Table Lands of South India. 321 



more accurate terms, which are quite as convenient as the old 

 term, and have the advantage of perspicuity, as well as of 

 conveying definite ideas. 



Rhomboidal Trap. 



Trap is another term in Petrology, which has been most 

 abominably misused. As it is now made use of by English 

 geologists, in a generic sense, for all submarine lavas, or 

 rocks of undoubted igneous origin, modified by unknown 

 agencies, without any respect to the mineral composition ; 

 perhaps therefore it would be best to reject the term altoge- 

 ther from a petrological definition, and leave it for the use 

 of geological theorists. 



In India, however, the term has been very universally used 

 as applied to rocks generally characterized by the presence 

 of hornblende ; and as in describing these protean rocks, a 

 general term for such as we cannot arrange with any par- 

 ticular species of rock, will be very convenient as a sort of 

 lumber room ; I propose therefore we should limit its use to 

 such rocks as have been described by Lieut. Baird Smith, as 

 forming dykes in hills of Palicondah and Amboor (Madras 

 Journal, vol. ix. page 287.) 



The colour is generally a sort of reddish brown, a little 

 speckled in appearance when viewed closely. The fracture 

 surface is dull and earthy, never in any degree granitic. To a 

 lens, the structure is to be seen composed of amorphous 

 grains of hornblende, without glistering faces, mixed up with 

 felspar in the greatest proportion, which partly isolates the 

 grains and causes the approach to a speckled appearance. 

 In structure, the masses are composed of a number of small 

 rhomboids, not generally larger than three or four inches in 

 the face. 



The cleavage is generally in faces parallel to the faces 

 of the rhomboidal prisms, and it breaks with difficulty in any 

 other directions. It is very tough, and is broken with dif- 

 ficulty. 



