In presenting to the public the Lepcha-English dictionary of the late General G. B. Mainwaring, it is incumbent on me to explain the state of the manuscripts he left behind and the part which the editor has performed in making them ready for the press. Owing to a many years' acquaintance with the language of a race, that was dying out, the General was imbued with predilection for his study, which plainly appears in his grammar'), the only work he could publish. In the preface to this grammar (pg. XXI) he promised that a Dictionary should follow "should his health and circumstances permit". The lamented General was prevented from exhausting his material by his decease.

The materials of the Lepcha-English dictionary are laid down in two very voluminous manuscripts, which represent the first draught or archetype of the work. They are written in large octavo in 703 pages of bluish and yellowish paper. They contain a huge collection of Lepcha-glosses which were augmented by revising the first entry again and again. The single words were written in the so-called Lepcha character but according to the European alphabet. But I must say to my great regret that no notice was at hand concerning the method and the sources from which the collections were derived, it was at first impossible to ascertain where the lost clue was to be taken up again. But in sifting the materials it could be stated, that the author had commenced his work by collecting oral and manuscriptinformation from the natives.

This plainly appears in the following passages. Sub v. bro (supply the question ,,what is in English pum-bróm-la?" Answer: ma-zů a-čùm a-tyak a-tím pům-bróm-là

1) A Grammar of the Róng (Lepcha) language, as it exists in the Dorjiling and Sikim hills. By Colonel G. B. M. Bengal Staff-Corps, Calc. Baptist Mission Press 1876.

yum-ba ("body short, head large is defined by p."). For further examples see under pók-yak; trám; rip, zo; lya, (under certain words the editor found great compilations of synonyms, the English definition he added himself). Certain lines of the above mentioned manuscript-elucidations must have been so indistinctly writton that wrong entries could not be avoided by Mainwaring himself; e. g. o-tson which he defined by "a union" should be "onion T. btson. The word jüm-tyát jüm-fyet (a spec. of rice) seems to be very dubious, should it be written nŭm-tyát num fyet? When M. Mainwaring jotted down his first notes in the above mentioned great book he was often misled himself by a peculiarity of the so-called L.-writing, by which the final consonants are not written in the same line with the initials but above them, in reduced or abbreviated forms. I found s. v. (a-)hrát: a-hrắt gam nól instead of a-hrát gal nón or tük-nóm where tük-nól should be written etc.

Even the native writers themselves are not masters of their pens in this point. The scientific names of birds were defined by the author according to Jerdon's well-known work'), as could be plainly proved from a quotation 8. v. să-hret čuk fo where "Je. 2, 277" was added. At length I could ascertain, that the author had augmented his collected materials by analyzing Biblical books and the Lepcha abbreviated legends concerning the famous founder of Lamaism, known in history under the name of Padmasambhava (Tib. U-rgyan-pa, Pad-ma 'byun-gnas, Lepcha Ta-se t'uk-bo t'ín). From these popular legends called Ta-še sun in Lepcha he had derived many quotations, the greatest part of which were transcribed without any translation or at least without sufficient philological definition.

The first half of the huge manuscript he had compiled in the above-mentioned manner was tolerably written, but the latter part from letter m -z was an almost illegible scrawl, the worst being the letters 8, s, t, ta, là, 2. It would be in vain to seek even the smallest blank space in these pages, every line of which was so underscored with additional matter, or corrections as to render the first entry scarcely visible.

The most illegible part was the English definition which was added in pale ink or in pencil. It must be observed that in this latter portion the author was. more and more influenced by his peculiar method of comparative philology, so that the definitions of the roots in many cases have had to be abandoned by the editor.")

The next state of the future dictionary must be called a huge volume in Imp. fol, consisting of 501 leaves, which seems to have been intended for immediate printing. The Lepcha-words were given here in Lepcha-characters, transcribed into Roman and rendered into English, they were arranged in the order of the Lepcha-alphabet as set down in Mainwaring's Grammar pg. (2). A considerable number of words were noted as borrowed from Tibetan by means of Csoma Korosi's Essay towards a dictionary of the T. language.

