August 2, 1907] 



SCIENCE 



135 



or less fresli water? For freshwater 

 springs, as Hitchcock and others have de- 

 scribed them, are not uncommon beneath 

 the ocean, and very commonly fresh water 

 may be found by digging in the sand of 

 ocean beaches. 



Such an inquiry would have been im- 

 possible in Darwin's time. But since his 

 time so many holes have been put down to 

 very great depths by churn or diamond 

 drill, in search of oil, gas, brine, artesian 

 water, etc., that it does seem fair to ask the 

 question. Yet we can hardly separate in 

 the inquiry the composition from the con- 

 centration. For in the proportion of the 

 elements, especially of the chlorine, seems 

 to be the stirest test of the character of the 

 water. I shall not try to prove this in 

 detail, as it depends on study of many 

 water analyses, but I may give you some 

 idea of the reason. Chlorine (CI) now ex- 

 ists in ocean water in excess of sodium, and 

 there is, as we shall see, good reason for 

 believing that that excess has been even 

 greater than now in times past. 



By an excess of chlorine I mean that 

 after chlorine has been set aside enough to 

 combine with all the sodium that the an- 

 alysis of a water shows, there is still chlo- 

 rine left which is usually understood to be 

 combined with the potassium, calcium and 

 magnesium present. 



On the other hand, in waters that are 

 mineralized by leaching, whether from 

 granites of New England or the alkali 

 plains of the west, and in all rivers with 

 very few exceptions, explicable often by 

 contamination with manufacturing wastes, 

 there is sodium enough to combine with CI 

 and then some left. There is hardly any 

 rock, but the unique Stassfurt deposit from 

 which one can imagine chlorides other than 

 sodium chloride to be leached. Nor, on 

 the other hand, is chlorine in solution easy 

 to get out again. The rare and valuable 



silver ore, hornsilver, a few significant vol- 

 canic chlorides, quite soluble, and transient 

 around active volcanoes, and the deposits 

 of dried up oceans and lakes, and a few 

 minerals of no quantitative importance like 

 apatite and sodalite, complete the list of 

 chlorine minerals. 



To get results of any value, one must 

 take all the available analyses of a region, 

 see what the surface waters are, how they 

 seem to be affected in composition as one 

 follows them in depth, eliminate waters 

 that may come from desiccated seas, gyp- 

 sum beds, or salt beds, or may be erroneous 

 and contaminated. Gradually I find my- 

 self forming some idea of what the buried 

 water must have been. I miist frankly 

 own that from my studies in that state 

 with which I am best acquainted, Michigan, 

 I was first inclined to think that there was 

 no sign of earlier, weaker ocean waters. 

 On the contrary, deep wells from perhaps 

 every geological horizon seem to have 

 stronger brines than the ocean. 



Upon taking a wider view I have come 

 to change my mind, and think it is not alto- 

 gether accidental that the water at She- 

 boygan, Wisconsin, has at a depth of 1,340 

 feet 10 parts per thousand of solids, of 

 which 4.3 are CI; that the deepest well at 

 Cincinnati, 1,245 feet, has 11 parts per 

 thousand, of which 6 are CI; that Litton 's 

 very careful analysis of the water from a 

 well 2,200 feet deep at St. Louis (the water 

 mainly at 1,515 feet in the Magnesian lime- 

 stone) should give 8.791 per thousand, of 

 which 4.1 are chlorine, and that near us at 

 Montreal in rocks of similar age in a 1,500- 

 foot well (at 1,190 feet) Adams found 7.57 

 per thousand, of which 2.46 are chlorine. 

 In Europe the only well at all similar I 

 have yet found is one reported by Struve 

 at St. Petersburg (658 feet through Silu- 

 rian rocks to the granite) with 3.89 per 

 thousand, of which, however, 2.3 are chlo- 



