^{page 337 note 1} There is a form which seems to be different from the literary, viz., probably for , (38 v.) .

^{page 337 note 2} Although this form seems to be quite Pehlevi-like (cf. C. Salemann, op. cit., 303), it is very common in Khorasan (äftídum).

^{page 338 note 1} In the peasant language in many districts of Khorasan this form seems to be practically never used, and the direct verbal construction is invariably preferred. Instead of the people will say : ú ki raf (with this ‘ post-positional’ ki), etc. It seems that even good colloquial avoids it.

^{page 339 note 1} As we will see later, however, there are instances (common in the old language) in which the full form of infinitive is used in such cases.

^{page 339 note 2} There is no doubt that in the colloquial this ‘ apocopated’ infinitive is the only one used. Many special suffixes are mentioned as being in use in the dialects (as in the Central group, cf. W. Geiger, op. cit., 399–400, and D. Lorimer, 466), but it is necessary to be extremely cautious in treating them as such. The idea of an infinitive is too abstract for the practical and unsophisticated peasants, and therefore they easily add the personal pronominal suffixes, etc., to make it more concrete. The suffix -mun, perhaps, is — of the 1st pers. plur., and kertmun means ‘ our doing’, etc., but in the case of -un or -iun it is necessary to note that they (apparently originally the suffixes of the plural) occasionally form a sort of abstract name: arusúin, i.e. ‘wedding’, churåghún ‘evening’ (i.e. period from the beginning of darkness and till the peasants go to bed, 6–7 p.m. to 8–9 p.m., when lamps are used—åkhiri churåghú umad ‘ he came late in the evening’).

^{page 339 note 3} The examples of this use of the present stem are very numerous in the colloquial: furush, bedeh (hich bedéyi nedårúm), gir (de pul begir uma), etc.

^{page 340 note 1} Nafaḥāt, 116, .

^{page 340 note 2} For its meaning see the list of rare words, E, No. 4.

^{page 340 note 3} May it be possible that these forms are similar to the verbal nouns in some Caspian dialects, also with a suffix -a ? (cf. W. Geiger, op. cit., 369, in the dialect of Tālish).

^{page 341 note 1} There are many similar participles, as guyå, dånå, binå, etc., in the colloquial, but they seem to be treated, as often in the literary language, as ordinary adjectives.

^{page 341 note 2} This form of the present participle can be regarded as non-existent in the more simple kind of colloquial.

^{page 341 note 3} The present participle generally is as rarely used in the simple colloquial as the infinitive. This form (in -anda) is also borrowed from the literary language and treated as an adjective. Cf. W. Geiger, op. cit., 399, and D. Lorimer, op. cit., 465.

^{page 342 note 1} Cf. P. Horn, op. cit., p. 153. It is very interesting that while the original base of this form is not popular at all in the colloquial, this future participle is very common. It has various shades of meaning, but a sense of the future seems to be essentially inherent in it.

^{page 342 note 2} About it see later on, section 6.

^{page 342 note 3} Cf. Salemann, op. cit., p. 314. Examples of a similar use of the preterite with the suffix -i are numerous in the old language, and P. Horn, op. cit., 151–2, quotes several of them, as (Shāhnāma, Vullers, 1672) , etc.

^{page 342 note 4} Naf., 142, .

^{page 343 note 1} Naf., 227, gives .

^{page 344 note 1} In the earlier language the expression is common, but I cannot remember many examples like the one given above in the text. (The contemporary colloquial equivalent of this expression is balki, rarer kåshki).

^{page 344 note 2} 2 It is somewhat strange to find that the authors of the more recent grammars invariably call this form conditionalis or conjunctive, etc. It may be taken for certain that even if it recalls the ancient conditionalis, etc., it is in the Modern Persian only optative or rather ‘ precative’.

^{page 344 note 3} Similar formations are not uncommon to the vulgar form of the present colloquial, and can be regarded as rather impersonal.

^{page 346 note 1} It is difficult to say how much the modernization of the text by the later scribes is to be blamed for the disappearance of hamī. A book, which was probably written at exactly the same time as the Ṭabaqāt, in the district of Jām, a dependency of Herat, by the celebrated saint Aḥmad-i-Jām, shows also a tendency to avoid this archaic form. It is , often referred to above. At the same time, the books composed before and even after that period, but which reached us in older copies, prefer to write hamī very often, as is the case with the Tadhkira of ‘Aṭṭār, from Nishapur, Asrāru’t-tawḥīd by Ibn-Munawwar, who apparently wrote it in Abīward or Tūs, the translation of Ṭabari' Tafsīr (see above, note on p. 23), made probably in Bukhara, etc. Therefore, it is impossible to come to definite conclusions, and it would be more cautious to believe that the copyists have done much to damage the texts in this connexion.

