GENESIS OF THE ALKALINE ROCKS 129 
have a specific gravity of about 2.54. Obviously the assumption 
that they would sink at all needs further justification. 
The hypothesis concerned faces objections on chemical grounds. 
Table II gives the average of the analyses of 20 typical basanites 
(col. z) and the corresponding averages for 20 quartz dionites (col. 2) 
and 12 granodiorites (col. 3). 
TABLE II 
I 2 3 
SIO FR roe separ 44.41 59-47 65.10 
Oe Waits Sea een ene 1.56 0.64 0.54 
INTE OM Sita pete weenie 15.81 16.52 15.82 
eLO Nae Renae 4.66 2.63 1.64 
1 el O Vipera are ee ees a 5.85 4.11 2.66 
Mn Ores saan 0.14 0.08 0.05 
JIN Tred OR ea al 8.20 Bes 2G) 
CaO Were en ass Ay LU 10.12 6.24 4.66 
INGOs ee aoa 3.81 2.98 3.82 
KOU ae ahs RNC Sa Di QO 1.93 2.20 
EEO RE Gay cnet ae 2.42 1.30 I.09 
BAO Sie OMA ui cveiorD cite 0.65 0.26 0.16 
100.00 100.00 100.00 
Inspection of the table shows that the settling-out of “‘plagio- 
clase crystals and femic minerals” as well as quartz could not 
produce a basanitic magma from that of either quartz diorite or 
granodiorite. For instance, the potash should be much higher in 
the basanite unless it be assumed that much biotite had settled out; 
but if it had, the iron oxides and magnesia should be proportionately 
diminished. Yet these components of the femic minerals are much 
more abundant than in either of the postulated parent magmas. 
Nor is it likely that the lime could be relatively so much increased 
in the mother-liquor after the manner suggested, that is, by the 
sinking (or rising) of plagioclase crystals and femic crystals. 
Absence of foyaitic types in most batholiths and stocks.—Again the 
theory suffers statistically as regards the occurrence of alkaline 
rocks. If, as Bowen holds, the cooling of a great body of granite 
magma should normally generate a foyaitic submagma as an end 
product, then all or most granite batholiths and most of the largest 
stocks should exhibit foyaitic phases. They do not. 
