28 THE INDIAN ANTIQUAEY. [Jaxitabt, from a distance. The chief sources of confusion are the Sabsean forms for the letters— I. h, d 3 3, s. II. a, i, s, o, n. HI. j and «j. IV. n, », ? n. V.i and s. VI. 1 and d. VU. * and p. VIH. 2 and c. IS. p and d. An attentive collation of texts only can elicit the true lection. (See p. 26.) There is Teason to believe that, besides the monumental, another more manageable cursive form of writing also developed itself: the in- scriptions of B e 1 e d A r h a b, of B e 1 e d Nehm, of Silyam, but principally the graffiti ofJebel Sheyhan, which contain so many strange signs, bear witness to this. It is even possible that a portion of these signs are due to the combination of two or three tetters tor their unusual forms. That the Sabsean characters allowed of ligatures is proved by the existence of numerous monograms where one ion trait serves to unite three and even four letters. Like many other nations, the Sabsans also used ornamental letters, of which Reveral specimens exist. In the Museum of the Bombay Branch of the Royal Asiatic S there is among the Sabfean inscriptions one n iili a lar**e ornamental initial enclosed in a quadran- gular frame cut round it, leaving the Idler i in relievo, with three ornamental cavities in its body ; and in another much smaller slab one trait unites several letters. The Sabeean orthography is very sparing in tin; designation of vowels. The letter a never l.t,i[ ilncally denotes a vowel 5 with ran- except i"ns i ami 1* are rendered by » and 1 at the end of words only. There exists, on the contrary, a great tendency to elide these in the body of words, even when they are radicals, or when they represent an element of grammatical flection. Thus we meet with zr.^ ( U . 624, - ». jdm ( // . ] to (Op, iv. 1). tnn (//• &89)| instead of the usual orthography, Dan. pw, craa, dtti. Sometimes the sa-iptio defectiva is adopted where the exist- ence of & iliphi liong is certain ; thus, for instance, the word H a d r a m | n t ifl nearly always spelt no-En ; likewise rp* (Os. xvin. 5) for tp*». • The words are generally separated by a per- pendicular line ; this, however, is often omitted in inscriptions written with cursive characters, which aggravates the difficulty or interpretation. thi.s mark of separation is too close Jr. the adjoining characters, and is apt to give rise to mistakes. For instance, the representative of e - standing too near after the perpendicular of separation, may with it be read as ■>, 1; if it be after ^ a it will make with it the letter u a, and if it be immediately after & 3 the supposed coales- cence will represent >J a ; and lastly if it precedes ^ x both together will make p j. The end of the inscription is sometimes indicated by certain ornaments ; there are also two or three signs to indicate the beginning, especially in long texts sculptured on large edifices. Inscriptions of small extent destined to attract the attention of thepublicare enclosed by one or two letters. ( i->. j * The Verb. The voices winch have hitherto been authen- ticated are the following seven : — 1. Original form ^5 Qal : — toi, n% vbc. ]<;.)?. 2. Energetic form fee Pa'el : — r-rc, -m. 3. Reflective form bxn Tafa'el (tafav!) . ■*>-. I Internal reflective form fer© Pat 'a I :- iarc, mrp, S*rc- 5a. Causative form 'rccn Hafel: rnm, vwn. '^n, Tin, Wn, rnsn. ■wco Safel : finro, tixj, an(i)D, ks(i)c, ^::r (Muueanand Hadramaut dial.). 6- Hi fleet ive causative form : tare Satfttl; w^rc, lime, ■nhrc 7. Causative and reciprocal form Smsn Hifl- fa'ol : r2rr;n. '1'h q Qeri is the principal voire, from which the other voices are deri veil, either by im modtfiaatinna of the rath* M I, OTby the aid of Oertsin letters added externally, writing shows only the skeletons of words, we are notable i" point oat the various details of tli root with reference to the vowels. Act ingly we do not know whether 1 radical was pronounced with the vowels w, /, u, as in the majority of the Semi tin languages, or whether it was affected by tho sheva, as in Ethiopia. Thanks to tho usage of separately prom rag the duplicated liquids, it is possible to discover the existence of the Fa'el among the voices derived by the internal modification of the root : mc, j::?, rrt> (H 188, 2) ; the p " All thin I Lave ftbowu on p. 20.
