TlIAP. I\'. :l.] MKTAMORPHIC UOCKS. 99 



at this place is their comparative freedom from calcedonic incrustations, 

 AA'hich are very rare here. 



In several places on the top of the ridge the travertin-like incrusta- 

 tion presents very strikingly the appearance of having been deposited 

 around small springs, which seem to have discharged their water-like 

 fountains on all sides. Similar phenomena, but still better marked, 

 were observed near Tripunguly, in Trichinopoly district, some 30 miles to 

 the south-east, and of which we shall have to speak presently. 



North of the Cajareputty magnesite, are several beds of compact talc 

 or steatite, associated Avith beds of amphibolite or hornblende rock, and 

 abounding in small acicular crystals of bright green actinolite. This 

 spot lies just within the Trichinopoly district. 



Very similar, indeed almost identical in appearance, with the phe- 

 Travertin of Tripim- i^omena at Cajareputty just described, are those 

 ^" ^' presented to view on the high ground half way be- 



tween the two villages of Tripunguly and Theerampolliam, 10 miles 

 north-west by north from Trichinopoly. Here the surface of the gneiss 

 rock is in various places covered by a pale brown compact travertin, tra- 

 versed by bands and veins of a white mineral (in all probability of 

 magnesite,) showdng by its mode of occurrence that it had been formed 

 from the overflow of small springs, probably (from the similarity of the 

 minerals thej^ yielded,) but so many branches of one important spring 

 rising from unkno\vn depths. 



There would appear to have been six orifices whence water flowed, 

 whether they be considered as separate springs, or only as various branches 

 of one main spring. The first of these extinct spring mouths is on a 

 waste piece of high ground to the right of, and on the path leading from, 

 Tripunguly to Theerampolliam, and is about half way between the two 

 villages. Abovit 200 yards north, across a depression which is cultivated, 

 lie the other springs, or rather sites v.diere springs have been closeK- 

 clustered together, and here it is that the travertin formation of the 



(321) 



