THE NONMETALLIO MINERALS. 395 



Aocordino^ to the reports of the State ^eolooist' this eartli, in its 

 air-drv coiulition, has the following' composition: 



Loss at red heat 1(). 50 



Silit-a 20. GO 



Ferric oxide (3. 03 



Maiitjaiiic oxide 0. 75 



AluiniiKi 20. 40 



Lime 8. 06 



Magnesia 4. 58 



Carl)onie acid 10. 38 



Sulphurie acid 6. 55 



Phosphoric acid 2. 43 



Nitric acid 3. 50 



Chlorides of alkalies and loss 0. 32 



100. 10 



The researches of Muntz and Marcano '^ have shown that the soils as 

 well as the earth from the floor of caves, in Venezuela and other por- 

 tions of South America may be rich in calcium nitrate to an extent 

 quite unknow^n in other countries. 



Origin. — The source of the nitrates, both of caves and of the Chilean 

 pampas has been a subject of considerable discussion. There appears 

 little doubt but the deposits in caves and those disseminated in soils 

 are due to the nitrifying agencies of Ijacteria acting upon organic matter 

 whereby the organic nitrogen is converted into nitric acid which inmie- 

 diately combines with the most available l)ases, l)e the}^ of lime, soda, 

 or potash. The accumulation of the niter in caves is probably due, as 

 suggested })y W. H. Hess (see Bibliograph}^), to the retention by the 

 clay of the nitrates brought in from the surface by percolating waters. 



In other words, the caves serve merely as receptacles, or store- 

 houses, for nitrates which had their origin in the surface soil. The 

 Chilean nitrate beds are considered by Muntz and Marcano as having a 

 very similar origin. The material being soluble is gradually leached 

 out from the soils in which it originated and drained into inclosed 

 salt marshes or inland seas where a double decomposition takes place 

 between the sodium chloride and calcium nitrate, whereby sodium 

 nitrate and calcium chloride are produced. That such a double 

 decomposition may take place has been shown by actual experiment. 



This is not widely different from the view taken also by W. Newton.^ 



After discussing ])riefly theories previously advanced including 

 Darwin's theory of derivation from decomposing seaweeds accumu- 

 lated on old sea beaches, and the even less plausible one of its deriva- 

 tion from guano, he goes on to show that the plain of Tamarugal 



' Geological Report of Indiana, 1878, p. 163. 



^Comptes Rendus de I'Academie des Sciences, CI, Paris, 1885, p. 1265. 



HJeological Magazine, III, 1896, j). 339. 



