At BHBS Rhet 216:5 the form ܐܝܡܝܬܐ with var. ܐܝܡܢܬܐ might profitably be read as ܐܝܡܟܝܘܬܐ or the like, i.e. "the attribute of being from somewhere in particular." This would be a typical "Bar Hebreacism!" Watts' "demigods" seems highly unlikely, in spite of its presence in the Greek parallel at Book II - Chapter 22.
Whether or not ܡܢ precedes seems to be an idiolectal rather than semantic choice.