DKCCAN TIUP IRON-CLAY. 211 



foot of the great Jamboti ridge, and pass south-eastward forming, near 

 Ambgaon and Chapoli (Champolee), a well-defined plateau capping 

 the extreme southern promontory of the great Deccan trap area over- 

 looking the deep and romantic ravine of the Mahadayi (Maadwee) river, 

 (see page 3}. As seen from the edge of this ravine, the iron-clay beds 

 appear to be represented on its southern side by a similar set forming a 

 plateau around Gauli (Gowulee) . To the northward of Ram Ghat this 

 set, which may for convenience be called the " watershed series," is 

 represented in the plateaux near the source of the Ghatprabha already 

 alluded to, and also in the valley of the Haranhashi river near 

 Mahadeogarh and Narayangarh (Narraingurh) . 



Two remarkable caves were observed in these lower iron-clays, 



one a little to the west of Sadda (Suda), the 

 Caves in the iron-clay. 



other about a mile to the north-north-west of 



Holund. Both occur below the surface of wide level spreads of 

 rock, and present no other openings but large rudely circular holes 

 12 to 15 feet in diameter and about as deep. Neither is accessible 

 without ladders, owing to the '^ oubliette'^ character of the mouth, 

 and torches are also required, for want of which I was unable to 

 explore them, as time did not admit of my re-visiting either place. 

 According to the villagers, both are frequently tenanted by wild beasts, 

 bears or tigers; the Sadda cave indeed was said to be a permanent 

 tigers^ den, and my guides so greatly objected to attempting any descent 

 into it, that it looked as if they had told the truth. It has two open- 

 ings at no great distance apart, and is said to be very extensive. These 

 caves have in all probability other lateral exits for the water which 

 drains into them. They are probably due to the action of subterraneous 

 streams formed during the tremendous bursts of the south-west monsoon. 

 None of the openings seen had the character of swallow-holes, but 

 seemed due to the roof of the subterranean passage having fallen in 

 locally. 



( 211 ) 



