376 Journal of the Asiatic Society of Bengal. [N.S., XVII, 
Still many more appear in more modern periods: mantishakh, 
garza, gulbang, palahang, 
Some years ago, craiee mesidaictatiy. I came across an inter- 
sting document, particularly valuable in its bearing on these 
didn It was a fragment of a secret code of a religious 
community which fouciihed at least 400 years ago, probably 
much earlier. Although only about one hundred words could 
be extracted from the fragment for examination, many idioms 
are found there which can be at once recognised as favourite 
expressions of the present-day gypsies in Persia. 
THe MANUSCRIPT 
Several years ago, a Manuscript was offered to me for 
purchase at Qarshi, in Bukhara. It was a volume of some two 
hundred folios, in quario, containing an interesting collection of 
philosophic and Sufic treatises, mostly rare, apparently selected 
by a man of learning and of discriminating taste in these matters.' 
Unfortunately I did not succeed in my endeavours to purchase 
this interesting volume. The owner, by some obscure process of 
reasoning had come to the conclusion that the MS. was nothing 
less than an autograph of Avicenna himself, several of whose 
treatises were included in it. For this reason he demanded 
a really fantastic price, and it was waste of time to argue 
with him regarding the falsity of his pretensions. All I could 
obtain was his consent to take the precious book on loan for 
one evening for a closer examination. The copy possessed all 
the typical features of manuscripts dating from the end of the 
1 It epee in mpage to short Lain iio (1) A short treatise by 7 
Farabi (d. A.H./950 oe (2) Several small works of Abia ‘Ali Sin 
(Avicenna) ‘a. =< A.H. ATA D.). The treatises in duesiot are ‘hiciee 
ned by Brockelmann, Gesch. d. Ar. Lit., I, p. 455, Nos. 24, 37 and 
extremist shade, i.e.: (5) A letter by ‘Aynu’l-Qudat § Hamadaet (oroste 
533 A.H./1139 A.D.) ; oe A Treatise by Shihabu’d-Din Mouse: SurAs 
Magqtil (executed 587 A.H./1191 A.D.), see Brock., ibid., 438, called J al 
Alwahu’l-‘Imadiyya (rather rare); (7) The well-known eps by 
c 
ges 
of Mirsadu’l-‘ibad, d. a 650 A.H. fing A.D); (10) A version of 
Jawidan-khirad, Ghend > pions the works of al-Jahiz, on a supposed 
Pahlawi original attributed to ae (This version is quite different from 
that of on Kashi, which is in Persian, and not in Arabic, as is the 
neti in question). There rooties also many other interesting extracts and 
no 

Cn 


