^{page 178 note 1} I take this opportunity of recording my entire disbelief in the Hon. G. Campbell and Mr. Hyde Clarke's fantastic idea that the term Khás as applied to themselves by this Hindu race has anything whatever to do with Kasyapa, Kashmir, the Caucasus, or any primitive Aryan word at all. It is a pure Arabic word which, ever since the days of the first Muhammedan land-settlement of India, has been the regular technical word for “special, principal, select.” How many hundreds of villages are there in India which have this word affixed to them, to distinguish them from places of similar name, but less importance? The Nepalese justly arrogate to themselves, by the use of this word, the first rank among the many tribes of the country. They are the “royal race,” par-excellence.

^{page 180 note 1} Both my interpreters were Thapas. I only know of the other two clans from Mr. Hodgson's note, above quoted.

^{page 183 note 1} The prefix mi found with all words referring to parts of the human body is the Tibetan word for ‘man.’ Hodgson is in error in treating it as a radical.

^{page 185 note 1} According to Jaeschke (Tib. Gr. p. 28, § 39), this word is only used in “respectful and elegant speech,” except in Eastern Tibet (and especially the province of Ü, where Lhasa is situated), where it sounds lā-pa.

^{page 186 note 1} The Magars are well aware of the fact that their language deviates much from the “Bhot,” (as Tibetan is called in the hills), and even go so far as to use this fact as an argument in favour of their alleged Rajput descent. They tell me that their immediate neighbours, the Gurungs, speak a language far more like “Bhot” than they do.

^{page 186 note 2} These numbers refer to those in sec. 2, § 1.

^{page 186 note 3} But still ke in Tirhut and Purneah Hindi.

^{page 188 note 1} I give the Sanskrit characters as a key to the pronunciation, as far as they go.

^{page 189 note 1} Foucaux, Grammaire Tibétaine, p. 3.

^{page 190 note 1} Foucaux, Gram. Tibet. p.3, 1. 12, and p. 8, para. c.

^{page 191 note 1} Like the Sankrit Tatpurushah.

^{page 191 note 1} This affix mán is probably connected with the Tibetan mo affixed to female nouns.

^{page 195 note 1} See on this subject Caldwell's Dravidian Grammar, p. 338 et seqq.

^{page 196 note 1} See these words in the Vocabulary.

^{page 196 note 2} This form is not used with words beginning with the vowel á.

^{page 197 note 1} The affix ang is sometimes added without the l to express present time, as pasang, ‘it happens,’ for pas-lang.

^{page 203 note 1} Or from Tibetan rang, ‘self,’ ‘very,’ ‘indeed.’

^{page 205 note 1} Hodgson erroneously miryárus, which is used sensu obscæno.

^{page 206 note 1} Nákun, respectful form of pronoun of 2nd person, takes the respectful form of the verb, ráni instead of ráná, which is used in commanding an inferior as in sentence 1.

^{page 206 note 2} Lit. ‘Of you the name what is?’ The verb is generally placed last, however long the sentence, and personal pronouns first.

^{page 206 note 3} Kulale, contracted from kulag ále.

^{page 206 note 4} Lesá would have been more correct, but ále was probably used in consequence of its occurring in the question. Observe that Tannung has no affix. This is probably omitted, because Tannung-ang would sound badly, or in imitation of the Hindi idiom, e.g., ghar jáo, where ko is omitted.

^{page 206 note 5} Le for ále. The final k in kurik is pronounced like g, and the s of los almost like sh.

^{page 206 note 6} Literally, ‘Of five days a road (it) is.’ The affix of the genitive is appended only to the latter of the two words ‘five days.’

^{page 206 note 7} For Ngá-ke. Case affixes are often omitted in long sentences. Nungke le, lit. ‘to go it is,’ i.e., ‘I must go.’

^{page 206 note 8} Imang, lit. ‘In my house,’ we should say, ‘into.’

^{page 206 note 9} Jyáke, pronounced jaïke. ‘to eat,’ lit. ‘give me to eat.’ Here the affix ke, ‘to,’ common to nouns and verbs explains itself. Compare Hindi k háne ko do.

