August 2, 1907] 



SCIENCE 



139 



eentration of the open ocean. As the area 

 of sedimentary rocks increases relatively 

 to that of igneous rocks the character of 

 the supply of salts furnished by the rivers 

 must change.^ On the whole accumulation 

 was probably more rapid relatively to the 

 area of land surface, at early dates, when 

 evaporation was more powerful. It is 

 doubtful if the land was so well forest 

 clad and protected from erosion as later. 

 Moreover, salts would be less likely to be 

 thrown oiit when the ocean was weaker and 

 not so saturated. 



The broad conclusion of a gradual ac-^ 

 cumulation of salt in the sea, particularly 

 sodium, is confirmed by so many inde- 

 pendent lines of evidence as to have a very 

 small probable error. 



Let us review the arguments for accumu- 

 lation of sodium before leaving this part 

 of the subject. 



1. The argument from the erosion of 

 continents developed by Mackie. 



2. The argument from the excess of 

 sodium in river water, worked up recently 

 by Dubois after Murray and Joly. 



3. The argument from the shortage of 

 sodium in the average sedimentary rock 

 as compared with the average igneous rocks 

 from which it is derived, as developed by 

 Van Hise, Clarke and Mead. 



4. The argument which may be drawn 

 from a study of the early buried waters as 

 suggested by Hunt. 



5. And finally the argument which as 

 "we have seen follows from the composi- 

 tion of the vital fluid and Quinton 's law. 



These various arguments depend on vari- 

 ous lines of facts, so that one may be 

 fallacious without disproving the other. 

 On the other hand, the stronger lend 

 strength to the weaker. For instance, I 



' In early times the supply of chlorine seems 

 to have been relatively more rapid, being derived 

 from the leaching of the Keewatin rocks. 



doubt if any one from mineral water 

 analyses alone would be likely to feel much 

 assurance. The errors and defects in the 

 chemical work, in the collection of samples, 

 and in the circulation and preservation of 

 the waters, are too great. 



The one thing that seems most assured 

 regarding the buried waters is that the 

 early ones were relatively richer in calcium 

 chloride. 



The many published, and dozens of yet 

 unpublished, determinations upon our 

 Michigan waters, have put that in my mind 

 beyond question. 



The analyses of the coastal plain post- 

 Paleozoic waters deserve farther study. 

 The saltest well of the analyses recently 

 collected by Smith in Alabama from the 

 Tertiary has 30.5 parts per thousand with 

 11.47 sodium against 18.52 CL, while the 

 ratio of magnesium to chlorine is .1224: 

 .247. 



This is not far from what we might ex- 

 pect from a buried sea water of this age, 

 but on the other hand among the waters 

 from the Cretaceous there is none that ap- 

 pears to be at all unmixed sea water. The 

 saltest, and at the same time one of the 

 deepest (IVa) has only 8.57 per thousand 

 and Na:Cl::2.999:4.538 = 0.66, while Mg: 

 Ca::.0431:.1396. 



The analyses collected by Veatch in 

 Louisiana, where we know that rock salt 

 beds occur, show in many cases greater con- 

 centration of salt than the present ocean. 

 They also show very little of sulphates, but 

 they show a ratio of magnesium to calcium 

 quite different from the Saginaw brines 

 which mark them as relatively recent. 



Leaving now the tedious numerical part 

 of our subject, we may ask ourselves what 

 bearing this has upon the difficulty to 

 which we referred at the beginning. 



The vital fluid seems to date or preserve 

 the ocean composition, concentration and 



