EARTH'S CRUST — WASHINGTON. 277 



2 to nearly zero. Of the many rocks of all kinds that I have analyzed, 

 there has not been a single one that did not contain titanium, in some 

 cases in very small, but always in easily determinable, quantity. This 

 is also the experience of Doctor Hillebrand, 7 and probably of every 

 other experienced analyst of rocks. 



The maximum for phosphorus pentoxide (P 2 5 ) is but a little above 

 16 per cent in some highly unusual rocks from Sweden and Virginia, 

 that are composed largely of apatite, with titaniferous magnetite or 

 rutile. In few rocks, however, is it above 3 per cent, and its general 

 range is from about 1 per cent to zero. It does not seem to be present 

 so constantly as titanium (or manganese), as one occasionally meets 

 with a rock that shows no trace of it, though this may be because 

 of the more delicate tests for the other two. 



Manganese, as manganous oxide (MnO), is present in practically 

 every rock that has been analyzed, but its maximum is much lower 

 than those of titanium and phosphorus oxides. Some, if not most, 

 of the high figures reported for it are almost certainly due to analyt- 

 ical errors, and the highest recorded figures that are trustworthy are 

 1.90 and 1.46 in two rocks from Bahia, Brazil. Its general range is 

 from 0.3 per cent to nearly zero. 



The other minor constituents that are readily determinable, and 

 many of which are indeed determined in good analyses, are quite 

 varied. The list is as follows: Carbon dioxide (C0 2 ), zirconia 

 (Zr0 2 ), chromium sesquioxide (Cr 2 3 ), vanadium sesquioxide 

 ( V 2 O s ) , the " rare earths " ( (Ce, Y) 2 3 ) , nickel oxide (NiO) , strontia 

 (SrO), baria (BaO), lithia (Li 2 0), sulphur as both sulphide (S) 

 and sulphur trioxide (S0 3 ), chlorine (CI), and fluorine (F). To 

 these might be added boron, cobalt, copper, glucinum, lead, molybde- 

 num, nitrogen, and zinc, which, however, are present almost always 

 in such extremely small amounts, or the analytical difficulties are so 

 great for the separation of the small quantities in which they occur, 

 that their determination is rarely attempted. 



The maxima and ranges of some of these may be briefly stated. 

 Carbon dioxide may be present, as a component of a few minerals 

 (as in primary calcite and cancrinite), in some unaltered rocks; but 

 its presence is generally due to alteration. In one calcite trachyte 

 from Spain its amount is 7.69 per cent, and in cancrinite rocks it may 

 reach about 1.70, the carbonate minerals in these being apparently 

 primary. But it is generally considered as a measure of the alteration 

 of the rock by weathering. 



Zirconia is much less abundant than the closely related titania 

 and, though it reaches a maximum of nearly 5 per cent in some Green- 

 land rocks, in general it seldom is over 1 per cent, is usually much 



'Hillebrand, W. h\, U. S. Geol. Survey Bull. No. 700, p. 25, 1919. 



