446 Journal of the Asiatic Society of Bengal. [December, 1914. 



or ew: bacho = bacheft (P. becheha), koldo = koldew , etc. As a 

 rule the final syllable ab is also eu or ew: khew = ¥. Khwab, 

 ishtew = P. shitabeh, just as in Qaini. In the last a is pro- 

 nounced , though very seldom , as I : e.g. — dishte = dashte. E and 

 i are often deeper than in Qaini. U is often sounded in the 

 place of e or i and vice versa: dlz = ~P. dur, zid = ~P. zud: sur = 

 P. sir, etc. This rule is one of the most common to all the 

 dialects of Iranian languages, observable as well in Pashtoo as 

 in Baluchi and Kurdish. 



There is a short o, which can be called a peculiar Gypsy 

 sound having no corresponding one in local Persian idiom. 

 I believe in many cases this o = the suffix of substantives, cor- 

 responding to that of English Gypsy o, as in senuto, etc. But 

 very often on the end of a word it is pronounced longer, and 

 then is hardly different from o = a. Only in the middle of 

 words it can be observed properly, as in jodo, goro, gelor, etc. 



Changes of the consonants in real Gypsy words are almost 

 untraceable, owing to the deplorably scanty remnants of their 

 vocabulary. Some changes which appear there can only be 

 exceptions. Thus : z = k— P. bazi = G . boki ; s = k — P. busiden 

 = G. bakiden. t = sh,V. keshtden,G. katoiden, etc. In Smart 

 and Crofton's ''Dialect of the English Gypsies" (Lond. 1875) 

 there are many phonetic changes given, but many of them, as 

 far as I know, are not observed in Qaini Gypsy. Such changes 

 as k— f ; kh— f, b— d, etc. I never met. The change of the 

 labials into gutturals is very common to the Iranian languages 

 and in some of Persian dialects, e.g. that of Biabunak, it is a 

 rule: gis = Vera, bist ; gecho = P. becheha etc. 



Some changes in Qaini Gypsy are very strange : dehevnden 

 P. dewiden; menon = P. nan; kutagon = P. kuja, etc. 



Syn harmonic tendencies of Qaini are not common to the 

 ( »yP s y> although there are sometimes the traces of assonance. 



Euphonically are used n and w, sometimes y: suni = P- 

 suyi ; ki ne or kiwe = P . ki est ? palu num. = ~P.pahluy em, etc . But 

 hiatus is also used very freely. 



Accentuation is not so strong as is Qaini. 



The transposition of syllables and of their elements is 

 almost a rule in this Gypsy dialect. 



IV. 

 Morphology. 



Here as well as in the vocabulary I will only note the 

 differences between Gypsy and Persian." 



Nouns.— The gender is not distinguished as in Persian. 

 But there are many traces of it in the most common termina- 

 tions of two chief forms: o (6 or o) and i, corresp. to Masculine 

 and Feminine suffixes in Engl. Gypsy. 



