUMBER. 289 would be applied to the actor who was seen performing it, or to the place or time in which he performed it. As the actor is not always known, the pronoun thus acquired would also naturally be assigned to positions in space and time. Thus true pronouns, prepositions, and adverbs would be formed to express all spatial relations^ This seems to be the true reason of the fact, that some of the commonest Chinese verbs coincide in sound (though usually differing in tone) from the most ancient and widely spread pronouns. Since Bopp's time all philologists seem to agree in accepting the view that case suffixes are of pronominal origin. Yet it may not be considered superfluous to remark, in proof of the pronominal origin of the accusative in m, so widely spread in Sanscrit, Latin, German, and English, 1 that the corresponding Turanian accusative suffixes gi, i, ni, a, etc., are all easily reduced to demonstrative roots. The Greek, Zend, and Sanscrit languages were spoken by nations in very near relations with Semitic peoples, and none of the other Indo-European races have had so full a development of the dual as these three. "We can then only regard the dual number as of Semitic origin. It does not appear in the Hamite languages. Thus we are shut up to this hypothesis. The Sanscrit mark of the nominative plural is h, corresponding to the Greek, Latin, and English s. 1 Compare the English him, whom, them. 19 290 china's place in philology. The Mongol has s and d. Perhaps all meet in the Hebrew th, For s and t are interchangeable letters. The genitive plural in m, so extensively used in Sanscrit, Zend, and Latin, may be referred to the Hebrew plural in im, and ultimately to the demon- strative in m. Bopp finds the demonstrative ma in the Greek /nev and the old Latin emem. I would add the Chinese men, " some one," the Siamese and Malay second personal pronoun men, mu, and the European words mutti, much, many, magnus. So the Chinese ta, " many," da, " great," * may be referred, with some probability, to the demonstrative root t. As the Semitic dual is formed from the plural by slightly altering the suffix, that is, by changing im to dyim, or th to thdyim, so the Sanscrit dual is formed from the plural by changing, e.g., as to au (Bopp, § 206) in the nominative, am to oh in the genitive, and so on. Gender. The triple distinction of gender, as masculine, femi- nine, or neuter, found in Sanscrit and other Indo- European languages, we may suppose to have originated among a Semitic or Hamitic people, and to have been carried on to its completion by the Indo-Europeans. If the Hamites were not sufficiently imaginative to personify natural objects, the credit of this creation 1 In Mandarin to and ta. GENDER OF NOUNS. 291 must be allowed to the Semites, of whose tendencies to view nature with a poetic eye we have such abundant proofs. But the mythological creations of the Egyptian mind (unless they sprang from Shemite teaching) may well suggest that the gift of imagin- ation was shared by some at least of the Hamites. The mark of the feminine in old Egyptian was t, and this agrees with the Hebrew feminine- ending th, some- times shortened to h. (See Ges. Heb. Gr., § 79.) The Indo-Europeans were likewise highly imagin- ative, and they adopted with avidity from both Hamites and Semites their personifications, alike in grammar and in mythology. They also carried forward the distinction of genders to its completion by adding a third form, the neuter. To the ancient Hebrew, while his language was in course of formation, inanimate objects were by the poetic faculty endowed with life and distinguished as masculine or feminine. Strong and powerful objects appeared as masculine. Those which are easily asso- ciated with weakness and timidity were regarded as feminine (Ges. Heb. Gr., § 105). But strength and power can be attributed to few things, and conse- quently the majority of the names of inanimate objects are feminine. Abstract ideas, offices, and collectives are usually feminine. Objects seized upon by the imaginative nations as suitable for mythological personification are in Hebrew 292 china's place in philology. nearly all masculine. Cloud, rain, morning, tree, heaven, sun, moon, river, mountain^ light, are examples. Among words occasionally feminine are evening, sun, fire, cloud, wind. Of these the last is rarely masculine. Name, blood, city, are masculine. In Sanscrit sun, moon, soul (atman), head, mountain, tree, evening, are masculine. Earth, night, light, life, heaven, river, are feminine. Dawn, mind (manas), blood, honey, deed, water, gift, are neuter. When the Greek and Latin languages made the moon feminine, they departed from the usage of the Hebrew and Sanscrit. In all the four languages life is feminine. River is masculine in Hebrew, Greek, and Latin. In Sanscrit it is feminine. Wind is feminine in Hebrew, but masculine in Greek (cive/mos) and Latin. The Greek has also a neuter word, irvevfia. Comparison of Adjectives. If Bopp's explanation of the Sanscrit comparative degree in tar a, as derived from tar, "to pass beyond," is open to any doubt, I would suggest that it should be considered whether the Mongol demonstrative fere may not have originated it. It has its source in the primitive root t'a, "other," "he." Its force would be, after the word good, for instance, " that other is good." The Manchu ere, "this," would furnish an explanation of the Latin or, in melior, "better," and the Mongol ene might be adduced to throw light PRONOUNS. 293 on the Greek comparative cov, in /caXkiav, " more beautiful." The Mongols say Pimu, for " such," Puilin bogda, for "extremely wise and holy," and demile airiben baina, for " there are very many." Here demile means "very," baina, "to be," and airiben, "many." Perhaps an explanation of the Sanscrit superlative in tama may be found in this last form. In Latin the sim in carissimus, "most dear," and pessimus, "the worst," may be the same word with the t changed to s. Personal Pronouns. First Person. Aham, " I." The Chinese nga, and Mongol na in namai, " me," etc. The forms in m are identical with the Mongol bi, "I," minu, "of me," but instead of being limited to the nominative and genitive, they are extended to all the cases. The accusative mam has the demonstrative in m for its final letter. The instrumental mayd has the Chinese ,£J[ yi, " to take," as its suffix, or, in other words, the demonstrative in i. The dative mahyam has the Chinese ^ | " give to," and the demonstrative m for its suffix. The ablative mat has the demonstrative in t for its ending. The genitive mama and locative mayi have respectively the demonstrative in m and the Chinese J£ yu, "at," "in," for their ending. The plural vayam is apparently the Chinese ^ yu, "I," and the English we. The oblique cases in the 294 china's place in philology. plural are, accusative asmdn, indicative asmdbhih, dative asmabhyam, ablative asmat, genitive asmdkam, locative asmdsu. Bopp regards the initial a as meaning "I," and the whole of the remainder as demonstrative. 1 I would draw attention to the modern Chinese plural suffix jpj men in women, " we," the causative pet, the dative pet, the demonstrative elements k, t, m, and the preposition ^ tsai, ze, as throwing light on these forms. Second Person. — As in the first person, there is a mixture of three roots, aha, ma, wa, corresponding to the na, hi, mi, of Mongol, so in the second we have twa, yu, corresponding to the Mongol c'hi, Pa. In the first person the speakers of Sanscrit selected na and mi, and made no use of hi, except in the substantive verb. In the second they neglected c'hi, and made use of Pa in the singular and the Chinese nu in the plural. The n is lost, as is the case in the Latin vos and English we. Third Person. — The nominative sah is found in the Chinese J^ ts'i and jtjf si, " this." The old form of both these words was si. The ta of the accusative tarn, instrumental tena, dative tasmai, ablative tasmdt, genitive tasya, locative tasmin, is the Chinese fg di, £ ti, fj$ ti, 38 te, "this," and flj, Pa, "that," "other," with the Mongol Pere, " that." The inserted m in 1 In the Shanghai dialect " we " is ngu ni, " I you " ; " you " is nung na, " you he." PRONOUNS. 295 three of the oblique cases resembles that of the Mongol second person, which has in the dative and locative c'himador, accusative c'himai, indicative c'himaber, &hi- maloga, ablative c'hima ec'he. This similarity becomes still closer when it is remembered that the second person usually takes origin from the third. The Greek av and Mongol si (old form of c'M) help to connect the Sanscrit tvam, Latin tu, English thou, with the Chinese si and ti, " this," " him." Demonstrative and Belative. The demonstrative ayam, " this/' accusative imam, is the Chinese ffi i, and Mongol ino, ano, and ene. Compare also the Mongol im, " such a," "so," as in im yehe, "so large." The Sanscrit relative is formed from this demon- strative, as the Latin qui and the English who are derived from the demonstrative in k. The Turanians, like the Chinese, are without a full relative, and the appearance of this feature in the Indo-European lan- guages must be attributed to Semite influence. The Hebrew asher, " who," " which," may be compared with the Chinese Pj| shut, "who?" "who- ever," ^ shi, " this." The old forms of these words would be zhi and perhaps zhid. They are found in the oldest remains of Chinese literature. I suppose the demonstrative to have been first, then the interrogative, and lastly the relative. Since the 296 demonstrative and interrogative are paired together, as words alike in form in so many languages, there can be no doubt of their identity of origin. How easy is the transition from the one to the pother may be seen, for instance, in that ? as distinguished from that. But when the interrogatives that ? and who ? and which ? are formed from the demonstratives that and he, sl transition just as easy changes the interrogative into a relative, and advances language on the path of progress another stage. A tone of the voice divides the word that ? when it asks a question, from that, when it points to some object ; and a change to another position in the sentence distinguishes the relative that from the demonstrative that, e.g., That watch, That watch ? and The watch that he made. The history of the formation of all relatives was very much like this. The Hebrew relative asher, then, may be supposed to have come out of the Chinese demon- strative and interrogative root in zhi, unless it be formed from the demonstrative in t, by change of t to sh. The reason why the Eastern Asiatic nations did not adopt a relative with full powers is found in the nature of their grammar. The subordinate sentence must in their languages come before the principal one. A sentence whose nominative is a relative pronoun is with them a subordinate sentence, and speech cannot in their languages expand itself by a series of subor- PRONOUNS. 297 dinate or circumstantial clauses coming after that which contains the nominative and principal yerb. It was the triumph of Semitic grammar, by simply drawing back the verb to the beginning of the sen- tence, to leave the way open for a concatenation of clauses to follow, which might commence at discretion with conjunctions or the relative pronoun. This afforded a facility and easy sequence to the expression of thought, which is unknown in Eastern Asia. The Indo-Europeans took from the Semites this feature, and hence the origin of the relative pronoun in their grammar. Interrogative Pronoun. The interrogative leak, ham, kena, etc., is the same word as the Mongol hen, "who?" This is proved by the related interrogative adverbs, viz., kati, "how many?" kada, "when?" corresponding to the Mongol heden, "how many?" hejiye (j for d), "when?" The Sanscrit relative yah appears in Mongol as one of the interrogatives, and is conjugated like other adverbs. We find the forms yambor, "what?" yago, " what ?" yagahin, yagahinem, yagahihu, yagonhihu, "how?" yagahiju, "how?" Reflexive Pronoun. The Sanscrit reflexive pronoun swa is the Latin se, 298 sui, suus, and our self. It agrees with the Chinese §f tsi, zi, "self," "spontaneously," "from." The Chinese have another reflexive, g, ki, which seems to be connected with the demonstrative in g and k, viz., gi and kit. Verbs. Derivatives. The desiderative and intensive forms of verbs redu- plicate the first letter of the root. This is a principle we find in Mongol and Turkish. The Sanscrit suso- bhish is " to desire to shine," and sosuby, " to shine very brightly." In Mongol c'habc'hagan is " very white," habhara is " very black." The resemblance, though only partial, is worth attention. Complete similarity in all points is not required in order to prove consanguinity of language. Else why is the conju- gation of the Greek verb so different from the Sanscrit in many respects ? As the Sanscrit has a causal, a passive, a desiderative, and an intensive, among its derivatives, so the Mongol has a causal and a collective. When Sanscrit gram- mar was formed, the passive had not become^ a voice, but was, as in Mongol, simply a derivative. Here is evidence of consanguinity. Derivative syllables immediately follow the root, after them come the marks of mood or tense, and finally those of person. Consequently derivatives are the oldest, then come the mood and tense marks, and PERSONAL ENDINGS. 299 the personal endings, the most recent in formation, stand last. i Personal Endings. The personal endings in Sanscrit verbs, as in the present twi, si, ti, may be compared with those belonging to certain Tartar languages which border on the Indo- European area, viz., the Turkish and the Buriat- Mongol. The more distant languages, such as Mongol Proper, Japanese, Tamil, have not personal endings. We conclude, therefore, that the marks for the persons sprang into existence after the Mongol and other older branches of the Turanian family had left their original seats in Western Asia, and before they were followed by the Turks. The Turks did not leave the vicinity of the Arian mother-stem till the principle of the relative pronoun had been introduced into their lan- guage from the Semitic, and they themselves had communicated the personal endings to the Arians, or received them from that race. The Turkish relative is the interrogative in k, and as such it agrees with the Hebrew *3, occasionally used as a relative, and with the Latin and English relative qui and who. In the Turkish personal endings, as they are at present (Davids' Grammar), we find the elements urn, first person ; sen, second person. In the third person of the present tense we find nothing. The syllables urn, un, 1, are the marks used in the preterite. 300 china's place in philology. TURKISH PRESENT TENSE. SINGULAR. PLURAL 1. deugurum, I strike. deuguruz, we strike. 2. deugursen, thou strikest. deugwsiz, you strike. 3. deugur, lie strikes. deugurler, they strike. The antiquity of the Turkish is shown in the absence of the initial s and inserted r found in the English equivalent strike, SANSCRIT PRESENT TENSE. SINGULAR. DUAL. PLURAL. 1. karomi, I do kurvah kurmah. 2. Jcaroshi kuruthah kurutha. 3. karoti kurutah kurvanti. Where the original elements are not too much decayed, we see in these two examples the identity of the marks of person. In the first person singular and plural m is the distinguishing mark. It is dropped, however, in the Turkish plural, where the suffix it of biz, " we," alone remains. In the second person s stands in the Turkish singular and plural. It has changed to P in the Sanscrit plural, reminding us of the Mongol Pa, "ye." In the third person the Sanscrit prefers the demonstrative root t, while the Turkish adopts that in i or o, as in the preterite dengdi, "he struck," where the second d marks past time. The Turkish plural ler is probably formed from ol, TENSE MARKS. 301 "he," equivalent to the Latin tile. An r is added and the initial o is dropped. The idea of marking past time by a prefixed a in Sanscrit and e in Greek, having no prototype in Turanian languages, may with probability be traced to Semitic influence. The creative power of Semitic grammar is centred on the beginning of the word and sentence, and in Turanian grammar on the end. The vowel in the Semitic past tense is a, as in bar a, "he created." The two Sanscrit preterites of bhu, " to become," are abhavam and babhuva. Kartum, " to do," has three preterites, all having the vowel «, viz., akaravam, chakara (ch for k), akdrsham. I suppose, therefore, that a has in it a past force, and it may be compared with the Chinese g, i, " already," which seems to be the root of Turanian and Indo-European preterites in u (Tamil), and ui (Latin), 1 since a some- times changes to ♦. 1 Bopp regards the augment in a as a privativum, and views it as an expression of the denial of presence. This view has involved him in some difficulties, and brought him into collision with more than one philologist. For example, how shall we explain the Greek augment in e, which bears no likeness to the a of negation ? Bopp says the a in the Sanscrit aug- ment had already lost its negative force, and had become a sign of past time, before it passed into Greek as the augment in e. Buttmann supposes it to be a broken-down form of the consonantal augment, regarding irvitTov as a shortened form from rervirrov. This, however, does not look very probable, and, in respect to Bopp's opinion, it is surely better first to make wider researches in kindred families of language, in hope of dis- covering the true origin of the augment. To explain it as the a of negation should be only a dernier ressort. 302 china's place in philology. The Sanscrit future in td is probably connected with the infinitive in turn, and the Latin supine in turn, tu. I suppose its origin to be in the preposition to, the mark of the English infinitive, and the Chinese chi, "to a place," in old Chinese ti. If this be true, it is also formed ultimately from the demonstrative root in t, if at least that be not rather regarded as itself previously a verb of motion towards. The Sanscrit future in sya is conveniently referred to the Chinese J|§ tsiang, siung, an auxiliary word used in giving a future tense to verbs. It means primarily, " starting from the side of," " side," hence " to lead a division." The Mongols have indeed a future in sogai, used for the first person singular and plural, which may be formed from the old Chinese sik, "give," first used as a precative and then as a future. The former etymology is the more probable in appearance. The future participles seem to have connexion with the Turanian conjugation. They are formed with the suffixes (1) tavya, (2) aniya, (3) ya. Thus, from bhicj, " to eat," is formed bhoktavya, " edible." JBhoJamja, bhojya, mean the same thing. We are strongly inclined here to identify the first of these forms with the Mongol suffix t ( o, t'ai, Pel, in heregt'ei, " necessary," idelt'ei, " to be eaten," dort'ai, " willing," Johist'ai, "ought to be." The Latin gerund dicendas, "to be said," also bears features of resemblance. In Manchu CONDITIONAL MOOD. 303 the similar form ends in rangga, e.g., ararangga, " that which is to be written," from arare, "to write." The form in ya seems to be connected with the future indicative of the Mongol verb, which is formed by the same syllable. The potential in ya may be compared with the Mongol future in ya. The Sanscrit potential has usually the idea of fitness (Williams' Sans. Gram., p. 199), and is sometimes a softened imperative. The Mongol future is also used imperatively, as in yabiya, " let us go." So the Latin potential in e or i, as in amem, " I may love," sis, " thou mayst be," may also be explained. The Sanscrit conditional in sya seems to be identical with the Mongol conditional suffix so, as in bolbeso, "if it be so." There can be little doubt that it is the Chinese sik, "to give." The Latin conditional con- junction si finds also here a convenient etymology, and is then seen to be parallel to our own word if, derived from give. I see no reason why we should not hope to be able at some time to go further back and identify the conditional in s ultimately with the demon- strative in s. Such simple ideas as giving, going, coming, carrying, have attached to them sounds which are like the common demonstratives. Thus, in addition to examples mentioned on a former page, ti is used for " to arrive at a place," " him," " to." Gip is " to give," and it is also "to arrive at a place." Bed, "this," "that," is also "to carry in the hand." Kit, 304 china's place in philology. gid, "he," is perhaps the European verb gad, "to go," Eussian chod, Sanscrit gati, "going," gata, "gone." Si, " this," is in old Chinese " to move from one place to another," and in Mongol, under the form ac'hiraho, it means "to carry," and under that of ec'hihu "to go." It is also in the West the verb of existence sum, asmi, esse. Further, the demonstrative zhi, zhet, "this," may be compared with shed, "to let go," shoot, and such like verbs. Verb as Substantive. Infinitive. Participles. As the Sanscrit infinitive in turn is apparently formed of the demonstrative in t, and the accusative in m, so in Mongol the infinitive in hu resembles the accusative in gi, and in Turkish the infinitive in mek seems to be formed from the demonstrative in m. The participle in t, as in bodhat, " knowing," is like the Mongol gerund in ged, which in colloquial is pro- nounced ed, thus, medeged or meded, " knowing." The two roots, budh, med, are, there can be little doubt, the same word. The participle in amdna may be compared with the Mongol colloquial gerund in man. This form is not given in Schmidt's Grammar. Its use is parallel to that of the gerunds in ged an&Ju. The passive past participle in ta may be compared with the Mongol gerund inju, of which the equivalent old form is du. As the Sanscrit form is often used AUXILIARY VERBS. 305 indicatively as a perfect, so is it with the Mongol. The substantive comes first, and then the indicative. The verb is fundamentally a substantive, and gerunds, participles, and infinitives, lie at the base and constitute the foundation of the Turanian verb, e.g., ochogder medeji, " I knew it yesterday," where the Chinese tsok, " yesterday," is seen in the first of the words, and ji, the colloquial form of the gerund ju, in the second. It may be objected that this Mongol gerund is active, and the Sanscrit form ta, now compared with it, passive. I would then suggest a comparison with the Mongol adjective in t'o, t'ai, as in heregt'ei, "necessary," moriPai, " possessed of a horse." Bopp states that the passive participial suffix ta forms in Sanscrit possessive adjectives out of substantives, as p'alitds, " gifted with fruit" (§ 835). So in English we say "horned cattle," forming a possessive adjective from "horn," just as the Mongols would say uburt'ei, " horned," from ebur or ubur, " horn." Auxiliary Yerbs. The substantive verb as, " to be," in English am, art, are, was, appears in Mongol without s. The root there- fore is a ; which means "being," and is also the ultimate root of aham, the first personal pronoun. The idea of being is derived from that of personality, and the oldest expression of personality is found in this pronoun a. 20 306 The second Sanscrit auxiliary verb is kri, "do," karomi, " I do." In Mongol a very common verb is hi, "do," himoi, "I do," or "be does," hibe, "be did," etc. Tbe tbird Sanscrit auxiliary verb is bhu, " become," " be," bhavitum, " to become," bhava, " become," abhavam, " I was," or " I was becoming." This verb is in its Mongol form distinguished as neuter and causative. The root bu is neuter, "be." The insertion of I makes it equivalent to our word " do," taken intransitively as in bolomoi, "it will do." The past participle bologsen means " completed," and is used as an auxiliary to express the accomplishment of the action of any verb. "What proof can be more convincing than the existence of these auxiliary verbs of the essential identity in origin of the Sanscrit and Mongol languages ? But the same proof holds good also for the Turkish and Tungusic stocks. It is only when we come to the Japanese and Dravidian branches that this system of identical auxiliary verbs diminishes from three to one. The verb a for existence keeps its place everywhere. Hence it appears that the original Tartar language, which was split into Turk, Mongol, Manchu, Finnish, etc., immediately preceded the Sanscrit in the linguistic development of the world. Adverbial Suffixes. T in Sanscrit is d in Mongolian. Thus among ADVERBS. 307 the adverbs of place, atra, "here," tatra, "there," correspond to the Mongol ende, "here," t'ende, 1 "there." The suffix in the two languages is identical. D in Sanscrit is equivalent to j or d in Mongol. Thus kadd, "when?" is the Mongol hejiye, "when?" Ekaddy "once," is in Mongol nigodaga or nigodd, "once," from nig, " one." Tadd, "then," is in Mongol t'eduile, " then." The suffix vat in suryavat, "like the sun," from mrya, " sun," may possibly be the Mongol adeli, "like." The initial v was originally not consonantal. The Latin is idem. Negative Adverbs. The negative na, ne, nehi, is derived from the same source as the Japanese negative. That source will have been some Turanian language in South-western Asia. The negative md is found in Chinese, in the Tartar languages, in Tibetan, and in the Semitic family. It is used over nearly the whole of Asia, but, except in Greek, is little employed in Europe. Time, Manner, Comparison, Place. Adya, "to-day," ei now," may be compared with the Mongol edoge, "now." Evam, eva, "so," "thus," are suggestive of identity with the Mongol yim, "thus," " so." Kwa, " where," is the Mongol hamiga. 1 The Mongol e is the same in sound as the Sanscrit «. 308 china's place in philology. Prepositions. The absence of prepositions to mark the relations of nouns is peculiar, among the Indo-European languages, to the Sanscrit branch. The Romans used "in," "ex," " ab," etc., as the English now use "from," "in," "to," etc. It is a specialty of the Sanscrit, and of the triple- branched Turanian system, to employ case suffixes instead of the more ancient prepositions found in the Chinese, the Semitic, and the Himalaic systems. The Greeks, loving freedom, early threw off the yoke of this Turanian law. The speakers of Sanskrit never did so. In Homer the adverbial case suffixes are used with the prepositions. In later Greek the adverbial case suffixes are not found. They have given place to prepositions, as afterwards the cases of nouns also became needless over much of the European area, and were exchanged for the primeval prepositions which seem to be ever engaged in recovering their long lost dominion. In Sanscrit the prepositions are only used in compounds as inseparable prefixes, and here the nearest Turanian type to which in this respect it can be compared is the Dra vidian. Compounds. When in Sanscrit words are compounded, con- nective letters are not used, and the resulting whole is treated as a single word. Thus, for " moonlight " chandraprabhd is used. In Mongol it would be saranu SANSCRIT COMPOUNDS. 309 gerel, where nu is the genitive case. The Tartar languages have an aversion to naked compounds, and prefer to introduce, as here, the genitive suffix. This I believe to be a comparatively modern tendency. The Sanscrit acts here according to the true ancient principle for the compounding of words by simple juxtaposition, as found in Chinese. The Tartar lan- guages appear to have acquired the habit of inserting case suffixes, and other particles, between words which would otherwise coalesce into compounds, since they were separated from the Japane'se and Dravidian branches. Hence, in regard to the way of forming compounds, the Hindoo principle must be compared with that existing in older stems, e.g., in Chinese yue Hang, " moon light." In Japanese and the Dravidian languages the crude forms or roots are likewise placed side by side without connecting particles. Japanese tsuki akara, " moon light." In the Greek and Latin languages, as in lunce lumen, the genitive suffix is, as in Mongol, carefully inserted. Hence the Tartar race remained in juxtaposition with the forefathers of the Greeks and Latins later than the time when the speakers of the Sanscrit and Dravidian idioms were in a position to exercise an influence upon each other. The compound gurusishyau means " master and scholar." There is no conjunction. Au is the sign of the dual. Guru, " teacher," and sishya, " scholar," are co-ordinate nouns — roots standing together without 310 china's place in philology. connective, as bakshi, "teacher," shabi, "scholar," might do in Mongol. But the Mongol is without the dual mark, unless hoyol, u the two," be added, as is sometimes done. The want of a conjunction is in accordance with the custom in all eastern Asiatic languages. In the compound maranavyadhisokah, marana is " death," vyddhi is " sickness," and sokdh is " sorrow." These three nouns are written together without a conjunction, forming one huge word, which in Sanscrit syntax is treated as a single substantive. It may be compared with the Chinese sheng lau ping si, "birth, old age, sickness, and death," in Mongol t'urehu, ot'olhu, obc'hinhii, uhuhn. The four Chinese substan- tives become in the Tartar idiom four infinitives. How thoroughly they are regarded as substantives appears from the fact, that in the Buddhistic language common in Mongolia they are known as the durben dalai, " four seas." The addition of the connecting conjunction in more western languages is proof of the influence of Semitic grammar. The aggregation of substantives without conjunctions is a circumstance in Sanscrit which shows how completely that language rests, in regard to its linguistic principles, on the speech of more eastern races. The resemblance may be noticed in all sorts of compounds. In this part of grammar Sanscrit looks SANSCRIT COMPOUNDS. 