420 
J. HOFtNELL ON 
From a consideration of the details given above a certain number of facts and 
conclusions of importance emerge, to wit : — 
{a) In all cases the fragments of bangles and of chank shells appear to have been 
surface finds. In several cases this is definitely stated and in the remainder wherever 
no statement of horizon is given, the context points to a like provenance. From this 
it follows that association with neolithic artifacts in itself has little value or significance ; 
both neoliths and chank fragments are practically indestructible by atmospheric 
weathering agencies, and their association may merely connote the fact that particular 
surface areas have suffered little or no denudation or change since neolithic times, 
whereby the broken implements and discarded ornaments of a later age have mingled 
with those of an earlier one. Or it may be the result of the artifacts of different ages 
having been weathered out of different alluvial strata in such way that they come 
eventually to lie together at a lower level of the original ground or else in some newer 
river deposit into which floods may have rolled them. 
( b ) The facts already noted that all sections of chank shells, working pieces as 
well as wastage scraps, show cleanly sawn surfaces as verified by examination of the 
originals now in the Madras Museum, and that these surfaces show series of striae often at 
two or more angles to one another , are sufficient to negative the tentative suggestion made 
by Mr. Bruce Foote [loc. cit ., vol. I, p. 81, and vol. II, p. 132) assigning a neolithic origin 
to the workmanship. Neither serrate nor biserrate chert flake saws, however delicately 
made, could possibly produce such cleanly sawn sections as we see represented in the 
collection. The aid of thin metal saws must be invoked, and it is most significant that 
in two instances (Ambavalli in Kathiawar and Muski in the Raichur Doab) fragments of 
iron knives were found associated with the remnants of chank working-sections. In 
several other cases (Srinivaspur in Mysore, Havaligi Hill in Anantapur, and Bastipad 
in Kurnul) pieces of iron slag were found in association. 
As the working sections of chank shell retain visible evidence of being sawn by 
means of a metal (iron) saw, and as iron fragments are frequently associated with them , 
the evidence is to me satisfactory that the age of the former cannot possibly be neo- 
lithic ; knowledge of the manufacture of iron into somewhat elaborate tools — saws, files 
and drills — must have been possessed by the bangle makers. This would appear there- 
fore to rule out the early iron age, when iron weapons and tools were of primitive design. 
Incidentally this conclusion is likely to affect the estimate of age accorded to the 
potsherds so frequently associated with fragments of chank bangles and to render 
doubtful their identification as neolithic or even of early iron age. 
( c ) Three sites alone give other than negative evidence in regard to age. These 
are Gudivada in Kistna District, Valabhipur in Kathiawar and Mahuri in Gujarat. 
The remains at the first named are indubitably Buddhistic while the occurrence of a 
figurine of a bull with a double garland round the hump points distinctly to an age 
when the adherents of Brahmanism were in the land, holding in especial reverence 
Siva’s sacred bull. Most important find of all was that made in the ruins of Valabhipur, 
for the history of this old city is fairly well known ; the dates of many of the great 
events that happened there are known and the descriptions of two Chinese Buddhist 
