516 OKIEWTAL ELEMENTS OF CULTURE IN THE OCCIDENT. 



Sedillot," is ii mere play of fancy and can not 1)6 followed up historic- 

 ally, unless one should introduce as an argument for it the iii of the 

 rock inscri})tion of the tenth century at Tor, near the Gulf Suez, writ- 

 ten in Roman fashion immediately before the Arabic numerals.* On 

 the other hand that the Divani or Siyaqat ciphers, which show some 

 similarities to the common Aralnc numerals, arose from the ahl)revi- 

 ated Arabic numerals, was already correctly recognized Ijy De 8acy.'' 

 Nevertheless the Indian origin of the Arabic luimerals, acknowledged 

 by the Aral)s themselves, can now be considered as al)solutely certain. 

 Prinsep's view,'' that the numerals of India developed from the 

 initial letters of the corresponding numlx'rs, does not take into account 

 the possi})ility of accidental similarity l)etween simple signs. As 

 regards India ciphers 1-3, the formation through the corresponding 

 number of strokes would at tirst suggest itseUV but foreign influence 

 is i)ro])able. But our tirst task, laying aside conjectures, shoukl be to 

 collect all the datii on the ciphers.' ^^'oepcke's•'^ assumption of ii migra- 

 tion of the Indian ciphers during the second or third century to Alex- 

 andria is wholl}' improbable, and also devoid of an}' proof. Alexandria 

 had no need for foreign numerals, and the India system, being at that 

 time still without the symbol (zero), had no material advantages over 

 the anticjue. Rather, in their course to the west, we lirst meet the 

 India numerals among the Arabs, the earliest documentar}' proof being 

 furnished by the Fayum Papyrus, No. 798, of the Vienna collection of 

 260 A. H. (=873-74 A. D.).'* The west Arabic rather than the east 



«Sur I'origine ile nos fhiffres, lettre de M. L. Am. Sedillot Ti M. le i)rinee Baltha- 

 sar Boneoiiipagiii. Extvait de.s Atti dell' Accademia pontificia de'Nuovi Lineei, Toiiio 

 xviii-Anno xviii Sessione Va del 2 Aprile, 1865, Rome, 1865, esi^ecially p. 6. 



''Compare Zeitsehr. der Deutsch. Morgenl, Gesellseh. 50, 1896, p. 90. 



cGrammaire Arabe, 2d ed., vol. 1, Paris, 1831, p. 91 and plate viii. Exam])les of 

 the Siyaqat numerals are also found in the proceedings of the International Congress 

 of Orientalists at Vienna, 1888, Semitic Section, i)l. ii. Turkish MSS. of the Library 

 of the Deutsche Morgenl. Gesellseh., No. 5, of the old catalogue (II, p. 39). The 

 photograph of another Turkish document with Siakat ciphers, the original of which 

 is in the Ethn(igraj)liical Museum at Muni(;h, was sent me a few days ago by Dr. Buch- 

 ner. Count i^herhard von Miilinen, who lirst called my attention to this script and 

 who kindly iilaced at my disposal some modern examples of it, refers me to Ismail 

 Ghalil) bin E.dhem, Takvim-i-meskukiat-i-seldschukijje, Constantinople, 1309, p. 56. 



f/ Journal of the Asiatic Society of Bengal, vol. vii, part 1, Calcutta, 1838, p. 34S ff. 



'Compare Biihler, Indische Palaeographic, pi. ix. 



./'I refer to Euting's synopsis of the Aramean ciphers in his Nabatiiische Inschriften, 

 p. 96-97; Li(lzl)ar8ki, Handbuch dernordsemitischen Epigraphik, p. 198 ff. ; Gotthold 

 Gundermann, Die Zahlzeichen, Giessen, 1899; on the evolution of cii)her writing in 

 India, Buhler, Indische Palaeographie, pi. ix. 



f/Memoire sur la propagation des chiffres indiens: Journal Asiatique, Scrie vi. 

 Tome 1, 1863. 



^'Comp. Kara))acek, Fiihrer durch die Aussteliung (Papyrus Krziier/.og Kainer), 

 pp. 21 ()-2 17; U )V later proofs id. m Wiener Zeitschrift f iir die Kunde der Morgenlandes, 

 xi, 1897, p. 13. 