311 like an old Mongol using but sparingly its apparatus of case particles. E.g., srarga yata, " gone to heaven," svargang gd, " the Ganges of heaven." In Mongol T*engri dor garaksan, Tengrin Gangga ma run, where dor and n are locative and possessive. In Chinese, ancient or modern, the position of the verb, as standing before its noun, weakens the resemblance to Sanscrit, and throws into more prominent relief the essential identity of Sanscrit and Turanian syntax. The Sanscrit manda gala, "going slowly," is in Chinese man tscu, and in Mongol odan yabahu. In the last two of these languages this compound may take a genitive suffix and another noun, for instance, man, after it. The Sanscrit form is an adjective, of which the syntax is the same as if it were simple. "When such compounds occur as rdjaydmin, "that which goes to the king" (e.g., revenue), raja guru, "king's instructor"; rdjakala, "king's family," from kala, " family," " caste," the Chinese kia or ho, "house," "family"; rdjayhna, "regicide," from g/inat, " killing " ; rdj((danda, " punishment by a king," from danda, " punishment ; " Chinese analogy seems to require that the relation should in all cases be regarded as possesxire. Even where the English rendering re- quires from or by, as if the relation were ablative or instrumental, it is better to hold to the simplicity of primeval grammar, and explain all such instances on the principle of possessive dependence. Thus, 312 china's place in philology. " punishment by a king " is also rendered by " king's punishment/' without much forcing. By regarding the relation as possessive in all cases where in a compound the second noun depends upon the first, the analogy with Chinese grammar becomes perfect. Thus, wang ts'i, "king's son," wang tsung, " king's family," wang fa, " punishment by the king," wang hi, " land appropriated to the use of the king." The same law rules in all the languages from the Hindoo area eastward to the Japanese Islands, except in the eastern Himalaic and Malay region, where the Semitic inversion, which transposes the genitive, holds sway. The true reason why this inversion is impossible in Sanscrit is, that this language is in fact controlled by the same laws of position as the Turanian idioms. The modern Pekinese speaks of fu fu lia, for " husband and wife." Here fu, " husband," takes one intonation, and fu, "wife," another, while Ha is a contraction for Hang, "two," and corresponds to the dual suffix, which would by the Sanscrit grammarian be placed here. Could analogy be closer ? But com- pare the words themselves ; fa, " husband," is bharu, fuj^^w^ef^ is bhdryd, and in Greek ttoois is "husband." The proof of original connexion in language thus becomes still more clear. Examples may easily be collected from Mongol to show that the inserted particles are often omitted, and LAWS OF POSITION. 313 that the analogy thus brought to view may also be extended to the use of dual and plural suffixes. Thus c'has c'hagan, means " white as snow." If written in full, met'u or adeli would be added after c'has, " snow." This is exactly the Chinese shtet bale, " white as snow," and thebSanscrit MmasUala, " ice-cold." So also echige ehe hoiyogola, " the father and mother both." The resemblance to Chinese and Turanian idiom is carried also into what are called the relative com- pounds. Thus in mahadhanah purushah, " a man who has great wealth," mahd is "great," dhanah is an adjectival form of dhanam, " wealth." Native Sanscrit authors explain this usage as equivalent to the employ- ment of the relative in the genitive case. 1 "With this may be compared in Chinese a sentence such as ta Mo wen ch'i sh'i, " a scholar who has great learning," con- sisting of ta, "great," Mo-wen, ("learning and heard") "learning," cM, the possessive particle, sM, "scholar." In Mandarin the possessive ti is also used after adjectives, as in hau ti, "good." Compare also the Mongol yehe gabiya t'ai Mimun, or yeheu gabiyan humun, " a man who has great merit." Gabiya, " merit," is here made into an adjective by the suffix t'ai, which thus corresponds to the Sanscrit adjective suffix h. Laws of Position. In the Sanscrit and the Turanian languages, the 1 Williams' Sanscrit Grammar, p. 166. 314 laws of position are the same in several of the most important particulars. If we take the sentence, " And the children of Israel dwelt among the Canaanites," in Jud. iii. 5, we find the Hebrew order the same with the English. The San- scrit and Mongol both read, " Israel people Canaanites among dwelt." The Chinese would be " Israel people at Canaanite people midst lived/' or "Israel people lived Canaanite people among." Thus the translators of the scriptures at Calcutta (edition 1852) have adopted an order for the words exactly agreeing with the Mongol. Another example is, " Jabin king of Canaan, that reigned in Hazor " (Jud. iv. 2). As before, the Semitic order is as in English. The Sanscrit reads Hatsora- nivdsinah kindniyardjasya ydvinasya haste. Here nivd- sinah is an adjective, "residing in." Raja is "king." Sya is the possessive suffix. Haste is "hand" in the locative. Omitting the word "hand" and the possessive case preceding it, the Sanscrit reads, " Hazor residing Canaanite king's Jabin." The Mongol reads, " Hazor in ruling Canaanites' Jabin king," Hajor for ejelegsan Hanayan t'anu Jabin hagan. The translators have adopted in Calcutta the same order nearly as those who performed their work on the eastern shore of Lake Baikal in Siberia. The Mongol introduces the case suffix after Hazor, and gives the name Jabin before his title. These are the variations; otherwise LAWS OF POSITION. 315 the laws of position are identical. T'anu is the genitive plural. Hence the Sanscrit language has peculiar laws. The Greek and English both have the Hebrew order. In this part of grammar the Sanscrit is cut off from its proper relationship, and bears no close resemblance to any western language. It is in agreement with the eastern idiom of the Asiatic continent, with that of China and Japan, Mongolia, and Dravidia. This general agreement in syntax between the Sanscrit and Turanian types is subject to numberless exceptions. To make this plain, I here give two sentences out of the Hitopadesa, taken from Williams' Grammar. Asti, "there is," gautamasya mimes tapo- vane, "in the sage Gautama's grove of penance," Mahatapa ndma munih, "a sage named Mahatapah." Tena, " by him," dkramasannidhdne, " in the neigh* bourhood of his hermitage," mushika kdvakah, " a young mouse," Jcdka mukhdd bhrashto, " fallen from the mouth of a crow," drishtah, " was seen." Turanian syntax would require the verb asti, " there is," to be at the end of the first sentence, and the descriptive clause, "crow's mouth from fallen," to precede the noun " mouse little one," to which it refers. These two things excepted, the laws of arrangement are Turanian, as in "Gautama sage's penance garden in," " a Mahatapah named sage," " hermitage neigh- bourhood in," " mouse's little one," " crow mouth from 316 china's place in philology. fallen," and the position of the verb " was seen " at the end of the second sentence. The resemblance is still closer, inasmuch as drishtah is a participle used indicatively, which is a common phenomenon in Mongol grammar. Judged by syntax alone, Sanscrit and Mongol are sisters, just as Hebrew, Greek, and English, if tested in the same way, might, though the similarity is somewhat less close, also be called sisters. It was after the separation of the Chinese from the primitive stock, that the great Turanian inversion occurred, which placed the verb last, and thus originated the declension of nouns. The Turanians remained long enough in the west to bring with them in their wanderings the declension of substantives, the conjugation of verbs, and a syntax which places them in a midway position between China and the western world. And then the Sanscrit, the most easterly member of the Indo-European family, by its peculiar syntax, its principle of agglutination in compounds^ and its use of the participle, conveniently occupies the interval between Turania and Europe. A word upon the Zend. The absence of the Turanian order in Zend syntax is a sure indication of Semitic influence. Bopp gives the following sentence in Zend. Staumi, "I praise," maig'emcha, "the clouds," varemcha, " and rain," ya te kehr pern, " which thy body," vah'sayato, " make to grow," baresnus paiti gairinanm, /.IN!) SYNTAX. 317 " on the heights of the mountains." Here gain, the Sanscrit giri, and Mongol agola, " mountain, " occurs last, after its nominative. This is Semitic order, which is also prominent in the whole sentence. The Zend, in fact, has an accidence and vocabulary like the Sanscrit, but a syntax like the Hebrew. As in the modern Persian, Semitic words had also pushed their way into the Zend. Thus athr, " fire," is the Hebrew esh by the common change which takes place between sh and t. A few Zend words with old Chinese and Mongol equivalents are here appended. " Bad," Zend eghe, Chinese ah. " Flesh," Zend machshe, Mongol maha, Persian maso. "Not," Zend ma, Mongol bis/ti, bu, Chinese mo. " Ear," Zend goshtc, Chinese ngi, Sanscrit ghosha, Persian gosh. CHAPTER XIII. European Languages. — Latest and Grandest Development of Language. — The Alphabet. — Common Radical Syllabary of Chinese and European Languages. — European Radical Syl- labary. — The European Word. — Semite Influence seen in conjugational vowel changes, in doubled consonants, in Masculine and Feminine Terminations, and in Dual and Plural Numbers. — Turanian Influence seen in Moods and Tenses, and ln Compounds. — European Syntax. — Chinese Element. — Semitic and Turanian Elements. — Greek. — Tones in Chinese are Accents in Greer. — Common "Words in Greek, and Mongol. — Latin. — Resemblance of Latin Gerund and Supine to those of Tartar Languages. — List of Roots Com- mon to Latin, Chinese, and Mongol. — Latin Syntax more Turanian than the Greek. — Roman Family Relationships Suggestive of Connexion with Eastern Ideas. — Resemblance between Roman and Old Chinese Religious Beliefs. — Russian: The Best New Type of the Sclavonic Family.— Full Alpha- bet. — Abounds in Prefixes to Roots.— Examples of Syntax. — Anglo-Saxon. — The Syntax Turanian. — Anglo-Saxon and German have more of the Turanian Element than is seen in the English. — English Returns to Chinese and Primeval Syntax. — Cause of these Variations. — Resemblance of Anglo-Saxon Poetry to that of the Mongols. — Alliteration: Exchanged for Rhyme; Cause of this Change. — English. — List of Common "Words, Chinese and English. Old as are the European languages, evidenced by an unbroken series of literary works, dating from about the ninth century before Christ, they bear in their structure the marks of youth, if compared with the INDO-EUROPEAN LANGUAGES. 319 Turanian and Semitic families. Principles of grammar seem to have been early borrowed from both these families, and incorporated in European speech at a time when language was still plastic. Destined them- selves to be the dominant powers in the world's history, from the conquest of Babylon by Cyrus, they were before that epoch living as independent nations, occa- sionally subject to the Semitic rulers who from time to time were able to reduce some of the nearer among them under their sway. Semitic and Turanian in- quest at a very ancient date would impart Semitic and Turanian elements to the language of the conquered, and when these nations themselves invaded regions occupied previously by Semite and Turanian peoples, similar results would ensue. The Indo-European system, with the Chinese, Semitic, and. Turanian, would each branch directly out of the primeval trunk of language. Each would develope its special characteristics with a varying rate of rapidity. Civilization, the invention of writing, maturity in arts, and in political institutions, would soon harden the Chinese and the Semitic families into a distinctive moulded form. The Turanian and Indo- European would take a longer time to harden, and in them the language-forming power would continue for a longer period. They would remain in a plastic state till the polysyllable and the paradigms of substantive and verb forms were completed. 320 Rapidity of change in language is in proportion to the civilization of the people speaking it. In early times languages changed more quickly than now. Not many centuries would be required for the primeval development of the existing families. But once formed they would last for thousands of years. Yet there is a sense in which they may be said to stand in chrono- logical succession. As in geology it has been shown that the lesser ranges of mountains were first elevated and the highest last, so it has been in the history of language. Mono- syllabic speech preceded the dissyllable, as the dis- syllable preceded the polysyllable. The Alps, Andes, and Himalayas, were not thrown up till the lesser mountain systems were complete. They are more aspiring, they pierce the region of the clouds, they possess a greater variety of vegetation, present to the eye richer landscapes, originate larger rivers, and pro- mote the fertility of wider tracts of land than lesser mountain chains. So it has been with the Indo- European languages. They have aimed high in thought, enlarged the field of poetic feeling, advanced scientific inquiry, and led the human race forward on the path of civilized progress to an unequalled degree. The greatest thinkers in philosophy, the creative intellects of science, the most noteworthy conquerors and legislators, have been those whose speech was Indo-European. EtOPXAH LANGUAGES RKC1M. 32] As the cause of the greater elevation and raagnit of the more recent mountain chains may be probably referred to the increased thickness of the earth's crust, unci the consequent increased pressure on the liquid materials beneath, so the richness in forms and in capability for expressing human thought which marks the Indo-European languages, is due to the united action of the older linguistic families upon this last and noblest creation of the language-forming power. It is necessary, therefore, to review briefly the traces presented to view in European languages of the presence in a long distant time of strong Turanian and Semitic influence. The German, Russian, French, and English of the present day are descended from older forms of speech, which assumed their peculiar shape under this double formative influence. The Alphabet. A. The broad a in father belongs to all languages. It lias been replaced in modern Chinese by o. In western vocabularies it occupies less space in those which are modern than in those which arc ancient. Hence it abounds in Hebrew and Sanscrit. It has become u in some Chinese words, as in ff< } mu s " father," " mother," from the primeval ba, ma. It belongs to the pronouns, and represents the first person. As a suilix it is common in Greek and Sanscrit, as in the Greek names 322 china's place in philology. of the Phoenician letters. Beth, caph, teth, etc., became beta, kappa, theta, etc. In the inevitable softening down of language the consonantal finals t, p, took after them in Greek the vowel a. A has become o in Hebrew, so that lo, "not," is written X7, with aleph. In Tamil, Japanese, "and Mongol, a is much prefixed to roots, as in Japanese ame, " rain," Tamil mazhei, Latin madeo, " moisten." The short a of Sanscrit is understood whenever no vowel mark is used. This is an indication of the extensive ancient use of this vowel. B. The consonant b was the old representative of the Chinese p and / in the lower or sonant series. It is interchanged with m, as in Persian and Mongol boron, "rain," Japanese ame, Hebrew mayim, "water," Hebrew matar, "it rained." Also Mongol bi, "I," Turanian min, Latin me, mihi. Also bal, "honey," Latin mel. It is sometimes used in Greek for w, as in fiovkofjiai,, " I wish," Latin volo, velle. In Sanscrit it also took an aspirate occasionally like d and g, as it probably did in old Chinese in the first sonant tone, which in Mandarin has become aspirated p. CH. The compound letter ch is the modern equivalent of t in the eastern Asiatic languages, and of k in the THE ALPHABET. 323 Indo-European. In modern Chinese we also see ch originated from /, when standing before the letters u and i. All Chinese words in ch not thus formed recently from k, are derived from t or d, according as they belong to the surd or sonant series. The Sanscrit ch, aspirated and unaspirated, are both from k. The Tibetan ch, ch', is from s or sh. The Mongol ch aspirated is from t's, and this again from s or 6//, as c'hag, " time," Chinese zhi, c'hihgon, "stone," Chinese zhiag. The same thing occurs in Fuhkien, where shut, " water," is chui y also in Tibetan, as c'hu, "water," e'hi, "die," Chinese si. Perhaps a part of the Chinese words in sh formerly had c l h for sh. The guttural ch, as in loch, " a lake," in Europe, represents k in Chinese. Thus wcchse/n, " to change," may be seen to agree with the Chinese yik, " to change." The Turkish ch corresponds to the Mongol d. Thus kachj "how many," is in Mongol heden. The Chinese clt is also found to be j in Mongol, both having sprung out of an older d or t. The Japanese ch is in Mongol s, and is, strictly speaking, in. J). The letter d is in old Chinese the equivalent of the modern ch in the lower series, and of the modern Mongol j. Compare the Sanscrit kadd, "wlun:" Mongol hcjij/c. The lost final d of many Chinese roots ifl recovered 324 china's place in philology. in Japanese, as kudari, " to descend/' in Chinese Ma, ge, Kara, sadzki, "to receive," %§> sheu, zhud. D is often interchangeable with /, as in lacryma, haupvov, '* tear," longus, " long," Chinese J| dung, " long." D sometimes becomes t in western languages, as Chinese da, " earth," Latin terra. Dj takes the place of g in the Shanghai dialect before the vowels i, u. The same occurs in English, as in " bridge," from " brig," and several other words where g is final. E. The vowel e in " then," " there," is derived in Chinese from ya, and is scarcely used except as a modern final. In Mongol e is classed with u and u as female, while a, o, are male. These terms mean that when the root has, for instance, e for its vowel, the vowel of the added s}dlables shall be of the same class. The Mongol e is now enunciated as the Sanscrit short a, but in Hindoo words whose sound is transferred, as in Ganges, a is used by the Mongols for the short a ; thus they write Gangga (not Gengge). The European e is usually the Chinese i, as in yit i "one," Sclavonic yedin, Chinese nyit, "hot," Scotch het, German heiss. The English e, as in be, see, me, we all pronounce in UN. AI.I'IIAI open syllables with the sound i. The modern Chinese media] 0, as in p'01, "to be fit for," "to match," is found in European languages to be 1, as in Jif. But in this case the old form in Chinese is usually p'tt. The Chinese j has changed to 0, while the English c has changed b The Sanscrit e is found late in the series of vowels, forming with a supplement to the three chief vowels, Oj /', u. Similarly, the Greeks adopting the three leading vowels of the Phoenician alphabet a, u, /, alcph, 000, yod, proceeded to apply an aspirate he to represent the vowel e, as they used a guttural sign (ii/in for the vowel 0. The Semites were content with three vowels. The Indo-Europeans needed five. The Greeks did with the Semitic alphabet in the west what the inventors of the Devanagari did in the east. They took the chief vowels as they found them, and used new signs for vowels not represented. That at both ends of the Semitic ana, which once probably reached from the Mediterranean to the Indus, the tin. fi/rjj/, car, and yod, should haw been regarded as vowels, may be appealed to in proof that they were not originally consonantal signs, as some grammarians maintain, but true vowels. F. F is a new letter in Chinese. It proceeds from \ />, and p*. It is wanting in B£ongol and Tamil, as it is 326 china's place in philology. also in Sanscrit. In Greek it crept in as an aspirated p, gliding afterwards into (j>, pronounced like the Latin f. The Latin /came from jo, as did the English. The Semitic / probably also came from p. One sign £ is used for p and / in Hebrew, and the Greeks, adopting the Phoenician alphabet, used the same sign for the value p, which shows that at about the time B.C. 1000, this was its usual force. F is inserted in a few German words after a radical initial p, as in pflegen {pledge, pignns). In Japanese / is used for h when standing before u, and proceeds ultimately from p, b. G. The letter g is the old form of the modern k and k ( in the lower series. In Mongol the old Chinese g is found as g or h, e.g. hwun delehu, "to honour," from Chinese king. In Hebrew and in European alphabets, it precedes the corresponding surd letter k. In Latin g changed to dj before e, and this again became zh in French. The Mongol g sometimes corresponds to the Sanscrit k and the Greek %, as in gar, kara, x e fy> Du t also to the Sanscrit g and gh. The Persian g sometimes corresponds to the Chinese ni, as in gao, " cow," gosh, " meat," " flesh," " ear," as does the Greek in yvvq, " woman," if compared with ^f nio. An initial g is often dropped, as in if from give, and THE ALPHABET. 327 in the Piatt Deutsch, where gewesen, " been/' becomes yevesen. H. The Chinese h seems to be a modern letter formed from k, k', and g. For example, jgl hi, "joy," Latin gaudium* The final d is recovered from the phonetic, ^ kit, "luck," forming the upper part of the character. The Japanese h represents p or b. The Mongol h represents k, k l , g. In Greek it stands for s. In Latin it corresponds to the Chinese k and g, as hie, " this," Chinese ^ gi. The same is true in German and English, as kok JfJ, "high," hoch, "high." In the old middle dialect of China, as still spoken in the Sucheu and Hangcheu region, h is subdivided into a strong and weak aspirate. In the Mandarin dialect of north and west China, it coincides with s when it precedes i and u. In Zend and Persian, h occurs for s in hapta, " seven," etc. The Semitic heth, the Scotch ch in loch, is not used in the eastern Asiatic languages. The Semitic heth and he both correspond to g, and probably derived their origin from that letter. /. / is one of the three primitive vowels. In modern Chinese it sometimes becomes wei. This we learn from the Japanese, who call wei, " a seat," i. It is a prefix in Japanese and Tamil, as in iku, " how many," from 328 china's place in philology. ki, "how many?" The changes of vowels are too rapid to allow any general correspondence to be traced between the Chinese i and the European equivalent, or vice versa. J. The Chinese modern j is from ni, the Mongol from d. The Chinese j is zh, the Mongol is dj. The Sanscrit j is dj, and is derived from g, as ch from k. The Latin j was y, and sometimes dj, and has changed into zh in French, and into dj in English. The Mongol ujihu, " to see," jirehe, " heart," are the Latin videre and the Persian dili. The Semitic y is pronounced j by Europeans, as in Jehovah. The Sanscrit yuj, "join," in Chinese yok, is in Latin jung, and in Greek feiry, where dz is the Greek equivalent of dj. The Greeks could not pronounce ch or sh. The Arabic /, pro- nounced dj, is altered from an older g, as in jahannam, from Gehenna, " hell," just as dj has replaced g in the English words gender, genitive, etc., derived through the French from the Latin. Thus it appears that dj is primitive in no alphabet, but, like /and ch, is of recent origin, and was perhaps quite unknown in the early languages of the world. K. In modern Chinese, k before i and u has changed to ch. In the European languages, k changes to ch before THE ALPHABET. 329 all vowels, except o and u. In ancient Chinese, k changed to h, but was also itself changed from g. There are not wanting indications that the true primeval source of k was g. The original of the Hebrew % ki, "for," "that," and ItS, ko, "thus," is found in the Chinese j£ gi, "he." The Sanscrit k corresponds to the Mongol g } as kara, " hand," Mongol gar. The Japanese k also corresponds to the Mongol g, as in kado, " gate," Mongol egude, Chinese hu, gud. In Sanscrit, s occurs for k, as in sat a, " a hundred," as compared with centum. The Chinese change from k to A exists in Mongol, where the Sanscrit kat'ara y " hard," is found hat'o, Japanese kataku ; and in Europe, where collis became "hill," and collum, "neck," hah. These two words are in Chinese ngok, "hill," and kang, " neck," where the old finals both appear. In Russian, ch occurs commonly for k, as in chistiye, "pure," castus, the Chinese kit, "pure," "clean." In Tamil, the old k appears for the Chinese and Mongol h, as in karumei, "darkness," Mongol harangwei, Chinese hek. Kt appears as initial in the Greek ktclvco, "kill." Here the intermediate vowel has been dropped. The Hebrew is katal, " he killed," and the true root is kat, "to cut." The aspirated form of k appears in Sanscrit, Chinese, the Himalaic languages, in Corean, and in Mongol and Turkish. Pronounced as the k and h in the word 330 china's place in philology. inkhorn, but brought closer together. In Eastern Asia the aspirated and unaspirated k are separate letters. In Europe, on the other hand, if k is aspirated, it is the consequence of local or individual habit, and embraces all the instances. In the province where card is called k'ard, cold will also be called k'old, and so on. Z. The Latin / is found in Chinese usually as ch, coming down from an older d, as in JP- ch'ang, "long," old sound dung, Latin longus. So the Hebrew /, as in lakach, "to take," Greek Xcuyxdva), seems to be found in the Chinese t, as in tek, "to get." Compare also #37, "clothe," "put on," with the Mongol debel, "clothes." The frequent change of d to / perhaps indicates that the true origin of the letter is d. It is sometimes changed to n, as in the Mongol nog on, " green," Chinese lok, and the Latin nemus, " grove," Chinese Urn, Hebrew lo, " not," non, na. L is frequently inserted after an initial k, t, p, g } d, b, as in flat, pledge, black, as compared with patina, pactum, and the Sanscrit bahula, " black," and the Chinese bed, "spread out," bang, "pledge," mek, "black." If I occurs after an initial s in European languages it is radical, and the s not so. Thus, sloe, "a wild plum," is in Chinese li, "plum," and slachten, slay are in Chinese lok, " kill." 1 1 Slip, Latin labor, lapsus, German schleifen. THE ALPHABET. 331 Sometimes a connecting vowel is introduced between the initial and the inserted /, as in kclKvtttg), " hide," "cover," Sanscrit hub, Chinese hap. For caput, "head," the Russians have glava and golova. The insertion of I is common in some of the Himalaic languages, in Semitic, and in Indo-European lan- guages. It is avoided in Chinese and the triple- branched Turanian system. Hence in comparing roots it must be omitted from the European word before the Chinese or Mongol equivalent can be found. L is a favourite suffix in Turanian words, and a common third letter in Semitic trilateral roots ; as in Mongol gol, "river," Chinese ga. Hebrew ?3J?, ngagal, " revolved," from a biliteral root gak, which appears in circulus, circle, kvkXos, etc. L is sometimes inserted between the final consonant and the preceding vowel, as in our word old, Mongol ot'olju, " old," Latin vetus. It then sometimes takes a vowel, as in the Russian zoloto, u gold," where z is g. In the Cochin- Chinese and Siamese languages / takes the place of h. So also in the Malayo-Polynesian. 1 The Chinese / is usually r in the west, as rota, " wheel," Chinese lut, " a round thing." M. The letter m in Chinese corresponds to the m of western languages, as in mel, " honey," Chinese mit ; 1 Thus, lima, "five," may agree with the Hebrew hhamesh. 