A TYPE OF IGNEOUS DIFFERENTIATION 645 
contrasting series in a single mass cooling in a single chamber, and 
a study of the literature of other districts makes it seem likely that 
this double sequence is a common thing. 
The association of both granophyr and anorthosite as differenti- 
ates of normal gabbro is too common to have escaped observation. 
Each rock requires large bodies of magma for its production and 
each is of low specific gravity. Several geologists discuss these 
rocks as normal differentiates of gabbro, but no one has developed 
a satisfactory explanation of the manner in which the two rocks 
form. To besure, a line of descent can be stated so that both are 
mentioned, but the suggested origin would lead one to expect them 
in very different field relations from those at Duluth. Thus, when 
peridotite, troctolite, and olivine gabbro had separated from the 
magma, the feldspar should begin to grow more acid and the olivine 
less abundant; but no such relation appears. Later, considerable 
magma of dioritic composition might yield crystals of gabbroic 
nature while itself as a liquid approaching the composition of a 
granite. However, before it reaches the composition of such a 
granophyr as the red rock of Duluth a large amount of feldspar 
must have crystallized from a magma too acid to yield basic labra- 
dorite; that is, at some stage, a quantity of diorite and grano- 
diorite must have separated. No such rocks appear. Instead, 
there is a very narrow zone of granophyr diabase, in which the 
feldspars, to be sure, are zoned according to theory; but the 
zone is too small to yield the masses of red granophyr actually 
found. 
Thus the characteristic rock series at Duluth is neither of 
Bowen’s recognized series:' “‘gabbro, diorite, quartz diorite, grano- 
diorite, granite’; nor ‘“‘gabbro, diorite, syenite, granite.’ The 
series is rather (1) gabbro, (2) granophyr diabase, (3) granite, with 
the second member very small in bulk. This is believed to be a 
common sequence. Harker recently called attention to the general 
lack of intermediate rocks in such rock masses.? 
t Loc. cit. 
2A. Harker, ‘‘Differentiation in Intercrustal Magma Basins,” Jour. Geol., 
XXIV, 554. 
