________________
CHAPTER III]
BOWER MANUSCRIPT
xxxi
southern (g) (F. GI. No. 18, p. 82, Plate xi. 11. 10, 15) and the northern (n) (F. GI. No. 35, p. 153, Plate xxxii, 1. 11). Moreover, there is a peculiar form rů (h) and (p) which substitute two parallel strokes for the southern semi-circle, and this form appears to be common to both areas; for it is seen in the south at Junagadh (Lat. 21° 31': F. GI, No. 14, p. 59, Plate viii, 1. 10), as well as in the north at Bilsad (Lat. 27° 33'; F. GI. No. 10, p. 44, Plate v, 1. 11).
In the third place, there is the striking difference in the use of the two forms of the letter y, the old and the modern. In Parts I-III, as already observed, and as will be explained in detail in Chapter IV, the modern form of y is used optionally with its older three-pronged form; while in Parts V-VII that three-pronged form is used exclusively. The modern form of y originated in the norch, and its use never spread to the south.62
The obvious conclusion suggested by the foregoing evidence is that the persons who wrote Parts V-VII were natives of some place lying within the southern area. In the case of Part VI, at all events, this conclusion is confirmed by the other significant fact that the folios of Parts VI are numbered on their obverse sides (see Chapter II, p. xx). For, as Bühler has pointed out in the. Vienna Oriental Journal, Vol. VII, p. 261, the practice of numbering the folios on their obverse side is a peculiarity of Southern India. We have a good example of this practice, of a very early date, in the copper-plates of the Pallava king Sivaskanda Varman, and the Kôndamudi Plates of Jaya Varman, a contemporary of the Andhra kings Gautamiputra and Vašiştbiputra, who reigned about 113-137 A.D. These copper-plates may be seen in the Epigraphia Indica, Vol. I, pp. 4-6, Plates I-V.. Vol. V, p. 86, and Vol. VI, p. 315. At the same time, the place whence the writers of Parts V-VII came must have been somewhere near the border line of the two areas. This is indicated by the circumstance that the southern forms of é, ru and rû arc employed in conjunction with the northern form of m, exactly as in the inscriptions, above mentioned, at Eran' and Khôh, both of which places lie on the border line. While the writers of Parts V-VII appear to have come from some place near the southern limit of the northern area, the person who wrote Parts I-III must have come from somewhere near its northern limit, that is to say, from Kasbinir or Udyana. This is indicated by the occurrence in Part I) (fol. 27a, 1. 11) of the peculiar Sarada form of the letter k (Table I, No. 2 in Traverse 2). The Sarada script is peculiar to Kashmir, where it originated directly from the Gupta script in the course of the seventh century, and where it is still current, almost unchanged, to the present day. The Sarada forms of
Fig. 14. those letters which enter into the present enquiry are shown in the lower line of
1 3
* A Fig. 14.63 The upper line shows the corresponding letters in the script of the r Horiuzi Manuscript, which was written in the first half of the sixth century
Letters of the Horiuzi and Sarada scripts. (Anecdota O.xoniensia, Vol. I Part III, p. 64), Its script, therefore, was the immediate predecessor of the Sarada script. The
69 There is a further point of difference between Parts I-III and Parts V-VII. It concerns the shape of the initial vowel i. This point, however, is not decisive of locality, and will be discussed in the sequel, p. XXXVI.
63 These letters are extracted from a birch-bark manuscript in Sarada characters which was prescated to me by Dr. Stein in December 1898.
d