CHEMICAL COMPOSITION OF RIVERS AND LAKES 

 Table 81. — Mean composition of river waters of the world, in parts per million 



G41 



' Millequivalents of strongly ionized components. 



provided cognizance is taken of the different way in 

 which he expressed bicarbonate. It is, however, very 

 close to the weighted mean of Conway (1942, 1943) 

 which was based on Clarke's data but took into account 

 the relative abundance of dilute tropical rivers. 



MINOR CONSTITUENTS 

 GENERAL REMARKS 



This section deals with constituents represented in so 

 few of the general tables that they demand separate 

 treatment. A number of chemical elements do not 

 seem to have been detected in a single lake or river 

 water. They are tellurium, all the noble gases except 

 argon and radon, indium, thallium, scandium, yttrium, 

 the rare earths, hafnium, germanium, columbium, 1 

 tantalum, tungsten, rhenium, the platinum and palla- 

 dium metals, and actinium. 



FLUORINE, BROMINE, AND IODINE 



Numerous data for fluorine and a few for bromine 

 and iodine are in the tables of general analyses. There 

 is an extensive body of information about the fluorine 

 and iodine content of lakes and rivers because of the 

 medical significance of these elements. There is less 

 information available about bromine. 



Most fresh waters have less than the single part per 

 million of fluorine which is regarded as optimal for 

 health of human teeth. In concentrated waters the 

 content may be somewhat higher, but it is usually 

 limited by the low solubility of calcium fluoride. 

 Many analj T ses include more fluorine than should be 

 dissolved in the presence of accompanying calcium, 

 and it is probably generally true, as Kobayashi (1954) 

 has found, that an appreciable part of the analytically 

 determined fluorine is not present in simple solution. 

 Waters unusually high in fluorine are commonly 

 associated with vulcanism, igneous rocks, or apatite 

 deposits. 



The reader seeking data on fluorine in addition to 

 those included in the general analyses may consult the 



1 1 am informed by Dr. Heinz Lofflcr that he has detected minute amounts of 

 columbium in East African high altitude lakes. 



following papers for various parts of the world: 

 Chamberlain (1946); Cherkinskii and others (1953); 

 Gabovich (1952); Gandra (1950); Kobayashi (1954); 

 Kredba and Hamackova (1950); Krepogorskii and 

 Bogusevich (1953); Kubota (1952); Mose and Exner 

 (1952); Novokhatskii and Kalinin (1953); Paraje 

 (1950); Richard and Vialard-Goudou (1954); Tageeva 

 (1943); Tomic (1951, 1954); Van Burkalow (1948); 

 Vinogradov, Danilova, and Selivanov (1937); Walker 

 (1940); Wilson (1954); Mackereth and Heron (1954); 

 Juday, Birge, and Meloche (1938); Konovalov (1959). 

 Konovalov's paper is of particular interest because it 

 permits computation of the mean fluorine content of 

 the rivers of about 80 percent of the entire area of the 

 U.S.S.R. at 0.089 ppm. 



Reliable information on the bromine content of lakes 

 and rivers is rather scarce. Correns (1956), reviewing 

 the geochemistry of the halogens, accepts the single 

 value of Behne (1953) for the water of the Grosse Bode 

 as the best available estimate of the bromine content of 

 the river water of the world, but the mean content of 

 U.S.S.R. rivers may be computed from the data of 

 Konovalov (1959) to be 0.019 ppm — more than three 

 times Behne's value. This is almost certainly closer 

 to the global total. There are many other analyses, 

 but most of them are for saline lakes or for rivers 

 influenced by rock-salt deposits or industrial sewage. 

 Some of these data are shown in table 82. From this 

 information it is not possible to tell with assurance 

 whether the CI/Br ratio of lake and river waters departs 

 significantly from the marine ratio of 294:1, although 

 it is likely to be slightly higher than this. 



More data are available for iodine than for bromine . 

 but they have been collected for medical purposes and 

 their geochemical usefulness is somewhat limited. An 

 assortment of data for iodine and bromine that can be 

 used for present purposes is included in table 82. 

 Hutchinson (1957) believes that 0.2 ppb would be a 

 reasonable mean figure for lakes and rivers. This 

 seems rather low. The mean of 1.8 ppb of Goldschmidt 



