LOSSES IN RESIDUAL SOIL FORMATION 39 



removed. In this particular sample, 100 feet of rock 

 would produce only 2.34 feet of soil. It is not uncommon 

 in limestone soil regions, as Kentucky, Tennessee and 

 the Ozark region, to find soils forty and more feet in 

 depth, and, since the average limestone contains nearly 

 90 per cent of carbonate, these deep layers of soil must 

 represent some hundreds of feet of rock. This is the re- 

 sult almost entirely of solution by carbonated waters, 

 which gradually develop crevices and caverns in the 

 rock. 



Other types of rock, however, do not suffer such a 

 large amount of loss. The loss, of course, varies with 

 the character of the processes which are at work, as 

 has been pointed out in the case of granite and gneiss. 

 In Columns V and VI, a clay from diabase rock suffered 

 a loss of 39.51 per cent, and a basalt soil in France rep- 

 resented a loss of over 60 per cent. The latter are much 

 more basic than the granite or gneiss, and would there- 

 fore be more amenable to chemical decay. The soap- 

 stone, which results from the alteration of pyroxinite 

 rock, undergoes a loss of 52 per cent in the transition 

 to soil. In Columns XV to XVIII are given the calcu- 

 lated loss in changing from the average analysis of 

 igneous rocks to shale and sandstone respectively. As 

 was stated above, these latter are ancient soil material, 

 or potential soil material, and the figures given represent 

 an attempt to determine the average change which takes 

 place in the derivation of a shale or sandstone (corre- 

 sponding to clay or sand soil) from igneous rocks. These 

 calculations are, of course, less accurate than the pre- 

 vious figures on such loss, because these rocks have been 