^{page 206 note 10} Che le, ‘I cut,’ present, used with a future sense. There is much vagueness on the subject of time in the Magar mind.

^{page 207 note 1} Nákun is indeclinable. The sentence is literally ‘To you how many years have been’.

^{page 207 note 2} There is no reflexive pronoun, lit. ‘He of-him in-house is.’

^{page 207 note 3} Lit. ‘that cannot (be).”

^{page 207 note 4} Tsánre doubled for emphasis.

^{page 207 note 5} Personal pronoun put last for emphasis.

^{page 207 note 6} Dative case of hi, ‘what,’ lit. ‘for what.’

^{page 207 note 7} Lit. ‘with-me to go has it pleased ?’ This construction is used because the question implies will, ‘do you wish to go ?’

^{page 207 note 8} Short for jaïle. Vide sec. 5, § 3, 4.

^{page 207 note 9} For langhang-ang, the affix is omitted, perhaps to avoid repetition of sounds; but more probably because a Magar never inserts affixes, if he can make his meaning clear without them.

^{page 207 note 10} Rahang, a form of the present. Vide Sect. 5, § 3, Remark 3. The affix ang is common to nouns and verbs, and means ‘in,’ rahang therefore is ‘comingin.’ Observe that the verb is almost uniformly placed at the end of the sentence.

^{page 207 note 11} Di káske, ‘to feed with water.’ In the next sentence, dána kásá, ‘to feed with grain.’ The sentence shows how much use is made of Urdu words in Magar.

^{page 207 note 12} After verbs of motion the affix in ang is often used.

^{page 207 note 13} Lit. ‘Village.’

^{page 207 note 14} See sec. 5, remark 1.

^{page 207 note 15} Lit. ‘Here to sit pleasant not is.’

^{page 207 note 16} ‘To eat has been prepared or not prepared;’ bhe contraction for bhyá, past tense of bhyáke, ‘to prepare.’ This contraction is merely euphonic.

^{page 208 note 1} Kose for kos or kus, ‘which ?’ an emphatic e or i is constantly added to pronouns. It is a thegá or expletive, and is not essential to the meaning.

^{page 208 note 2} A'le, not le or lesá, as it is a question. See chap. 1, sec. 8, § 3.

^{page 208 note 3} Tsanke differs from leke just as ‘become’ differs from ‘be.’

^{page 208 note 4} Parchü, i.e. ‘from the other side’ (to wit, ‘of the valley’). All the Magar country being mountainous, a man from another village naturally comes across a valley.

^{page 208 note 5} Sign of the genitive omitted. See sec. 2, § 1, remark 2.

^{page 208 note 6} Kahek, ‘one day,’ is apparently indeclinable. In this sentence it is in the genitive case. It is contracted or corrupted from kat, ‘one,’ and ek, which probably means ‘day.’. I cannot find in Tibetan any word which seems connectable with this according to any canons of relation of sounds. Nor does any word exist in the hill languages which is like this. The Tibetan zhag is the nearest. I incline to the idea that the word ek is an expletive or meaningless addition which has acquired by habit the sense of ‘day.’ If any one should object to the word ‘meaningless,’ I would remind him that in the Turanian family there exist many generic affixes which are really of no use to the general meaning of the sentence but merely serve to mark the class of objects to which the word belongs. This is one of them. They say nisek, ‘two days,’ etc., though the word for ‘day’ is din in Magar.

^{page 208 note 7} Past tense of chike or cheke, ‘to cut, tear.’

^{page 208 note 8} Chini here evidently means ‘from.’ It is, however, the only instance I have found of its use. It has a suspicious resemblance to the Hindi chhín lena, ‘to take by force.’ It would he too much to assert, however, that this was the real origin of it.

^{page 208 note 9} Chü is here quite an expletive, and used similarly to ngoï chü. Perhaps something in the original meaning of the word may account for its having this quasi adjectival form.

^{page 208 note 10} Affix ang omitted.

^{page 208 note 11} Lash le, literally Hindi lagtá hai, which is quite untranslatable.

^{page 208 note 12} Lit. ‘Having occurred it is completed,’ —Ho chuká, Hindi idiom again.

