276 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



range from about 20 to about 10.- It is wholly absent only in the 

 " magmatic " ores and in some rocks that are composed entirely, or 

 almost so, of olivine. The two oxides of iron reach, of course, their 

 maxima in such rocks as the iron ores already spoken of; 6 the 

 highest figures recorded for Fe 2 3 being 88.41 (Sweden) and 62.39 

 (Ontario), while for FeO they are 34.58 (Sweden) and 32.92 (Min- 

 nesota). Their general range is from nearly 15 for each (FeO gen- 

 erally higher than Fe 2 3 ) and but little more than that for both 

 together in any one rock to less than one-half of 1 per cent. Iron is 

 seldom entirely absent. 



Magnesia reaches its maximum in the almost purely olivine rocks 

 (dunites) of North Carolina, 48.58, and of New Zealand, 47.38, but 

 its general range is from about 25 to much less than 1 per cent. 

 Lime is highest (22.52) in some pyroxenites of the Urals and almost 

 as high (about 20) in the anorthosites of Canada and elsewhere, but 

 it ranges in general from about 15 to nearly zero. 



Of the two alkalies, soda reaches a maximum of 19.48 in a rare 

 rock from Canada, and of 18.67 in another from Turkestan ; but its 

 general range is from about 15 per cent down to nearly zero. It is 

 hardly ever entirely absent. Potash shows a somewhat smaller range 

 than soda, its maximum being 17.94 in a recently discovered lava 

 from Italy, the next highest figure being 11.91, from Wyoming; but 

 in general it seldom gets above 10 per cent, ranging from that down 

 to zero. Its amount is generally less than that of soda. 



As regards water, the last of the major constituents, a few volcanic 

 glasses are known which, although perfectly fresh and undecomposed, 

 contain up to 8 or 10 per cent, and there are some fresh crystalline 

 rocks that contain from 3 to 5 per cent. Generally, however, if a 

 rock contains more than about 2 per cent of H 2 this can usually be 

 attributed to alteration, though few rocks are quite free from this 

 constituent. 



After the major come the "minor" constituents, which are almost 

 always present in very small amounts, seldom over 2 per cent for any 

 one, or rarely up to 5 per cent for all of them, in any one rock. Of 

 these minor constituents three are of special importance, partly be- 

 cause of their almost constant presence and partly because they are 

 generally present in largest amount. These three are titanium diox- 

 ide, phosphorus pentoxide, and manganous oxide, and all three should 

 be determined in a good rock analysis. 



Titanium dioxide (Ti0 2 ) reaches a maximum in some very rare 

 rocks from Virginia (69.67 and 65.90) and Quebec (53.35), but as a 

 general thing its percentage is seldom over 5, and is mostly from about 



8 It is a question whether all of these ore bodies are to be considered as really igneous 

 rocks, though some undoubtedly are. 