1) The birds of India I-III, Calcutta 1862 ff. 2) See Grammar, XI Note, XXII Note.

But many words of the first books, many quotations especially from Tă-de-sun were left out; the editor has incorporated them again into the work, whenever he had anything like certainty, that these quotations had been correctly delineated or could be correctly defined.

But in this copy also I sought in vain for any attempt at arranging the matter according to etymological principles. Nearly all the various spellings represented by the native manuscripts, nearly all the differently written prefixes and their combinations had been incorporated in extenso, so that the dictionary even if at any time finished in this way would have increased to an enormous extent. So the author got tired of his work. The portion he finished under such circumstances included all from k to br with the exception of the consonants combined with 7 (kl, gl, pl, fl). For the remaining part of the work corresponding to fol. 270—437 of the present printed dictionary the editor had to have recourse to the sources of the above-mentioned first draught, which was in this portion in a hopeless condition.

The following sources were at hand:

1. Reconstruction of the illegible words from quotations in the former plainly written portion of the work;

2. The etymological method, by which the various derivations (disjecta membra) could be collected under one root;

3. The books which Mainwaring must have excerpted.

This proceeding necessitated digressions and offered many difficulties, but it was the only way to make the work clear. After repeated attempts to make immediate use of Mainwaring's so-called fair copy, I set to work to copy the whole matter of both the transscripts word for word on to a separate slip of paper as far as I could make out the author's owing to his indistinct writing and still vaguer definition. These slips were arranged according to etymology and in the order of the Lepcha-alphabet after a corresponding English-Lepcha-dictionary had been worked out. Then I analyzed the Lepcha-books which were courteously lent to me by British Government.

The result was:

1. I could settle the orthography of the language following the principles of the printed Biblical versions.

2. I could give a more correct definition of the Buddhistical terms quoted by M. from Ta-se-sun. I have analyzed two separate renderings of the book together with a great portion of the block-printing edition of Pad-ma-t'an-yig in Tibetan (fol. 1-146). This original book helped me to the Tibetan spelling of the nomina propria and other terms and the T. forms could then be translated into the corresponding Sanskrit').

1) e. g. rum-lyan in Tă-še-suů: T, hla'i gnas or bde-ba čan Skt, devaloka or sukhavati.

3. I could settle more correct definitions of the L. words (roots), than Mainwaring's peculiar method had assigned them. In most parts I found the oldest notes in his first entry to be correct.

The dictionary is sufficient for reading the above-named books which were Mainwaring's own sources; all the nomina propria of beasts and plants, all Tungbor expressions are entirely M.'s work.

It must be remembered, that in certain cases a difference arose between the assumed authority of orthography (the Biblical translations) and the etymology. Then I abandoned the former inserting quotations from the respective various spellings. This was done especially in the case of the i-vowel combined with consonant. That vowel seems to have a peculiar inclination to sound like yi after consonant. The question where to write ya, ye, e so well-known to all students of modern Indian manuscripts was decided according to etymology; I wrote nyan causative of nan, nyăn from T. nyan(-ba); but nyen milk: all three roots are written nyån (or even nyan) and nyen in Tă-še-sun-Mscpts. promiscuously, etc. etc.

Words marked M (without number) are not certified by written authority, but I would not eliminate them after finding that many words in M.'s first notes left out by himself in the second copy proved correct. So I think that future research and fresh material may afford the needed confirmation; e. g. s v. suk-grup s. chest M. Question: must "chest" be understood in s. of "box" or of "thorax"? sũñ-kỏ 8. a hoe M. Qu: can it be a separate spec. of "hoe" or should it be corrected into n-kẩn? In the first pages I was misled by some definitions given by M. e. g. jom (to be overclouded) seems to be incorrect. All the fantastic etymological matter added by the author is now removed e. g. s. v. klyen (cylindrical) decreasing, to diminish M. cfr. Germ. "klein" etc. or kup (little, a child) efr. English „cub" and similar "result of power of letters".

