118 REGINALD A. DALY 
in admitting assimilation on the great scale, yet believe a small 
amount of syntexis to be capable of upsetting equilibrium in a 
primitive magma and so initiate its marked differentiation. These 
authors are so fortified with reasons for their faith that it is clearly 
expedient to keep an open mind concerning syntexis and to define 
“differentiation” accordingly. 
Absence of fractional crystallization in most basaltic sills. — 
Another general objection to the pure-fractionation hypothesis is 
that it seems to prove too much. Numberless basaltic (diabasic, 
gabbroid) sills show by their wide extent that they were compara- 
tively fluent during injection. Presumably the interior portion of 
each sill of notable thickness remained fluent for some time (hours, 
days, or weeks) after injection. During this interval the magnetite, 
olivine, augite, or calcic plagioclase of early crystallization would 
sink if the basalt regularly behaved like Bowen’s crucible melts, on. 
which he so largely bases his theory. Strong differentiation through 
the sinking of crystals was observed in his artificial melts after 
these had stood only a few minutes or at most a couple of hours. 
If, therefore, the experimental analogy were good, the average 
basaltic sill, five to fifty meters thick and hence slowly chilled, 
should exhibit differentiation by gravity. According to Bowen’s 
argument, diorites or even quartzose phases might be regularly 
expected near the roofs and ultra-femic phases near the floors. 
These deductions do not match the facts; the chemical homo- 
geneity of most basaltic sills is almost perfect. The gravitative 
differentiation of basaltic magma is manifestly slow and, in a sense, 
difficult. Very thick sills, like that of the Palisades of New Jersey, 
may preserve magmatic life long enough for an appreciable sinking 
of specifically heavy crystals toward the bottom, but in general 
it looks as if some “‘contamination”’ were necessary before ordinary 
injected basalt breaks up into contrasted phases. The ‘“‘con- 
taminating”’ materials may be either exotic juvenile gas or resurgent 
gas or liquid. 
Comparative homogeneity of femic phases in differentiated sills 
and laccoliths —Where gravitative differentiation has taken place 
in sill or laccolith, the femic phase is commonly rather uniform 
from top to bottom, except at the layer transitional to the over- 
