B72 HENRY S. WASHINGTON 
the actual withdrawal of the amount of silica entering the albite 
molecule from the possibility of chemical combination with potash, 
that is, by the early crystallization of the sodic feldspars, at least 
in great part, thus bringing about the formation of leucite at a later 
stage. Such a withdrawal of silica by soda is rendered possible by the 
fact that labradorite has a greater tendency toward crystallization 
than have orthoclase and leucite,* as shown by the general truth of 
Rosenbusch’s order of crystallization as far as regards labradorite, 
orthoclase, and leucite, which immediately concern us. Specific 
instances of this are presented by the Italian lavas, in the mantles 
of orthoclase about cores of anorthite and labradorite,? which have 
also been observed elsewhere, and in the inclusions of labradorite 
in large leucite phenocrysts.’ 
The objection may be raised to this explanation that the more 
calcic members of the soda-lime feldspar series, those with more 
anorthite than Ab,An,, contain less silica than does leucite, and that, 
furthermore, it is precisely these more calcic members, especially 
labradorite, which are most apt to occur in connection with leucite, 
as is shown in the case of the Italian leucitic rocks. It might therefore 
seem to follow that the early crystallization of labradorite (Ab,An,), 
with a silica percentage of only 51.4 less than that of either leucite 
or orthoclase, would bring about in the remaining portion of the 
magma an increase of the silica percentage relative to that of potash, 
so that the early separation of labradorite ought to tend to prevent, 
rather than induce, the crystallization of leucite. Such an objection, 
however, leaves out of account the relations of the soda to the silica, 
and overlooks the fact that in entering albite molecules at an early 
stage of the crystallization it binds more silica than it would do if 
it were not thus removed and crystallized later as nephelite. 
From the usually late crystallization of the alkali-feldspars it 
t Cf. C. Doelter, Petrogenesis (1906), p. 136. 
2H. S. Washington, Journal of Geology, Vol. IV (1896), p. 549; The Roman 
Comagmatic Region (1906), p. 30. 
3H. S. Washington, The Roman Comagmatic Region (1906), pp. 34, 91. Cf. E. 
Hussak, Neues Jahrbuch (1890), Vol. I, p. 168. It may be mentioned that labradorite 
is not listed as an inclusion in leucite by either Rosenbusch (Mikroskopische Physio- 
graphie, Vol. I, second half [1905], p. 32) or Iddings (Rock Minerals [1906], p. 
242). 
