values (negative slope). No correlations are observed for MI versus MnO and Sr. Figure 5 is a stacked plot of the elements that best correlate with the MI. Trends with positive slopes indicate that the element was not incorporated into the phases that fractioned from the magma, that is, the element was incompatible. Compatible elements (e.g., CaO and Ni) have negative slopes and were thus removed from the magma during fractionation.
The trends shown on Figures 4 and 5 suggest that the data represent a single suite. Some of the data which constitute trends on Figures 4 and 5, as well as subsequent diagrams, may prove to represent portions of other trends not yet delineated due to limited data.
96.0
88.0 FGS1
X
a MPI3 *FGS7 Z 80.0 
*FGS8
FGS2 FGS6 ,,.,.., -- MP12
Ll. / *MG6
72.0I HFQ FGS121
72.0 FGS14 FGS13 FGS3 I
FGS9# GS4
- 'OFGS10
/ t/1 64.0 I
J LTQ / HT FGS11
i, ,
G5 0, i
56.0
0.40 0.80 1.20 1.60 2.00 2.40
TI02 (wt.%)
Figure 4. Mafic index (Fe203*/Fe2O3* + MgO) plotted versus TiO2 (weight percent). Modified from
Weigand and Ragland (1970). HTQ is high-TiO2, quartz-normative; HFQ is high-Fe203,
quartz-normative.
14