Indian Antiquary Vol 40

Author(s): Richard Carnac Temple, Devadatta Ramkrishna Bhandarkar
Publisher: Swati Publications

Page 57

FEBRUARY, 1911.)
OLD INDIAN NUMERICAL SYMBOLS
The Taxila plate anl other inscriptions from t'e Panjab frontier give us the key of the Kharūsthĩ notation as far as the hundreds, so that our knowle Ige of the notation within this limit is probably corre:t.
The script is written from right to left, and in the notation the smaller elements are on the lett.
Oar information about the Kharūsthi writing will, possibly, be somewhat extendel in the near future, but, as far as our present knowledge goes, the Klarusthi notation appears to have little connection with the Indian notation proper. It is said that the script is derived from or allied to Aramaic and the two notations have close resemblances.
In the interpretations of the Kharüsthi notation our earlier orientalists made the usual mistakes-eg., Cunningham read 3:3' insteal of 20 + 20 + 20 ( = 60).
III.
The notation that was in general use in India in early times, and persisted until quite recently has been variously termed the Brahmi, Sanskrit, old Nagari, and old Indian notation. It is a non-place-valne notation with special symbols for the numbers one to ten, twenty, thirty ,
... a hundred and a thousand. The numbers 11 to 19, 21 to 29, etc., are expressed by the symbol for the tens followel hy symbol for the units. Two hundred and three hundreil are expressel by the symbol for 10) with the addition, respectively, of one or two horizontal strokes or books (see table II). Higher multiples of a hundred are denoted by the symbol for 100 followed by the corresponding units figure. The thousands, which occur very rarely, are treated in the same way as the hundreds. To express three hundred and ninety-four,' to the symbols for 100 are attached two horizontal strokes (or hooks) on its right side, and this is followel by the symbols for ninety and four in order, thus F O y . No symbol for zero was employed.
We have already pointed out some of the errors that the early orientalists fell into in dealing with this notation, but there are errors of another type that are more difficult to deal with. The results of the earlier investigators wore based almost entirely upon the evidence given by eye copies of inscriptions, and that found in comparatively modern manuscripts. The old fashioned copies of inscriptions were, indeed, a fruitful source of error in many ways and in particular with regard to the forms of numerical symbols. We now have, however, a body of mechanically reproduced inscriptions, which should give evidence as to the forms of the symbols sufficient to enable us to determine the system used with fair accuracy; and in the present note it is proposed to utilise this superior evidence and to exclude, as evidence, the old fashioned eye copies. This does not, however, make the task any easier : the old eye copies are often so delightfully clear and anambiguous, whereas the mechanical copies are as obscure and as difficult to read as the originals.
It is, of course, impossible to give here all the examples of the Brāhmä symbols that are available, but in all cases the sources of our information are indicated and the reader is referred to these sources for first-hand evidence. The earliest examples are taken from the Asoka inscriptions, following which the Nānāghāt, Kārle and Nasik inscriptions have been utilised. The Mathorā inscriptions and, later on, the Gupta inscriptions extend our evidence to the north, as do the Pallava plates and others to the south. Of great value also is the evidence afforded by coins and in particular by the coins of the western Kshatrapas. The sources here indicated may be considered to give representative examples which are, more or less, confirmed by incidental examples of other periods and places, and by the practice followed in the earliest manuscripts known to us.
In some cases the numerical symbols are accompanied by the equivalent expressions in words; other examples, but these are unfortunately of comparatively late dato, aro in series-as in pagination ; while a third class consists of isolated numbers, principally dates, and these, if the symbols are not of normal types, must be to some extent conjectural. The attached table is divided into sections corresponding to these three classes.
15 Ep. Ind., Vol. IV, p. 54; Arch. Suru., India, Vol. V, PI, XVI and Pl. XXVIII.