________________
CHAPTER V
BOWER MANUSCRIPT
xlvit
merely transitional stage, as we shall see presently,-was reached, the character would be written with a single movement of the hand. Beginning with the top of the medial straight line, the hand moved down to the base line, then upward and leftward, round the loop, back to the point of junction, and finally onward to the angle on the right. But it soon began to be observed that the letter could be written with greater speed, and with more economy of effort, if the downward movement of the hand was carried at once to the loop on the left without touching the base line at all. This slight change produced what is practically the modern form, as shown in (i) and (k). Thus, there were now three forms: the old, the transitional, and the modern. The old form persisted in the Gupta script of the southern area. The transitional form arose in the northern area about the middle of the fourth century A.D., and disappeared about the end of the sixth century. The modern form arese practically at the same time as the transitional form ; but it gradua''y extruded the latter; and it persists to the present day in the slightly modified Nagari form of the letter which only projects the perpendicular below the base line.
The transitional and modern forms, or, to use an inclusive and more convenient term, the "new form" of y was, so to speak, invented in the western portion of the northern area. Thence it gradually spread over the eastern portion. This may be seen clearly from the epigraphic records of the Gupta period. See Fig. 20.
Fig. 20. It first appears in the year 372 A, D in the stone pillar inscription of Vishnuvardhana at Bijaya gadh 5
ar (Long. 77° 20'), in érêyê, (a) (F.GI., No. 59, p. 252, Plate xxxvi C, 1. 4), and about 400 A.D. in the
c
a nd about 400 AD in the First appearance of the new form. rock inscription at Tusâm (Long. 76° 0'), in yoga (b), (F.GI., No. 67, p. 269, Plate xc, 1.3) The boundary of the two areas, as previously stated (Chap. III, p. xxvii), is E. Long 81°. In the eastern area the new form makes its first appearance in the stone inscription of Isvaravarman at Jaunpur (Long. 82° 43'), in anyavayê (c), (F.GI., No. 51, p. 228, Plate xxxii A. 1. 2). Unfortunately this inscription is mutilated, and its date, if there was any, is lost; but it belongs to the middle of the sixth century. The first dated inscription in which the new form is found, is that of Mahânâman, in 588 A.D., at Bôdhgâyâ (Long. 85° 2). Here both new forms, the transitional and modern, occur numerously; e.g. the former (d) in yukta, the latter (e) in yêna, (F.GI, No. 71, p. 274, Plate xliA, 1. 1).
For writing the single y, the new form appears to have come into use about the middle of the fourth century, but for the subscript y, as the second part of a compound letter, it was in use about three centuries earlier, from the beginning of the Indo- Fig. 21. Seythic period. See Figure 21. An example of the transitional form (a) of the subscript y, from a Kushana inscription, is shown in plate III, line 42. No. 3, of Bühler's Indian Palæography (in the Encyclopædia of Indo-Aryan a Research). Examples of the modern form of the subscript y (6) occur Subscript ya. numerously; e.g., in Kanishka's inscription of his seventh year, i.e., in the year 51 B.C., in the Epigraphia Indica, Vol. I, p. 391, No. XIX.82 It can hardly be doubted but that it was the economy of time and effort in writing the new form of y, which led to its adoption in the case of the subscript y. But in the case of the single y,
Fig. 22. there operated an additional reason. This is brought out very strikingly by a certain circumstance in the use of the new form in the Bower Manuscript. This is the circumstance that in writing the syllables yé, yai, yo, yau, the new form of y is employed whenever the vowel (é, at, 0, au) is made with a Vocalic superior and lateral lateral stroko, but the old form is used when the vowel is made
strokes. with a superior stroke. These vowels, namely, are indicated by attaching to the head of the
85 In the second line of the accompanying Plate. It is also shown in Buhler's Indian Palæography. Plate III, 1. 41, No. 5.--As to the Kushana dates, I follow Dr. Fleet's theory, which I now believe to be correct, that they are to be reckoned from 57 B.C., being dates of the so-called Sanvat Era,