332 china's place in philology. miles, " soldier," Chinese mo, " military." The final m of some European roots is represented by ng in Chinese, as KapirTU), " bend," Chinese ^ kung, " bow," " to bow." The Hebrew final m seems to correspond in the same manner to the Chinese final ng, as ram, " high," Chinese lung. The Greek m sometimes corresponds to the Mongol b and the Chinese p, as fia/cdpio?, " blessed," Mongol boyint'o, " happy," Chinese pok, " happiness." The Chinese m occasionally agrees with the English b, as in black, Greek yiteXa?, Chinese mek, "ink," Mongol behe. Final m has in modern Chinese become n. m The letter n is frequently interchanged with t as in eh, I'&o?, " one," unus, Chinese yit, " one." Final n in Chinese corresponds to final n in the west. Fundo, to " found," may be compared with the Chinese ^ pen, " root," "foundation." Chinese hen, " wheel," English " round," Chinese tan, " that which is stretched out," Latin tendo, " stretch." Final n is often dropped in Tamil, as in kuzhal, " tube," Chinese kwan ; Tamil kuri, kol, " stick," Chinese kan ; Tamil tdl, " sheet of paper," Chinese tan. The Tamil n final sometimes represents the Chinese t final, as in tan, " stand," Chinese dat, Indo-European stan, stad. THE ALPHABET. 333 "What we write ng is a separate letter related to k and g, as n is to t and d. It is initial in Chinese and Tibetan. The Chinese initial ng is apt to be omitted, as in wo, " I," formerly nga. The final is also often dropped, as in kwang, " light," Mongol gerel, Japanese akari, Latin gloria. So also neng, " able," Tibetan nupo, " one who is able." In Latin roots ng often replaces the final k, as in pingOy pinxi, pictum, pango, pepigi, etc. "What we write ni, is in Sanscrit and old Chinese regarded as a distinct letter belonging to the ch andy series. It has changed in modern Chinese to j. In Turkish and Mongol it is found as ¥ or g. For example, nin, "man," Turkish kHshi, Mongol humun, ni, " two," Turkish ik% niok, " flesh," Turkish gosh, ni, "ear," Turkish gosh, niok, "if," Turkish eger. The European avrjp, and homo, " man," seem to belong to this little knot of words. Compare also gleich, " like," "if," gracilis, "tender," yvvrj, "woman," yd\a, "milk," with the Chinese niok, "like," "if," niok, "weak," " tender," nia, " woman," niu, " milk." The Chinese n, ng, and ni, are on the whole usually found k, g, h, in Tartar and European languages. Compare ngic, " cow," Mongol uher, Latin vacca, German kuh. Nga, " I," ego, niuen, " origin," yivo<;, genus. Ngan, " eye," oculus. Some examples exist of n unaltered, as nehmen, 334 china's place in philology. nimm, " take," Chinese nim, " carry," " burden," "responsibility," in modern Chinese jen. For the Hebrew ayin see o. 0. The letter o, like the other vowels, is often prefixed to roots. Chinese Pi, "tooth," qSovs, oSovtos, dens. Compare in Malay orang, "man," with the Polynesian rang, " man," and the Chinese lang, " man." The Japanese say obui, "carry on the back," Chinese pet. In Turanian languages the prefixed vowel is the same as that of the root syllable. Mongol olos, " people," \ao?, leute. So in Chinese the colloquial word for " elder brother," is aha, where the prefixed vowel takes its quality from that of the root ka, the old word for "brother." This is a very old law of change, for it appears also in the Semitic ahh, "brother," ab, "father." In the triangle of the three primeval vowels a, i, u, the letter stands between a and u, and is liable to change into either of these vowels, or into the inter- mediate values 6, e, and the in " gone." The old Chinese has become u in the modern language. The modern Chinese has come out of a. The Mongol has in the eastern dialect the values o and the in " gone." The Chinese is usually the in " go." The old Chinese agrees with the European 0, as in THE ALPHABET. 335 rota, rotation, Chinese Ion, " wheel," " revolve," lot, " anything round," now changed to lun, lu. The Greek letter o was taken from the Phoenician ayin, of which the old sound was ng and g. Thus T\ty, " he sang," " he answered," " he spoke," is by Gesenius identified with cano, " I sing," but may as probably be compared with the Chinese =j ngen, "words," "to speak." So DM "congregate," must have been anciently pronounced gamam. It is identified by the same grammarian with the root in yd/jbos, cum, cumulus, which is the same with the Chinese Jgj/ gam, " collected." Thus the Hebrew ayin was first g, then it became ng, and was afterwards dropped or changed for a vowel, usually o. P. P in Chinese rests upon b as its base. No widely extended roots with the initial p are without repre- sentatives in the old sonant series. Thus pang, " to tie," has bog among the sonants, with the same meaning. Compare the European pack and pango. In the Mongol syllabary b is the normal form of this labial. The aspirated and unaspirated p grew out of b in the Chinese, Tibetan, and Sanscrit syllabaries. In the Semitic languages p, v, and /, appeared on the base of b. The Greeks, however, assigned the values b and p to the second and seventeenth Hebrew letters. We must therefore suppose b to be older in Hebrew 336 china's place in philology. than v, andp than/. All Latin, Teutonic, and Persian words in /, can in Chinese and Mongol only have equivalents in p. Thus fugio, fliehen, cpevyco, are in Chinese bik, in Mandarin pi. The Japanese equivalent for the European and Chinese p is h. The Egyptian has p for the Hebrew /in apta, "bird," Hebrew £)ij}. Q. The letter q is ku or kw, and its existence is a proof that the Phoenician alphabet was once syllabic, and perhaps it may be concluded that Cadmus made use of that alphabet partly at least as syllabic. 1 The Chinese kwa comes from an older ku, kwo from kok, and so on. Hence kw is modern, and of no use in tracing etymo- logies on the Chinese side. R. This letter has appeared recently in Chinese. It shares with j the possession of the inheritance of words once belonging to the lost initial ni> as j^ ni 3 "son," now called ur, or er, or rh, as it is differently written, Turkish ugli, Mongol hubegun. In Japanese r repre- sents the Chinese I. In Mongol and Tibetan as an initial it seems to indicate a Semitic origin ; for the 1 Professor Key, in English Cyclopaedia, Art. Q. Yet this hypothesis fails to explain why the Hebrew Jcuph is used as a final, as in pHJD f " was sweet." It may be remarked here that Jcuph in this example is a suffix, the root being mat, "honey," "sweet," the Chinese mit } and Greek /xe0u. THE ALPHABET. 337 words in which it is found, e.g., Tibetan rab, " high," Mongol airiben, " many," Mongol oregen, " broad," Mongol orosiyahu, "to be pitiful," as compared with y) f rab, "high," "great," "many"; im, rahab, "broad"; mH, ratsa, "treat kindly." R is inserted commonly in roots after the initial in the Himalaic, Semitic, and Indo-European languages, and before the final in the latter of these families, as in crow, Latin corvus, Sanscrit hdka, umbrella from umbella, sprache from speech, world from welt. Before com- paring roots, this inserted r must be everywhere first eliminated. The comparison can then be conveniently made. Thus sprache, speech. Omit the prefix s and the inserted r. Change the guttural ch into the k or g from which it sprang. The root is then pak, which in old Chinese means "to speak," and is so used in the modern Shanghai dialect. R is also a common suffix in Himalaic, Turanian, Semitic, and Indo-European languages, as in Tibetan charpa, " rain," from c l hu, " water," Hebrew kaphar, " to cover," Chinese hap. R and I are in many respects much alike. The European prefix r is in Chinese /. Thus, ros, regen, rain, are in Chinese lu, " dew," in old times lok. As / came from d, so also did r derive its origin from that letter. The Hebrew T\^, "weave," is texo in Latin and tek in old Chinese. I suppose both these words to have had formerly an initial d. Thus, $V\, 22 338 china's place in philology. rosh, "head," is Chinese dud, in modern times t'eti. The Semitic sh is commonly convertible with t, and was perhaps derived from it. The change, however, might be the other way. The Aramean, which used tarn for " there," is usually supposed to be newer than the Hebrew, which used sham. To judge from Chinese analogy, the most widely spread at the present time should be regarded as the newest of the Semitic lan- guages. Further, as Abraham came from the land of the Chaldees, the language-forms preserved in the Nineveh and Babylon inscriptions should be regarded as older than the Hebrew. If so, t might be the older form. R is also introduced as a second letter in Semitic roots. Thus, barueh, "the blessed," where the root bak agrees with the Chinese p>ok, "happiness." S. The letter s is freely introduced as a prefix before the radical initials k, t, p, I, m, n. Thus, small is the same word as minus, [iifepos, and the Chinese mi, "little." It sometimes comes in place of e, as squire from equerry, the Latin equites. In Sanscrit and Zend s stands for k, as in the old name Massagetae, where Massa is Mahd, or magnus, " the great Getae." In Latin s final stands for t, as in patior, passus. In Russian s final stands for k, as in sosat, " to suck." In Grerman it represents t, as in beissen, " to bite." THE ALPHABET. 339 In Hebrew both samech and sin (having each the value s) interchange with t. Thus, DD£, pasas, " diffuse," is in Arabic basat. The older form is t. The Chinese s corresponds to that of Europe in words such as su, in kau su, " to tell," old form sok, as compared with sagen, say. So also Chinese sat, "scatter," "sow," Latin sero, satus. This root is in Persian zed, as in ghemzeda, " heaviness-dispelling." So also Chinese sok, English seek. Sometimes the European s is recognized in the Chinese ts or t's. Thus sot, " a drunkard," is tsui or tsot, " intoxicated." The Chinese s becomes t in Cochin- Chinese, and generally in the Eastern Himalaic and Malay system. SH. This sound was not employed by the Greeks and Romans, and they did not, therefore, need an alpha- betic sign to represent it. In the Turanian languages it is also very sparingly used. As it is fully developed in the Chinese, Semitic, and Himalaic families, the cause of its non-appearance in Greek and Latin may be probably traced to Turanian influence. It has struggled back into existence in the French, where it appears as the representative of the Latin k, as in color, French chaleur. The Chinese sh is sh in Teutonic, Sanscrit, and Sclavonic languages, and s in Greek, Latin, and 340 china's place in philology. Mongol. Thus, schiessen, "to shoot," is the old Chinese shet, " an arrow." ^, ^, and !§•, in Mandarin, ski, she, she, and meaning " arrow," " shed," " let go." How many summers and winters have passed since the ancestors of the Teutons and of the Chinese parted from each other, each with their vocabulary of common words, such as to shed, to shoot, a shed, etc. ? It is mar- vellous that, after so many ages, Time's defacing fingers have not yet destroyed the traces of original identity. The German sch is often softened down from sk, as in schreiben, the equivalent of scribo and fypdcjxo. So the English sh comes often from an older k, as in " wash," Chinese ok, Mongol ogahu, " to wash," and perhaps the Greek vypos, vypalvco, "moist," "wet," "to wet," unless that comes from sole, our " soak." The Chinese sh is sometimes represented by the European k, as in eado, " to fall," in Chinese shwai, in the old form shat, and probably the same as the word " to let go," she, given above. For all verb-roots are capable of assuming the causative, transitive, intransi- tive, passive, and reflexive modes. In Semitic languages and in Chinese dialects, sh is apt to change to s and to t. The affinity of sh for k appears first in Sanscrit and then in Europe generally. At present, however, in the Mandarin change now gradually taking place of ki and kii, to chi and chu, as also in that of hi and hit, to shi and shu, we see the budding of a similar principle. ill l : ALFHAE We also sec sh pushing its conquests in the Turanian area, as in Manchu, whore it represents the M< aspirated ch. This cJi aspirated is iii the Mongol area the eastern representative of the northern and western ts. For the Buriats, Kalkas, and Kalmucks, all prefer fs, which appears to he the older and typical form. The eastern c'h may therefore he regarded as a of a tendency to introduce sh, appearing at the east end of Mongolia. Sh proper also occurs in [Mongol words commencing with si, which are softened into shi, as shidorogo, 1 "honest." Such is the law in Japanese also. The letter sh is thus seen in these three languages asserting its lost existence, and winning back its ancient dominion, as in France, Spain, and other portions of the Latin area it has also been seen to do. T. The letter t comes in very many cases from d. Thus the Japanese kita, "north," is in [Mongol hofai, "behind," "north," hy'em, "behind," "after." Here j is d, and the old Chinese word would he gud, in the modern clipped form heu. So trctvu, "to tread," ifl in Chinese dat. The modern ti, "brother," tau, "reason," are from older words do, do. The letter t occurs for the Chinese sh and s in the eastern llimalaie languages. 1 The Chinese shih, tl real," old form zhit. 342 china's place in philology. This letter early became t', a form which probably appeared in the transition from d. The Sanscrit, Tibetan, and Chinese, have it in addition to t and d. In Turkish, Mongol, and Manchu, it is the normal form of t; that is, every t is aspirated, and sounds like the union of h with t in " anthill." The Semitic form of this letter was th, as in our " thin," " thick," and the Greek in r/%u, " I place," which occurs for the Mongol d, as in OaXdcraT], dalai, "the sea." T often precedes s in Japanese and Chinese, and in such cases is often aspirated. This compound letter, when not aspirated, is the German z in zeit, " time," and when aspirated it is the Mongol t's softened by the eastern tribes into &h. The German z or ts is derived from t, as is the Hebrew. For example, *)*]'¥, tsor, Tyre, was by the Greeks, Latins, and Arameans, known as Tvpos, Tyre. Probably in this case t was the original sound, but this is not certain. U. The letter u, like the other vowels, is prefixed in Mongol to roots. In modern Mongol it takes the place of e, as in umun, " before," in old Mongol emun. It corresponds to the Greek ei, as njihu, " to see," elhov, ecSo/jiai. Here / stands for d, and hu is the sign of the infinitive. THE ALPHABET. 343 In Japanese and Tamil u is prefixed to roots, as in uma, u horse.' ' The modern Chinese u (wu) has lost m from before it in many instances, as in mo, " not," now wu. It has also taken a prefix before it very frequently ; for example, e in du, " sorrowful," modern sound ch'eu, Latin dolor, doleo. The old equivalent of the modern Chinese u was o. The modern diphthongs au, eu, iau, ieu, were formerly o, u, o, u. A modern form of u is u\ This vowel appears in no ancient alphabet, so far as can be known. Its place in the triangle of which the angles are the three primitive vowels, is between u and i. As it has replaced u in some Chinese and Mongol words, so has it done in many French words of Latin derivation. r. The letter v has taken the place of u in many Latin words, as vereor, " to fear," Chinese wei, Mongol aimoi. The Romans, however, pronounced it w. The Germans have also changed initial w to v, retaining the written symbol unaltered. The English are right to keep w in " was," " were," " will," for the equivalent words in Chinese and Mongol have no trace of v. The Germans write werden, ivollen, was, correctly, but they are wrong in the sound they give to the initial w. The Greek digamma was v or w, as in FeiKoac, viginti, 344 "twenty." Here the d of the full form duikosi was dropped, and u became F, and was afterwards lost. That the Sanscrit v was a vowel seems to be deducible from facts such as that the suffix of the dative case was ye or ve. The old Chinese v was the sonant form of /, and as such came from b. In modern Chinese it has become /. Its equivalent in western alphabets is b or p. In Hebrew, beth took v as one of its values. This was followed by the Greeks, who used their beta for words written by the Romans with v. So at present the Russians for the sound v write the sign b, herein imitating the Greeks. The Hebrew vav was formerly w. The Tamil v, like that of Sanscrit, represents the w of old Chinese, and the ui of Mongolian. The German v is an/, as in Vogel, "bird," Vater, "father." The English v is a German b, as in eben, even. Thus it appears that in Hebrew, old Chinese, English, and German, v rests on b and/. In Latin, Sanscrit, Tamil, and Russian, it is a mispronunciation of the primitive w. W. The letter w belongs as an initial to the old and new Chinese. In the old Hebrew alphabet it was vav, after- wards pronounced v. In Greek the Chinese w appears as a vowel. The root wan, " to bend," appears in the THE ALPHABET. 345 Latin vinam and Greek olvos, " wine," from vinea, " a vine," " that which bends." The English and Greek keep the primeval w. In Latin as now pronounced, and in French, it is changed to v. The European w, as preserved in the Saxon part of the English language, and in the ancient Latin, is represented in Chinese and Mongol by a vowel initial, which may be a, o, i, or one of the corresponding consonantal values, w, y. Thus vulgus, "the people," is in Mongol olos ; vacca, " cow," is uher ; video, " see," is ujimoi, etc. So also wail is in Mongol weilehu ; " was," " were," are in Mongol alio, " to be," weiledhu, " to do," (the substantive verb here assuming an active character,) and in Chinese wei, "to do," or "to be." The German wechseln, " to change," is in Chinese yik. Here the y is a modern prefix. A form still more modern is /, as in Mandarin. The Latins had vicis, "change," viz, vicissim, etc. In the Greek a/a? of irevTaias, " five times," we have the same root meaning " times," and it may be the origin of the Latin es in vicies, "twenty times," and the English ce in once, twice. The letter w is inserted after the initial of a root. This seems to occur through a tendency in the vowel u to become a consonant. Thus Jcu, "a melon," in Europe " cucumber," " gourd," etc., is in modern Chinese kwa. The vowel a is an addition, and u appears as w. So in our word sweat, w represents 346 the vowel u of sudor and sudo, and ea is inserted. The same thing takes place when the first letter is u, as in uge, the Mongol for " word." This corresponds to voco and vox in Latin. The vowel u becomes a consonant v, and o is inserted. So in the Chinese yue, " say," in the old form wat, the English equivalent is word, where r is inserted. The Latin is aio, " I say," the Greek avBrj, "voice." The Hebrew yadah, "praise," may be the same word. The original primeval root was pro- bably ad or ud. From this the Chinese formed Q and gljj yue and wei, both anciently called, we may suppose, ud, and afterwards yet and wat. In the mediasval dictionaries they are read yet, wei. A similar change took place in Semitic words. Thus the city Erech, in the old Hebrew Ark (where the initial vowel was perhaps intended to be repeated in the second syllable, so that it would read Arak), is now Waraka. The Arabs have prefixed w. The Latin form was Areca. It means "the long" city. The pointing of the grammarians, directing it to be read Erech, cannot represent a very ancient pronun- ciation. X The letter x and the ksh of Sanscrit represent com- binations of k and a sibilant. They are unknown in the languages of Eastern Asia. The Greek f took the place of the Hebrew Samech between n and o. Hence 1 III. AI.I'II \l 3 17 ire Learn that the base was looked on as .-?, and k ai an uldition. But its real value was nearer to /.• than to I. This is shown by such words as fwo?, " common," prhieh Lb the same as koiv6$. If, then, Palamedes and !iis associates, who are said in the time of the Trojan War to have added 6, f , (/>, ^, to the alphabet, proceeded 'o give this letter its position on the supposition that it is modified from s, they were mistaken. It is, in fact, formed by inserting B after /,-. Thus, typos, " dry," s convertible with o-^epo^, a/cripos, and ^epcro?, all meaning "dry." In a similar way y\r, j>si, is formed ■rom p by inserting g, and in ITcbrew U from / in the jame manner. Thus, ^u%v, " the breath," " soul," is :he Chinese p'ak, "the corporeal soul," which is distinct Tom the hwutl or gun, " the immaterial soul." This asl word by interchange of final n with final/) : jut, and maybe identified with hwei or /•« /, "ghost," 1 geist." So also yfrvxpos, "cold," is the Latin < rigidus f r being inserted in place of 8 in the more rn form. The Chinese and Tibetan are like the Semitic family n not taking t after h or />, hut only after /. Y. The letter y is j viewed Bfl a consonant. Vowel nitials liave a tendency to assume a consonantal form. >f the three prime TOWels, I takes//, '< tal \d >i akefl either //, as in the Tibetan yab, "father," from the 348 Semitic ab, or w, as in the Chinese wan, " bend," " circle," when compared with annus, " a year," or ng, as in Mandarin ngan, "rest," from an, the old form. The Chinese y is/ in Latin, as yik, "to throw," jacio, English jerk. In Greek it agrees with the unaspirated i, as I'&o?, " alone," Chinese yit, " one." Here, too, we see the probable origin of idem, "the same," "identical."' It is the Mongol adeli, "same," and the Chinese yit, " one." The vowel o also is apt to take y before it. The Mongol ogahu, " wash," is in Chinese yok. The Chinese y. and Latin j are in Sanscrit y and in • Greek f. This makes some confusion, for one of the most common values of j is d + zh, and it is formed from an older d with zh inserted, as Jupiter from Dili. In £Sei;?, a dialectic form of Zevs, the sibilant is placed before the initial d, instead of after. Hence j in Latin and z in Greek sometimes come from d, and at other times from y. In German j has the value y. In English y is used as in yoke, German jock, except in words of Latin derivation, as juvenile, which in Chinese is yen, old form yu or u. Z. The letter z may be connected with s, d, y, t, or k. Old Chinese words in z (that is, in the sonant series of s) have now become s. THE ALPHABET. 349 In the Hebrew vocabulary words with samech, tsade, and shin sometimes take zayin. Thus, ziir, sur, are both used for "to return." In Hebrew the occasional origin of z from d, like that of s, ts, sh, from d and t, may be shown by examples. Thus, KD, baza, " cut in two," is by Gesenius compared with the Sanscrit bhidh, "to cut." Compare also the Greek pi£a, " root," with radix. In Tibetan zang, "copper," zab, "deep," zar, "fork," seem to be allied with the Chinese dung, "copper," the European deep, and the Chinese Pa, "fork." The connexion of z with y has been already noticed in speaking of the Greek zeta under y. In German t has become z, and is then pronounced ts. In Russian final k in a root often becomes z, as in lizat, " to lick," German leeken. The letter zh is otherwise written j, as in the French jamais, jour. In Chinese it has grown from an older ni. It must be looked for in European vocabularies as n, h, h, etc., as stated under /. Common Radical Syllabary. The common roots of the Chinese and European languages consist of monosyllables. That all roots are monosyllabic was known by philologists as the result of the comparison made, in the first half of the present century, of European languages with those of Western 350 china's place in philology. Asia and India. But when the roots of European speech are compared with those of China, they assume a definite shape, at the knowledge of which philologists, while they hesitated to cross in their researches the Imaus and the Himalayas, could scarcely arrive. Roots may be first arranged in two groups, those which end in vowels and those which end in con- sonants. Among examples of the first are words such as a, used as " I," and as a verb " to be," ba, " father," ma, "mother." Of the second are bad, "divide," " other," kab, "cover," " head," nig, "hide," "black," dak, " cover," bang, " strike," " noise of slamming," kan, " tube," " straight stick," om, " dark," " shade." All the vowels interchange, but the chief lines of distinction are between a, i, u. Thus we have among the open syllables a triple division made by these primary vowels. The closed syllables in ultimate roots are chiefly formed by the six consonants g, d, b, ng, n, m. The initials are the three vowels, the six consonants just mentioned, with /, z, and zh. The syllabary, with these elements, would consist of, (1) three vowels; (2) eighteen biliteral syllables with consonant finals ; (3) twenty- seven biliteral combin- ations with vowel finals ; (4) 162 trilateral combinations with consonant initials, vowel medials, and consonant finals. In all there would, with these elements, be 210 combinations. COMMON RADICAL SYLLABARY. 351 This is the smallest number of syllables that we can allow for the common syllabary, unless we also eliminate /, by deducing it from d. If we add to the initials k, t, p, s, z, w, y, there will be twenty-one more biliteral combinations, and 126 more triliteral. In all 357. These are perhaps the most probable and convenient limits for the common syllabary. It would be unwise to extend the finals by adding k, t, y, for although in European roots the difference between k, t, p, and g, d, b, is recognized, it is not so in any Chinese dialect. The Shanghai people pro- nounce g final before a sonant and k before a surd. The difference depends on position, but the fact indi- cates the possibility of g, d, b, having once occupied an important place in the Chinese syllabary which was afterwards lost to k, t, p, as these have in their turn resigned their position at the ends of words in favour of the vowels. Examples will be here given tending to show that it cannot be learned from the European roots whether k, t, p, and g, d, b, all belonged to the primeval syllabary or not. The English reed, German rohr, is in Chinese lu and lut. The Latin rota, rotundas, and the Greek pvOpos, aptO/jios, with the English round and Latin arundo, " reed," all come from the same root. The finals t, d, nd, are found interchanging in European Ian- 352 china's place in philology. guages, while in Chinese, where lun is "wheel," and lut "anything round," n and t interchange. A fair inference is, that we cannot tell whether t or d was the original final, but that n and that final were inter- changeable before the Chinese language separated from the Indo-European. So with the Chinese yok, "to desire," when com- pared with the Latin acer, French aigre, and English eager, we cannot tell if k or g is the older. The arts of life had sufficiently advanced, when the Chinese separated from the Indo-Europeans, for the names of boats, of agricultural processes, of weaving, of houses, of the physician and the necromancer, to be the same. Take teoco, " to weave," Chinese tak, meaning " to weave cloth," or " to weave a hedge " of willow branches or bamboos. The corresponding European words do not appear ever to have d or g in them. Hence it may be. inferred that in this case t was the form of the initial previous to the separation of the races. The Russian is tikat or that, " to weave." The Greek has rev^o?, " a wall." At that distant time wooden cups were in use, which were called pat, jjffi the Sanscrit pdtra, the Latin patera. "A boat" was bat, and "an oar" was lut, iper/jbov. " A house " was ok, oIko<$. " To heal," and " a physician " were both it, larpos, Idofuu. " A magician " was ma, the Mongol bo, Persian magus, EUROPEAN RADICAL SYLLABARY. 