WATER OF STEAMBOAT SriilNGS. 349 



that a part of tlie sodium salt was neutral. This supposition also accords 

 with that made with reference to the silica The alumina was probably 

 present as an alkaline aluminate, but the (|uantity found was so small that 

 it did not appear worth while to compute its hypothetical compounds. In 

 Chapter XV it will be shoun that the trace of mercury in this water is 

 combined as a double sulj)hide with sodium, wh.ich is of the form IlgS, 

 wNa^S. The value of n in tliis case is probably four. No comments seem 

 needful on the state of combination of the other constituents of this water. 

 The suppositions made lead to the following scheme of composition as 

 the most probable prior to the action of the atmosphere u[)on the fluid : 



Probable comjwuilion of the water j/rior to oxidation. 



Grams. 



Ferrous carbouate, FeCO' 0. 0029 



MagueBiiim carbonate, MgCC 0.0099 



Calcic carbonate, CaCO'' 0.11577 



Calcic pbospbate, Ca^'P-O" 0.0137 



Potassic chloride, KCl 1. 9735 



Lithic sulphate, Li-SO* 0. 5(J50 



Sodic chloride, NaCl 14.1475 



Sodic sulphy drate, NaHS 0. 0358 



Sodic sulphate, Na^SO' 1. 1147 



Sodic bicarbonate, NallCO' 2.9023 



Sodic nionocarbonate, Na-CC 0.4314 



Sodic biborate, Na^B^C 3. 13G8 



Sodic quadrisilicate, Na-Si^O' 3. 9090 



Sodic sulphautimonide, Na-.SbS^ 0.0100 



Sodic sulpharsenide, Na'AsS^ 0.086G 



Alumina, AW 0.0025 



Sodium-mercurj- sulphide, HgS, «Na-S Trace 



Origin of the water. — Old residcuts iufomied me that the (piantity of water 

 flowing from the springs varies from year to yetjr, being greater in yeai-s of 

 heavy rain-fall than in dry seasons and greater in spring tlian in autumn. 

 If these statements be accurate, the supply must come from the surface 

 and no very long time can intervene between precipitation and return to 

 the surface. It is natural to supj)0se the great snowy range to be the source 

 of supply. The fissures underlying the range may afford a downward pas- 

 sage for the waters to the heated mass from wliich the basalt came, while 

 the fissui-es associated with the channel through which the basalt was ex- 



