Mineralogy of South America. 141 



of the Pacific Ocean ; and previously to having attained this ele- 

 vation, it may be supposed to have been occupied by a chain of 

 shallow lagoons broken up by promontories and islands ; and 

 these probably had for a long period tidal communication with 

 the main ocean, until, by the still greater upheaval of the whole 

 country, they became isolated from the sea and dried up, leaving 

 their original sea-bottoms covered by vast beds of salt, more or 

 less mixed with sand, and which are productive in nitrate of soda 

 on the slopes which now bound the present valleys and plains 

 and encircle the hills which represent the islands, bays, and 

 indentations of the ancient lagoons. 



Most descriptions of these nitrate-of-soda deposits leave the 

 reader under the impression that they occur like other sedimen- 

 tary strata, in more or less regular layers or beds of salt, sand, 

 clay, or gypseous marls, &c. This, however, is quite erroneous. 

 The nitrate of soda does not represent any bed in a series, but 

 in reality is only the representative of the outcrop of a bed of 

 salt (with which it is always more or less contaminated), or, in 

 other words, the bed of salt left on the bottom and sides of the 

 ancient lagoons, after their complete desiccation, becomes, at the 

 edge of the sloping shore or bank, more or less converted into 

 nitrate of soda by chemical action. In the centre or lower parts 

 of the salt-bed, nitrate is never found ; at the outcrop it is gene- 

 rally, if not always, found in more or less quantity — sometimes 

 almost on the surface itself, or covered by a few inches of sand, 

 but generally concealed by the upper part of the salt-bed being 

 still unaltered salt, and forming, as it were, a hard crust (the 

 " crosta " of the miners) over the soft white impure nitrate of 

 soda. 



It has been advanced that the nitrate of soda was the product 

 of the volcanic action developed in the higher mountains to the 

 east of the Pampa de Tamarugal, and that it had been dissolved 

 out and washed down into these lower valleys and plains, where 

 it had been deposited and dried up. 



All the circumstances of the locality, however, tend to reject 

 this hypothesis ; and surely in that case we ought to find the 

 nitrate deposits in the basins or lower parts of the valleys and 

 plains ; which is never the case in reality, as such low deposits 

 invariably consist of salt, alone or mixed with the gypseous com- 

 pounds usual to dried-up sea- water, whilst the nitrate-of-soda 

 deposits are found higher up, on the slopes of the hills which 

 surround such basins. 



A very attentive study of these deposits has quite confirmed 

 the view as to the origin of the nitrate of soda, expressed in 

 the report previously alluded to, and which supposes the nitrate 

 of soda to have been formed at the margin or fringe slopes of 