353 and Dravidian bagai. "A dog" was h'on, "a cow," gu. " A coverlid " and " to cover " were bed, the Latin pallium, and English bed. "To clothe" and "clothing" were wit, the Latin vestio, vestis. If boat and paddle (Latin batillus), and bowl and patera, are connected, it seems hopeless to expect that the original form of the initial, whether b or p, can now be ascertained with certainty. European Radical Syllabary. The European families while still one with the Hindoo prefixed s and sh to the initial consonant of many roots, and also inserted r or I after the initial in many more. To the six final consonants of the roots, which were originally h, t, p, ng, n, m, were added s, sh, r, I. Further, r and I were often inserted before the final consonant of the root. These processes were common to the Semitic and Indo-European systems. In the Semitic system the result was a vast formation of dissyllabic roots con- sisting of three letters each. Sibilant prefixes, the insertion of r and I, the duplication of certain letters, and the addition of r, I, s, sh, p, m, k, h, and perhaps others at the end, made that formation what it is. There is no trace of sibilant prefixes in the Turanian languages, nor of the insertion of r and I after the initial consonants of roots. In Eastern Asia sibilant prefixes occur only in Tibetan and Burmese, and the 23 354 CHINA'S PLACE IN PHILOLOGY. insertion of r and I only in these languages and in the eastern Himalaic family. We may, therefore, refer the sibilant prefixes and the insertion of r and I to Semitic influence. At least, these phenomena first make their appearance in that family, and the example was followed in the Indo-European and Himalaic systems. Examples : Hebrew shakab, 1 " he reclined," kvtttq), cubo, cumbo, English scoop. Hebrew sagab, " was high," gabab, " was high," shafat, " he judged," from pat or bad, " divide." Turanian influence on the European root appears in the finals, where r, I, s, are found as in Semitic. Examples : Mongol agola, Manchu aim (where the g of the root is dropped, as in colloquial Mongol), Sanscrit girt, Greek 6po, apcTTjv, " male," Mongol ere, Greek ttuXls, German burgh, berg, Mongol balgasun and huliij. Here the root is balg or balig, and the insertion of I is of Turanian origin. Khanbalik, the Turkish name of Peking in the time of Polo, mi " the city of the Khan." The European root shows a special and independent activity in its great extension of the sibilant pxefizet and of inflexions, and in the great variety of its initials and finals. The European Word. In assuming the polysyllabic form, the Eoropi word followed the Turanian analogy rather than the . i tic. This is signally manifest in the formation of derivatives, of case auffizeB, of the polysyllabic tenses and moods of verbs, and of the greater portion of the parti Yet the Semite influence ia very apparent in the introduction of strong preterites, doubled consonants, and all tense forms where the change of the vowel is a characteristic. The English preterite in 00, M, S, or o, etc., from a present in a, i, ete., may be accounted lor most satisfactorily in this way. This principle of change in the vowel — as in teethe, eodden, ttand, stood, sau/oi, geeandt — occurs Less prominently in Greek ami Latin, where Xeiira) lu ad ajmnjo 356 china's place in philology. spergo in disjiergo and other compounds. In the first of these examples the change fixes the tense, in the other it depends on laws of accent and quantity. In the Hebrew such changes distinguish tenses and moods, and so we find it in the Tibetan. We must suppose, then, that the ancestors of the Germans, Greeks, English, and Tibetians, adopted this mode of marking tense, mood, and conjugation, from the Semites. The distinction of masculine and feminine is also of Semite origin, and with it the idea of dual and plural numbers. The conception of mood and tense is chiefly Turanian. To this the Indo-European has, as its own contribution, added the distinction of voice, the augments of the past tenses, an increased number of tenses, and a very full development of the personal endings. A Greek verb has in its imperative the simple root, as found in all languages. Its particles and infinitive, past, future, conjunctive, and other forms, are Turanian. They are made by verbal and pronominal suffixes, in many cases identical with those used in Turanian languages. The theory of the conjugation of verbs rests on the mode of viewing the verb. It is regarded as a substantive, and the infinitive and participles were apparently first formed, the verb being here more concrete. From them came the past tenses of the indicative, in the manner already described in pre- ceding chapters. THE EUROPEAN WORD. 357 The formation of compounds reveals to us the principle of juxtaposition, as in the oldest stems. Thus, in XevfcoaroXos, " white- robed," the law of order is as in Chinese and Turanian. Where a preposition combines with a verb or noun to form a compound, the principle of order is Turanian, and not Chinese. Thus, /caraTrarea), " to trample down," is in Chinese chat hia, or, as the ancient sages would have pro- nounced it, dat ge, where ge, the equivalent of Kara, comes last. The English agrees with the Chinese, and the principle of arrangement is that of the juxta- position of two verbs in the order of time. But in English the word down, originally a verb, has become an adverb. In the Greek compound the principle of arrangement is Turanian. The word Kara, originally a verb, " to go down," and the same with the Chinese ge, or hia, and the Japanese kudaru, "go down," is here found in the position of the adverbs of space and direction, as in Mongol dotal tebi, "place below." So in English understand, in German verstehen. So in Russian nishodit, "to go down," where ni, "down," is connected with nijnie, "lower," as in the name Nijnie Novgorod, literally "the lower new city." The English down is the Latin de, the Chinese ti, "bottom," and the Mongol do, "below." Thus the Indo-European languages in their prepositional verb compounds use a Turanian law of arrangement, while giving to the pre- position a verb force which is peculiar to those languages. 358 European Syntax. The syntax of the European languages is a mixture. It contains Chinese, Semitic, and Turanian principles. The order of words is either natural or inverted. "Where it is natural, as in " William's son," " tall man," "William struck Thomas," "quickly fly," "with a stick William struck Thomas," the order is usually Chinese and primeval. Where it is inverted, as in " the son of William," " un ouvrier industrieux," " du sollst Gfott, deinen Herrn, anbeten," 1 " thou shalt worship God thy Lord," it is by principles derived from the more ancient Semitic and Turanian families. The post-position of the adjective, genitive, and adverb, is Semitic ; that of the verb is Turanian. The effect of Semitic influence is seen at its maxi- mum in the translations of the Scriptures made in the sixteenth and seventeenth centuries, e.g., in Luther's, we read, " Aller Augen, die in der Schule, sahen auf ihn." The English version reads, "And the eyes of all them that were in the synagogue were fastened on him." The relative clause in both cases is thoroughly Semitic. The only feature not Semitic is in the German, in the order of the words aller Augen, " eyes of all." The Mongol would be horal on dotora baiksan hwun bugudeger nidun yer teguni sirtabai, which might be translated word for word thus, "Synagogue's within being men all eyes-with him observed." All refers to 1 Luther's Bible, Luke iv. 8. GREEK. 359 men, and with to eyes. Him is the accusative after observed. The participle being is in the past tense, and here performs the duty of the relative. The same Semitic influence, however, appears in Homer. Thus, in ava% avSpoov, "king of men," the genitive comes last, and the adjective in Sovpl re /xaKpS, " with a long spear." Greek. The Greek seems to be specially founded on the Chinese in regard to tones. For what are the acute, grave, and circumflex accents but Chinese tones ? Yet we have found tones existing in the Himalaic lan- guages and also in a dialect of the Mongol. It is, therefore, uncertain from what source in particular the Greek accents are derived. Yocal sounds are necessarily either emphasized or slurred, even or inflected, high or low, long or short, or, in other words, admit of distinctions in emphasis, tone, pitch and quantity. In modern Europe the quick rising inflexion or tone is appropriated to ques- tions. When and where this began it is difficult to say. It would not be much needed in ancient Greek, for in that language very commonly interrogative words were placed at the beginning of sentences, and this inflexion was probably the proper sound of the acute accent. The grave accent would then be the quick descending inflexion heard in modern Europe 360 china's place in philology. as the tone of commands. The circumflex would be a combination of the rising and falling inflexion. Long and short quantity may be illustrated in Eastern Asia by the distinction between long and short tones in the south-eastern dialects of China, and that between long and short vowels in Sanscrit. The Greek circumflex was attached only to syllables long by nature. The Chinese circumflex or double inflexion may be applied to any word, according as it happens in the local habit of any dialect to be appro- priated to this or that tone class to which the word in question belongs. The occurrence of a long vowel in the last syllable of a Greek word necessitated the change of a circum- flex to an acute accent in the penultimate. Thus, evy€ became evya), and oho? became olvov. This change resembles the change of inflexion noticed some- times in Chinese compounds. In the Peking pronun- ciation of shui sheu, "sailor," the application of the lower slow rising inflexion (which properly belongs to them) to both words would be unpleasant to the ear. The first word takes instead the upper quick rising inflexion. In Pekinese, as in the ancient Greek, it is easier to pronounce a long inflexion before a short syllable than before a long one. Prefixes threw the accent back. Thus, tvittw, " I strike," became ervirre, " he struck." iraihevros " taught," became aTralhevros, "untaught." So in the LIST OF GREEK AND MONGOL COMMON WORDS. 361 Pekinese dialect, composition often deprives the last word of its emphasis. Thus, Yamen, " public office for despatch of business," which originally means " flag door," is emphasized on the first syllable, which keeps its proper tone, the upper quick rising inflexion, while the last syllable, requiring the same inflexion, is slurred over. A prefix in this instance deprives the word men, " door," of its tone. The special resemblance of Greek to Chinese and Mongol may be judged of by the following examples : avrSs, "he," "same," "himself," Mongol adeli, "same," Latin idem. oIkos, " house," Latin vicus, "village," Chinese ok, " house." 6pd(i>, "see," Mongol harahu, "see," "look at steadily." 6d\ao-(ra, " sea," Mongol dalai, " sea." tyevSos, "lie," Chinese put, "not," Latin falsus, English false. ep(o. Beat, ffi fa, bat, Latin batuo, Eussian bit. Bed, U? pei, bi, bit, " to spread," " a covering." Black, |H me, mek, " ink," " coal," " that which is black," Greek fxekas, Sanscrit malina. Boat, $g fa, bat, Anglo-Saxon bat, Eussian bot. Bow, ffi fu, bok, German bogen. 376 Break, ||§ pH, p f ik, " cleave with a hatchet," Latin frango^fractus, Sanscrit bhagna, "broken," Hebrew J?pS, "cleave." Bright, £j pe, bak, Sanscrit bhaj, "shine," Latin fulgeo. Burn, ^ fen, bun, Latin pruna, Greek irvp. But, boot (to add), $%. pei, pi, pit, Anglo-Saxon botan. Buy, bought, j| mai, mui, muh, Latin emo. Call, HIj* tow, #0, kok, Greek /caXeco. Can, " a cylindrical drinking vessel," Jg kwan, kan, ^ &aw, "a pipe." Hence anything long and round, as channel, through the Latin canalis, and cane, from canna, in Chinese jpfl ta, " a stalk." Certain, ^ kiue, kit, Latin certus. Chaste, gg /he, M, " clean," Latin castus, Greek Cough, jig &' " low," " bottom," Latin deorsum, Mongol ddra. LIST OF ENGLISH AND CHINESE COMMON WORDS. 377 Drag, draw, dray, j£ t'o, Pa, Latin traho, tracto, German Ziehen. Eager, ffi yu, yok, " to desire," Latin acer. Ear, hear, 3=p rh, ngi, Persian gosh, Sanscrit ghosha, Turkish kulak, Latin auris, German ohr, horen. p Eat, Pg wei, wid, tvat, Sanscrit annam, "food," Mongol idehu, " eat," Latin vescor. Elk, H» lu, lok, Eussian los, German elch. Embrace, Q pau, pok, Latin amplector, brachium, "arm," Greek 7n}%i/?. Fast, >gj pi, pit, Mongol bedu, ° firm," Greek ir&rros, Latin fides, Sanscrit bad, "to be steady," Hebrew /tO!p, "he trusted." Father, $&fu, bo, Latin, pater, Hebrew a b, Turkish baba. Flee, jjftfcpi, bi, bik } ~La,tmfugio, Greek fevyco, German fliehen, Russian biegat, "avoid," Hebrew ITl^. Fly, flit, Jg fei, pi, pit, Sanscrit patat, "bird," Greek TreT€ivo<$. Foetus, JJ p'ei, pH, p'it, Latin foetus. Fold (as in two-fold), fg pel, bei, bit (as in san pel, "three-fold"). Forth, fg fa, pat, " express," " go forth." Foundation, ;$* pen, pun, Latin fundamentum. Gather, Jf hivei, git. Give, ^ Arc, ^J9, German grefow. Glad, -jij A*, ArcY, Latin gaudeo, gratus, Greek yrjdco. Go, gang, ft hing, gang. Goose, $§ ngo, Russian gus, Mongol galagad. 378 Grip, grasp, 7^ kia, kap, "take under the arm or with tweezers," Latin capio, habeo. Gullet, P^ lieu, gu, Sanscrit gola, Latin gula, German hals, "neck." Hate, fj hwei, git. He, :& h% gi, Latin hie, Hebrew &On. Head, ^ kia, kap, "coat of mail," "first in rank," "cover," H kai, kap, "covering on the top," caput, KefyaXrj, kopf, haupt. Hem (as a substantive), $§5 kin, kim, "hem of a garment," "a boundary"; (as a verb) ij| kin, kim, "to prohibit," "restrain," Russian kaima. High, "^ kau, kok, Latin celsus, German hoch. Hollow, Jjj| hu, ku, Latin cavus, German koi\os. Hook, ^ keu, kok. Hoop (cooper), $S ku, kup. Horn, j| kio, kak, Latin cornu, Greek Kepas, Hebrew VX)^ Sanscrit sringa, Persian shag. Hot, |& je, nyit, Mongol halon, Latin calidus, German heiss. House, ^ kia, ke, Latin casa, Mongol gere. Humble, jf| kHen, k'im, Latin humilis. Hymn, B^ yin, gim, " to chant," Greek vfivos. I, ^ ngo, nga, Latin ego, German ich. Kick, gfl kio, kak, "foot," Welsh eic, "foot," ciciaw, " kick." King, ;g W», to, Welsh kun, " a chief," German Konig. LIST OF ENGLTSH AND CHINESE COMMON WORDS. 379 Lake, ^ che, dak, Latin lacus. Lamp, j;f{ Ian, lam, Greek Xapnras. Lath, g£ lie, lit, " to split/' " a rent," or " slit." Law, jg! li, li, leg, Latin &#, Greek \6yos. Leaf, ^ ^e, dip, "butterfly," (so called from its leaf- like wings,) jg tie, dip, " fold one thing over another," Mongol lap&hi, " leaf." Lick, ^ c'hang, dong, Latin /m^o, to', Greek Xe/%o). Long, J| c'hang, dong, Latin longus. Mill, jg mo, ma, " grind," Latin wofe. Mother, -g: ww, mo, Greek /jb^rrjp. Much, j|l mo, mo&, "abundant" (used in poetry). Muck, j% mo, mok, " dust." Paint, ^^ piau, pik, " to draw," " to adorn," Latin pictor, pice, pingo. Pair, |jjj p ( ei, p ( i, Latin js9#r. Part, glj ^e, jmY, 6#, " to part from," " different," Latin pars, portio, partio, Hebrew /*7?> " he divided." Paunch, JjJ fu, bok, German bauch. Peace, £J5 p'ing, bang, " even," " peace," Latin pax. Peel, )fc p ( i, ba, Latin pellis. Peg, pierce, Jlj " to pierce by setting on a spear." Compare the words prick, pick, spoke, poke, pike, with the Sanscrit pij, " kill," Greek m/cpos, Latin pungo, pugno, German fechten. Pledge, jg pHng, bang, " lean on," " proof," Latin pignus, German pflegen. Pot, $5 pei, put, " cup," Sanscrit pdtra. 380 china's place in philology. Prepare, fl§ pei, hi, Mongol belehu, Latin paro. Put (in put forth), fg fa, pat, " go forth," or " be put forth." Quiet, ^ hie, kit, Latin quies, quietis. Quoth, fjjf hwa, gwat, " say," " words," Sanscrit kath, Latin cedo. Reed, rod, |j lu, lut, German rohr, Latin arundo. Right, ]g clii, dik, " straight," Latin rectus, Greek BUaios, Sanscrit dakshina, " right." Ring, fjf $wgr, Jwgr, "collar." Round, fj| &m, "revolve," "awheel." Row, 3H fe«, fo, fo£, " a scull," German rwtfer, " oar," Latin renins, remigo, Greek eperfio?. Rude, j§» he, lod, Latin rudis. Rule, g| K% " to govern," %, Latin regula, rego, rex. Sad, '[$ few, efeo£, " sorry." Same, § 2'saw, ^saw, saw, "blend with," "be one with," Latin similis. Satisfy, j( sh'i, zhit, " full," " real," Latin satis. Say, gjp sw, sok, " tell," German sffaw, saga. Seed, ;jJH s#, satf, Latin sero, sator, Sclavonic syet, "sow." Seek, §^ so, sok, " seek," German suchen. Compare search. Self, fj taj, efee, zi, Latin se, German se$s£. Serve, Jp sA'i, *A£, Latin servo, servus, Sanscrit shach, " to serve," sri, " to serve." Set, |g s/*e, sM, Latin sisto, sedeo, Hebrew TV&. Shed, <§> she, shed, " cottage," §f s/^', sM, " house." LIST OF ENGLISH AND CHINESE COMMON WORDS. 381 To shed, Jg she, shed, "let go," "forgive." Shine, jtp§ shen, shin, German scheinen, Latin candeo. Shoot, ^ sh'i, shed, " arrow," " to swear," German schiessen. Sigh, Jfj, si, sik, German sorge, Sanscrit suka, " air," 'J wind." Sing, %% sung, zung, "to chant," £|[ sung, song, "to praise." Sister, jj$ ts'i, tsi, Latin soror. Slay, |g lu, lok, German schlachten, Anglo-Saxon sleahan. Small, {U wei, mi, Latin minutus, minor, Russian malo, Mongol baga. Smell, ^ wei, mi, " odour," Persian bui. Sot, JH to", £sw£, sot, " become intoxicated." Sound, ^ Psiuen, zien, " sound," " whole," " all," Latin sanus, German gesund. Sound (of voice), sheng, shang, Sanscrit sramana, " hearer," Latin sonitus. Split, jjlj pie, pit, " to separate." Spoke (of a wheel), $g /w, j?o&. Spread, $§ po, pat, "scatter," Latin pateo, "lie open," German breit, English broad, Mongol badaraho, "spread." Stand, stood, ^j ta, dat, "tread upon," Japanese tatta, " stand," Tamil tan, " stand," Latin sto. Step, gg t l a, dap, " to step," Russian stupat. Stick, $g c'^w, tf f 0&, "pierce," Latin s%o, German 382 Straight, jg ch'i, dik, Tamil takudi, " right, " Latin rectus, Greek 8tfcaio$, "just." Strike, fj ta, tang, Mongol tugsehu, " beat," Hebrew ym, " struck." Strong, J|£ chivang, tong. Suck, Pj£ su, tso, sok, Latin sago. Suet, 5$ su, sot, " fat about the entrails." Tablet, ^L and §lj cha, tap, " bamboo or wooden tablets." Take off, Jj| che, tak. Take on the person, JJJJ5 tai, tak. Tap, tapestry, Jg &ha, t'ap, " pierce," " prick," " embroider," German teppich, French tapis, English tap a tree or a barrel. That, the, this, J* ti, di, "this," Jf che, te, "this." Through, }g t'eu, t c ok, "thorough." Throw, $£ t'eu, du, dut, ^ tieu, to. To, JiJ tau, to. Tongue, ^jj c'hang, dung, "to taste," Latin lingo, lingua. Trickle, fjff ti, tik. Turn, K chwen, tun. Vain (that which is empty and unsubstantial), j^f yew, in, " smoke," j| yww, on, " cloud," Latin vanus, anima, English vanish. Wash, $J yu, yok, Mongol ogahu. We, ^ yw, iwi, "I." When, where, which, who, fpf ^0, ga, " what ? " H &, #a, " how many ? " LIST OF ENGLISH AND CHINESE COMMON WORDS. 383 Wicked, g§ ngo, ak. Wind, flj tvan, " a bending," " to bend." Wish, %jfc yu, yok, Latin volo. Word, yue, wat y " say," German wort. Yoke (that which connects), ^fj yo, yak, " agree," "agreement," Latin jugum, jungo, Greek Qvyos. This vocabulary of 153 words is taken almost exclu- sively from the Saxon part of the English language. The few words of Latin origin which occur might as well be placed in a Latin list, but as they form part and parcel of our English tongue they have also a right to be here. 1 The old pronunciation of the Chinese words is indis- pensable in the comparison, and has been inserted in one or two forms. Most of the words are such as belong to the pith and marrow of language, and are not unlikely to be really primeval. A considerable difference in meaning, such as occurs, for example, under the words "vain," "shed," "leaf," " shoot," is not a fatal objection to the identification of the words, because of the great lapse of time since the ancestors of the Chinese and English spoke a common language. The great advantage of the comparison of roots of 1 For further examples, see Professor Haldeman's Relations between the Chinese and Indo-European Languages, p. 13, and Chalmers's Origin of the Chinese. 384 china's place in philology. the European stock with those of the Chinese lies in the fact of the great antiquity of both. By lists such as those compiled by Eichhoff in his work on Compara- tive Grammar, English words are carried back to a period about two thousand years before the Christian era, because the Hindoo family cannot well have entered the Indian peninsula later, and the identification of the English and Sanscrit vocabularies is well established. But the Chinese vocabulary can be traced by the aid derived from the phonetic elements of the characters to a time equally ancient. During the lapse of four or five millenniums, the roots must be expected to appear not without some considerable modifications in the sense. When they are verbs in China, they may be nouns in England, and vice versa. The existence of these differences thus adds increased certainty to the identification. CHAPTER XIV. Conclusion. — Primeval Aryan Civilization as Known from Language. — The Common Civilization of Aryans and Chinese may be Known from Language in the same Way. — Activity of the Third Millennium b.c. — Ethnology of Genesis X. Compared with the Modern Distribution of Races. — Characteristics of Families : The Chinese, Order ; The Semitic, Life ; The Himalaic, Quietness ; The Turanian, Extension ; The Malayo-Polynesian, Softness ; The Indo- European, Elevation ; All of One Blood. — Proof from Polynesian and American Traditions. — Resume.— Duty of Christians to Asia. Sufficient proof has already been given that a vocabulary of common words is just as possible for Europe and Eastern Asia as for Europe and India. If language proves that the English race is akin to the Hindoo, it also shows that it is akin to the Chinese. Philologists have shown that historical data may be recovered from the common vocabulary of the Indo- European family. Before their separation into Hindoo and Persian, Goth and Sclave, Greek and Latin, the Aryan race had towns and fortified places, reared cattle and ploughed the ground. They possessed as domestic animals the horse, swine, ox, dog, sheep, and goat ; l 1 Whitney's Lectures on Language ; Max Muller's Lectures, first course, p. 223. 25 386 they built ships ; they wove cloth ; they lived in houses ; they mined the earth for metals; they counted to a hundred ; they recognized the social duties and the family bonds. Similar results flow, as shown in the last chapter, from ao examination of the European and Chinese common vocabulary. We find there words used in the west for the horse, ox, dog, and domestic fowl. The boat was known, but not the ship. Weaving was practised, and was called by the same name by the Chinese as by the Latins. Wheels and carts were in use. Corn was ground with mill-stones. Wooden bowls were employed for holding food. The processes of sowing and reaping were known by the same simple names. The same is true of some useful vegetable pro- ductions. The Arabic word for flax, kuttan, is like the Chinese hot. The old Chinese had three words for houses of different sizes, corresponding to the European ol/cos, cot, and shed. If a complete comparative vocabulary were drawn up for each division of the Indo-European family, including the Celtic, Lithuanian, and Armenian, we should be in possession of all the important words in the primitive language spoken at the time when in the earth " there was one language and one speech." Roots which have survived the destroying effects of time through four thousand years may be assumed to have lived through the preceding period without much difficulty. The ACTIVITY OF THE THIRD MILLENIUM B.C. 387 vitality of roots is most remarkable, and nothing brings it more vividly into view than the fact of their con- temporaneous existence through so many ages at the extreme ends of Europe and Asia. Perhaps five hundred roots would satisfy the wants of the first men. o The activity of the language- forming faculty was at its maximum during the period when the distribution of nations took place. At b.c 2000 most of the races were settled in the regions they now occupy. Since that time the language-forming faculty has limited itself to the evolution of new languages out of old ones. Before that epoch the formation of the families took place, and for this result a space of 1500 years is not too much. During the 1500 years which seem to have inter- vened between the Deluge and the final settlement of the races, bands of colonists were traversing every region of the vast inheritance assigned by Providence to the human family. The energy and enterprise revealed in the mighty emigrations of those times, were paralleled by an intense intellectual activity, which rapidly and unconsciously traced the outline of the linguistic systems which have ever since prevailed in the two continents of Europe and Asia. "What are now families were then languages, and they were cognate to one another as branches from the same stock. This time of busy activity is described in the tenth 388 and eleventh, chapters of the Book of Genesis, which constitute the most valuable record we possess for primeval ethnology. Independent investigation leads us to the same period, described in the Bible as that "when the earth was divided." The Confusion of Tongues at Babel marks the time when the families of language now existing became separated. Patient inquiry leads to the support of the Scriptural statement, and throws light upon it. It seems to refer specially to the separation of the Semites, Turanians, Indo- Europeans, and a part of the Himalaic race, for the rest of the families had probably already left the Mesopotamian region. The object of the compiler of the tenth chapter was ethnological as well as genealogical, for Mizraim's seven sons are rather, as the plural termination indicates, seven races, and Canaan is said to be the father of eleven races. As Cush had an eastern and western branch, so other races, usually located in the west, may also have an eastern habitat. The name Bod, common to several races in Eastern Asia, ought, as already said in a former chapter, to be compared with Phut, the name of the third son of Ham. The Confusion of Tongues was followed by the do- mination of the Semite language, from Elam in Western Persia to Lydia in Asia Minor, and from Assyria to Sheba and Ophir in the south of the Arabian penin- GENESIS TENTH AND MODERN ETHNOLOGY. 389 sula. Striking traces of Semite influence are found in the Zend, the Persian, and all the Himalaic languages. The race of Ham extended into Africa. It fringed the sea- coast from Arabia Felix to the Indus, following the line of Cushite settlements. It then seems to have spread eastward, including the area of the Bod stock and that of China. Modern research finds no place for the Turanians or the Malayo-Polynesians among the names of the descendants of the sons of Noah. 1 If they are to be included in the range of the tenth chapter of Genesis, it must be without the light of race names. The Scripture record is silent. To the inspired writers "they are the nations that sat in darkness ,, and "the uttermost parts of the earth." The links of connexion are lost, and they have created no ancient literature that might have served as a guide. The linguistic proof, however, remains to show that they are of the common human stock. The Turanians are most nearly connected with the Japhetic languages, as the Himalaic and Malayo-Polynesian are with the Semitic. Thus we seem to have the Japhetic influence in the northern half of Eastern Asia, as that of Shem and Ham in the southern half. In the Pacific Ocean, Japan 1 The word Mongol may be compared with Magog, and Togarmah with the Turks and Tungus. The race name of the Japanese is Wo, which, as not having a consonant in it, is most nearly like Javan. 390 represents Japheth, and the Polynesian Archipelago, with Australasia, combine to spread Semite principles of language. On the American continent, Turanian and Polynesian linguistic principles meet in the various Indian lan- guages. New combinations are formed. But the peculiarities of the languages have not been found sufficiently distinctive to form a thoroughly satisfactory division into families. Yet it has been generally agreed to classify them as northern, central, and southern. The characteristics of the six families of languages reviewed in the preceding chapters are, in the Chinese order, in the Semitic life, in the Himalaic quietness, in the Turanian extension, in the Malayo-Polynesian softness, and in the Indo-European elevation. The love of order shown by the Chinese in their political and social sphere is found also in their lan- guage. The musical effect of the tones on the ear is parallel to the rigid laws of arrangement in their syntax. Antiquity prevails over novelty, and mono- syllabism has retained its empire among them, through a conservative principle, which has thus, happily for science, secured to us a copy more like the original mother of languages than can be found in any other land. The accuracy of the Chinese picture of that lost tongue, which it is the highest duty of philology to restore, is in proportion to the restraining force CHARACTERISTICS OF FAMILIES. 