NOTE ON THE LATEFJTE OF ORISSA, 289 



For these reasons, we may safely conclude that the above hypothesis 

 is untenable. 



The second hypothesis by no means explains the fact of laterite being 

 confined to a certain definite level, as is the case in Orissa. Were it 

 true, we might expect to find Lithomarge on the hills, Laterite on the 

 low grounds, and it would only be natural to anticipate, that, in a great 

 number of cases, each rock would be found not associated with the 

 other. This by no means agrees with the facts observed, for so con- 

 stantly does Laterite occur wherever Lithomarge does, unless there is 

 evidence of the former having been denuded, that it becomes evident 

 that the existence of the latter is dependent on the presence of the 

 former, and all probabilities therefore point to the last hypothesis, viz. : 

 the derivation of the iron in Lithomarge from the Laterite, by the 

 percolation of surface waters, as the only theory in accordance with 

 the facts. And the quantity of peroxide of iron which saturates the 

 surface of the clay below Laterite in old sections is good additional 

 evidence of this passage of the iron from the upper into the lower 

 formation. 



But another point yielding valuable evidence upon this question is 

 the mode of distribution of the iron as regards 



How would the iron _ . . . ■ 



be distributed according the quantity Contained in various parts of each 

 to each hypothesis ? 



rock. 



Assuming the first hypothesis to be correct, we should expect, 

 either that the quantity of iron throughout the two formations, 

 from top to bottom, would be the same; or that it would diminish 

 regularly, from the surface downwards, to the water level of the 

 country. 



Assuming the second, the Laterite would probably be the richer 

 formation, since in the 'process of the disintegration of the mass, formed 

 by the decomposition of gneiss and other rocks and its redistribution by 

 water, the non-ferruginous portion, in consequence of its lower specific 



F 



