177 



it soluble to meteoric water. Many of these facts are brought out by Cum- 

 ings in his paper, "On the Weathering of the Subcarboniferous Limestones 

 of Southern Indiana," in the Proceedings of this society for 1905, pages 

 85-100. The great central area is practically level, owing to Tertiary 

 peneplanation, thus lessening the amount of run-off. The western part of 

 the area is overlain by the Huron formation, composed largely of porous 

 sandstone which absorbs precipitation and passes, a part of it, at least, 

 downward into the underlying Mitchell. The area as a whole is wooded, 

 which also tends to hold meteoric water rather than to give it up to such 

 surface drainage as exists. The area in Indiana lies in a section of coun- 

 try which is one of relatively great humidity. 



The individual layers of the formation are comparatively thin and are 

 generally separated by thin layers of impervious shale. This factor tends 

 to weaken the layers when a cave is formed beneath them and allows them 

 to collapse, thus giving the stream the opportunity of enlarging the cave 

 in a mechanical way by removing the debris. 



FORMATION OF CAVES. 



Limestone (CaC0 3 ) is only slowly and difficultly soluble in pure water, 

 but when water descends through the atmosphere as in rain and snow, a 

 certain per cent, of CO, is dissolved, forming ILC0 3 . This is enabled to 

 dissolve calcium carbonate, forming calcium bicarbonate thus : H 2 C0 3 + 

 CaC0 3 =CaH 2 (C0 3 ) 2 . The latter product remains in solution until evapo- 

 ration takes place. It is owing to this fact that stalactites and stalagmites 

 are formed in caves. 



Now when rain-water falls on an area such as that underlain by the 

 Mitchell limestone where the conditions favor a minimum amount of run- 

 off and evaporation, and where the greater amount of precipitation soaks 

 into the soil, it will tend to collect and flow downward through the most 

 available passages. Such passages are furnished by the above-mentioned 

 joint-planes. Where two of these joint-planes cross at right angles, the 

 passage will be freest and it is probably at such points that most of the 

 ground-water passes downward. This downward flow of water may be 

 arrested by several causes, four of which are most important. The joints 

 become tighter as they descend into the earth ; the level of ground-water, 

 where the flow in the joints is retarded, may be reached; an unusually 

 impervious layer of limestone or shale may be present; or what is prob- 

 ably most important, a level corresponding to that of the local base level 