With respect to the Tibetan words, which are introduced into written Lepcha I first thought of making a separate index of the various spellings with their L. synonyms if even such exist. But when I endeavoured to read the printed and written books, I found it impossible to do so. The part which these borrowed words play in literature is just as indispensable to it as Pâli or even Sanskrit to Burmese. It cannot be denied, that the dictionary seems overloaded with number of similarly written or sounding roots (see under ke, se, ši, bóù, in the latter case the genuine L. word signifies exactly the opposite meaning of the Tibetan one), but the scholar who wishes to be able to understand a manuscript or even to find without difficulty the real meaning of certain terms or phrases in Exodus or the Gospel of John, will acknowledge the usefulness of my method by his own experience. Whatever seemed to be superfluous I have removed, e. g. a long list of the various Tibetan names written in L. characters of Buddhistic gods and Bodhisatva's (Amitabha, Manjuçri, Avalokiteçvara) etc. Further the method I followed was necessary under the following considerations:

Many Tibetan words are used by Lepcha's in a different, derived sense e. g.

Preface.

8. v. *pȧk-čo T. bag-čags; the T. roots are subject to Lepcha-formative rules; e. g. kyor from kor; năm-ảyin from ảyin etc.; the L. and Tibetan languages though not immediately cognate are so closely connected in many roots that it seems impossible to define exactly, what is borrowed and what is cognate in origin e. g. s. v. yả and ši 1, nók 2 etc., fiù (blue) etc. Certain incorrect words were removed from this portion of the work e. g. hum s. offering derived from húm pat; many words identified with T. were corrected e. g. *hra-mik acc. M. fr. T. Kra-ma and mik,* should be T. dra-mig etc., den-zün the scales of justice fr. T. bde and rdsun “truth and falsehood". But I am sure that this difficult question has not been settled by me in all cases. For the Buddhistical terms I added fresh matter as above mentioned e. g. 8. v. rum, mui, lyan, tsur, tu-še, sun-hlyo, myel, gyan etc.

The editor is responsible for the arrangement of the longer articles in the work e.g ku, lút, minh, lớn, li, nói, mat etc., and has added numerous quotations authenticating a the derivation of the definition. Compare the original article mat as it is printed in M.'s Outline Grammar 128-130 and the articles mat 1 and 2 in the dictionary. It is clear, that two different roots are confounded apud M. if no more. In the same manner no 1-4 formed one article, li 1 10 one article, ki (1–5 and -ki) si one single articlo apud M. For la 1 the editor alone is responsible.

He has inserted all the postpositions mentioned in the grammar and all the prefixes o.g. -nùn; -ún; -a; -să; -là; -ka; a-; ta- with the idea of rendering texts easier and enabling the reader to find in the dictionary grammatical matter if needed. I hope also that in certain cases the rules of L-derivation may be better defined than is the case in the grammar. The etymological method remained also the standard for shortness. All derived and compound words were as already mentioned incorporated under their respective roots, when at the same time the prefixes were inserted with quotations referring to the roots for the sake of convenience.

The following considerations may not be out of place.

I. Reduplication may be called a principal part of Lepcha-grammar. Roots are reduplicated:

9, i,

A. by repeating the initial consonant of the root with inherent å; but ♫, h are represented in reduplication by k; f d by t, p fb by p, ny by n, č, č, j ts, x tz by 8, v by f. Seo under kă-, tā-, på- etc.

B. by repeating the whole root with u vowel, the consonants of reduplication are treated as noted under A.

These forms are to be sought in the dictionary as follows: fă-vi-lă see under root vi; fu- refers to the root; kūñ-kón-lă see under kón; kun- refers to the root kóй; pă-plyu-lă see under plyu etc.

C. if other vowels are used to form the reduplicated or geminated root or different initial consonant is introduced into the second part, the word must be sought under the form which seemed to be the root e. g. yep-på yap-på see under yap; yep referring to the root yap; šól-là mól-iă see under šól; mól-là referring to šól. II. There are in Lepcha certain prefixes of undefined etymology which however seem to correspond to the so-called mute consonants in Tibetan. By prefixing them

1