________________
lxxxvi
THE INDIAN ANTIQUARY
[CHAPTER VIII
groups vifi, kāņa saja, pañchi, chuñchuna, and khari; and nearly the whole of the thir! class, only two variations (243 and 412) being preserved. What the cause of this mutilation whether intentional or other, may have been is not apparent.
At the end of the Pajaka-kévali manuscript, No. 70 of the Deccan College (viz. A in the list on page 214, in the Appendix to Part V), there is an appendix written in the modern Gujarati vernacular language, which explains the modus operandi in this kind of cubomancy. It runs as follows:
Tathae sakandvali-ni pasi nakh'va-ni viddhi lakhii chhaïll påsô sakan jié, tiharai:i 3 464
u!!W.. ! Cvi! 1-2.6 mondain, made in anii I b pagadâ.i padé dhuri, tá 200 ganii trami pagadan padê pehelünk, tá 300 kahii chyar pagadain padé, to 400 ganii phani pâsô bijîvár nahkhii tiharai pagadai pado, tô ek dink ek'da ganiiim bê pagadu i pade, tô 2 traņi pade, tô 3 chyar padai, tô 4 im triji-vár pani jánavu in pachhê pehelun saikadun | anai biji triji-bâr-nâ a:hk ékattha kijaż | jetalâ âmé, tetalâ upari a ak jóinai sakan joie etalê pehalusi êk padé - pachhi bé pade 1 pachhi triji-bar traņi pade to 123, ek sê naï trivisno âink thai im pehelu bé pade pachhé ék pa lé pachha trani paded to 213, bé saīsi nai tér-nô â nk awaï eni ritaïsin javusi sahil
This may be thus translated: "The mode of throwing the divination die (pasí, singular) is as follows. When the die is wanted for an oracle (Skr, sakuna), it must be thrown three times, and the first cast must be counted as hundred. Thus, if one pip (pagadam, sing.) falls, it counts 100; if two pips (pagada i, plur.) fall, they count 200; if three pips fall in the first cast, they represent 300; if four pips fall, they count 400. Next, the die (pasi sing.) is thrown for the second time. Then, of the pips that fall, one counts as the figure aik) 1; similarly if two pips fall, they are 2; if three fall, 3; if four fall, 4. In the same way, the cast of the third time must be understood. Finally, the hundred of the first throw, in th figures (áink) of the second and third, must be placed together. Whatever combined) figure results, upon that the oracle must be pronounced. Thus, if first one falls, next two fall, next at the third throw, three fall, then it is the combined) figure 123, one hundred and twenty three. Similarly, if at the first (cast) two fall, next one falls, next three fall, the result is the figure 213, two hundred and thirteen. This is the correct manner of proceeding."
It is clear from this explanation that in the ancient Indian art of cubomancy only a single die was used; and that the die indicated only the four numbers, respectively represented by 1, 2, 3, 4 pips on four different facets. A die in the form of a tetrahedron would satisfy these conditions ; but the existence of a tetrahedral di at any time is, I believe, an unheard of thing. It seems probable, therefore, that the die was one of that elongated kind, with four long sides and two rounded ends, which is known as talus or astragalus, or knucklebone, and on which the four long sides were marked. with pips. If the die had the ordinary cubical form, two of its six equal sides would have borne no pips; and then there would have been the not infrequent chance of one of the two unmarked facets turning up in any of the three consecutive casts. In such a case, o course, the throws would have lind to be repeated tiir some pid-markedyacet arhed up the the explanation above-quoted does not seem to contemplate the occurrence of such an eventuality, which is not even alluded to. At the same time there occurs in the Introduction to the manual in Part IV (1.3, on page 192) an obscure phrase which may point to the die having had the form of a six-sided cube. There the dice are described as kumbhakdri-matanga yukta, lit., "joined with a kumbhakâri and a matanga." This may mean marked with the . .. Lumbhalant nnother woman for the girl kumbhakâri). and mátaiga, or eleph
