THE NORM IN IGNEOUS ROCKS 83 



ilmenite (il), fluorite (ft), sodalite (so), noselite (no), and calcite 

 (cc), at the outset. (See note at end of paper.) The formula for 

 sodahte is 3(Na20.Al203.2Si02).2NaCl and that for nosehte is 

 2(Na20.Al2 03.2Si02)Na2S04. Sodahte is therefore hkethe nephehte 

 molecule taken three times with 2NaCl added. Where CI occurs in 

 notable amount with P2O5 present it is first allotted to apatite (see 

 Analysis I), and the remainder then takes half as many units of 

 Na^O for 2NaCl in sodalite. One-half the number of units of CI is 

 the key. By multiplying this number, .010, by 969 (the molecular 

 weight of sodalite), the percentage weight of sodalite (so) is obtained. 

 Where SO3 is present it takes an equal amount of NagO for Na^SO^ 

 in noselite (no). The number of units of SO3 is the key. By multi- 

 plying this number, .009, by 699 (the molecular weight of noselite), 

 we get the percentage weight of noselite. In this calculation the 

 nephelite molecules in sodalite (15), and in noselite (9), have been 

 lumped in with nephelite. CO 2 when present in a rock usually points 

 to a weathered condition in the rock, the mineral then being present 

 as an alteration product. Where CO2 occurs in a rock and is not a 

 product of alteration, calcite is an original mineral constituent. In 

 Analysis O we allot to 6 units of CO^ as many units of CaO for calcite 

 (cc), CaO.COj, the ratio between CaO and CO^ being 1:1. 



In the foregoing examples of calculation a shortage of SiO^ was 

 met by distributing Na2 0.Al2 03 between albite and nephelite, after 

 making orthoclase with K2O.AI2O3; or the shortage was provided 

 against, after making nephelite, by distributing K2O.AI2O3 between 

 orthoclase and leucite. With SiOa too low for either of these alter- 

 natives we may allot K2O.AI2O3 to leucite and Na2 0.Al2 03 to 

 nephelite, using up in this way a relatively small amount of SiO^. 

 This is the procedure in Analysis P, where we make leucite, nephelite, 

 anorthite, magnetite, diopside, and olivine, only to find that we have 

 run over by 74 units of SiO^. In making diopside we used 302 SiO^. 

 If now we take CaO from diopside, turn the MgO and FeO thus set 

 free into more olivine, and use the lime (with the requisite amount of 

 SiO^) in ackermanite (am), 4Ca0.3Si02, which by its formula 

 uses up less SiO^ for the same amount of CaO than diopside does, we 

 can do away with the SiO^ deficit. When the ratios of CaOiSiOj in 

 diopside, 1:2 (or 4:8), and in ackermanite, 4f3, are considered, it 