391 which among the Chinese has always hindered develop- ment. That restraint has been caused partly by a feeling of art, which pleased itself with simple triumphs and the retention of the antique ; and partly from want of the poetic impulse, which in more western regions has had so powerful an influence on the advance of language. The principle of life characterizes remarkably the Semitic languages. The Koran and the Bible are replete with poetic expression. The people among whom these books originated were accustomed to look on the world with the poet's eye. This impulse was imparted to them by their possession of early revela- tion ; and its effect was to modify, first, their language, and afterwards their literature, by rapid transitions, personifications, and the breaking up of natural order, so as to place them in complete contrast to the linguistic and literary development of Eastern Asia. The poetic spirit of the Semites probably originated the Indo- European mythology, as it did the more imaginative part of the Indo-European languages. Where the distinctions of gender are found in nouns, there will also be found male and female divinities with names and genealogies. The same feeling for personification, perhaps, has impressed on the Hamitic languages and systems of thought whatever features they possess of a kindred kind. The Hamites were a materialistic race, working patiently at trades and land cultivation. They 392 were farmers and artificers, and they appear to have originated writing. With such tastes they would not create the mythology which prevailed in Babylon and Egypt. Semite influence may be pointed to as a more likely source of their religious ideas, as it would be also of much of their grammar. The Himalaic peoples from Tibet to Cochin- China are characterized in their language, and in their history, by nothing so much as quietness. They have founded no institutions, originated no arts. They have received without giving. Their religion came from India and is Indo-European. Their arts were borrowed from China. The Tibetians have taken some elements in their language from the Semites, others from the Turanians, and others, again, from the Chinese. The Himalaic race are more thoroughly Buddhist than any other linguistic family. A contem- plative religion, opposed to activity, pleased them because it agreed with their natural disposition. Its effect on them has been to confirm them in their quiet ways of thinking. They can never produce any im- press on history till they abandon this inactive and gentle religion. The Turanian race has played in the world a much more important part than those who reside east and north of the Himalayan Mountains. In the fifth century, the Huns under Attila were named " The scourge of God " ; and in the thirteenth century, the CHARACTERISTICS OF FAMILIES. 393 Mongols were the conquerors of all Asia and the dread of all Europe. Occupying Siberia, North Europe, Japan, Tartary, and South India, they won for them- selves a good title to the name of Japheth " the extender! 1 In harmony with this name are the characteristics of their language. They founded the polysyllable and the most widely used system in the world of cases, tenses, and moods. They thus added immensely to the progress of language, by the simple process of ap- pending syllables to roots by agglutination. The language- forming faculty then applied itself to the crys- tallization of these polysyllabic forms into the grammat- ical paradigms which belong to the several languages respectively of the Turanian and Indo-European stocks. The characteristic of the Malayo-Polynesian lan- guages is softness. The primitive monosyllable became a dissyllable by the enervating effect of climate. The initial consonant formed the first syllable and the final the second. Agglutination proceeded on the same principle to work out the Oceanic polysyllable. Every- thing favoured an easy pronunciation adapted to a race accustomed to lassitude and contented to deteriorate. A people having a very soft language can never elevate themselves unless under new conditions, such as the introduction of Christianity. The last of the series, the Indo-European, is remark- able for elevation. This system is built on those that went before, and in many respects combines and per- 394 CHINA S PLACE IN PHILOLOGY. fects their peculiar excellences. The topmost branches of the tree of language, those that spread widest and aspire highest, are the Indo-European. It is this race that has led the mind of the world in science and philosophy, and its language constitutes the most fitting vehicle for the transmission of scientific and philosophic thought. The monosyllabic languages are the lower branches, thick and of great length, but with no bend upwards. The dissyllabic modes of human speech are higher and are turned heavenwards. The earlier poly- syllabic languages have a vast extension, but not much upward curvature. The chief beauty of the tree is in its higher foliage. Here are seen the greatest variety of picturesque effects, the most vigorous growth, the most elegant forms, the most imposing altitude. All the branches, however, upper or lower, proceed from one trunk. " God hath made of one blood all nations of men for to dwell on all the face of the earth." When the European goes into the other continents of the world, as traveller, colonist, missionary, and civilizer, he meets everywhere with men of the same race. " But what have we in common with the Turanians, with Chinese, and Samoyedes ? Yery little it may seem : and yet it is not very little, for it is our common humanity. It is not the yellow skin, or the high cheek-bones, that make the man. Nay, if we look but steadily into those black Chinese eyes, we shall find that there, too, there LANGUAGE SHOWS THE RACES TO BE ONE. 395 is a soul that responds to a soul, and that the God whom they mean is the same God whom we mean, however helpless their utterance, however imperfect their worship." 1 Language proves them to be one with ourselves. The black, the yellow, the copper- coloured, and the brown races come of one stock. If the yellow and the white can by linguistic proofs be shown to be one, the presumption will be strong that the same is true of all. The evidence is more accessible in the case of the yellow race than of the rest, because they have an ancient literature and a writing by means of ideographic signs, of which the phonetic values are known. With the less civilized races we have not this advantage. Their languages are perpetually changing, and we cannot recover their ancient forms. But if the differences between a white and a yellow skin, an upright and a receding forehead, a Caucasian and a Mongolian head, a large blue eye, set deep, and a small black eye, set on the surface, are not conclusive against consanguinity, so neither must it be allowed that a black or red skin, proves descent from a different Adam. If Adam were the progenitor of Caucasians only, as held by M'Causland, 2 we should not find European roots existing in abundance in the Chinese vocabulary. 1 Max Miiller, Lectures on the Science of Religion. 2 Adam and the Adamites, 1864. Dr. M'Causland has felt strongly the force of the stone hatchet argument. But the right way to proceed is rather to make mutual concessions in chronology. 396 china's place in philology. Nor should we meet the old type of the Aryan pro- nouns and the Aryan system of accidence in Turanian languages still spoken in Tartary and Siberia. If the Polynesians were not of Asiatic origin, we should not find proofs of their syllabic system being based upon an old Asiatic syllabary and their laws of syntax all formed on Asiatic models. Man cannot retain his civilization and morality when isolated — he will cease to practise old arts, he will forget facts once familiar to him, his religious ideas will become dim, his range of thought will in each successive century grow more limited, and he will fall into habits which are immoral and debasing. That the Polynesians are now inferior to the Japanese and Chinese is the effect of their distant wanderings, and is an argument for the propriety of offering to them early the blessings of religious and moral teaching, with instruction in the arts of civilized man. The religion of the Polynesians is more like that of the Brahmans than of the Buddhists, and there were probably communicated to them, in early times, from India, some features of the Hindoo faith. Who can Tiki be but Sakra ? What can be the paradise of Tiki, as believed in by the Samoans, but the thirty- third heaven of Sakra ? At any rate s is changed into t quite commonly in the cognate languages spoken on the Birman peninsula. Yet the Samoan belief in a Supreme God, called Tangoloa reminds us strongly of RESUME. 397 the Mongol and Turkish faith in Tengri, and that of the Chinese in Tien. The addition of the two consonants g, r, is Turanian, and it was apparently from the Turanians, therefore, that faith in the Supreme Being under this name was derived. The worship of an- cestors, common in the South Seas, would be learned from the Chinese ; while the human sacrifices, which also existed among them to a frightful extent, 1 must be viewed as Turanian, — for in some parts of India the aborigines are, under British eyes, only beginning to allow this practice to fall into desuetude, — or they are Semitic, and are of the same origin as the sacrifices to Moloch condemned in the Old Testament. Let the reader now recall the successive steps of this investigation from the commencement. The old insti- tutions of China were shown to be like those of the renowned cities of ancient Mesopotamia. It was stated that the remarkable similarity in arts, usages, and ideas, existing among the races that lived near the Yellow Eiver, the Euphrates, and the Nile, indicated that they sprang from a common source. After briefly glancing at the geographical areas of the families of languages spoken in Asia, a sketch was drawn of their most general features, as constituting a rough picture of the world's primeval language. The roots are recoverable in a monosyllabic form. They were chiefly imitations of natural sounds, and were 1 Williams's Missionary Enterprizes. 398 increased by the aid of the principle of the association of ideas. Special divine aid was afforded to primeval man in the task of forming for himself a language. In the chapter on the Chinese language, after it had been shown that the conditions of the situation would be best suited by supposing the Chinese to have left Western Asia about 5,000 years ago, and yet subse- quently to the Deluge of Noah, the mode of recoveriDg the primeval Chinese syllabary from the phonetic element of the characters was described. The syntax, so accordant with nature and innocent of inversions, was seen to be of the most primeval type. The next step in the progress of language was taken in the formation of the Semitic language. The people who used Semite speech added a consonant to the root, introduced prefixes to mark conjugations and moods, invented a plural and dual number, originated genders among nouns not properly masculine or feminine, and revolutionized the syntax. In speaking of the languages used in the region south of China, it was seen that while their tones, their syllabaries, and their vocabulary, connect them closely with China, their syntax links them remarkably with the Semitic type. While this is the case with the Siamese, the Cochin- Chinese, and the Miau aborigines in China itself, some still more striking Semitic charac- teristics belong to the Tibetan language. Though its tones, roots, and radical syllabary show it to be akin to RESUME. 399 the Chinese, and its syntax and case suffixes prove its relationship to the Turanian type, its mode of conju- gating verbs and its consonantal prefixes are Semite, and seem to point for their origin to a time earlier than the Aryan occupation of India and Persia, which drove J;he Semites and Turanians from their neighbourhood on the west and south. The Japanese received special attention as the oldest of the polysyllabic languages in Asia, and it was shown how case particles grew into existence by agglutination, the syllables made use of for this purpose being words existing as separate roots in Chinese and other lan- guages. The second division of the polysyllabic Turanian system was described as the Dravidian. The growth of the verb by agglutination was here traced, and a growing resemblance in vocabulary and grammar to the western type found to be perceptible. The greatest likeness and nearest kinship between the Indo-European languages and the threefold Tu- ranian type was proved to exist in the Tartar, of which Mongol was taken as the best representative. Here it was shown that the pronouns and substantive verbs, declension of nouns, and verb conjugation of western speech, rest chiefly upon the Tartar branch of the Turanian family as their source and foundation. The Malays and Polynesians have a syllabary and vocabulary which was evidently once continental. The 400 china's place in philology. Malay and Siamese are specially connected with each other, while Chinese influence in the principles pre- vailing in the Polynesian languages is very perceptible. These islanders retain traces of a lost civilization, which conies more prominently to view on the American continent. Language and religious beliefs alike point to Southern Asia as the source from which came the tribes that inhabit Australia, Polynesia, and the civilized portion of the American tribes. The sudden expansion of language observable in Sanscrit, as compared with the preceding systems, indicates the commencement of a new era of develop- ment, characterized by unparalleled richness of forms. This new advance proceeds on principles already existing in older systems. In introducing gender in nouns, and sex in mythology, Semitic example was followed. So, also, the prefix of sibilants in the root and the insertion of r and I after the initial of the primeval syllable seem to have come from the same source. But in all the newer portions of the Sanscrit grammatical formation we find laws prevailing which also characterize Turanian languages. Case suffixes, the verb, and the syntax, bear united testimony to this statement. But there is a more highly wrought appearance in the forms. Agglutination has become inflexion. Boot and suffix are fused into a closer union. The advance in analytical acuteness, which was after a few centuries to culminate in the creation of kiM 401 Hindoo philosophy, is first soon in llio mil. divisions of the verb paradigms. The adjective was now for the first tune declined like the substantive, and the relative pronoun began to cxort some of that j which it has more fully assumed in the Eur< | languages. AYhen the speech of ancient and modern Europe was brought under our review, it was found, as in [ that the principles of older languages v king underneath the surface. Bui they appeared in combinations suited to the mental conditions of the successive races who have in that favoured i wrought out such a marvellous history in the poll social, and in I sphere. Greet in the vicinity of the ancient empires, was abl< derive from them tl. idly nurturing these, she was seen to develope wit) hing rapidity those creations in po< history, and philosophy, which the world will 1 to admire. A language and literature tii'ul and complete as the Great could ne\ originated but from the happy com bin ation of fruitful principles, derived from the pre-existing s of Language and thought. The Latin, the T the Sclavonic forms of langine special elements, contributed in varied proportion B the same sour* In all these lai early ol 402 later ones, and new phenomena exemplify over again what took place long ago. When we say, " Alfred the Great," we use a French idiom, dating from the Norman Conquest ; and among our Saxon idioms, old and new, forming the major part of the language, Turanian modes of expression may be pointed out, which at some distant time, when our ancestors lived near the Caspian, found their way into colloquial use in some similar manner. For the English, " and came before him," the Anglo-Saxon G-ospels have, in Mark vi. 33, " and him beforan comon." These words are exactly in the order of Manchu and Mongol syntax. How great are the linguistic accessions to European speech received from far Asia has been shown by exam- ples of common words. They are enough to make plain that the vocabularies of the east and west are essentially the same. This identity dates from a time previous to the settlement of the Chinese in China and the Mongols in Mongolia. Philology may here safely take her stand, and add a chapter of illustration to the sacred record, where it treats of the division of the earth and the planting of nations. It is the duty, as it is the destiny, of the nations of Europe to give back to the east the treasure of heavenly lio-ht which they once received from it. To Asia they owe the first impulses to thought, the earliest lessons in the arts, the invention of writing, and the price- less deposit of divine revelation. "Freely ye have Dl TV OP THE WEST TO ENLIGHTEN THE 1ABT. 409 received," says the Saviour, " freely giye.' 1 'J higher menial elevation and their richer sto* knowledge fit them to be the instructors of the old world; and to this undertaking Divine Providence is leading them by unmistakable signs. England has received the rule of India for this purpose, that she may become the teacher and evangelizer of India. Commerce and war have opened the gates of China, that Christian truth may enter them. All new i therefore, should be welcome that tend to show that the Chinese are one with us in origin, and that their history, their institutions, their language even, deriye their source, as ours do, from Western Asia. Let the kindly sympathy of the west for the east be the more called forth as the proofs of common brotherhood are accumulated. THE 8TKP1I' LINGUISTIC PUBLICATIONS 01 TRUBNER & CO., 8 and 60, PATERNOSTER ROW, LONDON. B.O. Ahlwardt. — The Divans oi the Six Am mm Abab* Poi re, Enrj ' 'Antara, Tarata, Zuhair, 'Algama, and Imruolgais; chiefly according I M8S. of Paris, Qotha, and Leyden, and the collection of their I : with a complete list of the various readings of the Text. Edited by W, Amu Professor of Oriental Languages at the University of Geifswald. 8vo. pp, 340, sewed. 1870. 12*. Alabaster. — The Wheel of the Law : Buddhism illustrated from Siamese Sources by the Modern Buddhist, a Life of Buddha, and an account of the Phra Bat By Henby Alabaster, Esq., Interpreter of n Consulate-General in Siam ; Member of the Koyal Asiatic Society. I 8vo. pp. lviii. and 324. 1871. lis. Alcock. — A Practical Grammar of the Japanese Lavgi Rutherford Alcock, Resident British Minister at Jeddo. 4to. pp, 81, sewed. 18*. Alcock. — Familiar Dialogues in Japanese, with English and French Translations, for the use of Students. By Sir ELOTHBBIOBD Ai pp. viii. and 40, sewed. Paris and London, 1863. 5*. Alger.— The Poetry of the Orient. By William Roohsbi n 8vo. cloth, pp. xii. and 337. 9s. Alif Lailat wa Lailat. — The Arabian Nights. 4 vols. ito. pp. I 493, 442, 431. Cairo, a.h. 1279 (1862). £2 This celebrated Edition of the Arabian Nights la now, for the fir.-t time, Oflsnd at I which makes it accessible to Scholars of limited mi Andrews. — A Dictionary oi the Hawaiian Language, t<> whiol appended an English- Hawaiian Vocabulary, and a Chronological Table of Remarkable Events. By Lokkin AndRVWB. B?0. pp. 560, doth, i. 1 I Arabic, Persian, and Turkish Books (A Catalogue of Printed in the East. Constantly for sale by Trubner and Co., B sad 60, Paternoster Row, London. Contents. — Arabic. Persian, and Turkish Books printed in Bgj Arabic Books printed in Oudh.— Persian Literature printed in Oa Editions of the Koran printed in Oudh. — Arabic Hooks printed at Bombay.— Persian Books printed at Bombay.— Arabic Literature printed at Tunis.— Arabic Literature printed in Syria. Ilium, pp, 68. Price La, Asher. — On the Studi ob Modebn Languages or G English Language in particular. An Essay. By David A-iii u, Ph.h. I pp. viii. and 80, cloth. - Js. Asiatic Society. — Journal of the Royal Asiatic S liKirviN \m» [hbxand, from the Commencement I pl.t. in 20 Vols. Bvo., with many Plates. I as follows:— Nos. l to 1 1. fo.each; No. 16, oh ; No. 17, 2 Parts, is. each N Vols. 1. to IX. -Vol. X.. Tart 1, on.; Pari 2. M .. Part 1,6*,; Part 2 not published. VoL XII., 2 MIL, 2 Parte, St \ oL XIV., Pari L, Pari 1, 6a.; Pan 2, with Maps, 10«. \ oL M XVI L. 2 Parte, 6«. i ich. Vol. X \ 1 1 1 . 2 P to 4, 16a.— Vol. XX., 8 Parte, it, sach. 1000. 0.10.7 1. 2 Linguistic Publications of Triibner t ion at KeU-TtUlg kwan, m North < lima. Wylie. — III. The Hrhat Sanhita ; or, Complete System ol Translated from Sanskrit into English by Dr. II. Kern. — IV. The Pon India. By Charles E. Gover.— V. The Poetry of Mohamed Rabadnn, ol Hon. Lord Stanley of Alderley.-VI. Essay on the Creed and Customs ol Charles 1'. Brown.— VII. On Malabar, Coromandel, Quilon, etc. I'.vC. P. I the Treatment of the Nexus in the Neo-Aryan I anguages of India. By John i\. some Remarks on the Great Tope at SanehL I from Mathura. Translated by Professor J. Dow soil.— Note to the Math lira Major-General A. Cunningham. Asiatic Society. — Transactions of tiie Royal Asiatic Sa Great Britain and Ireland. Complete in 3 vols. !t<>.. 80 Plates similes, etc., cloth. London, 1827 to 1835. Published at JL ( J 5» . : reduced to £3 3.9. The above contains contributions by Professor Wilson, C. C. Ilamrhtt.n. rison, Colebrooke, Humboldt, Dorn, Orotefend, and other eminent Oriental .scholars. Auctores Sanscriti. Edited for ihe Sanskrit Tr\i r the supervision of TsBODOB GoLDSTOCKBB. Vol. 1., containing the Nyaya-Malfi-Vistara. Parts I. to V., pp. 1 to 400, large 1 to each part. Axon. — The Literature of the Lavcasbzbi Dialect. A Biblio- graphical Essav. Py William E. A. Axon, F.R.S.L. 1 1870. Lv. Bachmaier. — Pasigbaphicai Dictionary and Gt&amkail By A BaciImaikr, President of the Central Pasigraphical Society at Munch. cloth, pp. viii. ; 26 ; 100. 1870. 8*. 6c/. Bachmaier. — Pasigeaphisohes Woetbebucb ztm Gtbbbauohi puj deutschr Spkacuk. Verfasst von Avion Ba< iim.ui.u, Voreitiendem -h-s Central- Vereins fur Pasigraphie in Munchcn. lSino. cloth, pp. riii 120. 1870. 2*. 6d. Bachmaier.— Diction \vi ki: Pasioeaphique, peecsde' db i Redige par Antoi.m: BagkMAIBB, President dfl la Socictc (Vntrule de Pasi- graphie a Munich. lSuio. doth. pp. vi. 26 ; 168; l •">•'. 1870, 5 Ballad Society's Publications. — Subscriptions — Small paper, guinea, and large paper, three guineas, per annum. 186 1. Ballads peom MLanttsceipts. Vol I. Ballads on the eondition of England in Henry VHP's and Kdward VI.*I Eleigni in.'. : ite of the Clergy, Monks, and Friars), on Wolsey ami Anne I at I. Edited by P. J. FuBirn u.i . EC 2. Ballads peom Majtusoeipts. Vol. 1. Pan .. [In the J 3. Ballads peoe Manuscripts. Vol. II. Pari 1. The I' Pittance. By Richard Williams. Edited bj . .J.l'i Ml M L Svo. 186 4. The Roxbtjeghe Ballads, Vm\ 1. With W. CHAPPBLL, Esq., P.8.A., author of "Popular Mum. Time," etc., etc., ami with copies of the Original Woodoata RuDoi.ru Blind end Mr. W. II. Boons, end engraved by Mr J. H. Bimbault end Mr. Boom. B?o. 4 Linguistic Publications of Trubner fy Co. Ballad Society's Publications— continued, 1870. 5. The Roa^burghe Ballads. Vol. I. Part II. With short Notes by W. Chappell, Esq., F.S.A., and with copies of the Original Woodcuts, drawn by Mr. Rudolph Blind and Mr. W. H. Hooper, and engraved by- Mr. J. H. Rimbault and Mr. Hooper. Ballantyne. — Elements of Hindi and Braj Bhaka Grammar. By the late James R. Ballantyne, LL.D. Second edition, revised and corrected Crown 8vo., pp. 44, cloth. 5s. Ballantyne. — First Lessons in Sanskrit Grammar ; together with an Introduction to the Hitopadesa, Second edition. By James R. Ballantyne. LL.D., Librarian of the India Office. 8vo. pp. viii. and 110, cloth. 1869. 5*. Bartlett. — Dictionary of Americanisms : a Glossary of Words and Phrases colloquially used in the United States. By John R. Bartlett. Second Edition, considerably enlarged and improved. 1 vol. 8vo., pp. xxxii. and 524, cloth. 16s. Beal. — Travels of Pah Hian and Sung- Yen, Buddhist Pilgrims from China to India (400 a.d. and 518 a.d.) Translated from the Chinese, by S. Beal (B.A. Trinity College, Cambridge), a Chaplain in Her Majesty's Fleet, a Member of the Royal Asiatic Society, and Author of a Translation of the Pratimoksha and the Amithaba Sutra from the Chinese. Crown 8vo. pp. lxxiii. and 210, cloth, ornamental, with a coloured map, 10s. 6d. Beal. — A Catena of Buddhist Scriptures from the Chinese. By S. Beal, B.A., Trinity College, Cambridge; a Chaplain in Her Majesty's Fleet, etc. 8vo. cloth, pp. xiv. and 436. 1871. 15s. Beames. — Outlines of Indian Philology. With a Map, showing the Distribution of the Indian Languages. By John Beames. Second enlarged and revised edition. Crown 8vo. cloth, pp. viii. and 96. 5s. Beames. — Xotes on the Bhojpuri Dialect of Hind£, spoken in Western Behar. By John Beames, Esq., B.C.S., Magistrate of Chumparun. 8vo. pp. 26, sewed. 1868. is. 6d. Bell. — English Yiseble Speech for the Million, for communicating the Exact Pronunciation of the Language to Native or Foreign Learners, and for Teaching Children and illiterate Adults to Read in few Days. By Alexander Melville Bell, F.E.I.S., F.ll.S.S.A., Lecturer on Elocution in University College, London. 4to. sewed, pp. 16. Is. Bell. — Visible Speech; the Science of Universal Alphabetics, or Self- Interpreting Physiological Letters, for the Writing of all Languages in one Alphabet. Illustrated by Tables, Diagrams, and Examples. By Alexander Melville Bell, F.E.I.S., F.R.S.A., Professor of Vocal Physiology, etc. 4to., pp. 156, cloth. 15s. Bellew. — A. Dictionary of the Pukkhto, or Pukshto Language, on a New and Improved System. With a reversed Part, or English and Pukkhto. By H. W. Bellew, Assistant Surgeon, Bengal Army. Super Royal 8vo., pp. xii. and 356, cloth. 42s. Bellew. — A Grammar of the Pukkhto or Pukshto Language, on a New and Improved System. Combining Brevity with Utility, and Illustrated by Exercises and Dialogues. By H. W. Bellew, Assistant Surgeon, Bengal Army. Super-royal 8vo.,pp. xii. and 156, cloth. 21s. Bellows. — English Outline Vocabulary, for the use of Students of the Chinese, Japanese, and other Languages. Arranged by John Bellows. With Notes on the writing of Chinese with Roman Letters. By Professor Summers, King's College, London. Crown 8vo., pp. 6 and 368, cloth. 6s. Bellows. — Outline Dictionary, for the use of Missionaries, Explorers, and Students of Language. By Max Muller, M.A.,Taylorian Professor in the University of Oxford. With an Introduction on the proper use of the ordinary English Alphabet in transcribing Foreign Languages. The Vocabulary compiled by John Bellows. Crown 8vo. Limp morocco, pp. xxxi. and 368. 7$. 5d 8 and 6n, Paternosb r Ro \ La Benfey. — A Practical Gbamicae. o» the v r the use of Early Students. By Tiii:onoit Bi\in, Proftmor of Sanskrit in the University of Gdttingen. Second, revised and ciliary d, edition. Ro] pp. viii. and 290, cloth. 10*. i\d. Beurmann. — Vocabi lari o* ceb Tresl Li Written down by iMoKiiv. von lii-.i kmvnn. Published with A Grammatical Sketch. By Dr. A. Mi OX, Of the University of Jena, pp. viii. and 78, cloth. .';>. li-/. Bholanauth Chunder. — The Travels oe \ Bxrdoo ro fASioui i-nktsop Bengal and Upper India. By Bholanauth Chvndi :b, Member of tin- Society of Bengal. With an Introduction by J. Talhoyi Whe< I " The History of India." Dedioated, by permission, to Bis Exa 11. i Laird Blair Lawrence, G.C.B., G.C.S. I., Viceroy and Governor^ I India, etc. In 2 volumes, crown 8vo., cloth, pp. xxv. and 440, viii. and IK). Hl>. Bigandet. — The Life or Legend of Gattdaxa, the Buddha of the Burmese, with Annotations. The ways to Neibban, and Notice on the Phongyies, or Burmese Monks. By the Right Reverend P. BlGAXDBT, Bishop of Ramatha, Vicar Apostolic of Ava and Pegu. 8vo. sewed, pp. xi.. 588, and Bleek. — A Comparative Grammar Of South Aerioaj \V. H. I. Bleek, Ph.D. Volume I. I. Phonology. II. The Concord. Section 1. The Noun. 8vo. pp. xxxvi. and 322, cloth. 16s. Bleek. — Reynard in South Africa; or, Hottentot Fables. T- Iated from the Original Manuscript in Sir George Grey's Library. By Dr. W. H. I. Bleek, Librarian to the Grey Library, Cape Town, Cape oi Hope. In one volume, small 8vo., pp. xxxi. and i)4, cloth. J". 0 readings. pp. 116. ; 3. Paxcii LTANTRA ir. and hi. Edited, with Gr. J5iiiii.ii:, Ph. D. Pp. 86, 14, 2. 6«. 6A 4. Panciiai -an-i i: \ I. Edited, with Notes, by F. KlBLHOBiT, Ph.D. Pp. 1 1 K 68. 6*. 6d. 5. KXlxdXsa's K v<. ii i % a Ai s v. With the Commentary of Mallin.itha. Edited, with Notes, by Shank w, P. I'amhi. SI. A, Perl I. CaatOS 1 \ 6. KiiiDAflA's M \i.\vik \..m\mikv. Edited, with > I*. Pandit, M.a. 8*. Bottrell. — Traditions and Hearthslde Store By William lion kill (an old Celt). Demy 12mo. pp. vi 292, clol Boyce. — A Graiolab of i jreuA*». — By Williaj Botes, Wesleyan Missionary. Third Edition, aogna atedendim] Exercises, by William J. Davis, W< ry. L2mo. pp. \ii. and 164, doth. Bowditch. — Suffolk Suuwm is. |; v N. I. BowdxtoH. Third 8vo. pp. xxvi. and 7.')-'!, cloth. 7>. <>'/. Carpenter.— Tin: Last Days in Bnglantj oi m Rajas EU mmwiu j Koy. \]y Mary Cabpsntbb, of Bristol. With FIto Mini pp. 272, cloth. 7.s. 64 Carr.— e-o|$«r s r*_§ ) -0'oL?S'. A Collection o* Telw Translated, Illustrated, and Explained ; together with some Baa rbs printed in the Bernftgaii and Teloeu Characters. By Captain M. W.Cabb, Madras Staff Corps. One Vol. and Efapplemnt, royal 8vo. pp. *88and 148. Catlin. — O-Kee-Pa. A Religious Ceremony of the Mamlans. By George Catlin. AVith 13 Coloured Illustrations. 4to. pp. GO, bound in cloth, gilt edges. 14a-. Chalmers. — The Origin or the Chinese; an Attempt to Trace the connection of the Chinese with Western Nations in their Religion, Superstitions, y Arts, Language, and Traditions. J Jy John CbALMBBS, A.M. Foolscap Bi cloth, pp. 78. 2*. Cxi. Chalmers. — The Speculations ojt Metaphysics, Polity, lot op " The Old Phxlosofhbr" Lad Tbzb. Translated Iron the Chinese, with an Introduction by John Chalmers, M.A. Foap. 8ro. cloth, x\. and 62. 