G44 



DATA OF GEOCHEMISTRY 



0.0003 to 0.002 ppm, with a mean of 0.00116. The 

 Na/Rb ratios ranged from 1,525 to 11,100 and the 

 ratio of the mean values was 3,578. This is very 

 similar to the findings of Borovik-Romanova already 

 referred to, with an average concentration of 0.0016 

 ppm and a ratio of 4,166. The K/Rb ratio found by 

 both of these workers is slightly in excess of 1,000, or 

 more than 10 times as great as that found by Schmidt. 

 Twenty-seven samples of water from major rivers of 

 North America had a mean rubidium content of 0.0017 

 ppm (W. H. Durum, written communication, 1960). 

 Rubidium seems to be considerably scarcer in the 

 hydrosphere, from which it is removed biologically 

 and probably chemically as well, than it is in the 

 litkosphere. 



CESIUM 



The only analyses for cesium appear to be six deter- 

 minations for rivers in Japan by Yamagata (1951b). 

 He found a range of cesium content between 0.00005 

 and 0.0002 ppm. The Na/Cs ratio ranged from 9,300 

 to 89,400 and the ratio of the mean contents was 

 31,900. 



BERYIiliTOM 



Beryllium appears to have been determined only by 

 Maliuga and Makarova (1956), who found 10 ppm of 

 total dry residue in both the River Il'kikan and the 

 River Gazimura, and by the U.S. Geological Survey, 

 which found between 0.1 and 1 ppb in the Atchafalaya 

 River, Louisiana (W. H. Durum, written communica- 

 tion, 1960). 



STRONTIUM 



The strontium content of lakes and rivers has been 

 studied most extensively by Oduni (1950, 1951, 1957), 

 who found that the Sr/Ca ratio reflected the geologic 

 environment, at least in part. It was high in the 

 presence of evaporite deposits, pegmatites, volcanic 

 rocks, fresh coral limestones, and limestones precipitated 

 directly from sea water. Lower Sr/Ca ratios were 

 found in association with consolidated limestones, re- 

 placed limestones, dolomites, nonvolcanic mafic igneous 

 rocks, and humid climate. A selection of Odum's data, 

 together with those of several other authors, is pre- 

 sented in table 85. Bristol Dry Lake is a locality 

 where celestite concretions occur, and the strontium 

 content of the Bristol water sample, which was very 

 concentrated and came from a drainage ditch in the 

 lakebed, is probably close to the maximum to be 

 expected in lake waters. 



Additional information about strontium in water 

 may be found in table 19, in earlier editions of this 

 work, in the papers of Odum and Lohammar cited in 

 table 85, and in papers by Braidech and Emery (1935), 

 Borovik-Romanova, Korolev, and Kutsenko (1954), 



Maliuga and Makarova (1956), Grushko and Shipitsyn 

 (1948), Nichols and McNall (1957), Horr, (1959), 

 and Skougstadt and Horr (1960). 



BARIUM 



Bowen (1948) found 10 ppb of barium in water from 

 Linsley Pond, but could not detect it in hard waters 

 from Connecticut. Braidech and Emery (1935) found 

 larger quantities, between 30 and 1,000 ppb. The 

 element has also been determined by Grushko and 

 Shipitsyn (1948) and by Maliuga and Makarova 

 (1956). The global Ca/Ba ratio would be about 1,500, 

 accepting Bowen's figure as representative of lakes and 

 rivers, or between 15 and 500, accepting the results of 

 Braidech and Emery. The most representative set of 

 data appears to be unpublished: 34 samples from major 

 North American rivers had a mean content of 54 ppb, 

 suggesting a Ca/Ba ratio of about 400 (W. H. Durum, 

 written communication, 1960). 



