278 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



less, and is quite absent from most rocks. It forms, by the way, one 

 of the most striking illustrations of the correlation of the occurrence 

 of different elements in different kinds of rocks, as will be brought 

 out later. 



Baria and strontia are very commonly present, though they are 

 seldom determined in analyses made outside of the United States, 

 Canada, and Australia. In almost every case the amount of baria 

 is much greater than that of strontia, this being an exception to a gen- 

 eral rule as to the occurrence of related elements, to be mentioned 

 later. They both reach their maxima in certain exceptional, highly 

 potassic rocks of Wyoming, of about 1 per cent for baria and 0.3 

 for strontia ; though usually baria is present up to but a few tenths 

 of one per cent, and strontia in hundredths. 



Sulphur, as sulphides, is present up to about 9 per cent in a peculiar, 

 pyrrhotite-bearing rock from Maine, and probably in similar amounts 

 in some sulphide ore bodies of magmatic origin in Norway, which 

 have not been fully investigated. But, as a rule, its amount is seldom 

 over 1 or 2 per cent, and is usually in tenths of a per cent. The 

 highest figures for sulphur trioxide are about 2.5 per cent in rocks 

 from Apulia and Kamerun, and somewhat lower on Tahiti, but 

 these are exceptional, and it is usually present only in tenths or 

 hundredths of a per cent. Much the same can be said of chlorine, 

 the highest figures for which are those of a rock from Turkestan 

 (about 7), one from Quebec (4.47), and one from French Guinea 

 (2.80). It is present in many rocks, especially lavas, but only in a 

 few tenths or hundredths of a per cent. 



Chromium sesquioxide is known to be present up to about 4 per 

 cent in some ores from Greece, which are probably of magmatic 

 origin, and is reported as between 2 and 3 per cent in some undoubt- 

 edly igneous rocks from Baden. But these are highly exceptional, 

 and about 0.5 may be taken as its usual maximum. It is generally 

 entirely absent. Vanadium sesquioxide is always present in much 

 less quantity and is usually quite absent. The oxides of the rare 

 earth metals, chiefly ceria and yttria, reach a maximum of 1.79 in a 

 rare type of rock from Madras, 0.6 in one from Sweden, and 0.4 in 

 one from the islet of Rockall, but the usual maximum is only one 

 or two tenths of 1 per cent. They are less often determined than 

 they should be. Nickel oxide is present in some rocks up to about 

 0.2 per cent. The maximum amount of each of the other minor con- 

 stituents may be placed at not over 0.5 per cent, and they are almost 

 always found only as one or two tenths, or still more often as 

 hundredths, of a per cent, or are absent. Indeed, for most of the 

 minor constituents the quantities usually yielded by analysis are so 

 small as to be significant only as to their actual presence or absence. 