4> • Chalmers. — An English and Cantonese Pocksi-Dictionabt, for tin- use of those who wish to learn the spoken language of Canton I John Chalmkhs, M.A. Third edition. Crown 8vo., pp. iv. and 14ti. Hong Kong, 1871. 15 s . Charnock. — Ltjdus Patbonytoctjb ; or, the Etymology of Chuious Sur- names. By RlOKABD Sn rin.N Ciiaunock, Ph.D., F.S.A., F.K.Ci.S. Crown 8vo.. pp. 182, cloth. 7.v. G^. Charnock. — Verba Nominally ; or Words derived from Proper Nan By Richard Sri rn in Charnock, Ph. Dr. P.S.A., etc. 8vo. pp. 826, cloth. 14*. Charnock. — Tin; Peoples oi Transylvania. Founded on a Paper read before Thk ANTHROPOLOGICAL Boon i\ Of London, on the Ith of May, 1869. By RiOHABD SiiiniN Ohabnock, Ph.D., F.8.A.., F.B.Q.S. Deny 8vo. pp. 36, sewed. 1870. 2s. iid. Chancer Society's Publications. Subscription, two guineas per annum. 1868. FimtStr Canterbury Tales. Part I. I. The Prologue and Knight's Tale, in 6 parallel Texts (from tl named below', together with Tables, showing theOron slee, and their varying order in 88 MSS. of the Tales, and in the printed editions, and also Specimens from sever il MSS "Moveable Prologues" of the Canterbury Tales, The Shipmau't Prologue, and Franklin's Prologue, irhen moved from their right pieces, and of the substitutes for then II. The Prologue and Knight's Tale from lh< MS. III. ,, >, •, if m n n tl.n-vMt ,, IV. „ „ „ „ ■■ Cambrid V. „ ,, N „ , gorj and. VI. „ „ „ ,, •■ •• Petwortli .. VII. „ Kos. II. to VII. arc separate Textl of the 6-Text edition of the Canterbury Tales, Part I. 8 Linguistic Publications of Trubner 8f Co. Chaucer Society's Publications — continued. 1868. Second Series. On Early English Pronunciation, with especial reference to Shak- spere and Chaucer, containing an investigation of the Correspondence of Writing with Speech in England, from the Anglo-Saxon period to the present day, preceded by a systematic notation of all spoken sounds, by means of the ordinary print- ing types. Including a re-arrangement of Prof. F. J. Child's Memoirs on the Language of Chaucer and Gower, and Reprints of the Rare Tracts by Salesbury on English, 1547, and Welsh, 1567, and by Barcley on French, 1521. By Alexander J. Ellis, F.R.S., etc., etc. Part I. On the Pronunciation of the xrvth, xvith, xvnth, and xvinth centuries. Essays on Chaucer; His Words and Works. Part I. 1. Ebert's Review of Sandras's Etude sur Chaucer, considere comme Imitateur des Trouveres, translated by J. W. Van Rees Hoets, M.A., Trinity Hall, Cambridge, and revised by the Author. — II. A Thirteenth Century Latin Treatise on the Chilindre: "For by my chilindre it is prime of day " {Shipmannes Tale). Edited, with a Trans- lation, by Mr. Edmund Brock, and illustrated by a Woodcut of the Instrument from the Ashmole MS. 1522. A Temporary Preface to the Six-Text Edition of Chaucer's Canterbury Tales. Part I. Attempting to show the true order of the Tales, and the Days and Stages of the Pilgrimage, etc., etc. By F. J. Furnivall, Esq., M.A., Trinitv Hall, Cambridge. 1869. First Series. VIII. The Miller's, Reeve's, Cook's, and Gamelyn's Tales : Ellesmere MS. IX. „ „ „ „ „ „ „ Hengwrt „ X. „ „ „ „ „ „ „ Cambridge,, XL „ „ „ „ „ „ „ Corpus XII. „ „ ,. „ „ „ „ Petworth „ XIII. „ „ „ „ „ „ „ Lansdowne,, These are separate issues of the 6-Text Chaucer's Canterbury Tales, Part II. 1869. Second Series. English Pronunciation, with especial reference to Shakspere and Chaucer. By Alexander J. Ellis, F.R.S. Part II. 1870. First Series. XIV. The Miller's, Reeve's, and Cook's Tales, with an Appendix of the Spurious Tale of Gamely n, in Six parallel Texts. Childers. — Khuddaka Patha. A Pali Text) with a Translation and Notes. By R. C. Childers, late of the Ceylon Civil Service. 8vo. pp. 32, stitched. Is. 6d. Childers. — A Pali-English Dictionary, with Sanskrit Equivalents, and with numerous Quotations, Extracts, and References. Compiled by R. C. Childeks, late of the Ceylon Civil Service. [In preparation. ChJoniqneDE Abou-Djafar-Mohammed-Ben-Djartr-Ren-Yezid Tabari. Traduite par Monsieur Hermann Zotenberg. Vol. I. 8vo. pp. 608. Vol. II, 8vo. pp. ii. and 252, sewed. Is. 6d. each. ( To be completed in Four Volumes.) Clarke. — Ten Great Religions : an Essay in Comparative Theology. By Jamks Freeman Clarkb. 8vo. cloth, pp. x. and 528. 1871. 14*. Colenso.— Eirst Steps in Zulu-Kafir : An Abridgement of the Ele- mentary Grammar of the Zulu- Kafir Language. By the Right Rev. John "W. Colenso, Bishop of Natal. 8vo. pp. 86, cloth. Ekukanyeni, 1859- 4*. 6d. Colenso. — Zulu-English Dictionary. By the Right Rev. John W. Co- lenso, Bishop of Natal. 8vo. pp. viii. and 552, sewed. Pietermaritzburg,1861. 15*. Colenso. — First Zulu-Kafir Reading Rook, two parts in one. Ry the Right Rev. John W. Colenso, Bishop of Natal. 16mo. pp. 44, sewed. Natal. 1*. Colenso. — Second Zulu-Kafir Reading Rook. By the same. 16mo. pp. 108, sewed. Natal. 3*. 8 and 60, Paternoster Row, I. Colenso ,— Foubtb Zi i.i -K' \i n EtiuDnre 8vo pp. 160, cloth. Natal, 1869. 7*. Colenso. Three Native Aoeonnts of the Visits of the Bisho] in September and October, 1859, to Upmande, King of the Zulu* ; with i oatory Notea and a Iitersl Translation, and a GL i the Zulu employed in the same i designed for the oae of Student* of l By the Right Kev. John W. Colenso, Bishop of Natal. I< cover. Natal, Maritsborg, I860. I Coleridge. — A Glossaeiax brnsxtothe Printed English literature of the Thirteenth Century. By Ui.kbekt Colbuds 2*. Cxi. Colleccao de Vocabulos e Frases usados na Provincia de 8 do Rio Grande do Sul, no Brasil. 12mo. pp. 32, sewed, la, Contopoulos.— A Lexicon op Modeeh &BEEi-EHOLiaH a.m. Bveuni Modbbm Gbsbx. By N. Contopoulos. Parti. Modern Greek-English. 8vo. cloth, pp, Part II. English-Modern Greek. 8vo. cloth, pp. 582. Cunningham. — The Ancient Geogeapht of Indu. I. The Buddhist Period, including the Campaign! of Alexander, and the Travel* of II By Alexander Cunntnoh am, Major -General, Royil I tired). With thirteen .Maps. 8to. pp. u. 590, cloth. 1870. 2 Cunningham. — ks Essai on the Asian AacHn exhibited in the Temples of Kashmere. By Captain (now M I Alexander Cunningham. 8vo. pp. 8b", cloth. With seventeen large foldine Plates. 18*. Cunningham.— Tin- 15 in i.sv Topes; or, Buddhist IConam utral India: comprising a brief Historical sketch of the Etiee, Progreaa, and Decline of Bnddhiam; with an Account of the Opening and Examination of the Groups of Topes around Bhilsa. By Brer.-Majof Alexander Cnnninf Bengal Engineers. Illustrated with thirty-three Plates. Bto. pp cloth. 1854. 21*. Delepierre. — l!i:vn: Akalttiqtje des Ouvbaois GJceits in I depuis lea Temps Anciena, jnaqn'an lix 1 *"* Sieele. Par on Bibliophile Bake. Small 4to. pp. 508, stiff covers. 1868. 80s. Delepierre. — Essai Histoeiqtte bt Bibliogbaphiqui bub un \i Par Octave Delepierre. 8to. pp. 24, sewed. With 15 pages ol W 1870. 8». 6rf. Dennys. — China and Japan. A complete Ghride to the Open r<*rtsof those countries, together with Pekin, Yeddo, Hong Kong, and Macao; fbrming a Guide Book and Vade Mecum for Travellers, Merchants, and I general; with 66 Maps and Plans. Bj Wu, Pbsobbxcb Man H.M.'i Consular Service; N. 15. Dbnnyb, late B.M.'s Consular S i Charles Kim;, Lieut. Royal Marine Artillery. Edited by N. li. I>. In one volume. 8vo. pp, 600, cloth. £2 Digest of Hindu Law, from the Replies of the Bhsstris in the BerersJ Courts of the Bombay Presidency. With an Introduction. Note-, and \ dix. Edited by Raymond \\V.-t and Johann I iler. VoL 1. Bto, cloth. £3 8s. Vol. [I. 8vo. pp. v. 118, doth, i Dohne. — A Zulu-Kafir Diotionaet, etymologioally explained, with copious Illustration! and examples, preceded bj in introduction mi tin- Zulu- Kafir Language. By the Rev. J. L. Duiim. I, pp. dii. md 118, sewed. Cape Town, 1867. 81s. Dohne. — The Foub Gospels in Zulu. I'-\ the Rer. J. I. I 1 Missionary to the American Board, C.F.M, Bto. pp. 208, doth. I burg, 1866. Doolittle. — An Enqlisb am. Chzbxsi Diotionaet. Bj thi Justus Doolittle, China. 10 Linguistic Publications of Trilbner <$* Co. Early English Text Society's Publications. Subscription, one guinea per annum. 1. Early English Alliterative Poems. In the "West-Midland Dialect of the Fourteenth Century. Edited by R. Morris, Esq., from an unique Cottonian MS. 16s. 2. Arthur (about 1440 a.d.). Edited by F. J. Eurntvall, Esq., from the Marquis of Bath's unique MS. 4s. 3. Ane Compendious and Breue Tractate concernyng ye Office and Dewtie of Kyngis, etc. By William Lauder. (1556 a.d.) Edited by F. Hall, Esq., D.C. L. 4s. 4. Sir Gawayne and the Green Knight (about 1320-30 a.d.). Edited by R. Morris, Esq., from an unique Cottonian MS. 10s. 5. Oe the Orthographie and Congruttle of the Beitan Tongue; a treates, noe shorter than necessarie, for the Schooles, be Alexander Hume. Edited for the first time from the unique MS. in the British Museum (about 1617 A.d.), by Henry B. Wheatley, Esq. 4s. 6. Lancelot of the Laik. Edited from tbe unique MS. in the Cam- bridge University Library (ab. 1500), by the Rev. Walter W. Skeat, M.A. 8s. 7. The Story of Genesis and Exodus, an Early English Song, of about 1250 a.d. Edited for the first time from the unique MS. in the Library of Corpus Christi College, Cambridge, by R. Morris, Esq. 8s. 8 Morte Arthure ; the Alliterative Yersion. Edited from Robert Thornton's unique MS. (about 1440 a.d.) at Lincoln, by the Rev. George Perry, M.A., Prebendary of Lincoln. 7s. 9. Animadversions uppon the Annotacions and Corrections of some Imperfections of Impressiones of Chaucer's Work.es, reprinted in 1598 ; by Francis Thxnne. Edited from the unique MS. in the Bridgewater Library. By G. H. Kingsley, Esq., M.D. 4s. 10. Merlin, or the Early History of King Arthur. Edited for the first time from the unique MS. in the Cambridge University Library (about 1450 a.d.), by Henry B. Wheatley, Esq. Part I. 2s. 6d. 11. The Monarche, and other Poems of Sir David Lyndesay. Edited from the first edition by Johnf Skott, in 1552, by Fitzedward Hall, Esq., D.C.L. Part I. 3s. 12. The "Wright's Chaste "Wife, a Merry Tale, by Adam of Cobsam (about 1462 a.d.), from the unique Lambeth MS. 306. Edited for the first time by F. J. Furnivall, Esq., M.A. Is. 13. Seinte Marheeete, J?e Melden ant Martyr. Three Texts of ab. 1200, 1310, 1330 a.d. First edited in 1862, by the Rev. Oswald Cockayne, M.A., and now re-issued. 2s. 14. Kyng Horn, with, fragments of Floriz and Blaunchenur, and the Assumption of the Blessed Virgin. Edited from the MSS. in the Library of the University of Cambridge and the British Museum, by the Rev. J. Rawson Lumby. 3s. 6d. 15. Political, Religious, and Love Poems, from the Lambeth MS. No. 306, and other sources. Edited by F.J. Furnivall, Esq., M.A. 7s. f 6d' 16. A Tretice in English breuely drawe out of \ book of Quintis essencijs in Latyn, \> Hermys )> prophete and king of Egipt after J? flood of Noe, fader of Philosophris, hadde by reuelacioun of an aungil of God to him sente. Edited from the Sloane MS. 73, by F. J. Furnivall, Esq., M.A. Is. 17. Parallel Extracts from 29 Manuscripts of Piers Plowman, with Comments, and a Proposal for the Society's Three-text edition of this Poem. By the Rev. W. Skeat, M.A. Is. 18. Hali Meidenhead, about 1200 a.d. Edited for the first time from the MS. (with a translation) by the Rev. Oswald Cockayne, M.A. Is. 8 and 60, Paternoster .Roto, London. 1 1 Early English Text Society's Publications— continued. 19. The Monarche, and other Poems of Sir David Lyndesay. Part II., the Complaynt of the King's Papingo, and other minor Poems. Edited from the First Edition by F. Hall, Esq., D.C.L. 3*. 6d. 20. Some Treatises by Richard Rolle de Hampole. Edited from Robert of Thornton's MS. (ab.1440 A.D.),byRev. George G. Perry, M. A. 1$. 21. Merlin, or the Early History of King Arthur. Part II. Edited by Henry B. Wheatley, Esq. 4s. 22. The Romans of Partenay, or Lusignen. Edited for the first time from the unique MS. in the Library of Trinity College, Cambridge, by the Rev. W. W. Skeat. M.A. 6s. 23. Dan Michel's Ayenbite of Inwyt, or Remorse of Conscience, in the Kentish dialect, 1340 a.d. Edited from the unique MS. in the British Museum, by Richard Morris, Esq. 10*. 6d. 24. Hymns of the Virgin and Christ ; The Parliament of Devils, and Other Religious Poems. Edited from the Lambeth MS. 853, by F. J. Furnivall, M.A. 3*. 25. The Stacions of Rome, and the Pilgrim's Sea- Voyage and Sea- Sickness, with Clene Maydenhod. Edited from the Vernon and Porkington MSS., etc., by F. J. Furnivall, Esq., M.A. Is. 26. Religious Pieces in Prose and Verse. Containing Dan Jon Gaytrigg's Sermon ; The Abbaye of S. Spirit ; Sayne Jon, and other pieces in the Northern Dialect. Edited from Robert of Thorntone's MS. (ab. 14G0 a.d.), by the Rev. G. Perry, M.A. 2*. 27. Maniptjltjs Vocabtjlortjm : a Rhyming Dictionary of the English Language, by Peter Levins (1570). Edited, with an Alphabetical Index, by Henry B. Wheatley. 12s. 28. The Vision of William concerning Piers Plowman, together with Vita de Dowel, Dobet et Dobest. 1362 a.d., by William Langland. The earliest or Vernon Text ; Text A. Edited from the Vernon MS., with full Collations, by Rev. W. W. Skeat, M.A. 7*. 29. Old English Homilies and Homiletic Treatises. (Sawles Warde and the Wohunge of Ure Lauerd : Ureisuns of Ure Louerd and of Ure Lefdi, etc.) of the Twelfth and Thirteenth Centuries. Edited from MSS. in the Brit- ish Museum, Lambeth, and Bodleian Libraries ; with Introduction, Transla- tion, and Notes. By Richard Morris. First Series. Part I. 7s. 30. Piers, the Ploughman's Crede (about 1394). Edited from the MSS. by the Rev. W. W. Skeat, M.A. 2*. 31. Instructions for Parish Priests. By John Myrc. Edited from Cotton MS. Claudius A. II., by Edward Peacock, Esq., F.S.A., etc., etc. h. 32. The Babees Book, Aristotle's ABC, Urbanitatis, Stans Puer ad Mensam, The Lytille Childrenes Lytil Boke. The Bokes of Nurture of Hugh Rhodes and John Russell, Wynk) n de Worde's Boke of Kervynge, The Booke of Demeanor, The Boke of Curtasye, Seager's Schoole of Vertue. etc., etc. With some French and Latin Poems on like subjects, and some Fore- words on Education in Early England. Edited by F. J. Furnivall, MA., Trin. Hall, Cambridge 15s. 33. The Book of the Knight de la Tofr Laitdhy, 1372. A Fall Book for his Daughters, Edited from the Harleian MS. 17u'l, by Thomas Wright, Esq., M. A., and Mr. William Rosmtkr. 8a. 34. Old English Homilies and Homilk lie Tit i:\ri -is. (Sawles Warde, and the Wohunge of Ure Lauerd: Ureisuns of Ure Louerd and of Ure Lefdi, etc.) of the Twelfth and Thirteenth Centuries. Edited from MSS. in the British Museum, Lambeth, and Bodleian Libraries; with Introduction, Trans- lation, and Notes, by Richard Morris. First Strw. Part 2. 12 Linguistic Publications of Triibner $* Co. Early English English Text Society's Publications— continued. 35. Sib David Lyndesay' s Woeks. Paet 3. The Historie of ane Nobil and Wailzeand Sqvyer, William Meldrum, umqvhyle Laird of Cleische and Bynnis, compylit be Sir Dauld Lyxdesay of the Mont alias Lyoun King of Armes. With the Testament of the said Williame Mel- drum, Squyer, compylit alswa he Sir Dauid Lyndesay, etc. Edited by F. Hall, D.C.L. 2s. 36. Meelin, oe the Eaely Histoey of King Aethtje. A Prose Romance (about 1450-1460 a.d.), edited from the unique MS. in the University Library, Cambridge, by Henry B. Wheatley. With an Essay on Arthurian Localities, by J. S. Stuart Glexxie, Esq. Part III. 1869. 12s. :37. Sle David Lyndesay' s Woeks. Part TV. Ane Satyre of the thrie estaits, in commendation of vertew and vitvperation of vyce. Maid be Sir David Linjjesay, of the Mont, alias Lyon King of Armes. At Edinbvrgh. Printed be Eobert Charteris, 1602. Cvm privilegio regis. Edited by F. Hall, Esq., D.C.L. 4*. 38. The Yision of William conceening Piees the Plowman, together with Vita de Dowel, Dobet, et Dobest, Secundum Wit et Resoun, by William Langlaxd (1377 a.d.). The " Crowley" Text; or Text B. Edited from MS. Laud Misc. 581, collated with MS. Rawl. Poet. 38, MS. B. 15. 17. in the Library of Trinity College, Cambridge, MS. Dd. 1. 17. in the Cambridge University Library, the MS. in Oriel College, Oxford, MS. Bodley 814, etc. By the Rev. Walter W. Skeat, M.A., late Fellow of Christ's College, Cambridge. 10s. 6d. 39. The "(test Hystoeiale" of the Destetjction of Teoy. An Alliterative Romance, translated from Guido De Colonna's " Hystoria Troiana." Now first edited from the unique MS. in the Hunterian Museum, University of Glasgow, by the Rev. Geo. A. Paxtox and David Doxaldsox. Part I. 10s. 6d. 40. English Gilds. The Original Ordinances of more than One Hundred Early English Gilds : Together with the olde usages of the cite of Wynchestre ; The Ordinances of Worcester ; The Office of the Mayor of Bristol ; and the Customary of the Manor of Tettenhall- Regis. From Original MSS. of the Fourteenth and Fifteenth Centuries. Edited with Notes by the late Toulmix Smith, Esq., F.R.S. of Northern Antiquaries (Copenhagen). With an Introduction and Glossary, etc., by his daughter, Lucy Toulmix Smith. And a Preliminary Essay, in Five Parts, Ox the History axd Developmext of Gilds, by Lujo Brextaxo, Doctor Juris Utriusque et Philosophise. 21s. 41. The Minoe Poems of William Latjdee, Playwright, Poet, and Minister of the Word of God (mainly on the State of Scotland in and about 1568 A.D., that year of Famine and Plague). Edited from the Unique Originals belonging to S. Christie-Miller, Esq., of Britwell, by F. J. Furnivall, M.A., Trin. Hall, Camb. 3s. 42. Beenaedtjs de Cuea eei Pamvliaeis, with some Early Scotch Prophecies, etc. From a MS., KK 1. 5, in the Cambridge University- Library. Edited by J. Rawsox Lumby, M.A., late Fellow of Magdalen College, Cambridge. 2s. 43. Ratis Baying, and other Moral and Religious Pieces, in Prose and Verse. Edited from the Cambridge University Library MS. KK 1. 5, by J. Rawsox Lumby, M.A., late Fellow of Magdalen College, Cambridge. 3s. 44. Joseph of Aeimathie : otherwise called the Romance of the Seint Graal, or Holy Grail: an alliterative poem, written about a.d. 1350, and now first printed from the unique copy in the Vernon MS. at Oxford. With an appendix, containing "The Lyfe of Joseph of Armathy," reprinted from the black-letter copy of Wynkyn de Worde ; " De sancto Joseph ab Arimathia," first printed by Pynson, a.d. 1516 ; and " The Lyfe of Joseph of Arimathia," first printed by Pynson, a.d. 1520. Edited, with Notes and Glossarial Indices, by the Rev. Walter W. Skeat, M.A. 5s. 8 and 60, Paternoster Row, London. 13 Early English Text Society's Publications— continued. 45. King Alfred's West- Saxon Version of Geegoey's Pastoral Care. With an English translation, the Latin Text, Notes, and an Introduction Edited by Henry Sweet, Esq., of Balliol College, Oxford. Part I. lO.v. Extra Series. Subscriptions — 'Small paper, one guinea ; large paper, two guineas, per annum. 1. The Eomance of William: of Paleene (otherwise known as the Romance of William and the Werwolf). Translated from the French at the command of Sir Humphrey de Bohun, about a.d. 1350, to which is added a fragment of the Alliterative Romance of Alisaunder, translated from the Latin by the same author, about a.d. 1340 ; the former re-edited from the J3 unique MS. in the Library of King's College, Cambridge, the latter now first edited from the unique MS. in the Bodleian Library, Oxford. By the Rev. Walter W. Skeat, M.A. 8vo. sewed, pp. xliv. and 328. £l 6s. 2 r On Eaely English Peonunctation, with especial reference to Shakspere and Chaucer ; containing an investigation of the Correspondence of Writing with Speech in England, from the Anglo-Saxon period to the present day, preceded by a systematic Notation of all Spoken Sounds by means of the ordinary Printing Types ; including a re-arrangement of Prof. F. J. Child's Memoirs on the Language of Chaucer and Gower, and reprints of the rare Tracts by Salesbury on English, 1547, and Welsh, 1567, and by Barcley on French, 1521. By Alexander J. Ellis, F.R.S. Part I. On the Pronunciation of the xivth, xvith, xvnth, andxvuith centuries. 8vo. sewed, pp. viii. and 416. 10*. 3. Caxton's Book of Cuetesye, printed at Westminster about 1477-8, A.d., and now reprinted, with two MS. copies of the same treatise, from the Oriel MS. 79, and the Balliol MS. 354. Edited by Frederick J. Furni- vall, M.A. 8vo. sewed, pp. xii. and 58. 5s. 4. The Lay of Havelok the Dane; composed in the reign of Edward I., about a.d. 1280. Formerly edited by Sir F. Madden for the Roxburghe Club, and now re-edited from the unique MS. Laud Misc. 108, in the Bodleian Library, Oxford, by the Rev. Walter W. Skeat, M.A. 8vo. sewed, pp. lv. and 160. 10s. 5. Chatjcee's Teanslation of Boethitjs's " De Consolatione Philosophie." Edited from the Additional MS. 10,340 in the British Museum. Collated with the Cambridge Univ. Libr. MS. Ii. 3. 21. By Richard Morris. 8vo. 12s. 6. The Eomance of the Cheveleee Assigne. Ee-edited from the unique manuscript in the British Museum, with a Preface, Notes, and Glossarial Index, by Henry H. Gibbs, Esq., M.A. 8vo. sewed, pp. xviii. and 38. 3s. 7. On Eaely English Peonunciation, with especial reference to Shakspere and Chaucer. By Alexander J. Ellis, F.R.S. , etc., etc. Part II. On the Pronunciation of the xm th and previous centuries, of Anglo-Saxon, Icelandic, Old Norse and Gothic, with Chronological Tables of the Value of Letters and Expression of Sounds in English Writing. 10s. 8. Qtjeene Elizabethes Achademy, by Sir Humpiikf.y Gilui A Booke of Precedence, The Ordering of a Funerall, etc. Varying Versions of the Good Wife, The Wise Man, etc., Maxims, Lydgate's Order of Fools, A Poem on Heraldry, Occleve on Lords' Men, etc., Edited by F. J. Furnivall, M.A., Trin. Hall, Camb. With Essays on Early Italian ami German Books of Courtesy, by W. M. Rossetti, Esq., and E. Osw.vid, Esq. 8vo. 13s. 14 Lingidstic Publications of Trubner ^ Co. Early English Text Society's Publications— continued. 9. The Fraternttye op Yacarondes, by John Awdelet (licensed in 1560-1, imprinted then, and in 1565), from the edition of 1575 in the Bodleian Library. A Caueat or Warening for Commen Cursetors vulgarely called Vagabones, by Thomas Harman, Esq,uiere. From the 3rd edition of 1567, belonging to Henry Huth, Esq., collated with the 2nd edition of 1567, in the Bodleian Library, Oxford, and with the reprint of the 4th edition of 1573. A Sermon in Praise of Thieves and Thievery, by Parson Haben or Hyberdyne, from the Lansdowne MS. 98, and Cotton Vesp. A. 25. Those parts of the Groundworke of Conny-catching (ed. 1592), that differ from Harman' s Caueat. Edited by Edward Viles & F. J. Furnivall. 8vo. Is. 6d. 10. The Fyrst Boke of the Introduction" of Knowledge, made by Andrew Borde, of Physycke Doctor. A Compendyous Regyment of a Dyetary of Helth made in Mountpyllier, compiled by Andrewe Boorde, of Physycke Doctor. Barnes in the Defence of the Berde : a treatyse made, answerynge the treatyse of Doctor Borde upon Berdes. Edited, with a life of Andrew Boorde, and large extracts from his Breuyary, by F. J. Furnivall, M. A., Trinity Hall, Camb. 8vo. 18s. 11. The Bruce ; or, the Book of the most excellent and noble Prince, Robert de Broyss, King of Scots : compiled by Master John Barbour, Arch- deacon of Aberdeen, a.d. 1375. Edited from MS. G 23 in the Library of St. John's College, Cambridge, written a.d. 1487 ; collated with the MS. in the Advocates' Library at Edinburgh, written a.d. 1489, and with Hart's Edition, printed a.d. 1616 ; with a Preface, Notes, and Glossarial Index, by the Rev. Walter W. Skeat, M.A. Part I. 8vo. 12s. 12. England in the Reign of King Henry the Eighth. A Dialogue between Cardinal Pole and Thomas Lupset, Lecturer in Rhetoric at Oxford. By Thomas Starkey, Chaplain to the King. Edited, with Preface, Notes, and Glossary, by J. M. Cowper. And with an Introduction containing the Life and Letters of Thomas Starkey, by the Rev. J. S. Brewer, M.A. Part II. 12*. (Fart I., Starkey' s Life and Letters, is in preparation. 13. A Supplicacyon for the Beggars. Written about the year 1529, by Simon Fish. Now re-edited by Frederick J. Furnivall. With a Supplycacion to our moste Soueraigne Lorde Kynge Henry the Eyght (1544 a.d.), A Supplication of the Poore Commons (1546 a.d.), The Decaye of England by the great multitude of Shepe (1550-3 a.d.). Edited by J. Meadows Cow per. %s. Edda Saemundar Hinns Froda — The Edda of Saemund the Learned. From the Old Norse or Icelandic. Part I. with a Mythological Index. 12mo. pp. 152, cloth, 3s. 6d. Part II. with Index of Persons and Places. By Benjamin Thorpe. 12mo. pp. viii. and 172, cloth. 1866. 4s. ; or in 1 Vol. complete, 7s. 6d. Edkins. — China's Place in Philology. An attempt to show that the Languages of Europe and Asia have a common origin. By the Rev. Joseph Edkins. Crown 8vo , pp. xxiii. — 403, cloth. Edkins. — A Vocabulary of the Shanghai Dialect. By J. Edelns. 8vo. half-calf, pp. vi. and 151. Shanghai, 1869. 21s. Edkins. — A Grammar of Colloquial Chinese, as exhibited in the Shanghai Dialect. By J. Edkins, B.A. Second edition, corrected. 8vo. half-calf, pp. viii. and 225. Shanghai, 1868. 21*. Edkins. — A. Grammar of the Chinese Colloquial Language, com- monly called the Mandarin Dialect By Joseph Edkins. Second edition. 8vo. half-calf, pp. viii. and 279. Shanghai, 1864. £1 10s. Eger and Grime ; an Early English Romance. Edited from Bishop Percy's Folio Manuscript, about 1650 a.d. By John W. Hales, M.A., Fellow and late Assistant Tutor of Christ's College, Cambridge, and Frederick J. Furnivall, M.A., of Trinity Hall, Cambridge. 1 vol. 4to. (only 100 copies printed), bound in the Eoxburghe style, pp. 61. Price 10s. 6d. 8 and 60, Paternoster Row, London. ] Eitel. — Handbooi fob tbm Bnmm oj Chimh Buddhmi Rev. E.J. lii hi., of the London Missionary Society. Crown Hvo. pp. \iii., 224, cloth, 18*. Eitel. — Tiikke Lectures on Buddhism. By the Bar, K J, Kmi.. (Jn I Elliot. — Thi Histoey of India, as told by Its own Historiam. Muhammadan Period. Edited from the Posthumous Papers of the lata Sir II. M. ELLIOT, ELC.B., East India Company's Bengal Civil E John DOWBON, M.E.A.8., Staff College, Sandhurst. Vols. I. and II. With a Portrait of Sir H. M. Elliot. 8vo.pp xxxii. and 542, x. and 580, cloili. L8 Vol. III. 8vo. pp. xii. and 627, cloth. 2 h. Elliot. — Memoirs on the History, Folk-Lore, and DlftTBlBUTlO] the Races of the North Western Provinces of India; being an amplified Edition of the original Supplementary Glossary of Indian Terms. By the late Sir Henry M. Elliot, K.C.B., of the Hon. East India Company's Bengal Civil Service. Edited, revised, and re-arranged, by John Bl M.H.A.S., Bengal Civil Service; Member of the German Oriental Society, of the Asiatic Societies of Paris and Bengal, and of the Philological Boo London. In 2 vols, demy 8vo., pp. xx., 370, and 39G, cloth. With two Lithographic Plates, one full-page coloured Map, and three large coloured folding Maps. 36s. Ellis. — The Asiatic Affinities of the Old Italians. By Bobebx Ellis, B.D., Fellow of St. John's College, Cambridge, and author of " Ancient Routes between Italy and Gaul." Crown 8vo. pp. iv. 156, cloth. 1870. 5*. English and Welsh Languages. — The Influence of thi Bioiibb am» Welsh Languages upon each other, exhibited in the Vocabularies of the two Tongues. Intended to suggest the importance to Philologers, Antiquaries, Ethnographers, and others, of giving due attention to the Celtic Branch of the Indo-Germanic Family of Languages. Square, pp. 30, sewed. 1869. 1 1. Etherington. — The Student's Grammar of tin; Enrol L By the Rev. W. Etherington, Missionary, Benares. Crown 8vo. pp. \ xlviii. cloth. 1870. 10s. 6d. Ethnological Society of London (The Journal of the). Edited by Professor Huxley, F.R.S., President of the Society ; George I F.R.S. ; Sir John Lubbock, Bart., F. R. S. ; Colonel A. Lank Pox, Hob Thomas Wright, Esq., Hon. Sec; Hyde Clakke, Esq.; Sub-Editor; and Assistant Secretary, J. H. Lamprey, Esq. Published Quarterly. Vol. L, No. 1. April, 1869. 8vo. pp. 88, sewed. 3s. Contents.— Flint Instruments from Oxfordshire tad the Isle of Thanet. (Illustrated.) By Colonel A. Lane Fox.— The Westerly Drifting of Nomads. By 11. II. Howorth.— On thi Shilling. By Hyde Clarke.— Letter" on a Marble Armlet. By EL W. Edwardl On I I Spear from Lough Gur, Limerick. (Illustrated.) By Col. A. Lane Fax.— On Chi:. BvW. II. Black.— Proto-ethiiie Condition of Asia Minor, l'.y Hyde Clarke. -On Stone Im- plements from the Cape. (Illustrated.) By Sir J. Lubboek. — Cromlechs and Megalitilic Structures. By II. M. Westropp.— Remarks on Mr. Westropp's Paper. By Colonel A Fox.— Stone Implements from San Jose. By A. Stcffcns.-On Child-bearing in Australia New Zealand. By J. Hooker, M.D.— On a" Pseudo-cromlech on Mount Alexaml By Acheson. — The Cave Cannibals of South Africa. By I.ayland. — Keviews Malay Archipelago (with illustrations); Fryer's Hill Tribes of India (with an illu-.tr Reliquiae Aquitaniciv, etc.— .Method of Photographic Measurement of the Human Frame (with an illustration). By J. II. Lamprey.— Notes and Queries. Vol. I., No. 2. July, 1869. 8vo. pp. 117, sewed. S*. Contents.