454 Scientific Intelligence. 



type, in tbat the whole interval from the Cambrian (or rather the 

 Saint Peter) sandstone to the Carboniferous is not very thick. 

 Thus very many of the deeper Missouri wells may draw both from 

 the Calciferous and the Lower part of the Pennsylvanian. The 

 Calciferous or St. Peter is, says Shepard, the most important 

 water bearer, and is classed as Cambro-Ordovician. 



This relatively thin column between the St. Peter and the 

 Coal-measures is in part due to one or more unconformities. 

 These imply that Missouri was out of water during long periods 

 of the Paleozoic. We also find that there are considerable local 

 distui'bances, so that, for instance, the beds from which the deep 

 St. Louis wells draw their water come to the surface within a 

 few miles. Thus both during Paleozoic times and more recently 

 considerable circulation may have taken place. Missouri, then, 

 is not a very good state from which to draw safe conclusions 

 as to buried waters. Nevertheless Shepard also notes indications 

 of " fossil brines." 



I have arranged the analyses in order according to the total 

 solids. The strongest water given has a little over two per cent 

 solids. I have then (tig. 1) drawn a line from an abscissa X 

 corresponding to this concentration down to an arbitrary ordi- 

 nate. Opposite points on this line corresponding to the strengths 

 of the various waters I have drawn lines connecting an abscissa 

 corresponding to the chlorine with one corresponding to the 

 sodium for the same analysis. If the waters were dilutions 

 of one water, or simply compounds of two waters, straight lines 

 could be drawn through the ends of these lines. 



With the chlorine this is nearly but not absolutely true, lead- 

 ing one to infer that there is one strong water, strong in solids 

 and also strong in chlorine, and that the waters with which it 

 has been diluted have relatively almost insignificant amounts of 

 chlorine. 



With the sodium this is not so much so. The ratio of sodium 

 to chlorine is quite variable, sometimes more and sometimes less 

 than in salt. In the sodium there are at least three types of 

 water that must be assumed, — a saline one with over 2'5 per cent 

 solids, a fresh one practically pure, and an alkali water such as is 

 characteristic of the far west and is best represented by the 

 Haver well, p. 69, strong in sodic sulphate or carbonate. Other 

 fairly weak wells show the same predominance of sodium. 



These arid land waters are factors which I have hitherto neg- 

 lected. If the stronger analyses were derived by solution of salt 

 from a weaker buried ocean water containing the calcium chlor- 

 ide, we might expect the chlorine to form a larger proportion of 

 the total solids in the weaker brines. Of this there is little or 

 no sign. 



On the whole, therefore, the Missouri analyses, if compatible 

 with the theory of the chemical evolution of the ocean at all, 

 would indicate a greater degree of concentration in the Paleozoic 

 than the figures in my address of last July, and hence much 



