THE NONMETALLIC MINERALS. 237 



been made with regard to the relationH between the ore and the country-roc-k ; and 

 nothing has yet been obnerved which warrants the conchision that the ore if fol- 

 lowed to sufficient depth, will be found inter-bedded with the underlying' forma- 

 tions, or even that it will be found occupying cavities in the limestone— although the 

 latter is quite possible. 



Concerning the origin of these deposits the author says: 



No eruptive rocks, either ancient or modern, are found in the vicinity of the 

 latter, nor are there any rocks in this region which, by weathering, could yield 

 bauxite as a residual product. Hence, any satisfactory explanation of the origin of 

 these deposits must give the source from which the material was derive<l, the means 

 by which it was transported, and the process of its local accumulation. 



As already stated in describing the stratigraphy of the region, the ore is associated 

 with the Knox dolomite or with calcareous sandy shales immediately overlying the 

 dolomite. The Connasauga, consisting of 2,000 feet or more of aluminous shales, 

 invariably underlies the dolomite at greater or less distance beneath the ore-bearing 

 regions, and is probably the source from which the alumina was derived. 



The faults of the region have been briefly described. Undoubtedly such enormous 

 dislocations of the strata generated a large amount of heat. The fractures facilitated 

 the circulation of water, and for considerable periods the region was probably the 

 seat of many thermal springs. These heated waters appear to have been the agent 

 by which the bauxite was brought to the surface in some soluble form and there 

 precipitated. 



The chemical reactions by which the precipitation was effected are not well under- 

 stood, and the conditions were not such as can be readily reproduced in the labora- 

 tory. Of the few soluble compounds of aluminum which occur in nature, only the 

 sulphate and the double sulphate of jjotash and alumina need be considered. 



The oxygen contained in the meteoric waters percolating at great depths through 

 the fractured strata would readily oxidize the sulphides disseminated in the aluminous 

 shales. Sulphates would thus be formed by a process strictly analogous to that com- 

 monly employed in the manufacture of alum. Probably the most abundant product 

 of the process in nature was ferrous sulphate. Some sulphate of aluminum must also 

 have been formed together with the double sulphate of potassium and aluminmn, 

 especially in the absence of sufficient potash to form alum with the whole. 



In its passage from the underlying shales through several thousand feet of dolo- 

 mite the heated water must have become highly charged with lime, in addition to the 

 ferrous and aluminous salts already in solution. But calcium carbonate reacts upon 

 aluminum sulphate and to some extent also on alum, forming a gelatinous or floc- 

 culent precipitate which consists of aluminum hydroxide and the basic sulphate. 

 This reaction may have taken place at great depth and the resulting flocculent pre- 

 cipitate may have been brought to the surface in suspension. From analogy with 

 pisolitic sinter and travertine now forming, such conditions would appear to be 

 highly favorable for the production of the structures actually found in the bauxite. 

 The precipitate was apparently collected in globular masses by the motion of the 

 ascending water, and constant changes in position permitted these to be coated with 

 successive layers of more compact material. Finally, after having received many 

 such coatings, the pisolites were deposited on the borders of the basin, and the 

 interstices were filled by minute oolites formed in a similar manner or by the floc- 

 culent precipitate itself. Slight differences in the conditions prevailing in the sev- 

 eral springs, such as concentration and relative proportion of the various salts in 

 solution, also temperature and flow of the water, would produce the variation in the 

 character of the ore observed at different points. 



The bedding observed in the bauxite-deposits may have been produced by the 

 successive layers deposited on the steeply inclined outlet of the basin. After the 