— Ordinary Meeting, March !), isi;*» (held at the Museum of Professor Huxley, F.K.S., President, in the Chair. Opening Address of the t»i the Characteristics of the population of Central and South India (Illustrate! Elliot. — Oo the Races of India as traced in existing Trio B and By G. Campbell, Esq.— Remarks by Mr. .lames Pergusson.- Remarks bj Mr. Walter Dendy. — Ordinarv Meeting, January 23rd, lSiiit. Professor Huxley, l.K lent, in the Chair. On the Lepohas. By Dr. A. Campbell, late superintendent historic Arcluvologv of India (Illustrated]. By Colonel Meadow* I M.U.I. A., etc.— Appendix I. Extract from description of the Pandoo I Babington, Esq. (Head before the Literary Society of Bombay, December 21 in V olume III. of the Society's Transactions .- Appendix II, Extract from alette: now Colonel, A. Doria, dated Camp Katannieh, April 12th. 1858.— On some of I Tribes of the North Western frontier of India. By Major Fosbcry, Y.c -On 1 16 Linguistic Publications of Trubner . 7a. Bel. Vol. I. Part 2.— Africa (North of the Tropic of Capricorn). 8vo. pp. 70. 2s. Vol. I. Part 3.— Madagascar. 8vo. pp. 24. Iff. Vol. II. Part 1.— Australia. 8vo. pp. iv. and 44. Is. CuJ. Vol. II. Part 2.— Papuan Languages of the Loyalty Islands and New Hebrides, compris- ing those of the Islands of Nengone, Lifu, Aneitum, Tana, ami others. 8vo. p. 12. iul. Vol.11. Part 3.— Fiji Islands and Rotuma (with Supplement to Fart II., Papuan Lan- guages, and Part I., Australia). 8vo. pp. 34. U, 18 Linguistic Publications of Trubner $* Co. Vol. II. Part 4.— New Zealand, the Chatham Islands, and Auckland Islands. 8vo. pp. 76. 3s. U. Vol.11. Part 4 {continuation).— Polynesia and Borneo. 8vo. pp. 77-154. 3s. Gd. Vol. III. Part 1.— Manuscripts and Incunables. 8vo. pp. viii. and 24. 2s. Vol. IV. Part 1.— Early Printed Books. England. 8vo. pp. vi. and 266. Grey. — Maobi Mementos: being a Series of Addresses presented by the Native People to His Excellency Sir George Grey, K.C.B., F.R.S. With Introductory Remarks and Explanatory Notes ; to which is added a small Collec- tion of Laments, etc. By Ch. Oliver B. Davis. 8vo. pp. iv. and 228, cloth. 12s. Green. — Shakespeare and the Emblem-~Weitebs : an Exposition of their Similarities of Thought and Expression. Preceded by a View of the Emblem-Book Literature down to a.d. 1616. By Henry Green, M.A. In one volume, pp. xvi. 572, profusely illustrated with Woodcuts and Photolith. Plates, elegantly bound in cloth gilt, large medium 8vo. ^1 lis. 6d ; large imperial 8vo. £2 12s. 6d. 1870. Griffith. — Scenes pbom the Ramayana, Meghadtjta, etc. Translated by Ralph T. H. Griffith, M.A., Principal of the Benares College. Second Edition. Crown 8vo. pp. xviii., 244, cloth. 6s. Contents. — Preface — Ayodhya— Bavan Doomed— The Birth of Bama— The Heir apparent— Manthara's Guile— Dasaratha's Oath — The Step -mother— Mother and Son— The Triumph of Love— Farewell?— The Hermit's Son— The Trial of Truth— The Forest— The Bape of Sita— Bama's Despair — The Messenger Cloud— Khumbakarna— The Suppliant Dove— True Glory — Feed the Poor— The Wise Scholar. Griffith. — The Bamayan of Valmiki. Translated into English verse. By Ralph T. H. Griffith, M.A., Principal of the Benares College. Vol. I., containing Books I. and II. 8vo pp. xxxii. 440, cloth. 1870. 18s. Vol. II., containing Book II., with additional Notes and Index of Names. 8vo. cloth, pp. 504. 18s. Grout. — The Isiztjltj : a Grammar of the Zulu Language ; accompanied with an Historical Introduction, also with an Appendix. By Rev. Lewis Grout. 8vo. pp. lii. and 432, cloth. 21s. Hang. — Essays on the Sacbed Language, Wbitings, and Religion of the Parsees. By Martin Haug, Dr. Phil. Superintendent of Sanskrit Studies in the Poona College. 8vo. pp. 278, cloth. [Out of print. Haug. — A Lectube on an Obiginal Speech of Zoboasteb (Yasna 45), with remarks on his age. By Martin Haug, Ph.D. 8vo. pp. 28, sewed. Bombay, 1865. 2s. Haug. — Outline of a Gbammab of the Zend Language. By Mabtin Haug, Dr. Phil. 8vo. pp. 82, sewed. 14s. Haug. — The Aitabeya Bbahmanam of the Eig Yeda : containing the Earliest Speculations of the Brahmans on the meaning of the Sacrificial Prayers, and od the Origin, Performance, and Sense of the Rites of the Vedic Religion. Edited, Translated, and Explained by Martin Haug, Ph.D., Superintendent of Sanskrit Studies in the Poona College, etc., etc. In 2 Vols. Crown 8vo. Vol. I. Contents, Sanskrit Text, with Preface, Introductory Essay, and a Map of the Sacrificial Compound at the Soma Sacrifice, pp. 312. Vol. II. Transla- tion with Notes, pp. 544. £3 3s. ■ Haug. — An Old Zand-Pahlayi Glossaby. Edited in the Original Characters, with a Transliteration in Roman Letters, an English Translation, and an Alphabetical Index. By Destur Hoshexgji Jamaspji, High-priest of the Parsis in Malwa, India. Revised with Notes and Introduction by Martin Haug, Ph.D., late Superintendent of Sanscrit Studies in the Poona College, Foreign Member of the Royal Bavarian Academy. Published by order of the Government of Bombay. 8vo. sewed, pp. lvi. and 132. 15s. Haug. — An Old Pahlayi-Pazand Glossaby. Edited, with an Alpha- betical Index, by Destur Hoshangji Jamaspji Asa, High Priest of the Parsis in Malwa, India. Revised and Enlarged, with an Introductory Essay on the Pahlavi Language, by Martin Haug, Ph.D. Published by order of the Government of Bombay. 8vo. pp. xvi. 152, 268, sewed. 1870. 28s. 8 and 60, Paternoster Row, London. 19 Hang. — Essay on the Pahlavi Language. By Maetin Haug, Ph. D., Professor of Sanscrit and Comparative Philology at the University of Munich, Member of the Royal Bavarian Academy of Sciences, etc. (From the Pahlavi- Pazand Glossary, edited by Destur Hoshangji and M. Haug.) 8vo. pp. 152, sewed. 1870. 3s. 6d. Haug. — The Religion of the Zoeoasteians, as contained in their Sacred Writings. With a History of the Zend and Pehlevi Literature, and a Grammar of the Zend and Pehlevi Languages. By Martin Haug, Ph.D., late Superin- tendent of Sanscrit Studies in the Poona College. 2 vols. 8vo. [In preparation. Heaviside. — American Antiquities ; or, the New World the Old, and the Old World the New. By John T. C. Heaviside. 8vo. pp. 46, sewed. 1*. 6d. Hepburn. — A Japanese and English Dictionary. With an English and Japanese Index. By J. C. Hepburn, A.M., M.D. Imperial 8vo. cloth, pp. xii., 560 and 132. 51. 5s. Hernisz. — A Guide to Conveesation in the English and Chinese Languages, for the use of Americans and Chinese in California and elsewhere. By Stanislas Hernisz. Square 8vo. pp. 274, sewed. 10s. 6d. The Chinese characters contained in this work are from the collections of Chinese groups, engraved on steel, and cast into moveable types, by Mr. Marcellin Legrand, engraver of the Imperial Printing Office at Paris. They are used by most of the missions to China. Hincks. — Specimen Chaptees of an Assteian Geammae. By the late Rev. E. Hincks, D.D., Hon. M.R.A.S. 8vo., pp. 44, sewed. 1*. History of the Sect of Maharajahs ; or, Yallabhachaeyas in Westeen India. With a Steel Plate. 8vo. pp. 384, cloth. 12*. Hoffmann. — Shopping Dialogues, in Japanese, Dutch, and English. By Professor J. Hoffmann. Oblong 8vo. pp. xiii. and 44, sewed. 3*. Hoffmann. — A Japanese Geammae. By J. J. Hoffmann, Ph. Doc, Member of the Royal Academy of Sciences, etc., etc. Published by command of His Majesty's Minister for Colonial Affairs. Imp. 8vo. pp. viii. 352, sewed. 12s. 6d. Historia y fundacion de la Ciudad de Tlaxcala, y sus cuatro cave- ceras. Sacada por Francisco de Loaiza de lengua Castellana a esta Mexicana. AfTo de 1718. Con una Traduccion Castellana, publicadopor S. Leon Reinisch. In one volume folio, with 25 Photographic Plates. [In preparation. Howse. — A Geammae of the Ceee Language. With which is com- bined an analysis of the Chippeway Dialect. By Joseph Howse, Esq., F.R. G.S. 8vo. pp. xx. and 324, cloth. 7s. 6d. Hunter. — A Compaeative Dictionaey of the Languages of India and High Asia, with a Dissertation, based on The Hodgson Lists, Official Records, and Manuscripts. By W. W. Hunter, B.A., M.R.A.S., Honorary Fellow, Ethnological Society, of Her Majesty's Bengal Civil Service. Folio, pp. vi. and 224, cloth. £2 2s. Ikhwanu-s Safa. — Ihkwanu-s Safa ; or, Beothees of Pueitt. De- scribing the Contention between Men and Beasts as to the Superiority of the Human Race. Translated from the Hindustani by Professor J. Dowsox, Stall College, Sandhurst. Crown 8vo. pp. viii. and 156, cloth. 7*. Inman.— Ancient Faiths Embodied in Ancient Times; or, an attempt to trace the Religious Belief, Sacred Rites, and Holy Emblems of certain Nations, by an interpretation of the names given to children by Priestly authority, or assumed by prophets, kings and hierarchs. By Thomas Inman, M.D., Liverpool. 2 vols. 8vo. pp. 1. and 1028, cloth, illustrated with numerous plates and woodcuts. ,£3. Inman. — Ancient Pagan and Modeen Cheistian Symbouhm Km and Explained. By Thomas Inman, M.D. (London), Physician to the Royal Infirmary, Liverpool. 8vo. pp. xvi. o'S, stiff covers, with numerous Illustrations. 1870. 5s.' 20 Linguistic Publications of Trubner 8f Co. Jaeschke. — A Short Practical Grammar of the Tibetan Language, with special Reference to the Spoken Dialects. By H. A. Jaeschke, Moravian Missionary. 8vo «ewed, pp. ii. and 56. 2s. 6d. Jaeschke. — Romanized Tibetan and English Dictionary, each word being re-produced in the Tibetan as well as in the Roman character. By H. A. Jaeschke, Moravian Missionary. 8vo. pp. ii. and 158, sewed. 5s. Julien. — Syntaxe Nouvelle de la Langtje Chlnoise. Vol. I. — Fondee sur la position des mots, suivie de deux traites sur les particules et les principaux termes de grammaire, d'une table des idiotismes, de fables, de legendes et d'apologues traduits mot a, mot. 8vo. sewed. 1869. 20s. Vol. II. — Fondee sur la position des mots confirmee par l'analyse d'un texte ancien, suivie d'un petit Dictionuaire du Roman des Deux Cousines, et de Dialogues dramatiques traduits mot a mot, par M. Stanislas Jolien, de l'lnstitut. 8vo. pp. 436, sewed. 1870. 20s. Justi. — Handbuch der Zendsprache, yon Ferdinand Jtjsti. Altbac- trisches Woerterbuch. Grammatik Chrestomathie. Four parts, 4to. sewed, pp. xxii. and 424. Leipzig, 1864. 24s. Kafir Essays, and other Pieces; with an English Translation. Edited by the Right Rev. the Bishop of Grahamstown. 32mo. pp. 84, sewed. 2s 6d. Kalidasa. — Kaghutansa. By Kalidasa. No. 1. (Cantos 1-3.) With Notes and Grammatical Explanations, by Rev. K. M. Banebjea, Second Professor of Bishop's College, Calcutta ; Member of the Board of Exam- iners, Fort -William; Honorary Member of the Royal Asiatic Society, London. 8vo. sewed, pp. 70. 4s. 6d. Kern. — The Brhat-Sanhita ; or, Complete System of Natural Astrology of Varaha-Mihira. Translated from Sanskrit into English by Dr. H. Kern, Professor of Sanskrit at the University of Leyden. 8vo. pp. 50, stitched, Part I. 2s. [ Will be completed in JSine Parts. Khirad-Afroz (The Illuminator of the Understanding). By Maulavi Hafizu'd-din. A new edition of the Hindustani Text, carefully revised, with Notes, Critical and Explanatory. By Edward B. Eastwick, M.P., F.R.S., F. S. A. , M.R. A. S., Professor of Hindustani at the late East India Company's College at Haileybury. 8vo. cloth, pp. xiv. and 321. 18s. Kidd. — Catalogue of the Chinese Library of the Royal Asiatic Society. By the Rev. S. Kidd. 8vo. pp. 58, sewed. Is. Kielhorn. — A Grammar of the Sanskrit Language. By E. Klelhoen, Ph.D., Superintendent of Sanskrit Studies in Deccan College. Registered under Act xxv. of 1867. Demy 8vo. pp. xvi. 260. cloth. 1870. 10s. 6d. Kistner. — Buddha and his Doctrines. A Biographical Essay. By Otto Kistner. Imperial 8vo., pp. iv. and 32, sewed. 2s. 6d. Koran (The). Arabic text, lithographed in Oudh, a.h. 1284 (1867). 16mo. pp. 942, bound in red goatskin, Oriental style, silver tooling. 7s. 6d. The printing, as well as the outer appearance of the book, is extremely tasteful, and the characters, although small, read very easily. As a cheap edition for reference this is preferable to any other, and its price puts it within the reach of every Oriental scholar. It is now first imported from India. Laghu Kaumudi. A Sanskrit Grammar. ByVaradaraja. With an English Version, Commentary, and References. By James R. Ballantyxe, LL.D., Prin- cipal of the Snskrit College, Benares. 8vo. pp. xxxvi. and 424, cloth. £] lis. 6d. Lee. — A Translation oe the Balavataro : a Native Grammar of the Pali Language. "With the Romanized Text, the Xagari Text, and Copious Explanatory Notes. By Lionel F. Lee. In one vol. 8vo. {In preparation). 8 and 60, Paternoster Row, London. 21 Legge — The Chinese Classics. "With a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes. By Jam El I D.D., of the London Missionary Society. In seven vols. Vol. I. containing Confucian Analects, the Great Learning, and the Doctrine of the Mean. 8vo. pp. 526, cloth. £2 2s.— Vol. II., containing the Works of Mencius. Bva pp. 634, cloth. £2 2s.— Vol. III. Part I. containing the First Part of the Shoo-King, or the Books of T. Aug, the Books of Yu, the Books of Hea. the Books of Shang, and the Prolegomena. Royal 8vo. pp. viii. and 280, cloth. £2 2s.— Vol. III. Part II. containing the Fifth Part of the Shoo-King, or the Books of Chow, and the Indexes. Royal 8vo. pp. 281 — 736, cloth. £2 2$. Legge.— The Life and Teachings of Confucius, with Explanatory Notes. By James Legge, D. D. Reproduced for General Readers from the Author's work, "The Chinese Classics," with the original Text. Second edition. Crown 8vo. cloth, pp. vi. and 338. 10s. 6d. Leigh. — The Eeligion of the World. By H. Stone Leigh. 12mo pp. xii. 66, cloth. 1869. 2s. 6d. Leitner. — The Eaces and Languages of Dardistan. By G. W. Leitner, M.A., Ph.D., Honorary Fellow of King's College London, etc.; late on Special Duty in Kashmir. 4 vols. 4to. [In tJie Press. Leland. — Hans Breitmann's Party. With other Ballads. By Charles G. Leland. Eighth Edition. Square, pp. xvi. and 74, sewed. 1*. Leland. — Hans Breitmann's Christmas. With other Ballads. By Charles G. Leland. Second edition. Square, pp. 80, sewed. Is. Leland. — Hans Breitmann as a Politician. By Charles G. Leland. Second edition. Square, pp. 72, sewed. Is. Leland. — Hans Breitmann in Church. With other Ballads. By Charles G. Leland. With an Introduction and Glossary. Square, pp. 80 sewed. 1870. Is. Leland. — Breitmann Ballads. Four Series complete. Contents: — Hans Breitmann's Party. Hans Breitmann's Christmas. Hans Breitmann as a Politician. Hans Breitmann in Church. With other Ballads. By Charles G. Leland. With Introductions and Glossaries. Square, pp. 300, cloth. 1870. 4s. U. Leland. — Hans Breitmann as an Uhlan. Six New Ballads, with a Glossary. Square, sewed, pp. 72. Is. Leland. — The Breitmann Ballads. Complete in 1 vol., including Nineteen Ballads illustrating his Travels in Europe (never before printed), with Comments by Fritz Schwackenhammer. By Charles G. Leland. Crown 8vo. handsomely bound in cloth, pp. xxviii. and 292. 6s. Lesley. — Man's Origin and Destiny, Sketched from the Platform of the Sciences, in a Course of Lectures delivered before the Lowell Institute, in Boston, in the Winter of 1865-6. By J. P. Lesley, Member of the National Academy of the United States, Secretary of the American Philosophical Society. Numerous Woodcuts. Crown 8vo. pp. 392, cloth. 10*. 6d. Liherien hag Avielen ; or, the Catholic Epistles and Gospels for the Day up to Ascension. Translated for the first time into the Bin honkc of Brittany. Also in three other parallel columns a New Version of the same into Breizounec (commonly called Breton and Armorican) ; a Version ii to Welsh, mostly new, and closely resembling the Preton ; and a Version G \m k or Manx or Cernaweg ; with Illustrative Articles by Christoll TbrriSN and Charles Waring Saxton, D.D. Ch. Ch., Oxford. The Penitential Psalms are also added. Oblong 4to. pp. 156, sewed. 5s. Lobscheid. — English and Chinese Dictionary, with the Pnnti and Mandarin Pronunciation. By the Rev. W. Lobscheid, Knight of Francis Joseph, C.M.I.R.G.S. A., N.Z.B.S.V., etc. Folio, pp. viii. and 2016. In Four Parts. £8 8s. 22 Linguistic Publications of Trubner $* Co. Lobscheid. — Chinese and English Dictionary, Arranged according to the Radicals. By the Rev. W. Lobscheid, Knight of Francis Joseph, C.M.I.R.G.S.A., N.Z.B.S.V., &c. 1 vol. imp. 8vo. double columns, pp. 600, bound. £2 8s. Ludewig (Hermann E.) — The Literature of American Aboriginal Languages. With Additions and Corrections by Professor Wm. W. Turner. Edited by Nicolas Trubner. 8vo. fly and general Title, 2 leaves ; Dr. Lude- wig's Preface, pp. v. — viii. ; Editor's Preface, pp. iv. — xii. ; Biographical Memoir of Dr. Ludewig, pp. xiii. — xiv. ; and Introductory Biographical Notices, pp. xiv. — xxiv., followed by List of Contents. Then follow Dr. Ludewig's Bibliotheca Glottica, alphabetically arranged, with Additions by the Editor, pp. 1 — 209 ; Professor Turner's Additions, with those of the Editor to the same, also alphabetically arranged, pp.210 — 246; Index, pp. 247 — 256; and List of Errata, pp. 257, 258. Handsomely bound in cloth. 10s. 6d. Macgowan. — A Manual of the Amoy Colloquial. By Rev. J. Macgowan, of the London Missionary Society. 8vo. sewed, pp. xvii. and 200. Amoy. 1871. £1 Is. Maclay and Baldwin. — An Alphabetic Dictionary of the Chinese Language in the Foochow Dialect. By Rev. R. S. Maclay, D.D., of the Methodist Episcopal Mission, and Rev. C. C Baldwin, A.M., of the American Board of Mission. 8vo. half-bound, pp. 1132. Foochow, 1871. £4 4s. Maha-Vira-Charita ; or, the Adventures of the Great Hero Eama. An Indian Drama in Seven Acts. Translated into English Prose from the Sanskrit of Bhavabhiiti. By John Pickford, M.A. Crown 8vo. cloth. Manava-Kalpa-Sntra ; being a portion of this ancient Work on Yaidik Rites, together with the Commentary of Kumarila-Swamin. A Facsimile of the MS. No. 17, in the Library of Her Majesty's Home Government for India. With a Preface by Theodor G-oldstucker. Oblong folio, pp. 268 of letter- press and 121 leaves of facsimiles. Cloth. £4 4s. Manipulus Vocabulorum ; A Rhyming Dictionary of the English Language. By Peter Levins (1570) Edited, with an Alphabetical Index, by Henry B. Wheatley. 8vo. pp. xvi. and 370, cloth. 14s. Manning. — An Inquiry into the Character and Origin of the Possessive Augment in English and in Cognate Dialects. By the late James Manning, Q.A.S., Recorder of Oxford. Svo.pp. iv. and 90. 2s. Markham.— Quichua Grammar and Dictionary. Contributions to- wards a Grammar and Dictionary of Quichua, the Language of the Yncas of Peru; collected by Clements R. Markham, F.S.A., Corr. Mem. of the Uni- versity of Chile. Author of " Cuzco and Lima," and "Travels in Peru and India." In one vol. crown 8vo., pp. 223, cloth. £l. Is. Markham. — Ollanta: A Drama in the Quichua Language. Text, Translation, and Introduction, By Clements R. Markham, F.R.G.S. Crown 8vo., pp. 128, cloth. 7s. 6d. Marsden. — Numismata Orientalia Illustrata. The Plates of the Oriental Coins, Ancient and Modern, of the Collection of the late William Marsden, F. R.S., etc., etc., engraved from drawings made under his direction. 4to. pp. iv. (explanatory advertisement), cloth, gilt top. £l lis. 6d. Mason. — Burmah : its People and Natural Productions ; or Notes on the Nations, Fauna, Flora, and Minerals of Tenasserim, Pegu, and Burmah. By Rev. F. Mason, D.D., M.R.A.S., Corresponding Member of the American Oriental Society, of the Boston Society of Natural History, and of the Lyceum of Natural History, New York. 8vo. pp. xviii. and 914, cloth. Rangoon, 1860. 305. Mason. — The Pali Text of Eachchayano's Grammar, with English Annotations. By Francis Mason, D.D. I. The Text Aphorisms, 1 to 673. II. The English Annotations, including the various Readings of six independent Burmese Manuscripts, the Singalese Text on Verbs, and the Cambodian Text on Syntax. To which is added a Concordance of the Aphorisms. In Two Parts. 8vo. sewed, pp. 208, 75, and 28. Toongoo, 1871. £l 12s. 8 and 60, Paternoster Row, London. 23 Mathuraprasada Misra. — A Trilingual Dictionary, being a compre- hensive Lexicon in English, Urdfi, and Hindi, exhibiting the Syllabication, Pronunciation, and Etymology of English AVords, with their Explanation in English, and in Urdu and Hindi in the Roman Character. By Matimk\- prasada Misra, Second Master, Queen's College, Benares. 8vo. pp. xv. and 1330, cloth. Benares, 1865. £2 2s. Mayers. — Illustrations of the Lamaist System in Tibet, drawn from Chinese Sources. By William Frederick Mayers, Esq., of Her Britannic Majesty's Consular Service, China. 8vo. pp. 21, sewed. 1869. 1*. 6d. Medhurst. — Chinese Dialogues, Questions, and Familiar Sentk r literally translated into English, with a view to promote commercial intercourse and assist beginners in the Language. By the late W. H. Medhurst, D.D. A new and enlarged Edition. 8vo. pp. 226. 18s. Megha-Duta (The). (Cloud -Messenger.) By Kalidasa. Translated from the Sanskrit into English verse, with Notes and Illustrations. By the late H. H. Wilson, M. A., F.R.S., Boden Professor of Sanskrit in the Uni- versity of Oxford, etc., etc. The Vocabulary by Francis Johnson, sometime Professor of Oriental Languages at the College of the Honourable the East India Company, Haileybury. New Edition. 4to. cloth, pp. xi. and 180. 10s. 6d. Memoirs read before the Anthropological Society of London, 1863 1864. 8vo., pp. 542, cloth. 21s. Memoirs read before the Anthropological Society of London, 1865-6. Vol. II. 8vo., pp. x. 464, cloth. 21s. Merx. — Grammatica Syriaca, quam post opus Hoffmanni refecit Adalbertus Merx, Phil. Dr. Theol. Lie. in Univ. Litt. Jenensi Priv. Docens. Particula I. Royal 8vo. pp. 136, sewed. 7s. Particula II. Royal 8vo. pp. 137-388, sewed. 10s. 6d. Moffat. — The Standard Alphabet Problem ; or the Preliminary- Subject of a General Phonic System, considered on the basis of some important facts in the Sechwana Language of South Africa, and in reference to the views of Professors Lepsius, Max Miiller, and others. A contribution to Phonetic Philology. By Robert Moffat, junr., Surveyor, Fellow of the Royal Geogra- phical Society. 8vo. pp. xxviii. and 174, cloth. 7s. 6d. Molesworth. — A Dictionary, Marathi and English. Compiled by J. T. Molesworth, assisted by George and Thomas Candy. Second Edition, revised and enlarged. By J. T. Molesworth. Royal 4to. pp. xxx and 922, boards. Bombay, 1857. £3 3s. Morfill. — The Slaves : their Ethnology, early History, and popular Traditions, with some account of Slavonic Literature. Being the substance of a course of Lectures delivered at Oxford. By W. R. Morfill, M.A. [/// prejHtration. Morley. — A Descriptive Catalogue of the Historical Mam scripts in the Arabic and Persian Languages preserved in the Library of the Royal Asiatic Society of Great Britain and Ireland. By William H. Morley, M.R.A.S. 8vo. pp. viii. and 160, sewed. London, 1854. 2s. 6d. Morrison. — A Dictionary of the Chinese Language. By the Rev. R. Morrison, D.D. Two vols. Vol. I. pp. x. and 762; Vol. II. pp.828, cloth. Shanghae, 1865. £6 6s. Muhammed. — The Life of Muhammed. Based on Mohammed Ibn Ishak By Abd El Malik Ibn Hisham. Edited by Dr. Ferdinand Wusten- feld. One volume containing the Arabic Text. 8vo. pp. 1026. sewed. Price 21s. Another volume, containing Introduction, Notes, and Index in German. 8vo. pp. lxxii. and 266, sewed. 7s. 6d. Each part sold separately. The test based on the Manuscripts of the Berlin, Leipaio, C.otha ami Lejdeo Libraries, has been carefully revised by the learned editor, and printed with the utmost axaotfl 24 Linguistic Publications of Trubner §• Co. Muir. — Original Sanskrit Texts, on the Origin and History of the People of India, their Eeligion and Institutions. Collected, Translated, and Illustrated by John Mum, Esq., D.C.L., LL.D., Ph.D. Vol. I. Mythical and Legendary Accounts of the Origin of Caste, with an Inquiry into its existence in the Vedic Age. Second Edition, re-written and greatly enlarged. 8vo. pp. xx. 532, cloth. 1868. 21s. Vol. 11. A New Edition is in preparation. Vol. III. The Vedas: Opinions of their Authors, and of later Indian Writers, on their Origin, Inspiration, and Authority. Second Edition, revised and enlarged. 8vo. pp. xxxii. 312, cloth. 1868. 16s. Vol. IV. Comparison of the Vedic with the later representation of the principal Indian Deities. 8vo. pp. xii. 440, cloth. 1863. 15s. Vol. V. Contributions to a Knowledge of the Cosmogony, Mythology, Religious Ideas, Life and Manners of the Indians in the Vedic Age. 8vo. pp. xvi. 492, cloth, 1870. 21*. M tiller (Max). — The Sacred Hymns oe the Brahmins, as preserved to us in the oldest collection of religious poetry, the Rig-Veda-Sanhita, translated and explained. By F. Max Muller, M.A., Fellow of All Souls' College ; Professor of Comparative Philology at Oxford ; Foreign Member of the Institute of France, etc., etc. In 8 vols. Volume I. 8vo. pp. clii. and 264. 12s. 6d. Muller (Max). — A New Edition of the Hymns of the Bjg-Yeda or the Sanhita Text, without the Commentary of the Sayana. Based upon the Editio princeps of Max Miiller. Large 8vo. of about 800 pages. [In preparation. " The above New Edition of the Sanhita Text of the Rig-Veda, -without the Commentary of Sayana, will contain foot-notes of the names of the Authors, Deities, and Metres. It will be comprised in about fifty large 8vo. sheets, and will be carefully corrected and revised by Frof. F. Max Muller. The price to subscribers before publication will be 24s. per copy. After" publi- cation the price will be 36s. per copy. Miiller (Max). — Lecture on Buddhist Nihilism. By P. Max Muller, M.A., Professor of Comparative Philology in the University of Oxford ; Mem- ber of the French Institute, etc. Delivered before the General Meeting of the Association of German Philologists, at Kiel, 28th September, 1869. (Translated from the German.) Sewed. 1869. Is. Maphegyi. — The Album of Language, illustrated by trie Lord's Prayer in one hundred languages, with historical descriptions of the principal languages, interlinear translation and pronunciation of each prayer, a dissertation on the languages of the world, and tables exhibiting" all known languages, dead and living. By G. Naphegyi, M.D., A.M., Member of the " Sociedad Geografica y Estadistica" of Mexico, and " Mejoras Materiales" of Texoco, of the Numis- matic and Antiquarian Society of Philadelphia, etc. In one splendid folio volume of 322 pages, illuminated frontispiece and title-page, elegantly bound in cloth, gilt top. £2 10s. Contents.— Preface (pp. 2). — Introduction. — Observations on the Origin of Language (pp. 12). —Authors of Collections of tbe Lord's Prayer (pp. 8).— Families of Language (pp. 13).— Alpha- bets (pp. 25). The Lord's Prayer in the following languages (each accompanied by a trans- literation into Roman characters, a translation into English, and a Monograph of the language) , printed in the original characters. A. Aryan Family. — 1. Sanskrit. 2. Bengalee. 3. Moltanee. 4. Hindoostanee. 5. Gipsy. 6. Greek. 7. Modern Greek. 8. Latin. 9. Italian. 10. French. 11. Spanish. 12. Portuguese. 13. Celtic. 14. Welsh. 15. Cornish. 16. Irish. 17. Gothic. 18. Anglo-Saxon. 19. Old Saxon and Dano-Saxon. 20. English (4 varieties). 21. German (4 varieties). 22. Dutch. 23. Runic. 24. Wallachian. 25. Icelandic. 26. Danish. 27. Norwegian. 28. Swedish. 29 Lithuanian. 30. Old Prussian. 31. Servian. 32. Sclavonic. 33. Polavian. 34. Bohemian. 35. Polish. 36. Russian. 37. Bulgaric. 38. Armenian. 39. Armenian-Turkish. 40. Albanian. 41. Persian. B. Semitic Family.— 1. Hebrew. 2. Chaldee. 3. Samaritan. 4. Syriac. 5. Syrc~Chald«ic. 6. Carshun. 7. Arabic. 8. iEthiopic. 8. Amharic. C. Turanian Family. — 1. Turkish. 2. Hungarian. 3. Finnish. 4. Estonian. 5. Lap- ponian. 6. Laplandic (Dialect of Uma-Lappmark) . 7. Basque. 8. Javanese. 9. Hawaiian. 10. Maori (New Zealandic). 11. Malay. 12. Ceylonese. 13. Moorish. 14. Coptic. 15. Berber. 16. Hottentot. 17. Susuic. 18. Burmese. 19. Siamese. 20. Mongolian. 21. Chinese. 22. Kalmuk. 23. Cashmere. D. American Family.— 1. Cherokee. 2. Delawar. 3. Micmac. 4. Totonac. 5. Othomi. 6. Cora. 7. Kolusic. 8. Greenland. 9. Mexican. 10. Mistekic. 11. Mayu. 12. Brazilian. 13. Chiquitic. 14. Amaric. 8 and 60, Paternoster Row, London. Nayler. — Commonsense Observations on the E of Rules (not yet reduced to System in any work extant) regarding Tin-: ENGLISH LAMfl on the pernicious effects of yielding blind obedience to so-called authorities, whether Dictionary -Compilers, Grammar- Makers, or SfELLlNG-BoOK ManufaCtuhers, instead of examining and judging for ourselves on all ques- tions that are open to investigation ; followed hy a Treatise, entitled I'udmn- ciation made Easy ; also an Essay on the Pronunciation of Pbofib Names. By B. S. Nayler, accredited Elocutionist to the most celebrated Literary Societies in London. 8vo. pp. iv. 148, boards. 1869. 5s. Newman. — A Dictionary of Modekn Arabic — 1. Anglo- Arabic Dictionary. 2. Anglo-Arabic Vocabulary. 