§ 17.
— 11 —
With the numerals there is a good deal of fluctuation. The rule that only feminine numbers of the second decade,—because they end in the plural in ē,—are to be supplied with ◌̈, is seldom strictly followed. Numerals with ܪ generally take ◌̈; farther, all which end in ܢܢ,—in particular ܬܪ̈ܢܢ, ܬܱܪ̈ܬܷܢܢ "two". The plural sign is the rule in numerals which have a possessive suffix (§ 149).
C. Generally speaking, a tolerable uniformity is found,—and that in old manuscripts,—only in cases under A; in cases under B, these manuscripts often omit the sign ◌̈, where it should stand, and employ it instead in other cases, but without consistency, e. g. in the masc. of the finite verb, as ܐܫܟ̈ܚܘ "they (masc.) found"; ܕܢܬܩ̈ܕܫܘܢ "that they (masc.) may be sanctified".[1]
D. The position of the points ◌̈ was not thoroughly determined: most frequently they were permitted to rest upon the third or fourth letter from the end of the word. Much depends here on the fancy of the writer; the position most favoured is over those letters which do not rise high above the line. With the point of the letter ܪ the plural sign generally blends into ܪ̈, e. g. ܡܳܪ̈ܘܳܬܴܐ "lords" ; ܫܱܪܻܝܪ̈ܐ "true" ; still there are found also ܝܱ̈ܩܺܝܪܱܝ "revered", ܥܷ̈ܣܪܻܝܢ "twenty", ܩܾ̈ܘܪܝܴܐ "villages", and many others.
Upper and under line. § 17. Here and there a line over the letter is found as a sign of the want of a vowel, e. g. ܦ̄ܠܸܓܘ pe̊leγ "were divided", as contrasted with ܦܱܠܷܓܘ "distributed"; ܠܲܚ̄ܡܝ laḥm "my bread". Oftener this ◌̄ stands as a sign that a consonant is to be omitted in the pronunciation, e. g. ܡܕܝܼܢ̄ܬܵܐ me̊δītā "town", ܒܲܪ̄ܬ݂ bath "daughter", ܗ̄ܘܵܐ wā "was". The West-- ↑ The sign ◌̈ is even set improperly over words, which are singular, but look like plural, e. g. over ܠܱܝ̈ܠܷܐ "night" (sing. abs. st.) and over Greek words in ◌ܹܐ η like ܗܘ̈ܠܐ ὕλη.