108 THE AMERICAN MONTHLY [April, 



sure, aided by secondary outgrowths of quartz, optically contin- 

 uous with the original grains, till it is now composed of irregu- 

 larly-shaped granules of quartz for the most part in close contact 

 with each other, but intermixed with a small proportion of a 

 felspathic paste. This paste yields more readily to the action of 

 weather than quartz, and, though weathering does doubtless also 

 take place along the junction surfaces of the individual grains of 

 quartz, the disintegration of the rock appears to result almost en- 

 tirely from the decomposition of the felspathic paste. In such a 

 rock the development of flexibility will depend on the proportion 

 and distribution of the felspathic paste. If this be absent or only 

 present in small proportions, decomposition will not extend to 

 any depth, and there will be a mere film of decomposed stuff' on 

 the surface of the undecomposed rock ; if it is too evenly dis- 

 tributed, the individual grains of quai'tz will not be in sufficiently 

 intimate contact with each other, and the rock will weather into 

 a fine sand, easily washed away; if finally it should be suitably 

 distributed, but too great in amount, the voids left by its removal 

 will be so large that the quartz aggregates will not interlock with 

 each other. The number of conditions that must be fulfilled ac 

 counts satisfactorily for the rarity of ffexible sandstone, and for 

 its capricious distribution in rocks that have to all appearances 

 the same composition and structure. 



In 1 87 1 Mr. Fedden found at Charli, south of the Penganga 

 river in Berar, a rock which exhibits the same sort of flexibility 

 as that of Kaliana, but differs widely from it in ever}- respect ex- 

 cept that it is now, what the other once was, a sandstone. In- 

 stead of being composed almost entirely of quartz and owing its 

 cohesion to the metamorphism it has undergone, it is an ordinary 

 unmetamorphosed calcareous sandstone, at first sight not remark- 

 able for anything but its softness, and the large proportion — over 

 35 per cent. — -of carbonate of lime it contains ; but a closer exam- 

 ination will show that the calcareous cement has become crystal- 

 lized, and that the rock is now in reality composed of crystals of 

 calcite some ^-inch in diameter which include the odd 65 per 

 cent, of quartz and other granules as so much extraneous matter. 

 Microscopic inclusions of extraneous matter in crystals are com- 

 mon enough, and often form a considerable proportion of their 

 total bulk, and the celebrated case of Fontainebleau sandstones, 

 where the calcite crystals contain 90 per cent, of sand, shows to 

 what an extent this can be carried. In the case of the Charli 

 rock, the large proportion of crystalline structure of the cement 

 has had important results. It is a well-established fact that when 

 a crystal is attacked by a solvent it is not uniformly acted on, 

 but that solution acts more rapidly along certain planes which 

 penetrate the crystal in various directions. In the case of the 

 Charli rock this has taken place, and the numerous surfaces of 

 contact between the crystalline calcareous cement and the grains 

 of sand have been so many more surfaces of weakness along 



