94 



portions are nitrogen 67, oxygen 33, or about 2 to 1. Atmos- 

 pheric air containing these gases in the proportions, nitrogen 80, 

 oxjgen 20, or 4 to 1. 



The explanation j'ou will usually find is, that the atmospheric 

 air which is naturally contained in spring water, as derived from 

 rain water, has been deprived of its oxygen by various substances 

 in the earth, but as you cannot take out of a basket more than is 

 put in, and as these waters naturally evolve rather more than the 

 fourth of a cubic inch of nitrogen per pint, and still retaiir in 

 solution a certain amount of this gas, and originally rain or spring 

 water does not contain more than the fourth of a cubic inch of 

 the mixed gases, I do not see how this can account for the 

 observed result ; if it be so, the part is greater than the whole. 



The water supplied by rain to the earth contains not only 

 oxjgen and nitrogen, Anth carbonic acid, but also ammonia, that 

 is, nitrogen plus hydrogen 3, and the earth gives more. Assuming 

 that the water comes from about Bitton, and flows through the 

 Carboniferous strata, we know that coal is constantly giving off 

 both carbonic acid gas, sulphuretted and carburetted hydrogen, 

 therefore we should find these gases besides ammonia. 



Let us first account for the ammonia. I find by experiment 

 that if chloride of sodium, with ammonia, be exposed to the action 

 of carbonate of magnesia dissolved in water, with carbonic acid 

 gas, a transformation takes place, and as a result we get chloride 

 of sodium, carbonate of soda, chloride of magnesium, and the 

 ammonia disappears, leaving only nitrogen gas. 



The rationale I take to be this — 



The carbonic acid gas in the water dissolves the carbonate of 

 magnesia, and then appears in the water in the form of bi-car- 

 bonato of magnesia, an unstable salt; this being brought into 

 contact with chloride of sodium, from the great affinity of chlorine 

 for magnesia, and of soda for carbonic acid, gives rise to a doul)lo 

 decomposition, and chloride of magnesia and bi-carbonate of soda 

 are formed, but at the moment when the decomposition takes 