^{page 346 note 2} The book of Aḥmad-i-Jām, mentioned in the preceding note, which tries to use the highest standard of the literary language, very often, nevertheless, cannot avoid a similar order of words. This may prove that it was very common in the colloquial of that time, although it might be regarded as ‘ rustic’.

^{page 347 note 1} Cf. P. Horn, op. cit., p. 154.

^{page 347 note 2} This is not uncommon in the colloquial and the dialects: tå ze pizkhanda bare bike wagaw bandåkhtäi, ‘ when, smiling with the upper lip, thou hast thrown the cheek and mole into a dimple’ (from Nayyīr, in Sabzawari).

^{page 348 note 1} In Aḥmad-i-Jām's book it is used as freely as here. (Its presence in the Nafaḥāt may be explained as only the result of Jami's carelessness as to the style of his book.)

^{page 348 note 2} Personally I heard this suffix on one or two occasions in Southern Khorasan, but my materials, collected there, are not available to me at present.

^{page 349 note 1} It seems that these subdivisions of the ‘ classes’ of the suffix -i, as many other constructions of traditional grammar, are based on logical speculations, and have nothing to do with the real synthetic study of the language. P. Horn's attempt to revise these elaborate ‘ hair-splittings’ in a more scientific way cannot be called quite successful. It is difficult to agree with him (op. cit., 151–2) that this suffix has much optative sense in it, and still less with his opinion that it appears as a prefix (this is quite a different matter, and the prefixes e-, i-, etc., are connected with the dialectical particles ha, he, etc., similar in their use to bi).

^{page 349 note 2} Not found in the Nafaḥāt.

^{page 350 note 1} Not in the Nafaḥāt.

^{page 350 note 2} Naf., 121, . In the same way Jami changes all these uncommon forms if he finds it necessary to reproduce the sentences in which they occur. He is particularly fond of this suffix -i, and uses it even in those cases in which it is omitted in the Ṭabaqāt.

^{page 351 note 1} Naf., 183.

^{page 352 note 1} Naf., 183.

^{page 352 note 2} Naf., 312.

^{page 352 note 3} So in Nafaḥāt, 127.

^{page 353 note 1} This suffix, which quite probably, as P. Horn (op. cit., 152) thinks, is of the same nature as -ā of the vocative case, does not exist in the present colloquial and dialects (except in the expressions borrowed from the literary language, such as khudåyå shukr, etc.). Therefore, the example above, which undoubtedly belongs to the colloquial (or even a dialect) spoken at that time, is interesting evidence that this form was really ‘ living’ and used in speech.

^{page 353 note 2} Except in the case where they receive the prefix bi (if it is really so, cf. later on, subsection (b) of the present section).

^{page 353 note 3} This form is very strange, because it recalls those used in the dialects of Fars (see O. Mann, op. cit., 26, 37–8), while in Khorasan, at present, no traces of it are found (cf. W. Geiger, op. cit., 398, and D. Lorimer, op. cit., 461–2).

^{page 353 note 4} On the margin is added , which is probably the conjecture of one of the scribes.

^{page 353 note 5} The addition of the cohesive suffix -i to is quite common in the old language. It is particularly frequent in the old Persian Tafsīr (mentioned on pp. 20–1) in the library of the Asiatic Society of Bengal, Aa7. The same is the case with the language of the mentioned above, where even a plural form appears with the same suffix as on f. 33, etc. The instance of (3rd pers. sing., with the suffix -īd) mentioned here is the only one I came across in the Ṭabaqāt.

^{page 354 note 1} These forms seem not to be frequent.

^{page 354 note 2} Not rarely used in even the more ‘ rustic ’ forms of the colloquial at present (as well as the verb räftän) instead of shudan.

^{page 354 note 3} It is very difficult to suppose that the original ā was shortened and could then disappear. It becomes i or e in some Kurdish dialects, but the whole character of the language of the Ṭabaqāt does not permit to suggest such accidental possibilities of analogy.

^{page 355 note 1} In that language the 3rd pers. sing. sṭi is also used. It would be extremely interesting if the students of Soghdian could discover a definite difference in the use of both.

^{page 356 note 1} This is an extremely strange use of — which could be taken here for -īd as the cohesive suffix, if the word to which it was added were a verb. Therefore we must take it either for the form of the 2nd pers. or the 3rd pers. sing.

^{page 356 note 2} Naf., 141, .

^{page 356 note 3} Naf., 142, .

