30 George Dahl, 



lows no fixed laws\ Consequently the transposition of ^ and *) in 

 this word resulting in the form PpltO is not an unusual phe- 

 nomenon. The ending m was later regarded as plural. 



In the Arabic a somewhat similar process took place. The oldest 



form of the noun is very likely .JoX^d ? practically a transliteration 

 of tentor(ium). Next the n assimilated to the r of the last syllable 

 and the form^- U U came into being. Last of all the vowel of the 

 penultimate syllable was assimilated to the ' of the ultima, and the 

 form *Jowb was the result. This last is the most typically Arabic 



form of the three". This explanation of the probable history of the 

 word is confirmed by the fact that at the time Dozy wrote (in 1845) 

 the word was in different stages of its development in various coun- 

 tries'. In Syria the form tantoura was used; in Egypt, tartour; 

 and in Algiers, tortora. This illustrates also the fact that in Syria 

 each of these forms was used at one time or another, and probably 

 more than one form was in accepted use at the same time. It 

 explains, too, the persistence or recurrence of the older form Tan- 

 tfira in the name of the modern town, although Tortura is appar- 

 ently the more recent version of the name. These various forms of 

 the word seem to be used interchangeably, now one, now another, 

 being in current use. 



The derivation from the Latin *' tentorium" thus takes into 

 account the various' changing forms tanttir, tartur and torttir. 

 Levy's suggestion* that f^p^lD is derived from the rather far- 

 fetched "teretes", as well as Jastrow's^ proposal of *'turritum" 

 must be rejected. The derivation from "tentorium" has also this 

 superiority to the other suggestions — we can see that the name 



' E. g. aKv<l>oc becomes in the Talmud }<pD1D (S. Krauss, Griech. u. Lat., 

 etc., I, pp. 113 ff.). Cp. also Syr. NT''l")p from Kalddpiov, Arab. 



from ^T^ ' jj**.iL«4> from jjaaJsJc-* {fJ-era^a), ^^jOx\ from apaivLKov, and 



many others. 



2 The form tontura also cited by Dozy (loc. cit.) is simply a variant form in 

 which the assimilation of the vowel ' preceded that of the consonant » . 



3 Dozy, Vetements, pp. 262 ff. ^ 



4 Page 29. 



5 Ibid. 