3. Arabo-English Dictionary. By F. W. Newman, Emeritus Professor of University College, London. In 2 vols, crown 8vo., pp. xvi. and 376—464, cloth. £'1 Is. Newman. — A Handbook of Modern Arabic, consisting of a Practical ^Grammar, with numerous Examples, Dialogues, and Newspaper Extracts, in a European Type. By F. W. Newman, Emeritus Professor of University College, London ; formerly Fellow of Balliol College, Oxford. Post 8vo. pp. xx. and 192, cloth. London, 1866. 6s. Newman. — The Text of the Igtjvine Inscriptions, with interlinear Latin Translation and Notes. By Francis W. Newman, late Professor of Latin at University College, London. 8vo. pp. xvi. and 54, sewed. 2s. Newman. — Orthoepy : or, a simple mode of Accenting English, for the advantage of Foreigners and of all Learners. By Francis W. Newman, Emeritus Professor of University College, London. 8vo. pp. 28, sewed. 1869. Is. Notley. — A Comparative Grammar of the French, Italian, Spanish, and Portuguese Languages. By Edwin A. Notley. Crown oblong 8vo. cloth, pp. xv. and 396. 7s. 6d. Oriental Text Society.— {The Publications of the Oriental Text Society.) 1 . Theophania; or, Divine Manifestations of our Lord and Saviour. By Eu sebius, Bishop of Csesarea. Syriac. Edited by Prof. S. Lee. 8vo. 18-12. 15*. 2. Athanasitjs's Festal Letters, discovered in an ancient Syriac Version. Edited by the Rev. W. Cureton. 8vo. 1848. 1.5s. 3. Shahrastani : Book of Religious and Philosophical Sects, in Arabic. Two Parts. 8vo. 1842. 30s. 4. Umdat Aridat Ahl al Sunnat tva al TAMaaT; Pillar of the Creed of the Sunnites. Edited in Arabic by the Rev. W. Cureton. 8vo. 1 843 5. History of the Almohades. Edited in Arabic by Dr. R. L\ A. Dozy. 8vo. 184-7. 10s. 6d. 6. SamaYeda. Edited in Sanskrit by Rev. G. Stevenson. 8vo. 1843. 1 2*. 7. Dasa Kumara Charita. Edited in Sanskrit by Professor H. II. Wilson. 8vo. 1846. £1 4s. 8. Maha ViRa Charita, or a History of Rama. A Sanskrit Play. Edited by F. H. Trithen. 8vo. 1848. 15s. 9. Mazhzan tjl Asrar: The Treasury of Secrets. By Nizam i. Edited in Persian by N. Bland. 4to. 1844. 'tis. C)d, 10. Saiaman-u-Ubsal ; A Romance of Jami (Dshami). Edited in Persian by F. Falconer. 4to. 1843. 10s. 11. Mirkhond's History of the Atabeks. Edited in Persian by W. H. Morley. 8vo. 1850. 12s. 12. Tuhfat-ul-Ahrar; the Gift of the Noble. A Poem. By Jami (Dshami). Edited in Persian by F. Falconer. 4to. 1813. 10*. 26 Linguistic Publications of Trubner fy Co. Osburn. — The Monumental History of Egypt, as recorded on the Ruins of her Temples, Palaces, and Tombs. By Wjlltam Osburn. Illustrated with Maps, Plates, etc. 2 vols. 8vo. pp. xii. and 461 ; vii. and 643, cloth. £2 2s. Vol. I.— From the Colonization of the Valley to the Visit of the Patriarch Abram. Vol. II.— From the Visit of Abram to the Exodus. Palmer. — Egyptian Chronicles, with a harmony of Sacred and Egyptian Chronology, and an Appendix on Babylonian and Assyrian Antiquities. By William Palmer, M.A., and late Fellow of Magdalen College, Oxford. 2 vols., 8vo. cloth, pp. lxxiv. and 428, and viii. and 636. 1861. 12s. Pand-Namah. — The Pand-Namah ; or, Books of Counsels. By Adarbad Maraspand. Translated from Pehlevi into Gujerathi, by Harbad Sheriarjee Dadabhoy. And from Gujerathi into English by the Rev. Shapurji Edalji. Fcap. 8vo. sewed. 1870. 6d. Pandit. — A Pandit's Bemabks on Peofessoe Max Mullee's Teansla- tion of the " Big- Veda." Sanskrit and English. Fcap. 8vo. sewed. 1870. 6d. Paspati. — Etudes sue les Tchinghianes (Gypsies) ou Bohemlens de L' Empire Ottoman. Par Alexandre G. Paspati, M.D. Large 8vo. sewed, pp. xii. and 652. Constantinople, 1871. 28s. Patell. — Cotvasjee Patell' s Cheonology, containing corresponding Dates of the different Eras used by Christians, Jews, Greeks, Hindus, Mohamedans, Parsees, Chinese, Japanese, etc. By Cowasjee Sorabjee Patell. 4to. pp. viii. and 184, cloth. 50s. Pauthier. — Le Livee de Maeco Polo, Citoyen de Venise, Conseiller Prive et Commissaire Imperial de Khoubilai-Khaan. Redige en fran^ais sous sa dictee en 1298 par Rusticien de Pise ; Publie pour la premiere fois d'apres trois manuscrits inedits de la Bibliotheque Imperiale de Paris, presentant la redaction primitive du Livre, revue par Marco Polo lui-meme et donnee par lui, en 1307, a Thiebault de Cepoy, accompagnee des Variantes, de 1' Explication des mots hors d'usage, et de commentaires geographiques et historiques, tires des ecrivains orientaux, principalement Chinois, avec une Carte generale del' Asie par M. G. Pauthier. Two vols. roy. 8vo. pp. clvi. 832. With Map and Yiew of Marco Polo's House at Venice. £1 8s. Pazand. — The Book op the Mainyo-i-Khaed. The Pazand and Sanskrit Texts (in Roman characters) as arranged by Neriosengh Dhaval, in the fifteenth century. With an English translation, a Glossary of the Pazand texts, containing the Sanskrit, Rosian, and Pahlavi equivalents, a sketch of Pazand Grammar, and an Introduction. By E. W. West. 8vo. sewed, pp. 484. 1871. 16*. Percy. — Bishop Peecy's Folio Manuscexpts — Ballads and Bomances. Edited by John W. Hales, M.A., Fellow and late Assistant Tutor of Christ's College, Cambridge ; and Frederick J. Furnivall, M.A., of Trinity Hall, Cam- bridge ; assisted by Professor Child, of Harvard University, Cambridge, U.S.A., "W. Chappell, Esq., etc. In 3 volumes. Vol. I., pp. 610; Vol. 2, pp. 681. ; Vol. 3, pp. 640. Demy 8vo. half-bound, £4 4s. Extra demy 8vo. half-bound, on Whatman's ribbed paper, £6 6s. Extra royal 8vo., paper covers, on What- man's best ribbed paper, £10 10s. Large 4to., paper covers, on Whatman's best ribbed paper, £12. Perny. — Dictionnaiee Feanqais-Latin-Chinois de la Langue Manda- rine Parlee. Par Paul Perny. M.A., de la Congregation des Missions Etrangeres. 4to. pp. viii. 459, sewed. £2 2s. Perny. — Geammaiee Peatique de la Langue Mandaeine Paelee. Par Paul Perny, M.A., de la Congregation des Missions Etrangeres. [In the Press. Perny. — Peoveebes Chinois, Eecueillis et mis en oedee. Par Paul Perny, M.A., de la Congregation des Missions Etrangeres. 12mo. pp. iv. 135. 3s. 8 and 60, Paternoster Roiu, London. 27 Perrin. — English-Zulu Dictionary. New Edition, revised by J. A. Brickhill, Interpreter to the Supreme Court of Natal. 12mo. pp. 226', cloth, Pietermaritzburg, 1865. 5*. Philological Society. — Proposals for the Publication of a New Engi.imi Dictionary. 8vo. pp. 32, sewed. 6d. Pierce the Ploughman's Crede (about 1394 Anno Domini). Transcribed and Edited from the MS. of Trinity College, Cambridge, R. 3, 15. Col- lated with the MS. Bibl. Reg. 18. B. xvii. in the British Museum, and with the old Printed Text of 1553, to which is appended " God spede the Plough" (about 1500 Anno Domini), from the Lansdowne MS. 7>>'~. By the Rev. Walter W. Skeat, M. A., late Fellow of Christ's College, Cambridge, pp. xx. and 75, cloth. 1867. 2s. 6d. Prakrita-Prakasa ; or, The Prakrit Grammar of Vararuchi, with the Commentary (Manorama) of Bhamaha. The first complete edition of the ■° Original Text with Various Readings from a Collation of Six Manuscripts in the Bodleian Library at Oxford, and the Libraries of the Royal Asiatic Society and the East India House; with copious Notes, an English Translation, and Index of Prakrit words, to which is prefixed an easy Introduction to Prakrit Grammar. By Edward Byles Cowell, of Magdalen Hall, Oxford, Pro- fessor of Sanskrit at Cambridge. Second issue, with new Preface, and corrections. 8vo. pp. xxxii. and 204. 14*. Priaulx. — Qu2estiones Mosaics; or, the first part of the Book of Genesis compared with the remains of ancient religions. By Osmond de Beauvoir Priaulx. 8vo. pp. viii. and 548, cloth. 12s. Raja-Niti. — A Collection of Hindu Apologues, in the Braj Bhasha Language. Revised edition. With a Preface, Notes, and Supplementary Glossary. By Fitzedward Hall, Esq. 8vo. cloth, pp. 204. 21s. Ram Raz. — Essay on the Architecture of the Hindus. By Ram Raz, Native Judge and Magistrate of Bangalore, Corresponding Member of the R.A.S. of Great Britain and Ireland. With 48 plates. 4to. pp. xiv. and 64, sewed. London, 1834. Original selling price, £l 11*. 6d., reduced (for a short time) tc 12*. Rask.— A Grammar of the Anglo-Saxon Tongue. Prom the Danish of Erasmus Ra6k, Professor of Literary History in, and Librarian to, the University of Copenhagen, etc. By Benjamin Thorpe, Member of the Munich Royal Academy of Sciences, and of the Society of Netherlandish Literature, Ley den. Second edition, corrected and improved. l8mo. pp. 200, cloth, 5s. 6d. Rawlinson. — A Commentary on the Cuneiform Inscriptions of Babylonia and Assyria, including Readings of the Inscription on the Nimrud Obelisk, and Brief Notice of the Ancient Kings of Nineveh and Babylon, Read before the Royal Asiatic Society, by Major H. C. Rawlinson. 8vo., pp. 84, sewed. London, 1850. 2s. 6d. Rawlinson. — Outlines of Assyrian History, from the Inscriptions of Nineveh. By Lieut. Col. Rawlinson, C.B., followed by some Remarks by A. H. Layakd, Esq., D.C.L. 8vo., pp. xliv., sewed. London, 1852. 1*. Renan. — An Essay on the Age and Antiquity of the Book of Nabathjean Agriculture. To which is added an Inaugural Lecture on the Position of the Shemitic Nations in the History of Civilization. By M. BftMMT Renan, Membre de l'Institut. Crown 8vo., pp. xvi. and 148, cloth. St. M. Revue Celtique. — The Revue Celtique, a Quarterly Magazine for Celtic Philology, Literature, and History. Edited with the assistance of the Chief Celtic Scholars of the British Islands and of the Continent, and Con- ducted by H. Gaidoz. 8vo. Subscription, £l per annum. Ridley — Kamilaroi, Dippil, and Turrubul. Languages Spoken by Australian Aborigines. By Rev. Wm. Ridley, M.A., of the University of Sydney ; Minister of the Presbyterian Church of New South Wake, Printed by authority. Small 4to. cloth, pp. vi. and 90. 30s. 28 Linguistic Publications of Trubner §• Co. Rig- Veda. — A New Edition of the Hymns of the Rig-Yeda in the Sanhita Text, without the Commentary of the Sayana. Based upon the Editio princeps of Max Muller. Large 8vo. of about 800 pages. See also under Max Muller. [In preparation. Rig-Veda-Sanhita (The). The Sacred Hymns of the Brahmins, as preserved to us in the oldest collection of Religious Poetry, the Rig-Yeda- Sanhita, translated and explained. By F. Max Mulleu, M.A., Fellow of All Soul's College ; Professor of Comparative Philology at Oxford ; Foreign Member of the Institute of France, etc. . In 8 vols. Yol. I. 8vo. pp. clii. and 264-. 12«. 6d. Rig-Veda-Sanhita (The) : The Sacked Hymn's of the Beahm ans. Translated and explained by F. Max Muller, M.A., LL.D., Fellow of All Soul's College, Professor of Comparative Philology at Oxford, Foreign Member of the Institute of France, etc., etc. Vol. I. Hymns to the Maruts, or the Storm-Gods. 8vo. pp. clii. and 264. cloth. 1869. 12s. 6d. Rig- Veda Sanhita. — A Collection of Ancient Hindu Hymns. Con- stituting the First Ashtaka, or Book of the Rig-veda ; the oldest authority for the religious and social institutions of the Hindus. Translated from the Original Sanskrit. By the late H. H. Wilson, M.A., F.R.S., etc. etc. etc. Second Edition, with a Postscript by Dr. Fitzedward Hall. Yol. I. 8vo. cloth, pp. lii. and 348, price 21s. Rig-veda Sanhita.— A Collection of Ancient Hindu Hymns, constitut- ing the Fifth to Eighth Ashtakas, or books of the Big- Veda, the oldest Authority for the Beligious and Social Institutions of the Hindus. Translated from the Original Sanskrit by the late Horace Hatman Wilson, M.A., F.R.S., etc. Edited by E. B. Cowell, M.A., Principal of the Calcutta Sanskrit College. Vol. IV., 8vo„ pp. 214, cloth. 14s. A few copies of Vols. II. and III. still left. [ Vols. V. and VI. in the Press. Sama-Vidhana-Brahmana (The). With the Commentary of Sayana. Edited, with Notes, Translation, and Index, by A. C Burxell, M.R.A.S., Madras Civil Service. In 1 vol. 8vo. [In preparation. Scheie de Vere. — Studies in English ; or, Glimpses of the Inner Life of our Language. By M. Schele de Vere, LL.D., Professor of Modern Languages in the University of Virginia. 8vo. cloth, pp. vi. and 365. 10s. 6d. Schemeil. — El Mtjbtaeie; or, First Born. (In Arabic, printed at Beyrout). Containing Five Comedies, called Comedies of Fiction, on Hopes and Judgments, in Twenty-six Poems of 1092 Verses, showing the Seven Stages of Life, from man's conception unto his death and burial. By Emix Ibrahim Schemeil. In one volume, 4to. pp. 166, sewed. 1870. os. Schlagintweit. — Buddhism in Tibet. Illustrated by Literary Docu- ments and Objects of Religious Worship. "With an Account of the Buddhist Systems preceding it in India. By Emil Schlagixtweit, LL.D. With a Folio Atlas of 20 Plates, and 20 Tables of Native Prints in the Text. Royal 8vo., pp. xxiv. and 404. £2 2s. Schlagintweit. — Glossary of Geographical Teems feom India and Tibet, with Native Transcription and Transliteration. By Hermann de Schlagixtweit. Forming, with a " Route Book of the Western Himalaya, Tibet, and Turkistan," the Third Volume of H., A., andR. de Schlagintweit's "Results of a Scientific Mission to India and High Asia." With an Atlas in imperial folio, of Maps, Panoramas, and Views. Royal 4to., pp. xxiv. and 293. £4. Schlottmann. — The Monument of a Yictoet of Mesha, King of the Moabites. A Contribution to Hebrew Archssology by Dr. Koxstaxtlv Schlottmann, Professor of Theology at the University of Halle. Translated from the German. [In the Press. Shapnrji Edalji. — A Geaioiae of the Gujaeati Language. By Shapurji Edalji. Cloth, pp. 127. 10s. 6d. Shapurji Edalji. — A Dictionaey, Gejeati and English. By Shapeeji Edalji. Second Edition. Crown Svo. cloth, pp. xxiv. and STL 21s. 8 and 60, Paternoster Bow, London, 29 Sherring — The Sacked City op the Hindus. An Account of Benares in Ancient and Modern Times. By the Rev. M. A. Sherring, M.A., LL.D. ; and Prefaced with an Introduction by Fitzedward Hall, Esq., D.C.L. 8vo. cloth, pp. xxxvi. and 388, with numerous full-page illustrations. 21s. Smith. — A Vocabulary of Proper Names en - Chinese and English. of Places, Persons, Tribes, and Sects, in China, Japan, Corea, Assam, Siam, Burmah, The Straits, and adjacent Countries. By F. Porter Smith, M.B., China. 4to. half-bound, pp. vi.,.72, and x. 1870. 10s. 6d. Smith. —Contributions towards the Materia Medic a and Natural •History or China. For the use of Medical Missionaries and Native Medical Students. By F. Porter Smith, M.B. London, Medical Missionary in Central China. Imp. 4to. cloth, pp. viii. and 240. 1870. £1 Is. Sophocles. — A Glossary oe Later and Byzantine Greek. By E. A. ^Sophocles. 4to., pp. iv. and 624, cloth. £2 2s. Sophocles. — Eomaic or Modern Greek Grammar. By E. A. Sophocles. 8vo. pp. xxviii. and 196. 7s. 6c?. Sophocles.— Greek Lexicon of the Roman and Byzantine Periods (from B.C. 146 to a.d. 1100). By E. A. Sophocles. Imp. 8vo. pp. xvi. 1188, cloth. 1870. £2 8«. Steele.^ — An Eastern Love Story. Kusa Jatakaya : a Buddhistic • Legendary Poem, with other Stories. By Thomas Steele, Ceylon Civil Service. Crown 8vo. cloth, pp. xii. and 260. 1871. 6s. Stratmann.— A Dictionary of the English Language. Compiled from the writings of the xmth, xivth, and xvth centuries. By Francis Henry Stratmann. Second Edition. 8vo. Part I, price 10s. 6d. Stratmann.— An Old English Poem of the Owl and the Nightingale. Edited by Francis Henry Stratmann. 8vo. cloth, pp. 60. 3s. Syed Ahmed. — A Series of Essays on the Life of Mohammed, and Subjects subsidiary thereto. By Syed Ahmed Khan Bahador, C.S.I., Author of the " Mohammedan Commentary on the Holy Bible," Honorary Member of the Royal Asiatic Society, and Life Honorary Secretary to the Allygurh Scien- tific Society. 8vo. pp. 532, with 4 Genealogical Tables, 2 Maps, and a Coloured Plate, handsomely bound in cloth. 30s. Contents.— Preface and Introduction.— Essay on the Historical Geography of Arabia.— Essay on the Manners and Customs of the Pre-Islamic Arabians.— Essay on the various Religions of the Pre-Islamic Arabs, wherein it is inquired to which of them Islam bears the closest resem- blance, and whether by such affinity Islam is proved to be of Divine Origin or "A Cunningly Devised Fable." — Essay on the question Whether Islam has been Beneficial or Injurious to Human Society in general, and to the Mosaic and Christian Dispensations.— Essay on the Mohammedan Theological Literature. — Essay on the Mohammedan Traditions. — Essay on the Holy Koran. — Essay on the History of the Holy Mecca, including an account of the distinguished part enacted in connexion therewith by the Ancestors of Mohammed.— Essay on the Pedigree of Mohammed.— Essay on the Prophecies respecting Mohammed, as contained in both the Old and the New Testament.— Essay on Shakkisadar and Mcraj, that is, the splitting (open) of the Chest of Mohammed ; and his Night Journey.— Essay on the Birth and Childhood of Mohammed. The Boke of Nurture. By John- Russell, about 1460-1470 Anno Domini. The Boke of Keruynge. By Wynkyn de Worde, Anno Domini 1513. The Boke of Nurture. By Hugh Rhodes, Anno Domini 1577. Edited from the Originals in the British Museum Library, by Frederick J. Furni- vall, M.A., Trinity Hall, Cambridge, Member of Council of the Philological and Early English Text Societies. 4to. half-morocco, gilt top, pp. xix. and 146, 28, xxviii. and 56. 1867. 1/. lis. 6a?. The Vision of William concerning Piers Plowman, together with Vita de Dowel, Dobet et Dobest, secundum wit et resoun. By William Langland (about 1362-1380 anno domini). Edited from numerous Manu- scripts, with Prefaces, Notes, and a Glossary. By the Rev. Walter W. Skeat M.A. pp. xliv. and 158, cloth, 1867. Vernon Text ; Text A. 7s. 6d. 30 Linguistic Publications of Trubner §• Co Thomas. — Eaely Sassanian Inscription's, Seals and Coins, illustrating the Early History of the Sassanian Dynasty, containing Proclamations of Arde- shir Babek, Sapor I., and his Successors. "With a Critical Examination and Explanation of the Celebrated Inscription in the Hajiabad Cave, demonstrating that Sapor, the Conqueror of Valerian, was a Professing Christian. By Edward Thomas, Esq. 8vo. cloth, pp. 148, Illustrated. 7*. 6d. Thomas. — The Chronicles oe the Pathan Kings of Dehli. Illus- trated by Coins, Inscriptions, and other Antiquarian Remains. By Edward Thomas, F.R. A.S., late of the East India Company's Bengal Civil Service. With numerous Copperplates and Woodcuts. Demy 8vo. cloth, pp. xxiv. and 467. 1871. 28s. Thomas. — Essays on Indian Antiquities : following up the Discoveries of James Prinsep, with specimens of his Engravings, and selections from his Useful Tables, and embodying the most recent investigations into the History, Palaeography, and Numismatics of Ancient India. By Edward Thomas, late of the East India Company's Bengal Civil Service. In 2 vols. 8vo., profusely illustrated. [In preparation. Thomas. — The Theory and Practice oe Creole Grammar. By J. J. Thomas. Port of Spain (Trinidad), 1S69. One vol. 8vo. boards, pp. viii. and 135. 12s. ^ Thonissen. — Etudes see L'Histoire dtj Droit Crimlnel des Petjtles Anciens (Inde Brahmanique, E'gypte, Judee), par J. J. Thonissen, Professeur a l'Universite Catholique de Louvain, Membre de l'Academie Royale de Bel- gique. 2 vols. 8vo. pp. xvi. 248, 320, sewed. 1869. 12s. Thorpe. — Diplomatarium Anglictjm JEvi Saxonici. A Collection of English Charters, from the reign of King yEthelberht of Kent, A.D., DCV., to that of William the Conqueror. Containing: I. Miscellaneous Charters. II. Wills. III. Guilds. IV. Manumissions and Acquittances. With a Transla- tion of the Anglo-Saxon. By the late Benjamin Thorpe, Member of the Royal Academy of Sciences at Munich, and of the Society of Netherlandish Literature at Leyden. 8vo. pp. xlii. and 682, clotb. 1865. £l Is. Tindall. — A Grammar and Vocabulary oe the Namaqua-Hottentot Language. By Henry Tindall, Wesleyan Missionary. 8vo., pp. 124, sewed. 6s. Van der Tuuk.— Outlines oe a Grammar of the Malagasy Language By H. N. van der Tuuk. 8vo., pp. 28, sewed. Is. Van der Tuuk. — Short Account of the Malay Manuscripts belonging to the Royal Asiatic Society. By H. N. van der Tuuk. 8vo., pp.52. 2s. 6d Vishnu-Purana (The) ; a System of Hindu Mythology and Tradition. Translated from the original Sanskrit, and Illustrated by Notes derived chiefly from other Puranas. By the late H. H. Wilson, M.A., F.R.S., Boden Pro- fessor of Sanskrit in the University of Oxford, etc., etc. Edited by Fitzedward Hall. In 6 vols. 8vo. Vol. I. pp. cxl. and 200 ; Vol. II. pp. 343 ; Vol. III. pp. 348 ; Vol. IV. pp. 346 cloth ; Vol. V. pp. 392, cloth. Price 10s. 6d. each. [ Vol. VI. in preparation. Wade. — Yu-Yen Tztj-Erh Chi. A progressive course designed to assist the Student of Colloquial Chinese, as spoken in the Capital and the Metropolitan Department. In eight parts, with Key, Syllabary, and Writing Exercises. By Thomas Francis Wade, C.B., Secretary to Her Britannic Majesty's Legation, Peking. 3 vols. 4to. Progressive Course, pp. xx. 296 and 16 ; Syllabary, pp. 126 and 36 ;. Writing Exercises, pp. 48; Key, pp. 174 and 140, sewed. £4. Wade. — Wen-Chten Tzu-Erh Chi. A series of papers selected as specimens of documentary Chinese, designed to assist Students of the language, as written by the officials of China. In sixteen parts, with Key. Vol. I. By Thomas Francis Wade, C.B., Secretary to Her Britannic Majesty's Legation at Peking. 4to., half-cloth, pp. xii. and 455 ; and iv., 72, and 52. £1 16s. 8 and 60, Paternoster Roiv, London. 31 Wake. — Chapters on Man. With the Outlines of a Science of com- parative Psychology. By C. Staniland Wake, Fellow of the Anthropological Society of London. Crown 8vo. pp. viii. and 344, cloth. 7s. Gd. Watson. — Index to the Native and Scientific Names of Indian and other Eastern Economic Plants and Products, originally prepared under the authority of the Secretary of State for India in Council. By John Forbes Watson, M.A., M.D., F.L.S., F.R.A.S., etc., Reporter on the Products of India. Imperial 8vo., cloth, pp. 650. £1 lis. 6c?. Watts. — Essays on Language and Literature. By Thomas "Watts, late of the British Museum. Reprinted, with Alterations and Additions, from the Transactions of the Philological Society, and elsewhere. In 1 vol. 8vo. [In preparation. Webster. — An Introductory Essay to the Science of Comparative JTheology ; with a Tabular Synopsis of Scientific Religion. By Edward Webster, of Ealing, Middlesex. Read in an abbreviated form as a Lecture to a public audience at Ealing, on the 3rd of January, 1870, and to an evening congregation at South Place Chapel, Finsbury Square, London, on the 27th of February, 1870. 8vo. pp. 28, sewed. 1870. Is. Wedgwood. — A Dictionary of the English Language. By Hens- sleigh Wedgwood, M.A. late Fellow of Christ's College, Cambridge. Vol. I. (A to D) 8vo., pp. xxiv. 508, cloth, 14*. ; Vol. II. (E to P; 8vo. pp. 578, cloth, 14*.; Vol. III., Part I. (Q to Sy), 8vo. pp.366, 10*. 6rf.; Vol. III. Part II. (T to W) 8vo. pp. 200, 5s. Gd. completing the Work. Price of the complete work, £ 2 4*. " Dictionaries are a class of books not usually esteemed light reading; but no intelligent man were to be pitied who should find himself shut up on a rainy day in a lonely house in the dreariest part of Salisbury Plain, with no other means of recreation than that which Mr. Wedg- wood's Dictionary of Etymology could afford him. He would read it through from cover to cover at a sitting, and only regret that he had not the second volume to begin upon forthwith. It is a very able book, of great research, full of delightful surprises, a repertory of the fairy tales of linguistic science."— Spectator. Wedgwood. — On the Origin of Language. By Hensleigh Wedgwood, late Fellow of Christ's College, Cambridge. Fcap. 8vo. pp. 172, cloth. 3*. Gd. Wekey. — A Grammar of the Hungarian Language, with appropriate Exercises, a Copious Vocabulary, and Specimens of Hungarian Poetry. By Stgismund Wekey, late Aide-de-Camp to Kossuth. 12mo., pp. xii. and 150, sewed. 4s. Gd. Wheeler. — The History of India from the Earliest Ages. By J. Talboys Wheeler, Assistant Secretary to the Government of India in the Foreign Department, Secretary to the Indian Record Commission, author of "The Geography of Herodotus," etc. etc. Vol. I., The Vedic Period and the Maha Bharata. 8vo. cloth, pp. lxxv. and 576'. 18s. Vol. II., The Ramayana and the Brahmanic Period. 8vo. cloth, pp. lxxxviii. and 680, with 2 Maps. 21s. Whitney. — Atharva Veda PrAtic,akhya ; or, Caunakiya Catunulhva- yik&. (The). Text, Translation, and Notes. By William D. Whitney, Pro- fessor of Sanskrit in Yale College. 8vo. pp. 286, boards. 12s. Whitney. — Language and the Study of Language : Twelve Lectures on the Principles of Linguistic Science. By William DwiGHT Wiutm.v, Pro- fessor of Sanskrit, etc., in Yale College. Second Edition, augmented by an Analysis. Crown 8vo. cloth, pp. xii. and 504. 10^. Gd. Whitney. — TIittiriya-PrAticIkhya, with its Commentary, the Tribhashyaratna : Text, Translation and Notes. By W. D. Whitnky, Prof. of Sanskrit in Yale College, iNew Haven*. Svo. pp. 469. 1871. 25*. Wilkins. — The Bhagavat-Geeta ; or, Dialogues of Kreeshna and Arjoon. Translated by CHAB. Wilkins. A faithful reprint of thr now very scarce Original London Edition of 1785, made at the Bradaheet Press, New York. In one vol. Svo. Beautifully printed with old face type on laid paper. 261 copies were produced of this edition, of which only a few now remain, 12#. 32 Linguistic Publications of Trubner §• Co. Williams. — First Lessons in the Maori Language, with a Short Vocabulary. By W. L. Williams, B. A. Square 8vo., pp. 80, cloth, London, m 1862. 10*. Williams. — Lexicon Cornu-Britannicum. A Dictionary of the Ancient Celtic Language of Cornwall, in which the words are elucidated by copious examples from the Cornish works now remaining, with translations in English. The synonyms are also given in the cognate dialects of Welsh, Armoric, Irish, Gaelic, and Manx, showing at one view the connexion between them. By the Rev. Robert Williams, M.A., Christ Church, Oxford, Parish Curate of Llangadwaladr and Rhydycroesan, Denbighshire. Sewed. 3 parts, pp. 400. £2 5s. Williams. — A Dictionary, English and Sanscrit. By Monier Williams, M.A. Published under the Patronage of the Honourable East India Company. 4to. pp. xii. 862, cloth. London, 1855. £3 3s. Wilson. — Works of the late Horace Hayman Wilson, M.A., F.R.S., Member of the Royal Asiatic Societies of Calcutta and Paris, and of the Oriental Society of Germany, etc, and Boden Professor of Sanskrit in the University of Oxford. Vols I. and II. Also, under this title, Essays and Lectures chiefly on the Religion of the Hindus, by the late H. H. Wilson, M.A., F.R.S., etc. Collected and edited by Dr. Reinhold Rost. 2 vols, cloth, pp. xiii. and 399, vi. and 416. 21*. Wilson. — Works of the late Horace Hayman Wilson, M.A., F.R.S., Member of the Royal Asiatic Societies of Calcutta and Paris, and of the Oriental Society of Germany, etc., and Boden Professor of Sanskrit in the University of Oxford. Vols. Ill, IV. and V. Also, under the title of Essays Analytical, Critical, and Philological, on subjects connected with Sanskrit Literature. Collected and Edited by Dr. Reinhold Rost. 3 vols. 8vo. pp. 408, 406, and 390, cloth. Price 36s Wilson. — Works of the Late Horace Hayman Wilson. Vols. VI. VII. VIII, IX. and X. Also, under the title of the Vishnu Purana, a system, of Hindu mythology and tradition. Translated from the original Sanskrit, and Illustrated by Notes derived chiefly from other Pur anas. By the late H. H. Wilson, Boden Professor of Sanskrit in the University of Oxford, etc., etc. Edited by Fitzedward Hall, M.A., D.C.L., Oxon. Vols. I. to V. 8vo., pp. cxl. and 2C0 ; 344; 344; 346, cloth. 21. 12s. 6d. [Vol. VI. in the press. Wilson. — Select Specimens of the Theatre oe the Hindus. Trans- lated from the Original Sanskrit. By Horace Hayman Wilson, M.A.,F.R.S. Third corrected edition. 2 vols. 8vo. [Nearly ready. CONTEXTS. Vol. I.— Preface — Treatise on the Dramatic System of the Hindus— Dramas translated from the Original Sanskrit— The Mrichchakati, or the Toy Cart— Yikrama and Urvasi, or the Hero and the Nymph— Uttara Rama Cheritra," or continuation of the History of Rama. Vol. II.— Dramas translated from the Original Sanskrit— Malati and Madhava, or the Stolen Marriage— Mudra Rakshasa,. or the Signet of the Minister— Retnavali, or the Necklace— Appendix, containing short accounts of different Dramas. Wilson. — The Present State of the Cultivation of Oriental Literature. A Lecture delivered at the Meeting of the Royal Asiatic Society. By the Director, Professor H. H. Wilson. 8vo., pp. 26, sewed. London, 1852. Gd. "Wise. — Commentary on the Hindu System of Medicine. By T. A. Wise, M.D., Bengal Medical Service. 8vo., pp. xx. and 432, cloth. 7s. 6d. "Wylie. — ]S"otes on Chinese Literature ; with introductory Remarks on the Progressive Advancement of the Art ; and a list of translations from the Chinese, into various European Languages. By A. Wylie, Agent of the British and Foreign Bible Society in China. 4to. pp. 296, cloth. Price, 1/. 10s. Yates. — A Bengali Grammar. By the late Rev. W. Yates, D.D. Reprinted, with improvements, from his Introduction to the Bengali Language, Editedby I.Wenger. Fcap. 8vo., pp.iv. and 150, bds. Calcutta, 1864. 3s.6d. STEPHEN* AUSTIN AND SONS PRINTERS, HERTFORD. p % I mmSm mm 9^H BBgk S8 H 5*SH ■ «i #^a TGH bSBHBI^HBsmI 9^n Sen!