J. HORN ELL ON 
41G 
agate flakes as suggested by him (loc. tit., p. 81, Vol. I). Had such been the case 
the shell sections and the waste fragments cut off in the course of sectioning — the 
shoulder of the shell and the lip portion of the mouth whorl— would not exhibit the 
perfect regularity and evenness of sawn surface which they do. It is quite possible 
to cut a ring section from a chank-shell by means of a flint “ saw ” but the task is 
one involving prodigious difficulty and the waste pieces must necessarily be broken 
and chipped off in the process in consequence of the impossibility of cutting cleanly 
through the shell owing to the smallness of the tool and the thickness of its back. 
Only a thin blade, such as the employment of iron or steel permits, will perform the 
task of sawing off the shoulder or the lip portion in one continuous operation and 
without breaking off the waste parts piecemeal as the sawing progresses, as must 
necessarily be the case if a small thick- backed stone tool be employed. 
The hone found may conceivably have been used for rubbing down the thickness 
of the edge of the iron saw employed, as at Dacca to-day, or in sharpening the edge 
of the chisel-edged implement used in sharpening the teeth of the saw itself. 
Raichur Doab (S.-W. Hyderabad). 
The country lying between the Tungabhadra and the Kistna, the Raichur Doab, 
appears to have been thickly populated in prehistoric times by the same race as has 
left great numbers of implements scattered through the present districts of Kurnul, 
Cuddapah and Bellary. Three sites have yielded remains of bangles and of these 
one has undoubtedly been a manufacturing centre where the raw material has been 
cut up and worked into bangles for sale to the people of the district. This ancient 
factory was located near Maski, on the right bank of a tributary of the Tunga- 
bhadra. Exhibits 2783 — 63 to 2783 — 85 are typical chank workshop waste exactly 
similar to what I have seen in Dacca factories. There are examples of obliquely 
cut ff shoulder pieces ” with the apex purposely smashed in as in modern rejects from 
a Dacca factory, of fragments of the columella and of cut-off “ beaks.” Fifty- three 
fragments of worked bangles are also shown and a considerable proportion, 2783 — 25 
to 2783 — 35, exhibit traces of ornamentation in the form usually of cross grooves. 
Associated with these chank artifacts are numerous neolithic chert flakes, scrapers, 
and cores, also a fair quantity of old pottery which Mr. Foote regards as of iron age 
(P- 2 3 , Vol. I), though I am personally inclined to place the age considerably more 
recent for reasons to be stated later. The And of a broken iron blade, 2783(2, at this 
site is of the utmost importance ; it may well be the remnant of a small hand-saw 
such as is used to-day by Bengal workmen for roughing out much of the line decora- 
tion so characteristic of bangle ornamentation. 
Fragments of chank-bangles have also been found at two other places in Raichur 
Doab — at Kotegallu, Ring Sugur Taluq and at Rawalkonda. In both cases they 
were associated with neolithic implements, basalt celts and very numerous chert and 
agate flakes ; a considerable number of the Rawalkonda flakes have biserrate edges. 
At Kotegallu, the objects had been turned up during ploughing and at Rawalkonda 
they appear also to have been surface-finds ; hence in my opinion the significance 