^{page 209 note 1} Lit. ‘Me he will beat,’ having-said I fied. See sec. 5 § 3, remark 9.

^{page 209 note 2} Bhogdishü is for dishchü.

^{page 209 note 3} ‘By you seen, how many times was (it).’ Cf. p. 208, note 3.

^{page 209 note 4} The words bariá tsánd are a sort of optative form, ‘May it be propitious.’ This word bariá appears only to be used in this phrase: it probably is a corruption of the Hindi bhalá.

^{page 209 note 5} See sec. 5, § 3, remark 9, example 3.

^{page 209 note 6} Future of kahke, ‘to place.’

^{page 209 note 7} Lit. ‘You do not go-then.’ See sec. 7, § 2.

^{page 209 note 8} See sec. 5, § 3, remark 10.

^{page 209 note 9} Expletive.

^{page 209 note 10} Literally, ‘From here (one) sees not.’ Dangá is also used.

^{page 210 note 1} Lit. ‘Of me the birth there indeed was; of me the father living was.’ Na is an expletive: see chap. 1, sec. 8, § 2. Ále, though really a present, is here used for a past, there being no past tense of leke, ‘to be;’ in the same way lesá at the end of the sentence, which, probably from the influence of the preceding nasals, is pronounced nesá. Ko is a generic appellative added to nouns signifying relation. This is a strictly Turanian characteristic, and is found in Tibetan, Chinese, and the Hill languages in a far more elaborately developed form. Nguná is for ngu with the expletive na; the phrase nguná nesá is a longer and fuller form for ngu le.

^{page 210 note 2} See chap 1, sec. 2, § 4 (c).

^{page 210 note 3} Bhogdisá, past tense of bhogdishke. Verbs ending in ishke or dishke have a frequentative sense; also generally convey the idea of motion.

^{page 210 note 4} Lit. ‘You having-fled go, again having-caught he brings.’ The idea of ‘if’ is not expressed, being involved in the general idea of the sentence. See chap. 1, sec. 7, § 2. There is no passive, consequently the phrase, ‘You will be caught,’ has to be expressed by ‘(some-one) catches you.’ See chap. 1, sec. 5, § 3, remark 10.

^{page 210 note 5} Khosá, lit. ‘he stole.’ Passive again expressed by active.

^{page 210 note 6} Lit. ‘That event happening, how many days were.’ See sec. 2, § 3, rem. 9.

^{page 210 note 7} ‘Two-days were.’

^{page 210 note 8} s. thega. See sec. 8, § 4.

^{page 210 note 9} Khata is lit. ‘with.’ This sentence seems modelled on Hindi us se mángá, where se= ‘with’ or ‘from.’

^{page 211 note 1} Lit. ‘to do it is not,’ this form is used instead of the regular future, to convey a sense of absolute certainty— ‘to do it is not (possible or to be thought of, etc.)’

^{page 211 note 2} ‘Of that stolen property.’ Khuse-o is, I fancy, a genitive of the present tense, thus Khuske = ‘to steal;’ khus-le = ‘thou stealest;’ khus[l]e=id. gen. Khuse-o. Vide sec. 2, § 3, for remarks on the general declinability of every root in Magar, indifferently as a noun or verb, or both, at once as in this case.

^{page 211 note 3} ‘Twice to put down it is (necessary).’

^{page 211 note 4} Dinles, (from dinke, ‘to find’,) ‘you will find,’ i.e., ‘will have to,’ construction im tated from Hindi use páná, as baiṭhne páoge, ‘you will have to sit.’

^{page 211 note 5} Lit. ‘What regret ?’ toba is the Arabic word so common in Urdu.

^{page 211 note 6} Patta is put first for emphasis. ‘All thieves,’ etc. Ordinarily it would come after khus.

^{page 211 note 7} Pardile, ‘it is proper.’

^{page 211 note 8} To commit a crime: riskheke.

^{page 219 note 1} Shing means tree, and is added to the distinctive names of all sorts of trees.

^{page 219 note 2} About this letter my store of words begins to get scanty, as my Magars left me when I had got as far as F in a regular alphabetical enquiry for words. Henceforward I can only give the words I picked up in the course of my firstseries of lessons, which were chiefly directed to the grammar.