^{page 357 note 1} Similar forms are very common all over Persia, not only in the dialects (cf. W. Geiger, op. cit., 366, in the Caspian group, 398, in the Central dialects, about which also D. Lorimer, op. cit., 463, 464, and O. Mann, op. cit., 37–8, for the dialects of Fars), but in the colloquial of the cities as well. Except in the Caspian dialects, however, these forms usually convey the idea of the past tense.

^{page 357 note 2} At present it is one of the most prominent phonetic phenomena of Khorasani Persian, and the usual is frequently bi here.

^{page 357 note 3} If bi may be prefixed to the past tense of the substantive verb, why cannot it be added to the present as it is usually added to other verbs ?

^{page 357 note 4} Naf., 76,

^{page 358 note 1} Naf., 243,

^{page 358 note 2} Ibid., , etc. In this way Jami usually renders these peculiar forms.

^{page 359 note 1} Cf. Browne, E., “An old Persian Commentary,” JRAS, 1894, 438.Google Scholar

^{page 360 note 1} It seems to have been the usual form of expressions of this type ; cf. E. Browne, op. cit., JRAS, 1894, 440, and R. Nicholson, Tadhkira, vol. ii, introd., 11.

^{page 361 note 1} This seems to be quite common in the old language, and a similar confusion is still observed in Khorasan.

^{page 362 note 1} The final z in and is undoubtedly an archaism because it goes back to ž of the Pehlevi frāž and awāž (see C. Salemann, op. cit., 319). It is remarkable that it appears only before the initial vowels of the pronouns, as d, the old ṭ, and after δ in paδ = for . These forms, , are not common in other books, and therefore it was a kind of indirect evidence in favour of their being regarded as dialectic to find them in the book of Aḥmad-i-Jām, who uses them very freely.

^{page 362 note 2} In the present ‘ rustic’ forms of the Khorasani colloquial bär and wär seem usually to take the place of and .

^{page 363 note 1} This preposition may be regarded as peculiar to Fars, where it takes the place of the Khorasani wär. Its applications are manifold, as sit means (cf. O. Mann, op. cit., 52), si kåzerún instead of , etc. In Khorasan it is not very common and used only to show direction, as in suneshār , etc.

^{page 363 note 2} is used as commonly in Aḥmad-i-Jām's book, but I noticed some cases of as well.

^{page 364 note 1} Cf. above, n. 1 on p. 362.

^{page 366 note 1} Cf. below, in the present list, and also A, 3, p. 20. I cannot remember having ever seen or heard , of which this word may be a diminutive. On the other hand, the more natural reading of this expression would be in contradiction with the context.

^{page 367 note 1} In the rustic colloquial of Southern Kberasan similar formations are still in use, such as kuchulu, from kuchak (so it is spelt locally), kachalu (‘ bald ’, from kal), gululä (from the usual ghullä, ‘ bullet’), etc.

^{page 376 note 1} Apparently the same word is met with on two occasions in Professor Nicholson's edition of the Luma‘ (Gibb Mem. Ser., vol. xxii), the text, pp. 173, 215. It is spelt in the edition as , but it would be interesting to learn if the original MSS. admit the reading . It is for the specialists in the Semitic languages to decide whether this yata and the Italian word ghetto are the same, or go back to a common origin.

^{page 377 note 1} Naf., 36, .

^{page 377 note 2} Probably an instance of , something as or (colloquial, S. Kh.) låt-u-påt, etc.

^{page 378 note 1} Written

^{page 379 note 1} Perhaps would be better ?

^{page 379 note 2} Better ?

^{page 380 note 1} Better .

^{page 380 note 2} In the Nafaḥāt, 324, so far omitted.

^{page 380 note 3} Added .

^{page 380 note 4} N. omitted from the asterisk.

^{page 380 note 5} N. .

^{page 380 note 6} N. different order of words

^{page 381 note 1} N. omitted.

^{page 381 note 2} N. added .

^{page 381 note 3} N. .

^{page 381 note 4} N. (sic).

^{page 381 note 5} N. added .

^{page 381 note 6} N. instead of this .

^{page 381 note 7} N. omitted.

^{page 381 note 8} N. added .

^{page 381 note 9} N. added .

^{page 381 note 10} N. .

^{page 381 note 11} N. added .

^{page 381 note 12} N. .

^{page 381 note 13} N. .

^{page 381 note 14} N. .

^{page 381 note 15} N. omitted.

^{page 382 note 1} N. .

^{page 382 note 2} N. .

^{page 382 note 3} N. .

^{page 382 note 4} N. omitted.

^{page 382 note 5} N. omitted from the asterisk.